Apoptosis and bcl-2 expression in normal human endometrium

Human Reproduction vol.13 no.12 pp.3496–3502, 1998

Apoptosis and bcl-2 expression in normal human endometrium, endometriosis and adenomyosis

Rebecca K.Jones1,2, Roger F.Searle1 and Judith N.Bulmer2,3 Departments of 1Immunology and 2Pathology, University of Newcastle, The Medical School, Framlington Place, Newcastle upon Tyne, NE2 4HH, UK 3To

whom correspondence should be addressed at: Department of Pathology, University of Newcastle, Royal Victoria Infirmary, Newcastle upon Tyne, NE1 4LP, UK

Apoptosis has been implicated in the pathogenesis of several diseases and is partly regulated by bcl-2, which blocks the apoptotic pathway and promotes cell survival. Apoptosis and bcl-2 expression were examined in paired eutopic and ectopic endometrium from women with endometriosis (n J 30 samples) or adenomyosis (n J 15 samples) and compared with control endometrium (n J 30 samples). Apoptotic cells were detected using the dUTP nick-end labelling (TUNEL) assay for DNA fragmentation; bcl-2 expression was demonstrated with a streptavidin–biotin peroxidase immunohistochemical technique. Apoptotic cells were rare in eutopic, ectopic and control endometrium; there were no significant differences between subject groups nor between eutopic and ectopic endometrium. Stromal bcl-2 expression increased in the late secretory phase in control and eutopic endometrium in endometriosis; double labelling studies revealed that most stromal bcl-2F cells were leukocytes. Stromal bcl-2 expression in endometriotic foci was significantly increased compared with the paired eutopic endometrium, did not vary with menstrual cycle and included a significant population of non-leukocytic bcl-2F stromal cells. In contrast, stromal bcl-2 expression in adenomyosis remained at low levels and did not show significant cyclical variation. Glandular epithelial bcl-2 expression also varied with menstrual cycle phase and peaked in the proliferative phase; in contrast, surface epithelial bcl-2 expression increased in the late secretory phase. Elevated stromal bcl-2 expression in ovarian endometriotic lesions could have implications for the growth and survival of ectopic endometrial tissue at these sites. Key words: adenomyosis/apoptosis/bcl-2/endometriosis/endometrium/menstrual cycle

Introduction Programmed cell death and apoptosis are terms used interchangeably to describe a method of cell death which has been 3496

implicated in the pathogenesis of a wide range of diseases (Reed, 1994). Apoptosis, which is distinct from necrosis, is characterized by nuclear fragmentation, formation of membrane-encased apoptotic bodies containing organelles, and cell shrinkage (Kerr et al., 1994). The B-cell lymphoma/leukaemia2 gene (bcl-2) is a proto-oncogene which prevents apoptosis, promoting cell survival by blocking a final common pathway resulting in cell death, although its exact mode of action remains unknown (Vaux et al., 1988; Hockenbery et al., 1990; Hockenbery, 1994). Studies of apoptosis in human endometrium during the menstrual cycle are scarce, sometimes with conflicting results (Tabibzadeh et al., 1994; Yasuda et al., 1995; Harada et al., 1996; Kokawa et al., 1996; Tabibzadeh, 1996). However, there have been several studies of endometrial bcl-2 expression as an indirect indicator of apoptosis. in glandular epithelium, bcl-2 expression varied with menstrual cycle phase; expression peaked at the end of the proliferative phase and was absent during the secretory phase (Gompel et al., 1994; Otsuki et al., 1994; Koh et al., 1995; Tabibzadeh et al., 1995). In late secretory phase stroma, bcl-2 was expressed only by endometrial lymphoid cells (Gompel et al., 1994; Koh et al., 1995; McLaren et al., 1997). These findings suggested that the expression of bcl-2 is regulated by the steroid hormones which control endometrial changes during the menstrual cycle (Otsuki et al., 1994; Koh et al., 1995). Investigation of other cell death regulatory genes, including bax and bcl-x, have suggested that cyclic endometrial changes may be regulated by alterations in the ratio of bcl-2 and related proteins (Tao et al., 1997; McLaren et al., 1997). Endometriosis and adenomyosis are characterized by endometrial tissue in locations other than the uterine cavity, but their aetiology and pathogenesis remain unclear. Recent studies have demonstrated elevated oestrogen receptor expression and altered leukocyte populations in endometriotic lesions compared with eutopic uterine endometrium from the same patients (Jones et al., 1995, 1996; Fujishita et al., 1997). Elevated concentrations of oestrogen receptors in ectopic ovarian locations could lead to upregulation of bcl-2, thereby preventing stromal or epithelial cell apoptosis (Jones et al., 1995). Harada et al. (1996), however, did not detect bcl-2 expression in glandular epithelial cells from ovarian endometriotic tissue and numerous apoptotic cells were detected; in adenomyosis bcl-2 expression was comparable with eutopic endometrium and apoptotic cells were absent. Others have detected bcl-21 macrophages in ectopic endometriotic tissue (McLaren et al., 1997). The aim of the present study was to clarify and compare the expression of bcl-2 and apoptosis in eutopic and ectopic endometrium from women with either © European Society of Human Reproduction and Embryology

bcl-2 and apoptosis in endometriosis

endometriosis or adenomyosis with that in women without endometriosis throughout the menstrual cycle. Materials and methods Tissues All patient and control paraffin tissue blocks were retrieved from archive files in the Department of Pathology, Royal Victoria Infirmary, Newcastle upon Tyne. Ethical approval for the project was granted by the Newcastle Joint Ethics Committee. All samples were obtained with informed consent. Thirty cases of ovarian (n 5 28) and Fallopian tube (n 5 2) endometriosis were selected at different phases of the menstrual cycle (10 proliferative; 10 early secretory; 10 late secretory) using established histological dating criteria (Noyes and Hertig, 1950). Two tissue blocks were examined from each case: one from the uterine endometrium (eutopic) and one from the endometriotic focus (ectopic). Fifteen cases of extensive deep adenomyosis were examined with the corresponding uterine endometrium (five proliferative; five early secretory; five late secretory). The endometrium from all cases showed histological changes consistent with the menstrual dates provided. Any sample showing histological evidence of necrosis was excluded, as were menstrual phase samples. Endometriosis was often unsuspected prior to operation and was diagnosed histologically. Patients currently receiving hormone therapy were excluded, as were cases in which the endometriotic foci showed evidence of active inflammation, which was assessed by neutrophil polymorph and plasma cell infiltration. Thirty blocks of normal endometrium (10 proliferative; 10 early secretory; 10 late secretory) from hysterectomy specimens performed for non-endometrial pathology, such as leiomyomata or benign ovarian cysts, were included as controls. All tissues had been fixed in 10% neutral buffered formalin for 24–48 h and routinely processed into paraffin wax. Sections were cut at 3 µm and mounted on aminopropyltriethoxysilane (APES) coated slides. The separation of specimens into early (days 14–22) and late secretory (day 23 onwards) phases was based on previous studies which have shown maximal changes in the endometrial leukocyte populations in the late secretory phase (Morris et al., 1985; Bulmer et al., 1991; Jones et al., 1996). Immunohistochemistry Single immunohistochemical labelling Expression of bcl-2 was detected using a streptavidin–biotin peroxidase complex immunohistochemical technique with microwave pretreatment for antigen retrieval as described previously (Jones et al., 1995). Briefly, after blocking endogenous peroxidase activity, sections were microwaved in citrate buffer (pH 6), washed in 0.05 M Tris buffered 0.15 M saline, pH 7.6 (TBS) and then incubated sequentially in normal rabbit serum (NRS) (10 min), anti-bcl-2 diluted 1:50 in NRS (bcl-2/100/D5, Novocastra Laboratories, Newcastle upon Tyne, UK) (30 min), biotinylated rabbit anti-mouse immunoglobulins (Dako, High Wycombe, UK) diluted 1:500 in NRS (30 min), and finally with streptavidin–biotin peroxidase complex (Dako) (30 min). Sections were washed in TBS between each incubation. The reaction was developed with 3,39-diaminobenzidine (DAB) (Sigma Chemical Co., Poole, UK) containing 0.02% hydrogen peroxide. Sections were lightly counterstained with Mayer’s haematoxylin, dehydrated, cleared and mounted in DPX synthetic resin (Raymond A. Lamb Ltd, London, UK). Negative controls were performed for each tissue block by replacing the primary antibody with normal serum; tonsil was used as a positive control. Double immunohistochemical labelling Tissues from proliferative and late secretory phase endometriosis cases were first labelled with anti-bcl-2 using the streptavidin–biotin

peroxidase technique as described above and the reaction was developed with nickel-modified DAB (Sigma) to produce a black reaction product. After thorough washing in TBS and 10 min incubation in NRS, the sections were labelled with anti-leukocyte common antigen (LCA) to detect leukocytes (diluted 1:10; Dako) or anti-CD68 to detect macrophages (diluted 1:50; Dako) using an alkaline phosphatase anti-alkaline phosphatase (APAAP) method. The sections were incubated sequentially with the primary monoclonal antibody (30 min), rabbit anti-mouse immunoglobulins (Dako) diluted 1:25 in TBS (30 min) and APAAP (Dako) diluted 1:30 in TBS (30 min). Following this the reaction was developed with Vectared (Vector Laboratories, Peterborough, UK) to produce a red reaction product. A negative control was performed for each stage of the technique and single labelled sections were examined to check the distribution of staining and exclude spurious double labelling; tonsil was used as a positive control. Apoptosis detection system Apoptosis was assessed using the Apoptosis Detection System, Fluorescein (Promega, Southampton, UK) which utilizes the principle of TdT-mediated dUTP nick-end labelling (TUNEL) to detect fragmented DNA in apoptotic cells. Sections were treated according to the manufacturer’s recommendations. Briefly, sections were deparaffinized and rehydrated, permeabilized in proteinase K, refixed in 4% paraformaldehyde and treated with TdT incubation buffer at 37°C for 60 min. After thorough washing in PBS the sections were counterstained with 1 µg/ml propidium iodide (Sigma) in PBS (15 min) and washed in distilled water before mounting in fluorescencefree mountant (Raymond A. Lamb Ltd). Three positive control tissues were included in each assay: a poorly differentiated endometrial carcinoma, rat testis and normal human endometrium treated with DNase (5 µg/ml DNase type 1; Sigma Chemical Co.) for 10 min. Normal endometrium incubated with TdT incubation buffer without the TdT enzyme was also included to estimate non-specific binding and autofluorescence. Quantification Immunohistochemistry bcl-21 cells were identified by the presence of brown nuclear reactivity. The number of stromal bcl-21 cells was assessed in the upper part of the stratum functionalis at 3400 magnification using a 10310 mm graticule either as a number of positive cells in two high power fields, with at least 200 cells per field, or as a percentage of positive cells per minimum of 500 total stromal cells. Epithelial staining was assessed by examining both the percentage of positive cells and the intensity of staining in all glands and in surface epithelium. The percentage of positive cells was scored as 0, no positive cells; 1, 0–25% positive; 2, 25–50% positive; 3, 50–75% positive; 4, 75–100% positive. The intensity of staining was scored as 0, no staining; 1, weak staining; 2, moderate staining; 3, strong staining compared with reactivity in the tonsil positive control. Double-labelled sections were quantified in endometrial stroma by counting the number of single and double labelled cells in two high power (3400) fields. Double-labelled cells were identified as those cells with black (nickel-modified DAB) nuclear staining and red (Vectared) membrane reactivity. Statistical analysis was performed using the Mann–Whitney test with the conventional significance level of P , 0.05. Apoptosis detection system Slides were analysed by fluorescence microscopy with a wide-band excitation barrier filter suitable for analysing both green (fluoresceinlabelled fragmented DNA) and red (propidium iodide nuclear coun-

3497

R.K.Jones, R.F.Searle and J.N.Bulmer

Figure 1. Variation in the absolute number of bcl-2 positive stromal cells per two high power (3400) fields with menstrual cycle phase in control endometrium and eutopic and ectopic endometrium from women with either endometriosis or adenomyosis. Each bar represents the mean number of bcl-2 positive cells with the SEM.

terstain) fluorescence. Apoptotic cells were identified by the presence of red and green nuclear fluorescence in the same cell. Due to the small number of positive cells, a semi-quantitative assessment method was used: the entire section of endometrium, endometriosis or adenomyosis was examined at 3200 magnification and the number of apoptotic cells in surface epithelium, glandular epithelium and stroma was categorized as: 0, no positive cells; 1/–, fewer than one or two positive cells per field; 1, fewer than 10% positive cells per field; and 11, 10–25% positive cells per field.

Results bcl-2 expression in endometrial stroma Changes with menstrual cycle phase The number and percentage of bcl-21 stromal cells in control endometrium and eutopic endometrium in endometriosis significantly increased from the proliferative to the late secretory phase (control: number, P 5 0.041; eutopic: number, P 5 0.008; percentage, P 5 0.023) (Figure 1, Table I). Control and eutopic endometrium did not differ at any cycle phase. In contrast, there were no significant cyclical changes in eutopic endometrium in adenomyosis. bcl-2 expression did not vary with menstrual cycle phase in ectopic endometrium in endometriosis or adenomyosis (Figure 1, Table I). Comparison between eutopic and ectopic endometrium The number of bcl-21 stromal cells in proliferative and early secretory phase ectopic endometrium in endometriosis was significantly higher than in the paired eutopic endometrium from the same patient (proliferative, P 5 0.05; early secretory, 3498

Table I. Percentage of bcl-2 positive stromal cells Mean 6 SEM percentage bcl-2 positive cells Proliferative phase Control endometrium Eutopic endometriosis Ectopic endometriosis Eutopic adenomyosis Ectopic adenomyosis

1.8 0.8 4.1 2.3 0.7

6 6 6 6 6

0.4 0.3 1.4 0.5 0.3

Early secretory Late secretory phase phase 0.7 1.1 5.3 1.8 1.0

6 6 6 6 6

0.3 0.4 1.3 0.3 0.4

3.5 3.7 4.4 1.6 0.8

6 6 6 6 6

0.7 1.4 1.2 0.4 0.6

P 5 0.0009) (Figure 1). In the late secretory phase the difference was not significant (Figure 2a). Similar results were observed for the percentage of bcl-21 cells in eutopic and ectopic endometrium, but the difference was significant only in the early secretory phase (P 5 0.001) (Table I). In contrast, ectopic endometrium in adenomyosis contained lower numbers and percentages of bcl-21 stromal cells than the paired eutopic endometrium throughout the menstrual cycle, but this was significant only in the proliferative phase for the percentage of positive cells (P 5 0.046) (Figures 1 and 2b, Table I). In ectopic endometrium, the number and percentage of bcl-21 stromal cells was higher in endometriosis than in adenomyosis; this was statistically significant in the early secretory (number, P 5 0.024; percentage, P 5 0.004) and late secretory phases (number, P 5 0.05; percentage, not significant) (Figure 1, Table I).


Figure 2. bcl-2 expression in (A) ectopic endometrium in endometriosis in the late secretory phase showing high numbers of positive stromal cells; (B) ectopic endometrium in adenomyosis in the proliferative phase showing few positive stromal cells (arrows). Original magnification 3216. Scale bar represents 46.3 µm.

bcl-2 expression in endometrial epithelium Changes with menstrual cycle phase The percentage and staining intensity of bcl-21 cells in glandular epithelium decreased from the proliferative to the early secretory phase in both control endometrium (percentage, P 5 0.019; intensity, P 5 0.038) and eutopic endometrium in endometriosis (percentage, P 5 0.008; intensity, P 5 0.019); and from the proliferative to the late secretory phase in eutopic endometrium (percentage, P 5 0.006; intensity, P 5 0.019). There were similar trends in eutopic endometrium in adenomyosis, although the differences were not significant (Table II). No significant differences were noted between control and eutopic endometrium (Table II). In foci of endometriosis, the percentage and staining intensity of glandular epithelial bcl-21 cells did not vary with menstrual cycle phase. In contrast, in adenomyosis the percentage of bcl-21 cells in glandular epithelium declined from the proliferative to the late secretory phase (P 5 0.012), although the staining intensity did not alter (Table II). Unlike bcl-2 expression in glandular epithelium, the percentage of bcl-21 cells in surface epithelium increased in the late secretory compared with the proliferative (control, P 5 0.014; eutopic, P 5 0.041) and early secretory phases (control, P 5 0.016; eutopic, P 5 0.031) in both control endometrium and eutopic endometrium in endometriosis (Table II). In eutopic endometrium, surface epithelial bcl-2 staining intensity also increased in the late secretory compared with the proliferative (P 5 0.05) and early secretory phases (P 5 0.034) (Table II). In contrast, in adenomyosis surface epithelial bcl-2 expression reduced in the late secretory phase, although the differences were not significant (Table II). Comparison between eutopic and ectopic endometrium At all phases of the menstrual cycle, there were no differences in the staining intensity or percentage of bcl-21 glandular epithelial cells between eutopic and ectopic endometrium either in endometriosis or in adenomyosis (Table II). However,

Table II. bcl-2 expression in glandular and surface epithelium Proliferative phase Percentage bcl-2-positive Control endometrium glandular surface Eutopic endometriosis glandular surface Ectopic endometriosis glandular surface Eutopic adenomyosis glandular surface Ectopic adenomyosis glandular surface

Early secretory Late secretory phase phase

cells 6 SEM 2.4 6 0.5 0.5 6 0.5

0.5 6 0.4 0.7 6 0.4

1.3 6 0.5 3.6 6 0.3

2.8 6 0.5 0.9 6 0.5

0.6 6 0.4 0.6 6 0.5

0.6 6 0.3 2.8 6 0.6

2.2 6 0.6 NA

1.1 6 0.5 NA

1.9 6 0.6 NA

4.0 6 0.0 1.6 6 1.2

1.2 6 0.8 0.0 6 0.0

1.2 6 0.8 0.3 6 0.3

4.0 6 0.0 NA

3.0 6 0.6 NA

0.8 6 0.6 NA

0.3 6 0.2 0.4 6 0.2

0.6 6 0.2 1.3 6 0.2

0.3 6 0.2 0.3 6 0.2

0.3 6 0.2 1.0 6 0.2

0.7 6 0.3 NA

0.9 6 0.2 NA

0.4 6 0.2 0.0 6 0.0

0.4 6 0.2 0.3 6 0.3

1.0 6 0.3 NA

0.4 6 0.2 NA

bcl-2 staining intensitya 6 SEM Control endometrium glandular 1.0 6 0.2 surface 0.1 6 0.1 Eutopic endometriosis glandular 1.0 6 0.2 surface 0.3 6 0.2 Ectopic endometriosis glandular 1.0 6 0.3 surface NA Eutopic adenomyosis glandular 1.4 6 0.2 surface 0.6 6 0.3 Ectopic adenomyosis glandular 1.2 6 0.2 surface NA

NA 5 not applicable. aScored on a scale 0 (no staining) to 3 (strong staining).

compared with endometriosis, ectopic endometrium in adenomyosis contained a higher percentage of bcl-21 glandular epithelial cells in the early secretory phase (P 5 0.045) (Table II). 3499

R.K.Jones, R.F.Searle and J.N.Bulmer

Table III. Double immunohistochemical labelling for bcl-2 and LCA or CD68 bcl-2 LCA double labelling (mean 6 SEM)a bcl-21 Proliferative 1.3 6 0.6 phase, eutopic Proliferative 15.8 6 6.3 phase, ectopic Late secretory 4.5 6 2.2 phase, eutopic aMean

bcl-2 CD68 double labelling (mean 6 SEM)a

LCA1

bcl-21 LCA1

bcl-21

CD681

17.0 6 3.4

14.5 6 6.8

11.5 6 0.6

8.8 6 1.1

0.0 6 0.0

46.8 6 16.7

31.3 6 4.6

76.8 6 26.1 11.3 6 2.7

1.3 6 1.0

68.5 6 18.6

36.8 6 10.3

32.3 6 13.1

0.0 6 0.0

8.3 6 1.9

bcl-21 CD681

number of cells per two high power (3400) fields.

Comparison between glandular and surface epithelium Glandular epithelium in proliferative control and endometriotic eutopic endometrium contained significantly more bcl-21 cells than surface epithelium (control, P 5 0.028; eutopic, P 5 0.017). In contrast, in the late secretory phase, there was an increased percentage of bcl-21 cells in surface compared with glandular epithelium (control, P 5 0.033; eutopic, P 5 0.016); similar trends were observed for bcl-2 staining intensity (proliferative phase: control, P 5 0.025; eutopic, P 5 0.019; late secretory phase: eutopic, P 5 0.05) (Table II). In adenomyosis, both the percentage and intensity of surface and glandular bcl-21 cells reduced in the late secretory phase but the differences were not statistically significant (Table II).

and there was no difference between ectopic and the paired eutopic endometrium; moreover, the number of apoptotic cells did not vary with menstrual cycle phase. A similar lack of apoptotic cells was seen in foci of adenomyosis; three of 15 cases contained a small number (1/–) of apoptotic cells but there were no significant differences between eutopic and ectopic endometrium, or with cycle phase. Interestingly, in most endometrial samples, macrophages showed positive fluorescence due to the presence of fragmented DNA within their cytoplasm, although their nuclei were negative. Macrophages provided a good internal positive control for the assay but were not included in the quantification.

Double immunohistochemical labelling Most bcl-21 cells in proliferative and late secretory phase eutopic endometrium in endometriosis were LCA1; bcl-2 single labelled cells were scanty, but many LCA single positive cells were identified. In proliferative phase ectopic endometrium the numbers of LCA single positive cells and bcl-2/LCA double labelled cells were increased compared with the paired eutopic endometrium. There were also increased numbers of bcl-2 single positive cells. Only very occasional bcl-2/CD68 double-labelled cells were detected, although there were numerous single labelled cells of each type (Table III).

Discussion Apoptosis and bcl-2 expression have been quantified in normal endometrium, endometriosis and adenomyosis. Epithelial bcl-2 expression in normal endometrium was similar to that reported previously (Gompel et al., 1994; Otsuki et al., 1994; Tabibzadeh et al., 1995; McLaren et al., 1997). High bcl-2 expression in proliferative glandular epithelium suggests a close relationship to oestrogen production and is correlated with oestrogen receptor expression in endometrial epithelium (Snijders et al., 1992; Coppens et al., 1993; Jones et al., 1995). The significant increase in surface epithelial bcl-21 cells in the late secretory phase may reflect upregulation by luteal phase oestrogen production. Timed endometrial samples or samples with corresponding serum oestrogen measurements would be required to investigate this more thoroughly. Epithelial expression of bcl-2 in endometriotic foci showed similar, but not statistically significant, cyclical changes, suggesting that the endometriotic lesions respond to regulatory factors in the same way as epithelium in normal cycling endometrium. Ectopic endometrium contains significantly more oestrogen receptors than eutopic endometrium at all phases of the cycle (Jones et al., 1995; Fujishita et al., 1997). Therefore, if bcl-2 expression was regulated solely by oestrogen, ectopic endometrium would be expected to express high bcl-2 levels. Since there was no increase in endometriotic foci, bcl-2 expression may be influenced by factors other than oestrogen such as local or systemic cytokines (Park, 1996; Cohen et al., 1997). In contrast, Harada et al. (1996) reported a lower proportion of bcl-21 cells in ectopic endometrium, although these workers also found no significant differences

Apoptosis detection The majority of cells in the DNase treated normal human endometrium displayed nuclear reactivity, demonstrating that the fluorescein-12-dUTP had been incorporated at the 39-OH ends of the fragmented DNA (11). Similarly, tissue from the poorly differentiated endometrial carcinoma (1) and the rat testis (1) contained positive cells as well as the nuclei having formed small, round apoptotic bodies. In contrast, normal endometrium and eutopic and ectopic endometrium from women with endometriosis or adenomyosis rarely contained apoptotic stromal or epithelial cells. In one of 30 samples of eutopic endometrium from endometriosis and six of 15 samples from adenomyosis a 1/– categorization was given, with only occasional epithelial apoptotic cells being observed. These positive intraepithelial cells were small with round nuclei and were considered to be intraepithelial leukocytes; no apoptotic stromal cells were seen. In endometriotic lesions, only three of 30 cases contained apoptotic cells (1/–) 3500


in epithelial bcl-2 expression between eutopic and ectopic endometrium in adenomyosis. Double immunohistochemical labelling revealed that most stromal bcl-21 cells were leukocytes. The rise in bcl-2 expression in the late secretory phase is, therefore, concomitant with the increase in leukocyte numbers in the late secretory phase which is predominantly due to CD561 granulated lymphocytes (eGL) (Morris et al., 1985; King et al., 1989; Bulmer et al., 1991; Klentzeris et al., 1992; Jones et al., 1996), some of which express bcl-2 (Koh et al., 1995; Jones et al., 1998a). Ectopic endometrium, however, contains reduced numbers of eGLs and elevated numbers of CD31 CD81 T cells (Jones et al., 1996, 1998b), and it is thus likely that the bcl-21 cells in ectopic endometrium are T cells. Furthermore, bcl-2 expression in ectopic stromal endometrium did not vary with cycle phase which concurs with similar findings for leukocyte numbers in ectopic endometrium (Jones et al., 1996). Double labelling studies also demonstrated that some bcl21 stromal cells were not leukocytes; these cells were present at higher levels in ectopic endometrium. The presence of bcl21 stromal cells in ectopic but not eutopic endometrium suggests upregulated bcl-2 expression within endometriotic lesions which may reflect upregulated stromal oestrogen receptor expression levels (Jones et al., 1996; Fujishita et al., 1997). bcl-2 expression is believed to block a final common pathway during the process of apoptosis (Vaux et al., 1988; Hockenbery et al., 1990; Hockenbery, 1994); hence, bcl-2 up-regulation may prevent programmed cell death in ectopic endometrium, thereby promoting the development of endometriosis. bcl-2 expression and long-term survival of leukocytes in ectopic tissue may also contribute to disease pathogenesis: leukocytes can produce cytokines which may promote disease development or cause infertility. The results of this study, however, conflict with a recent report of bcl-21 macrophages in ectopic, but not eutopic, endometrium (McLaren et al., 1997). In previous studies, fewer macrophages were detected in ectopic endometrium (Jones et al., 1996). The present study therefore provides no support for the view that apoptotic-resistant macrophages are involved in the pathophysiology of endometriosis (McLaren et al., 1997). The variation in stromal bcl-2 expression between foci of adenomyosis and ovarian endometriosis strengthens the view that endometriosis in different regions of the body should be considered separately (Howell et al., 1994; Jones et al., 1995, 1996). Differing bcl-2 expression may alter the response to hormonal therapy and may account for the observation that the response to hormonal therapy in endometriosis varies between patients and between different lesions in any one woman (Schweppe et al., 1984; Metzger et al., 1988). Foci of adenomyosis contain significantly fewer leukocytes than endometriotic foci at all phases of the menstrual cycle (Jones et al., 1988b); hence, it is possible that the low bcl-2 expression reflects reduced leukocyte numbers. These data suggest that adenomyotic lesions resemble eutopic endometrium rather than endometriotic lesions. bcl-2 is only one molecule of a series involved in apoptosis inducing/inhibiting pathways (Kerr et al., 1994), some of which have been demonstrated in endometriotic lesions (Harada

et al., 1996; McLaren et al., 1997). The TUNEL assay allows direct detection of fragmented DNA by catalytically incorporating fluorescein-12-UTP at the 39-hydroxyl ends of fragmented DNA using terminal deoxynucleotidyl transferase (TdT) (Gavrieli et al., 1992). However, this method cannot distinguish between apoptotic and necrotic cells, as both contain fragmented DNA (Gold et al., 1994). In the present study, only occasional apoptotic cells were observed using the TUNEL assay and there was no necrosis as assessed by morphological criteria. Harada et al. (1996) also detected apoptotic cells in only five of 24 cases of eutopic endometrial epithelium, but apoptotic cells were detected in all cases of ovarian endometriosis. These discrepant results may reflect technical differences between the two studies. Harada et al. (1996) used an indirect avidin–biotin peroxidase TUNEL assay, whereas directly conjugated fluorescein–dUTP was used in the present study. Several studies have provided convincing evidence for steroid hormone control of apoptosis in uterine epithelium in other mammals (Nawaz et al., 1987; Pollard et al., 1987; Rotello et al., 1989, 1991), but the present findings suggest that this may not be the case in humans. The absence of apoptotic cells in the stroma of endometriotic lesions and adenomyosis is not surprising in view of the high levels of bcl-2 expression in endometriotic lesions and the nature of the disease. In conclusion, this is a fully quantified, statistically analysed study of bcl-2 expression in normal endometrium and eutopic and ectopic endometrium in endometriosis and adenomyosis. bcl-2 expression peaked in glandular epithelium in the proliferative phase and in stroma in the late secretory phase. No significant apoptosis was detected in normal endometrium or in eutopic or ectopic endometrium from women with endometriosis and adenomyosis. Whether the prolonged survival of bcl-21 lymphoid and non-lymphoid cells in ectopic endometrium plays a role in the aetiology of endometriosis is presently under investigation.

Acknowledgements We would like to thank Miss S.Jones and Mrs B.Innes for their expert technical help as well as the pathologists of the Department of Pathology, Royal Victoria Infirmary for their help in obtaining specimens. This work was supported by a grant from The Sir Jules Thorn Charitable Trust (9312A).

References Bulmer, J.N., Morrison, L., Longfellow, M. et al. (1991) Granulated lymphocytes in human endometrium: histochemical and immunohistochemical studies. Hum. Reprod., 6, 791–798. Cohen, S.B., Crawley, J.B., Kahan, M.C. et al. (1997) Interleukin-10 rescues T cells from apoptotic cell death: association with upregulation of Bcl-2. Immunology, 92, 1–5. Coppens, M.T., Dhont, M.A., De Boever, J.G. et al. (1993) The distribution of oestrogen and progesterone receptors in the human endometrial basal and functional layer during the normal menstrual cycle. An immunocytochemical study. Histochemistry, 99, 121–126. Fujishita, A., Nakane, P.K., Koji, T. et al. (1997) Expression of estrogen and progesterone receptors in endometrium and peritoneal endometriosis: an immunohistochemical and in situ hybridization study. Fertil. Steril., 67, 856–864.

3501

R.K.Jones, R.F.Searle and J.N.Bulmer Gavrieli, Y., Sherman, Y. and Ben-Sasson, S.A. (1992) Identification of programmed cell death in situ via specific labelling of nuclear DNA fragmentation. J. Cell Biol., 119, 493–501. Gold, R., Schmied, M., Giegerich, G. et al. (1994) Differentiation between cellular apoptosis and necrosis by the combined use of in situ tailing and nick translation techniques. Lab. Invest., 71, 219–225. Gompel, A., Sabourin, J.C., Martin, A. et al. (1994) Bcl-2 expression in normal endometrium during the menstrual cycle. Am. J. Pathol., 144, 1195–1202. Harada, M., Suganuma, N., Furuhashi, M. et al. (1996) Detection of apoptosis in human endometriotic tissues. Mol. Hum. Reprod., 2, 307–315. Hockenbery, D.M. (1994) bcl-2 in cancer, development and apoptosis. J. Cell Sci., 18, 51–55. Hockenbery, D., Nunez, G., Milliman, C. et al. (1990) Bcl-2 is an inner mitochondrial membrane protein that blocks programmed cell death. Nature, 348, 334–336. Howell, R.J., Dowsett, M. and Edmonds, D.K. (1994) Oestrogen and progesterone receptors in endometriosis: heterogeneity of different sites. Hum. Reprod., 9, 1752–1758. Jones, R.K., Bulmer, J.N. and Searle, R.F. (1995) Immunohistochemical characterization of proliferation, oestrogen and progesterone receptor expression in endometriosis: comparison of eutopic and ectopic endometrium with normal cycling endometrium. Hum. Reprod., 10, 3272– 3279. Jones, R.K., Bulmer, J.N. and Searle, R.F. (1996) Immunohistochemical characterization of stromal leukocytes in ovarian endometriosis: comparison of eutopic and ectopic endometrium with normal endometrium. Fertil. Steril., 66, 81–89. Jones, R.K., Searle, R.F., Stewart, J.A. et al. (1998a) Apoptosis, bcl-2 expression, and proliferative human activity in human endometrial stroma and endometrial granulated lymphocytes. Biol. Reprod., 58, 995–1002. Jones, R.K., Searle, R.F. and Bulmer, J.N. (1998b) Phenotypic and functional studies of leukocytes in human endometrium and endometriosis. Hum. Reprod. Update, in press. Kerr, J.F.R., Winterford, C.M. and Harmon, B.V. (1994) Apoptosis. Its significance in cancer and cancer therapy. Cancer, 73, 2013–2026. King, A., Wellings, V., Gardner, L. et al. (1989) Immunocytochemical characterization of the unusual large granular lymphocytes in human endometrium throughout the menstrual cycle. Hum. Immunol., 24, 195–205. Klentzeris, L.D., Bulmer, J.N., Warren, A. et al. (1992) Endometrial lymphoid tissue in the timed endometrial biopsy: morphometric and immunohistochemical aspects. Am. J. Obstet. Gynecol., 167, 667–674. Koh, E.A., Illingworth, P.J., Duncan, W.C. et al. (1995) Immunolocalization of bcl-2 protein in human endometrium in the menstrual cycle and simulated early pregnancy. Hum. Reprod., 10, 1557–1562. Kokawa, K., Shikone, T. and Nakano, R. (1996) Apoptosis in the human uterine endometrium during the menstrual cycle. J. Clin. Endocrinol. Metab., 81, 4144–4147. McLaren, J., Prentice, A., Charnock-Jones, D.S. et al. (1997) Immunolocalization of the apoptosis regulating proteins Bcl-2 and Bax in human endometrium and isolated peritoneal fluid macrophages. Hum. Reprod., 12, 146–152. Metzger, D.A, Olive, D.L. and Haney, A.F. (1988) Limited hormonal responsiveness of ectopic endometrium: histologic correlation with intrauterine endometrium. Hum. Pathol., 19, 1417–1424. Morris, H., Edwards, J., Tiltman, A. et al. (1985) Endometrial lymphoid tissue: an immunohistological study. J. Clin. Pathol., 38, 644–652. Nawaz, S., Lynch, M.P., Galand.P. et al. (1987) Hormonal regulation of cell death in rabbit epithelium. Am. J. Pathol., 127, 51–59. Noyes, R.W. and Hertig, A.T. (1950) Dating the endometrial biopsy. Fertil. Steril., 1, 3–25. Otsuki, Y., Misaki, O., Sugimoto, O. et al. (1994) Cyclic bcl-2 gene expression in human uterine endometrium during menstrual cycle. Lancet, 344, 28–29. Park, J.R. (1996) Cytokine regulation of apoptosis in hematopoietic precursor cells. Curr. Opin. Hematol., 3, 191–196. Pollard, J.W., Pacey, J., Cheng, S.V. et al. (1987) Estrogens and cell death in murine uterine luminal epithelium. Cell Tiss. Res., 249, 533–540. Reed, J.C. (1994) Bcl-2 and the regulation of programmed cell death. J. Cell Biol., 124, 1–6. Rotello, R.J., Hocker, M.B. and Gerschenson, L.E. (1989) Biochemical evidence for programmed cell death in rabbit uterine epithelium. Am. J. Pathol., 134, 491–495. Rotello, R.J., Lieberman, R.C., Purchio, A.F. et al. (1991) Co-ordinated regulation of apoptosis and cell proliferation by transforming growth factor

3502

beta 1 in cultured uterine epithelial cells. Proc. Natl Acad. Sci. USA, 88, 3412–3415. Schweppe, K.W., Wynn, R.M. and Beller, F.K. (1984) Ultrastructural comparison of endometriotic implants and eutopic endometrium. Am. J. Obst. Gynecol., 148, 1024–1039. Snijders, M.P., de Goeij, A.F., Debets-Te Baerts, M.J. et al. (1992) Immunocytochemical analysis of oestrogen receptors and progesterone receptors in the human uterus throughout the menstrual cycle and after the menopause. J. Reprod. Fert., 94, 363–371. Tabibzadeh, S. (1996) The signals and molecular pathways involved in human menstruation, a unique process of tissue destruction and remodelling. Mol. Hum. Reprod., 2, 77–92. Tabibzadeh, S., Kong, Q.F., Satyaswaroop, P.G. et al. (1994) Distinct regional and menstrual cycle dependent distribution of apoptosis in human endometrium. Potential regulatory role of T cells and TNF-alpha. Endocrinol. J., 2, 87–95. Tabibzadeh, S., Zupi, E., Babaknia, A. et al. (1995) Site and menstrual cycledependent expression of proteins of the tumour necrosis factor (TNF) receptor family, and BCL-2 oncoprotein and phase-specific production of TNF alpha in human endometrium. Hum. Reprod., 10, 277–286. Tao, X.J., Tilly, K.I., Maravei, D.V. et al. (1997) Differential expression of members of the bcl-2 gene family in proliferative and secretory human endometrium glandular epithelial cell apoptosis is associated with increased expression of bax. J. Clin. Endocrinol. Metab., 82, 2738–2746. Vaux, D.L., Cory, S. and Adams, J.M. (1988) Bcl-2 gene promotes haemopoietic cell survival and co-operates with c-myc to immortalize preB cells. Nature, 335, 440–442. Yasuda, M., Umemura, S., Osamura, R.Y. et al. (1995) Apoptotic cells in the human endometrium and placental villi: pitfalls in applying the TUNEL method. Arch. Histol. Cytol., 58, 185–190. Received on February 6, 1998; accepted on August 25, 1998