- Email: [email protected]
Expression of α4β1 and αvβ3 integrins in the endometrium of women using the T200 copper intrauterine device
FERTILITY AND STERILITY威 VOL. 74, NO. 6, DECEMBER 2000 Copyright ©2000 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A.
Expression of ␣41 and ␣v3 integrins in the endometrium of women using the T200 copper intrauterine device Ricardo Savaris, M.D.,a Cla´udio Galleano Zettler, M.D.,a and Arnaldo Nicola Ferrari, M.D.b Fundac¸a˜o Faculdade Federal Cieˆncias Me´dicas and Fundac¸a˜o Universita´ria de Endocrinologia e Fertilidade, Porto Alegre, Brazil
Objective: To investigate the expression of ␣4 and 3 integrin subunit levels in the endometrium of healthy women and copper intrauterine device (IUD) T200 users. Design: Case control study. Setting: An academic teaching hospital and a primary care clinic. Patient(s): Thirteen copper IUD users and 13 normal fertile women. Intervention(s): Timed endometrial biopsies during the mid-secretory phase (days 20 to 24). Main Outcome Measure(s): Histologic dating of endometrium and immunohistochemical staining intensity of ␣4 and 3, using the semiquantitative immunohistochemical score (HSCORE). Result(s): All endometrial biopsies consistent with menstrual dates were examined for integrin expression (3 and ␣4). No difference in ␣4 integrin expression was found between IUD users and controls in both luminal and glandular epithelium. In fertile controls, ␣v3 staining was present in 100% and 38.4% of glandular and luminal epithelium, respectively. In contrast, only 61.5% of the IUD users had any ␣v3 staining in the glandular epithelium and only 53.9% in the luminal epithelium. The intensity of immunoreactivity between the two groups (mean HSCORE) did not differ significantly. Conclusion(s): Proportionately, significantly fewer women using copper IUD had positive ␣v3 immunoreactivity in the glandular epithelium of mid-secretory endometrium. (Fertil Steril威 2000;74:1102–7. ©2000 by American Society for Reproductive Medicine.) Key Words: Contraception, integrins, IUD, reproduction, endometrium, implantation
Received December 14, 1999; revised and accepted June 12, 2000. Supported in part by Conselho Nacional de Desenvolvimento Cientı´fico e Tecnolo´gico, grant number 142161/1997-7, Porto Alegre, Brazil. Reprint requests: Ricardo Savaris, M.D., Rua Reis Louzada, 33, Porto Alegre, RS, 90630-130, Brazil (FAX: 55-51-3404514; E-mail: savaris@orion .ufrgs.br). a Fundac¸a˜o Faculdade Federal Cieˆncias Me´dicas. b Fundac¸a˜o Universita´ria de Endocrinologia e Fertilidade. 0015-0282/00/$20.00 PII S0015-0282(00)01600-9
1102
After oral contraceptives, the intrauterine device (IUD) probably is the most common contraceptive method used in the world (1). However, its mechanism of action is controversial and as yet not well understood (2, 3). It has been suggested that IUDs act by interfering with sperm transport (4), fertilization (5), and implantation (6, 7). It is also believed that IUDs stimulate an inflammatory response in the uterus (3). Integrins, a family of cell adhesion molecules, consist of transmembrane ␣ and  subunits and are involved in cell– cell and cell– substratum interactions (8). These cell adhesion receptors are involved in a variety of biological processes (9) and are widely expressed, including in human endometrium (10). Integrin expression profiles have been demonstrated to be feasible by standard immunohistochemical techniques (10 –12).
Three specific integrin subunits, ␣1, ␣4, and 3, show cycle-specific patterns of expression in the glandular and luminal epithelium in human endometrium during the menstrual cycle (13). The abrupt appearance of ␣v3 on day 20 of the menstrual cycle (14) is coincident with the opening of the window of implantation, as described by Hertig et al. (15); the disappearance of ␣4-subunit on cycle day 24 sets its closure (13). The coexpression of these three integrins subunits is related to the putative “window of implantation” that is thought to occur on cycle days 20 to 24 (16). The expression of endometrial ␣v3 is abnormal in women with endometriosis (17), hydrosalpinges (18), luteal phase defect (8), and unexplained infertility (19). Nevertheless, some authors did not find this abnormality in patients with endometriosis (20). Because of
the temporal pattern of their expression around the time of implantation and their absence in conditions related to infertility, these integrins might have an important role as a potential marker of uterine receptivity (21, 22). We hypothesize that contraception by copper IUD may occur via alteration in the uterine receptivity, specifically the integrin expression profile. We conducted the present study to determine whether copper IUDs induce abnormal integrin expression during the window of implantation.
MATERIALS AND METHODS Study Design The use of human tissue for research was based on informed consent and was approved by the institutional review board of the General Hospital of Porto Alegre (HCPA). A case control study was performed comparing the expression of ␣41 and ␣v3 integrins in the endometrium of copper IUD users from a primary health care unit, to fertile controls who were undergoing tubal sterilization or hysterectomy for benign conditions. The IUD users, normally cycling women 18 to 40 years old, were invited to participate in the study by letter. The criteria for inclusion were no use of hormones over the last 3 months, a fixed partner, proven fertility (parity ⱖ1), negative exams for sexual transmitted diseases (gonorrhea and chlamydia), no previous history of pelvic inflammatory disease or endometriosis, and a recent normal pap smear. All of the women who did not fulfill these inclusion criterion were excluded. In addition, the endometrial biopsies performed during the study that were not “in phase” (dyssynchrony of ⱕ 2 days based on the time of ovulation) were also excluded.
Tissue Acquisition All of the biopsies were obtained with a plastic flexible curette (Z-Sampler, BEI Medical Instruments, San Francisco, CA) during the mid-luteal phase 6 to 10 days after ovulation (LH ⫹6 ⬃ ⫹10), based on a measurement of the urinary LH surge (LH kit, Expomed Inc., San Antonio, CA). The endometrial samples were divided for histologic dating and immunohistochemical staining. Endometrial histologic dating was performed according to the criteria of Noyes et al. (23).
Immunohistochemistry and Monoclonal Antibodies
Immunoperoxidase staining of 8-m cryostat sections was performed using a technique described elsewhere in detail (10). Briefly, frozen tissue sections were placed on poly-L-lysine-coated slides and fixed in acetone at ⫺20°C for 10 minutes. After a rinse with phosphate-buffered saline (PBS) at pH 7.2 to 7.4, the endogenous peroxidases were quenched with a 30 minute incubation with 0.3% H2O2 FERTILITY & STERILITY威
(Sigma Chemical, St. Louis, MO) in absolute methanol (Mallinckrodt, Paris, KY). After a new rinse with PBS, unspecific sites were blocked with 2% horse serum albumin (Vector Laboratories, Burlingame, CA) for 30 minutes. Monoclonal antibodies against ␣4 (B-5H10) and 3 (SSA6), kindly provided by Prof. B. A. Lessey, were incubated overnight at ⫹4°C at the dilution of 1:2,500 and 1:500, respectively. After a rinse with PBS for 3 minutes, secondary antibody consisting of biotinylated horse anti-mouse (Vector) was incubated for 30 minutes at room temperature. Avidin-biotinylated horseradish peroxidase macromolecular complex (ABC, Vector) was then incubated on the sections for 30 minutes. The diaminobenzidine (DAB; Sigma) was used as a chromogen to complete the reaction. The samples were counterstained with hematoxylin and mounted. The resulting staining was evaluated on a Nikon microscope at low (⫻100) and higher (⫻400) magnification by two independent readers in a blind fashion who used the histologic score (HSCORE) described elsewhere (17). An HSCORE greater than 0.7 was defined as positive staining (17). In previous studies, the HSCORE has been shown to have relatively low intraobserver (r ⫽ 0.983; P⫽.00001) and interobserver (r ⫽ 0.994; P⫽.00001) variation (24).
Statistical Analysis and Sample Size Comparisons of the means of the HSCORE of both groups were performed with the use of the Student t test. Fisher’s exact test was used to compare nonparametric dichotomic data (IUD/controls ⫻ integrin positive/negative). The sample size was calculated based on a formula described elsewhere (25); it considered ␣ and  errors of 5% and 20%, respectively, variance of the expression of the integrins as 0.79 according to previous report (19), and a minimal difference between the HSCORE of 0.9. The result was 13 subjects in each group. Statistical analysis was performed using Excel 7.0 software (Microsoft Corporation, Redmond, WA) and EPI Info 5.01b software (Centers for Disease Control, Atlanta, GA).
RESULTS Letters were sent to 152 copper IUD users who were registered in a secondary health care that invited them to participate in the study. Twenty-four patients attended the first consultation and agreed to participate. From these 24 patients, 7 (29.16%) quit or changed their address, making follow-up impossible. Three (12.5%) were excluded for not fulfilling the inclusion criteria. Fourteen endometrial samples were obtained from the remaining patients. One biopsy from an IUD user was out of phase, so this patient was excluded from the study. Control samples were obtained from patients at a tertiary health care 1103
FIGURE 1 Immunohistochemical localization of ␣4 and 3 integrin subunits in the endometrium from controls and from copper IUD T200 users. (A), A positive staining for mouse anti-human 3-subunit in the endometrial gland (arrow) on day 8 after ovulation in a control patient’s menstrual cycle. (B), A positive staining for ␣4 in a gland in an IUD user (arrow) and negative expression in the lumen as expected (arrow head). (C), The same sample stained for 3. Observe the negative expression in the gland (asterisk) and in the lumen (arrowhead). (D), A patient with negative 3 in glands (asterisk) and lumen (arrowhead), but with a positive-staining stromal epithelium (white arrow). (Original magnification, ⫻200.)
Savaris. Endometrial integrin in IUD T200 users. Fertil Steril 2000.
center who had had a tubal sterilization or a hysterectomy for a benign condition (patient number 5, 36 years old, had a hysterectomy for menorrhagia). No difference was observed between the two groups when age, parity, and day of biopsy were compared (Table 1). No complications or displacements of the contraceptive devices were observed. Table 2 shows the individual HSCORE of the patients. As expected, ␣4 was not found in the luminal epithelium of both groups, but was present in 76.9% and in 69.2% of the glandular 1104 Savaris et al.
Endometrial integrin in IUD T200 users
endometrium of IUD users and controls, respectively. The 3 subunit had a similar distribution between both groups, although five samples from IUD users had an HSCORE of zero for the glandular 3 subunit. No statistically significant difference was found in the mean HSCORE expression of ␣v3 and ␣41 in the luminal and glandular epithelium at days LH⫹6 ⬃ ⫹10 of the cycle (Student t test; P⬎.05) between IUD users and controls, as shown in Figure 2. Although there were no statistically significant differVol. 74, No. 6, December 2000
FIGURE 2
FIGURE 3
Immunohistochemical staining (HSCORE) and distribution of endometrial specimens obtained from controls (n ⫽ 13) and IUD users (n ⫽ 13). Results are means ⫾ SD. 䊐 ⫽ IUD. ⫽ control.
Percentage of positive immunoperoxidase staining (HSCORE ⬎ 0.7) for ␣4 and 3 integrin subunits in luminal and glandular endometrium in copper IUD users (n ⫽ 13) and controls (n ⫽ 13). 䊐 ⫽ IUD. ⫽ control. *Fisher’s exact test ⫽ 0.01.
Savaris. Endometrial integrin in IUD T200 users. Fertil Steril 2000. Savaris. Endometrial integrin in IUD T200 users. Fertil Steril 2000.
ences in the percentage of positive immunoreactivity for luminal and glandular ␣41 and luminal ␣v3 (P⬎.05, 2) between both groups, there were significantly fewer women who had any glandular epithelium ␣v3 staining among the IUD users compared to controls (61.5% vs. 100%, P⫽.01, Fisher’s exact test) as shown in Figures 1, 2 and 3.
DISCUSSION Intrauterine devices (IUD) are one of the most effective existing contraceptive methods. IUDs are thought to function in part by diminishing uterine receptivity by virtue of their nature as foreign bodies. However, there are very few published studies investigating the cellular mechanism of action of the IUD. To our knowledge, there are no published studies on the effects of the copper IUD on the expression of ␣v3. The ␣v3 integrin is increasingly recognized and used as a marker of uterine receptivity (19, 22, 26). The abrupt appearance of ␣v3 on day 20 of the menstrual cycle (14) is coincident with the opening of the window of implantation,
TABLE 1 Sample characterization. Controls (n ⫽ 13) Mean ⫾SD
IUDs (n ⫽ 13) Mean ⫾ SD
P value
27.9 ⫾ 5.18 2.5 ⫾ 0.77 23.3 ⫾ 0.58
30.1 ⫾ 3.18 3.2 ⫾ 1.64 22.6 ⫾ 1.58
0.19 NSa 0.18 NSa 0.18 NSa
Age (y) Parity Day of biopsy a
NS: not significant.
Savaris. Endometrial integrin in IUD T200 users. Fertil Steril 2000.
FERTILITY & STERILITY威
whereas the disappearance of ␣4-subunit on cycle day 24 sets its closure (13). The coexpression of these integrin subunits is related to the putative “window of implantation” thought to occur on cycle days 20 to 24 (16). With the use of HSCORE, we found no difference in the expression of ␣41 and ␣v3 integrins in either the luminal or glandular endometrium between women using and not using copper IUD (see Fig. 2). A significant difference was observed in the percentage of samples staining negative for ␣v3 integrin in the endometrium of IUD users (see Fig. 3). A number of investigators have used the quantitative analysis of the immunoreactivity of integrins and their mean levels to compare different treatment groups (27), but its clinical relevance is a matter that still needs further evaluation. No cut-off has been demonstrated to segregate the positive and negative expressions of the integrin that correlate to a clinical function. We used an HSCORE of 0.7 as a cut-off to define positive vs. negative staining based on prior work by Lessey et al. (17). Their study determined, through receiver operating characteristic curve analysis, that the cutoff of 0.77 correlated well with the presence of endometriosis. The endometrium of the majority of IUD users (61.5%) showed comparable ␣v3 staining intensity to that of control endometrium, but the trend toward complete absence of staining in the setting of in-phase endometrium may be clinically significant and needs further investigation. The absence of ␣v3 staining may occur in two ways. It can be a result of a histologic delay (type I defect), or may occur when there is no histologic delay (type II defect) (27). Type II defect has been previously reported in women with hydrosalpinx (18) and endometriosis (17), both of which are associated with decreased fertility. The changes in integrin expression profiles in women with hydrosalpinx and endo1105
TABLE 2 Individual HSCORE for ␣4 and 3 integrin subunits in luminal and glandular endometrium of copper intrauterine device (IUD) users and in controls. IUD
Control
␣4 Patient 1 2 3 4 5 6 7 8 9 10 11 12 13
3
␣4
3
Gland
Lumen
Gland
Lumen
Gland
Lumen
Gland
Lumen
2.2 3 2.7 0.6 1.5 1.5 3.5 2.1 2.2 2.2 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0
2.9 3 0 2.8 2.7 0 0 1.5 1.5 2.2 0 0 2.3
2.6 3.5 0.2 2.5 2.4 0 0 1.4 1.6 3.5 0 0 0
1.7 0.8 0 0 0 1.5 1.9 0.4 0 3.4 0.2 1.1 2.7
0 0 0 0 0 0 0 0 0 0 0 0 0
1.5 2 2.3 1.2 2.7 2.4 1.2 2.6 2.4 2.3 3.5 1.2 2
0 0 3 0 1.1 0.6 0 2.7 0 2 2 0 2.7
Savaris. Endometrial integrin in IUD T200 users. Fertil Steril 2000.
metriosis may be analogous to the changes seen in some of the IUD users. The Type II defect in the ␣v3 staining pattern may arise secondary to local inflammatory responses induced by the presence of a copper IUD (28). IUD users show a tendency toward an increase in the neutrophil, mononuclear, and plasma cell type subpopulation of lymphocytes in the endometrium, unrelated to premenstrual changes (29). These cells might secrete some factor, such as basic fibroblast growth factor (bFGF), that is related to diminished expression of ␣v3 (30). Copper IUD releases free copper and copper salts, which may have a direct biochemical impact on the endometrium and lead to morphologic changes (31). Additionally, EDTA, a calcium chelator, has been shown to inhibit in vitro binding of ␣1, ␣6 integrins to the extracellular matrix (32). Other potential mechanisms leading to alteration in ␣v3 expression include a failure to down-regulate the expression of progesterone receptors (8). It was demonstrated that progesterone receptors are still expressed from day 25 onwards of the menstrual cycle in the presence of a IUD T200 (33). Our current study did not address the status of confounding factors such as hydrosalpinx and endometriosis in the study population because these diagnoses would have required surgical procedures that otherwise were not needed. Another limitation of this study is the relatively small number of the study participants, and these results should be confirmed with a larger population. The HSCORE values may become more significant if a larger study population is used, but the finding of decreased absolute staining among IUD users confirms the IUDs direct impact on uterine receptivity. 1106 Savaris et al.
Endometrial integrin in IUD T200 users
The clinical relevance of this difference is open for further investigation. The negative expression of ␣v3 in the glandular endometrium of 38.5% of IUD users adds new data to the discussion of whether the copper IUD may influence integrin expression and consequently act on the implantation site. The ␣v3 expression alone cannot explain the mechanism of action of the copper IUD, as over 50% of the IUD participants had a positive expression and the contraceptive efficacy of the copper IUD is almost 100% (1). Therefore, the action of copper IUD possibly may have multiple mechanisms, including alteration in uterine receptivity as seen in a subset of participants (38.5%) in this study.
Acknowledgments: The authors thank Prof. Bruce Lessey (Department of Obstetrics and Gynecology, University of North Carolina) for his laboratory help and expertise.
References 1. Rowe PJ. Anticoncepcion intrauterina: adelantos y perspectivas. In: Lo`pez G, Yunes J, Solı`s JA, Omran AR, eds. Salud Reproductiva en las Ame`ricas. Washington, D.C.: Organizacio`n Panamericana de la Salud, 1992:256 –95. 2. Ortiz ME, Croxatto HB, Hardin CW. Mechanisms of action of intrauterine device. Obstet Gynecol Surv 1996;51(suppl 12):S42–51. 3. Spinnato JA. Mechanism of action of intrauterine contraceptive devices and its relation to informed consent. Am J Obstet Gynecol 1997;176: 503– 6. 4. Tredway DR, Umezaki CU, Mishell DRJ, Settlage DS. Effect of intrauterine devices on sperm transportation in the human being: preliminary report. Am J Obstet Gynecol 1975;123:734 –5. 5. Alvarez F, Brache V, Fernandez E, Guerrero B, Guiloff E, Hess R. New insights on the mode of action of intrauterine contraceptive devices in women. Fertil Steril 1988;49:768 –73.
Vol. 74, No. 6, December 2000
6. el-Sahwi S, Moyer DL. The leukocytic response to an intrauterine foreign body in the rabbit. Fertil Steril 1971;22:398 – 408. 7. Ortiz ME, Croxatto HB. The mode of action of IUDs. Contraception 1987;36:37–53. 8. Lessey BA, Yeh I, Castelbaum AJ, Fritz MA, Ilesanmi AO, Korzeniowski P, et al. Endometrial progesterone receptors and markers of uterine receptivity in the window of implantation. Fertil Steril 1996; 65:477– 83. 9. Ruoslahti E, Noble NA, Kagami S, Border WA. Integrins. Kidney Int Suppl 1994;45:S17–22. 10. Lessey BA, Damjanovich L, Coutifaris C, Castelbaum A, Albeida SM, Buck CA. Integrins adhesion molecules in the human endometrium— correlation with the normal and abnormal menstrual cycle. J Clin Invest 1992;90:188 –95. 11. Ortiz ME, Valenzuela JP, Salvatierra AM, Croxatto HB. The expression of alpha(v) and beta3 integrin subunits in the normal human fallopian tube epithelium suggests the occurrence of a tubal implantation window. Hum Reprod 1998;13:2916 –20. 12. Be´liard A, Donnez J, Nisolle M, Foidart JM. Localization of laminin, fibronectin, E-cadherin, and integrins in endometrium and endometriosis. Fertil Steril 1997;67:266 –72. 13. Lessey BA, Ilesanmi AO, Lessey MA, Riben M, Harris JE, Chwalisz K. Luminal and glandular endometrial epithelium express integrins differentially throughout the menstrual cycle: implications for implantation, contraception, and infertility. Am J Reprod Immunol 1996;35:195–204. 14. Tabibzadeh S. Patterns of expression of integrin molecules in human endometrium throughout the menstrual cycle. Hum Reprod 1992;7: 876 – 82. 15. Hertig AT, Rock J, Adams EC. A description of 34 human ova within the first 17 days of development. Am J Anat 1956;98:435–93. 16. Lessey BA, Castelbaum A, Buck CA, Lei Y, Yowell CW, Sun J. Further characterization of endometrial integrins during the menstrual cycle and in pregnancy. Fertil Steril 1994;62:497–506. 17. Lessey BA, Castelbaum AJ, Sawin SW, Buck CA, Schinnar R, Bilker W, et al. Aberrant integrin expression in the endometrium of women with endometriosis. J Clin Endocrinol Metab 1994;79:643–9. 18. Meyer WR, Castelbaum AJ, Somkuti S, Sagoskin AW, Doyle M, Harris JE, et al. Hydrosalpinges adversely affect markers of endometrial receptivity. Hum Reprod 1997;12:1393– 8. 19. Lessey BA, Castelbaum A, Sawin SW, Sun J. Integrins as markers of
FERTILITY & STERILITY威
20. 21. 22. 23. 24.
25. 26. 27. 28. 29. 30. 31. 32. 33.
uterine receptivity in women with primary unexplained infertility. Fertil Steril 1995;63:535– 42. Hii LL, Rogers PA. Endometrial vascular and glandular expression of integrin alpha(v)beta3 in women with and without endometriosis. Hum Reprod 1998;13:1030 –5. Lessey BA. The use of integrins for the assessment of uterine receptivity. Fertil Steril 1994;61:812– 4. Tabibzadeh S. Implantation: from basics to the clinic. Ann N Y Acad Sci 1997;828:131– 6. Noyes RW, Hertig AT, Rock J. Dating the endometrial biopsy. Fertil Steril 1950;1:3–25. Budwit-Nowotny DA, McCarty KS, Cox EB, Soper JT, Mutch DG, Creasman WT, et al. Immunohistochemical analyses of estrogen receptor in endometrial adenocarcinoma using a monoclonal antibody. Cancer Res 1986;46:5419 –25. Jekel JF, Elmore JG, Katz DL. Sample size, randomization, and probability theory. In: Jekel JF, ed. Epidemiology, biostatistics and preventive medicine. Philadelphia: WB Saunders, 1996:159 –71. Ilesanmi AO, Hawkins DA, Lessey BA. Immunohistochemical markers of uterine receptivity in the human endometrium. Microsc Res Tech 1993;25:208 –22. Lessey BA. Integrins and the endometrium: new markers of uterine receptivity. Ann N Y Acad Sci 1997;828:111–22. Newton J, Tacchi D. Long-term use of copper intrauterine devices. Lancet 1990;335:1322–3. Johannisson E. Mechanism of action of intrauterine devices: biochemical changes. Contraception 1998;36:11–22. Klein S, Giancotti FG, Presta M, Albelda SM, Buck CA, Rifkin DB. Basic fibroblast growth factor modulates integrin expression in microvascular endothelial cells. Mol Biol Cell 1993;4:973– 82. Vessey M, Meisler L, Flavel R, Yeates D. Outcome of pregnancy in women using different methods of contraception. Br J Obstet Gynaecol 1979;86:548 –56. Grosskinsky CM, Yowell CW, Sun J, Parise LV, Lessey BA. Modulation of integrin expression in endometrial stromal cells in vitro. J Clin Endocrinol Metab 1996;81:2047–54. de Castro A, Gancedo GP, Contreras F, Lapena G. The effect of copper ions in vivo on specific hormonal endometrial receptors. Adv Contracept 1986;2:399 – 404.
1107
Expression of ␣41 and ␣v3 integrins in the endometrium of women using the T200 copper intrauterine device Ricardo Savaris, M.D.,a Cla´udio Galleano Zettler, M.D.,a and Arnaldo Nicola Ferrari, M.D.b Fundac¸a˜o Faculdade Federal Cieˆncias Me´dicas and Fundac¸a˜o Universita´ria de Endocrinologia e Fertilidade, Porto Alegre, Brazil
Objective: To investigate the expression of ␣4 and 3 integrin subunit levels in the endometrium of healthy women and copper intrauterine device (IUD) T200 users. Design: Case control study. Setting: An academic teaching hospital and a primary care clinic. Patient(s): Thirteen copper IUD users and 13 normal fertile women. Intervention(s): Timed endometrial biopsies during the mid-secretory phase (days 20 to 24). Main Outcome Measure(s): Histologic dating of endometrium and immunohistochemical staining intensity of ␣4 and 3, using the semiquantitative immunohistochemical score (HSCORE). Result(s): All endometrial biopsies consistent with menstrual dates were examined for integrin expression (3 and ␣4). No difference in ␣4 integrin expression was found between IUD users and controls in both luminal and glandular epithelium. In fertile controls, ␣v3 staining was present in 100% and 38.4% of glandular and luminal epithelium, respectively. In contrast, only 61.5% of the IUD users had any ␣v3 staining in the glandular epithelium and only 53.9% in the luminal epithelium. The intensity of immunoreactivity between the two groups (mean HSCORE) did not differ significantly. Conclusion(s): Proportionately, significantly fewer women using copper IUD had positive ␣v3 immunoreactivity in the glandular epithelium of mid-secretory endometrium. (Fertil Steril威 2000;74:1102–7. ©2000 by American Society for Reproductive Medicine.) Key Words: Contraception, integrins, IUD, reproduction, endometrium, implantation
Received December 14, 1999; revised and accepted June 12, 2000. Supported in part by Conselho Nacional de Desenvolvimento Cientı´fico e Tecnolo´gico, grant number 142161/1997-7, Porto Alegre, Brazil. Reprint requests: Ricardo Savaris, M.D., Rua Reis Louzada, 33, Porto Alegre, RS, 90630-130, Brazil (FAX: 55-51-3404514; E-mail: savaris@orion .ufrgs.br). a Fundac¸a˜o Faculdade Federal Cieˆncias Me´dicas. b Fundac¸a˜o Universita´ria de Endocrinologia e Fertilidade. 0015-0282/00/$20.00 PII S0015-0282(00)01600-9
1102
After oral contraceptives, the intrauterine device (IUD) probably is the most common contraceptive method used in the world (1). However, its mechanism of action is controversial and as yet not well understood (2, 3). It has been suggested that IUDs act by interfering with sperm transport (4), fertilization (5), and implantation (6, 7). It is also believed that IUDs stimulate an inflammatory response in the uterus (3). Integrins, a family of cell adhesion molecules, consist of transmembrane ␣ and  subunits and are involved in cell– cell and cell– substratum interactions (8). These cell adhesion receptors are involved in a variety of biological processes (9) and are widely expressed, including in human endometrium (10). Integrin expression profiles have been demonstrated to be feasible by standard immunohistochemical techniques (10 –12).
Three specific integrin subunits, ␣1, ␣4, and 3, show cycle-specific patterns of expression in the glandular and luminal epithelium in human endometrium during the menstrual cycle (13). The abrupt appearance of ␣v3 on day 20 of the menstrual cycle (14) is coincident with the opening of the window of implantation, as described by Hertig et al. (15); the disappearance of ␣4-subunit on cycle day 24 sets its closure (13). The coexpression of these three integrins subunits is related to the putative “window of implantation” that is thought to occur on cycle days 20 to 24 (16). The expression of endometrial ␣v3 is abnormal in women with endometriosis (17), hydrosalpinges (18), luteal phase defect (8), and unexplained infertility (19). Nevertheless, some authors did not find this abnormality in patients with endometriosis (20). Because of
the temporal pattern of their expression around the time of implantation and their absence in conditions related to infertility, these integrins might have an important role as a potential marker of uterine receptivity (21, 22). We hypothesize that contraception by copper IUD may occur via alteration in the uterine receptivity, specifically the integrin expression profile. We conducted the present study to determine whether copper IUDs induce abnormal integrin expression during the window of implantation.
MATERIALS AND METHODS Study Design The use of human tissue for research was based on informed consent and was approved by the institutional review board of the General Hospital of Porto Alegre (HCPA). A case control study was performed comparing the expression of ␣41 and ␣v3 integrins in the endometrium of copper IUD users from a primary health care unit, to fertile controls who were undergoing tubal sterilization or hysterectomy for benign conditions. The IUD users, normally cycling women 18 to 40 years old, were invited to participate in the study by letter. The criteria for inclusion were no use of hormones over the last 3 months, a fixed partner, proven fertility (parity ⱖ1), negative exams for sexual transmitted diseases (gonorrhea and chlamydia), no previous history of pelvic inflammatory disease or endometriosis, and a recent normal pap smear. All of the women who did not fulfill these inclusion criterion were excluded. In addition, the endometrial biopsies performed during the study that were not “in phase” (dyssynchrony of ⱕ 2 days based on the time of ovulation) were also excluded.
Tissue Acquisition All of the biopsies were obtained with a plastic flexible curette (Z-Sampler, BEI Medical Instruments, San Francisco, CA) during the mid-luteal phase 6 to 10 days after ovulation (LH ⫹6 ⬃ ⫹10), based on a measurement of the urinary LH surge (LH kit, Expomed Inc., San Antonio, CA). The endometrial samples were divided for histologic dating and immunohistochemical staining. Endometrial histologic dating was performed according to the criteria of Noyes et al. (23).
Immunohistochemistry and Monoclonal Antibodies
Immunoperoxidase staining of 8-m cryostat sections was performed using a technique described elsewhere in detail (10). Briefly, frozen tissue sections were placed on poly-L-lysine-coated slides and fixed in acetone at ⫺20°C for 10 minutes. After a rinse with phosphate-buffered saline (PBS) at pH 7.2 to 7.4, the endogenous peroxidases were quenched with a 30 minute incubation with 0.3% H2O2 FERTILITY & STERILITY威
(Sigma Chemical, St. Louis, MO) in absolute methanol (Mallinckrodt, Paris, KY). After a new rinse with PBS, unspecific sites were blocked with 2% horse serum albumin (Vector Laboratories, Burlingame, CA) for 30 minutes. Monoclonal antibodies against ␣4 (B-5H10) and 3 (SSA6), kindly provided by Prof. B. A. Lessey, were incubated overnight at ⫹4°C at the dilution of 1:2,500 and 1:500, respectively. After a rinse with PBS for 3 minutes, secondary antibody consisting of biotinylated horse anti-mouse (Vector) was incubated for 30 minutes at room temperature. Avidin-biotinylated horseradish peroxidase macromolecular complex (ABC, Vector) was then incubated on the sections for 30 minutes. The diaminobenzidine (DAB; Sigma) was used as a chromogen to complete the reaction. The samples were counterstained with hematoxylin and mounted. The resulting staining was evaluated on a Nikon microscope at low (⫻100) and higher (⫻400) magnification by two independent readers in a blind fashion who used the histologic score (HSCORE) described elsewhere (17). An HSCORE greater than 0.7 was defined as positive staining (17). In previous studies, the HSCORE has been shown to have relatively low intraobserver (r ⫽ 0.983; P⫽.00001) and interobserver (r ⫽ 0.994; P⫽.00001) variation (24).
Statistical Analysis and Sample Size Comparisons of the means of the HSCORE of both groups were performed with the use of the Student t test. Fisher’s exact test was used to compare nonparametric dichotomic data (IUD/controls ⫻ integrin positive/negative). The sample size was calculated based on a formula described elsewhere (25); it considered ␣ and  errors of 5% and 20%, respectively, variance of the expression of the integrins as 0.79 according to previous report (19), and a minimal difference between the HSCORE of 0.9. The result was 13 subjects in each group. Statistical analysis was performed using Excel 7.0 software (Microsoft Corporation, Redmond, WA) and EPI Info 5.01b software (Centers for Disease Control, Atlanta, GA).
RESULTS Letters were sent to 152 copper IUD users who were registered in a secondary health care that invited them to participate in the study. Twenty-four patients attended the first consultation and agreed to participate. From these 24 patients, 7 (29.16%) quit or changed their address, making follow-up impossible. Three (12.5%) were excluded for not fulfilling the inclusion criteria. Fourteen endometrial samples were obtained from the remaining patients. One biopsy from an IUD user was out of phase, so this patient was excluded from the study. Control samples were obtained from patients at a tertiary health care 1103
FIGURE 1 Immunohistochemical localization of ␣4 and 3 integrin subunits in the endometrium from controls and from copper IUD T200 users. (A), A positive staining for mouse anti-human 3-subunit in the endometrial gland (arrow) on day 8 after ovulation in a control patient’s menstrual cycle. (B), A positive staining for ␣4 in a gland in an IUD user (arrow) and negative expression in the lumen as expected (arrow head). (C), The same sample stained for 3. Observe the negative expression in the gland (asterisk) and in the lumen (arrowhead). (D), A patient with negative 3 in glands (asterisk) and lumen (arrowhead), but with a positive-staining stromal epithelium (white arrow). (Original magnification, ⫻200.)
Savaris. Endometrial integrin in IUD T200 users. Fertil Steril 2000.
center who had had a tubal sterilization or a hysterectomy for a benign condition (patient number 5, 36 years old, had a hysterectomy for menorrhagia). No difference was observed between the two groups when age, parity, and day of biopsy were compared (Table 1). No complications or displacements of the contraceptive devices were observed. Table 2 shows the individual HSCORE of the patients. As expected, ␣4 was not found in the luminal epithelium of both groups, but was present in 76.9% and in 69.2% of the glandular 1104 Savaris et al.
Endometrial integrin in IUD T200 users
endometrium of IUD users and controls, respectively. The 3 subunit had a similar distribution between both groups, although five samples from IUD users had an HSCORE of zero for the glandular 3 subunit. No statistically significant difference was found in the mean HSCORE expression of ␣v3 and ␣41 in the luminal and glandular epithelium at days LH⫹6 ⬃ ⫹10 of the cycle (Student t test; P⬎.05) between IUD users and controls, as shown in Figure 2. Although there were no statistically significant differVol. 74, No. 6, December 2000
FIGURE 2
FIGURE 3
Immunohistochemical staining (HSCORE) and distribution of endometrial specimens obtained from controls (n ⫽ 13) and IUD users (n ⫽ 13). Results are means ⫾ SD. 䊐 ⫽ IUD. ⫽ control.
Percentage of positive immunoperoxidase staining (HSCORE ⬎ 0.7) for ␣4 and 3 integrin subunits in luminal and glandular endometrium in copper IUD users (n ⫽ 13) and controls (n ⫽ 13). 䊐 ⫽ IUD. ⫽ control. *Fisher’s exact test ⫽ 0.01.
Savaris. Endometrial integrin in IUD T200 users. Fertil Steril 2000. Savaris. Endometrial integrin in IUD T200 users. Fertil Steril 2000.
ences in the percentage of positive immunoreactivity for luminal and glandular ␣41 and luminal ␣v3 (P⬎.05, 2) between both groups, there were significantly fewer women who had any glandular epithelium ␣v3 staining among the IUD users compared to controls (61.5% vs. 100%, P⫽.01, Fisher’s exact test) as shown in Figures 1, 2 and 3.
DISCUSSION Intrauterine devices (IUD) are one of the most effective existing contraceptive methods. IUDs are thought to function in part by diminishing uterine receptivity by virtue of their nature as foreign bodies. However, there are very few published studies investigating the cellular mechanism of action of the IUD. To our knowledge, there are no published studies on the effects of the copper IUD on the expression of ␣v3. The ␣v3 integrin is increasingly recognized and used as a marker of uterine receptivity (19, 22, 26). The abrupt appearance of ␣v3 on day 20 of the menstrual cycle (14) is coincident with the opening of the window of implantation,
TABLE 1 Sample characterization. Controls (n ⫽ 13) Mean ⫾SD
IUDs (n ⫽ 13) Mean ⫾ SD
P value
27.9 ⫾ 5.18 2.5 ⫾ 0.77 23.3 ⫾ 0.58
30.1 ⫾ 3.18 3.2 ⫾ 1.64 22.6 ⫾ 1.58
0.19 NSa 0.18 NSa 0.18 NSa
Age (y) Parity Day of biopsy a
NS: not significant.
Savaris. Endometrial integrin in IUD T200 users. Fertil Steril 2000.
FERTILITY & STERILITY威
whereas the disappearance of ␣4-subunit on cycle day 24 sets its closure (13). The coexpression of these integrin subunits is related to the putative “window of implantation” thought to occur on cycle days 20 to 24 (16). With the use of HSCORE, we found no difference in the expression of ␣41 and ␣v3 integrins in either the luminal or glandular endometrium between women using and not using copper IUD (see Fig. 2). A significant difference was observed in the percentage of samples staining negative for ␣v3 integrin in the endometrium of IUD users (see Fig. 3). A number of investigators have used the quantitative analysis of the immunoreactivity of integrins and their mean levels to compare different treatment groups (27), but its clinical relevance is a matter that still needs further evaluation. No cut-off has been demonstrated to segregate the positive and negative expressions of the integrin that correlate to a clinical function. We used an HSCORE of 0.7 as a cut-off to define positive vs. negative staining based on prior work by Lessey et al. (17). Their study determined, through receiver operating characteristic curve analysis, that the cutoff of 0.77 correlated well with the presence of endometriosis. The endometrium of the majority of IUD users (61.5%) showed comparable ␣v3 staining intensity to that of control endometrium, but the trend toward complete absence of staining in the setting of in-phase endometrium may be clinically significant and needs further investigation. The absence of ␣v3 staining may occur in two ways. It can be a result of a histologic delay (type I defect), or may occur when there is no histologic delay (type II defect) (27). Type II defect has been previously reported in women with hydrosalpinx (18) and endometriosis (17), both of which are associated with decreased fertility. The changes in integrin expression profiles in women with hydrosalpinx and endo1105
TABLE 2 Individual HSCORE for ␣4 and 3 integrin subunits in luminal and glandular endometrium of copper intrauterine device (IUD) users and in controls. IUD
Control
␣4 Patient 1 2 3 4 5 6 7 8 9 10 11 12 13
3
␣4
3
Gland
Lumen
Gland
Lumen
Gland
Lumen
Gland
Lumen
2.2 3 2.7 0.6 1.5 1.5 3.5 2.1 2.2 2.2 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0
2.9 3 0 2.8 2.7 0 0 1.5 1.5 2.2 0 0 2.3
2.6 3.5 0.2 2.5 2.4 0 0 1.4 1.6 3.5 0 0 0
1.7 0.8 0 0 0 1.5 1.9 0.4 0 3.4 0.2 1.1 2.7
0 0 0 0 0 0 0 0 0 0 0 0 0
1.5 2 2.3 1.2 2.7 2.4 1.2 2.6 2.4 2.3 3.5 1.2 2
0 0 3 0 1.1 0.6 0 2.7 0 2 2 0 2.7
Savaris. Endometrial integrin in IUD T200 users. Fertil Steril 2000.
metriosis may be analogous to the changes seen in some of the IUD users. The Type II defect in the ␣v3 staining pattern may arise secondary to local inflammatory responses induced by the presence of a copper IUD (28). IUD users show a tendency toward an increase in the neutrophil, mononuclear, and plasma cell type subpopulation of lymphocytes in the endometrium, unrelated to premenstrual changes (29). These cells might secrete some factor, such as basic fibroblast growth factor (bFGF), that is related to diminished expression of ␣v3 (30). Copper IUD releases free copper and copper salts, which may have a direct biochemical impact on the endometrium and lead to morphologic changes (31). Additionally, EDTA, a calcium chelator, has been shown to inhibit in vitro binding of ␣1, ␣6 integrins to the extracellular matrix (32). Other potential mechanisms leading to alteration in ␣v3 expression include a failure to down-regulate the expression of progesterone receptors (8). It was demonstrated that progesterone receptors are still expressed from day 25 onwards of the menstrual cycle in the presence of a IUD T200 (33). Our current study did not address the status of confounding factors such as hydrosalpinx and endometriosis in the study population because these diagnoses would have required surgical procedures that otherwise were not needed. Another limitation of this study is the relatively small number of the study participants, and these results should be confirmed with a larger population. The HSCORE values may become more significant if a larger study population is used, but the finding of decreased absolute staining among IUD users confirms the IUDs direct impact on uterine receptivity. 1106 Savaris et al.
Endometrial integrin in IUD T200 users
The clinical relevance of this difference is open for further investigation. The negative expression of ␣v3 in the glandular endometrium of 38.5% of IUD users adds new data to the discussion of whether the copper IUD may influence integrin expression and consequently act on the implantation site. The ␣v3 expression alone cannot explain the mechanism of action of the copper IUD, as over 50% of the IUD participants had a positive expression and the contraceptive efficacy of the copper IUD is almost 100% (1). Therefore, the action of copper IUD possibly may have multiple mechanisms, including alteration in uterine receptivity as seen in a subset of participants (38.5%) in this study.
Acknowledgments: The authors thank Prof. Bruce Lessey (Department of Obstetrics and Gynecology, University of North Carolina) for his laboratory help and expertise.
References 1. Rowe PJ. Anticoncepcion intrauterina: adelantos y perspectivas. In: Lo`pez G, Yunes J, Solı`s JA, Omran AR, eds. Salud Reproductiva en las Ame`ricas. Washington, D.C.: Organizacio`n Panamericana de la Salud, 1992:256 –95. 2. Ortiz ME, Croxatto HB, Hardin CW. Mechanisms of action of intrauterine device. Obstet Gynecol Surv 1996;51(suppl 12):S42–51. 3. Spinnato JA. Mechanism of action of intrauterine contraceptive devices and its relation to informed consent. Am J Obstet Gynecol 1997;176: 503– 6. 4. Tredway DR, Umezaki CU, Mishell DRJ, Settlage DS. Effect of intrauterine devices on sperm transportation in the human being: preliminary report. Am J Obstet Gynecol 1975;123:734 –5. 5. Alvarez F, Brache V, Fernandez E, Guerrero B, Guiloff E, Hess R. New insights on the mode of action of intrauterine contraceptive devices in women. Fertil Steril 1988;49:768 –73.
Vol. 74, No. 6, December 2000
6. el-Sahwi S, Moyer DL. The leukocytic response to an intrauterine foreign body in the rabbit. Fertil Steril 1971;22:398 – 408. 7. Ortiz ME, Croxatto HB. The mode of action of IUDs. Contraception 1987;36:37–53. 8. Lessey BA, Yeh I, Castelbaum AJ, Fritz MA, Ilesanmi AO, Korzeniowski P, et al. Endometrial progesterone receptors and markers of uterine receptivity in the window of implantation. Fertil Steril 1996; 65:477– 83. 9. Ruoslahti E, Noble NA, Kagami S, Border WA. Integrins. Kidney Int Suppl 1994;45:S17–22. 10. Lessey BA, Damjanovich L, Coutifaris C, Castelbaum A, Albeida SM, Buck CA. Integrins adhesion molecules in the human endometrium— correlation with the normal and abnormal menstrual cycle. J Clin Invest 1992;90:188 –95. 11. Ortiz ME, Valenzuela JP, Salvatierra AM, Croxatto HB. The expression of alpha(v) and beta3 integrin subunits in the normal human fallopian tube epithelium suggests the occurrence of a tubal implantation window. Hum Reprod 1998;13:2916 –20. 12. Be´liard A, Donnez J, Nisolle M, Foidart JM. Localization of laminin, fibronectin, E-cadherin, and integrins in endometrium and endometriosis. Fertil Steril 1997;67:266 –72. 13. Lessey BA, Ilesanmi AO, Lessey MA, Riben M, Harris JE, Chwalisz K. Luminal and glandular endometrial epithelium express integrins differentially throughout the menstrual cycle: implications for implantation, contraception, and infertility. Am J Reprod Immunol 1996;35:195–204. 14. Tabibzadeh S. Patterns of expression of integrin molecules in human endometrium throughout the menstrual cycle. Hum Reprod 1992;7: 876 – 82. 15. Hertig AT, Rock J, Adams EC. A description of 34 human ova within the first 17 days of development. Am J Anat 1956;98:435–93. 16. Lessey BA, Castelbaum A, Buck CA, Lei Y, Yowell CW, Sun J. Further characterization of endometrial integrins during the menstrual cycle and in pregnancy. Fertil Steril 1994;62:497–506. 17. Lessey BA, Castelbaum AJ, Sawin SW, Buck CA, Schinnar R, Bilker W, et al. Aberrant integrin expression in the endometrium of women with endometriosis. J Clin Endocrinol Metab 1994;79:643–9. 18. Meyer WR, Castelbaum AJ, Somkuti S, Sagoskin AW, Doyle M, Harris JE, et al. Hydrosalpinges adversely affect markers of endometrial receptivity. Hum Reprod 1997;12:1393– 8. 19. Lessey BA, Castelbaum A, Sawin SW, Sun J. Integrins as markers of
FERTILITY & STERILITY威
20. 21. 22. 23. 24.
25. 26. 27. 28. 29. 30. 31. 32. 33.
uterine receptivity in women with primary unexplained infertility. Fertil Steril 1995;63:535– 42. Hii LL, Rogers PA. Endometrial vascular and glandular expression of integrin alpha(v)beta3 in women with and without endometriosis. Hum Reprod 1998;13:1030 –5. Lessey BA. The use of integrins for the assessment of uterine receptivity. Fertil Steril 1994;61:812– 4. Tabibzadeh S. Implantation: from basics to the clinic. Ann N Y Acad Sci 1997;828:131– 6. Noyes RW, Hertig AT, Rock J. Dating the endometrial biopsy. Fertil Steril 1950;1:3–25. Budwit-Nowotny DA, McCarty KS, Cox EB, Soper JT, Mutch DG, Creasman WT, et al. Immunohistochemical analyses of estrogen receptor in endometrial adenocarcinoma using a monoclonal antibody. Cancer Res 1986;46:5419 –25. Jekel JF, Elmore JG, Katz DL. Sample size, randomization, and probability theory. In: Jekel JF, ed. Epidemiology, biostatistics and preventive medicine. Philadelphia: WB Saunders, 1996:159 –71. Ilesanmi AO, Hawkins DA, Lessey BA. Immunohistochemical markers of uterine receptivity in the human endometrium. Microsc Res Tech 1993;25:208 –22. Lessey BA. Integrins and the endometrium: new markers of uterine receptivity. Ann N Y Acad Sci 1997;828:111–22. Newton J, Tacchi D. Long-term use of copper intrauterine devices. Lancet 1990;335:1322–3. Johannisson E. Mechanism of action of intrauterine devices: biochemical changes. Contraception 1998;36:11–22. Klein S, Giancotti FG, Presta M, Albelda SM, Buck CA, Rifkin DB. Basic fibroblast growth factor modulates integrin expression in microvascular endothelial cells. Mol Biol Cell 1993;4:973– 82. Vessey M, Meisler L, Flavel R, Yeates D. Outcome of pregnancy in women using different methods of contraception. Br J Obstet Gynaecol 1979;86:548 –56. Grosskinsky CM, Yowell CW, Sun J, Parise LV, Lessey BA. Modulation of integrin expression in endometrial stromal cells in vitro. J Clin Endocrinol Metab 1996;81:2047–54. de Castro A, Gancedo GP, Contreras F, Lapena G. The effect of copper ions in vivo on specific hormonal endometrial receptors. Adv Contracept 1986;2:399 – 404.
1107