- Email: [email protected]
T-lymphocyte reactivity during the menstrual cycle in women
CLINICAL
IMMUNOLOGY
AND
IMMUNOPATHOLOGY
56, 130-134 (I!?%)
BRIEF COMMUNICATION T-Lymphocyte
Reactivity during the Menstrual in Women
Cycle
ANTHONY R. CAGGIULA,* CATHERINE M. STONEY,t KAREN A. MA-rrHEws,t JANE F. OwtzNs,t MARY C. DAVIS,* AND BRUCE S. RABIN* Departments of Wsychiatry and SPathology, University of Pittsburgh School of Medicine and Department of *Psychology, University of Pittsburgh, Pittsburgh, Pennsylvania 15260 Reactivity of blood lymphocytes to nonspecific mitogenic stimulation with phytohemagghrtinin (PHA) was measured in nine healthy, regularly cycling women at three phases of their menstrual cycles correspondingto peak levels of estradiol (midfollicular phase), peak levels of progesterone (midluteal phase), and the lowest levels of both hormones (menstrual phase). Sampling points were verified by radioimmunoassay of estrogen, progesterone, luteinizing hormone, and follicle-stimulating hormone. There were significant increases in reactivity associated with an increasing concentration of PHA and with autologous plasma vs AB plasma. However, no differences were found in reactivity to PHA over the three menstrual cycle phases and correlational analyses indicated no relationship between counts and any of the hormones measured. 01990 Academic
Press. Inc.
INTRODUCTION The reproductive hormones estradiol, progesterone, and testosterone can influence the immune system in both animals and humans (1, 2). Differences in circulating levels of these hormones may underlie sex differences in immune function (2,3) as well as functional changes in the immune system associated with pregnancy (4). Since major fluctuations in plasma concentrations of the ovarian steroids, estrogen and progesterone, also characterize the menstrual cycle, it is reasonable to ask whether cycle-related changes in these hormones are accompanied by alterations in immunologic function. Several reports suggest cyclic variations in white cell subpopulations (5) and titers of antibodies to Can&da albicans (6). In one preliminary report of two cycling women (7), T-cell responsiveness, as measured by in vitro lymphocyte reactivity to phytohemagglutinin (PHA), was low during ovulation and high during menstrual bleeding. However, no information was provided on the method of determining cycle phase or on the levels of ovarian hormones associated with each sample point. Adequate procedures for validating cycle phase are essential to the successful execution of any research involving the menstrual cycle (8). In the present paper, nine women were followed for one complete cycle. Reactivity of blood lymphocytes to PHA was measured during the menstrual, follicular, and luteal phases. Cycle phase was determined by diary records and an 130 0090-1229/90 $1.50 Copyright 0 1990 by Academic Press, Inc. AII tights of reproduction in any form reserved.
BBIEF
COMMUNICATION
131
ovulation test kit and verified by circulating levels of estradiol, progesterone, luteinizing hormone (LH), and follicle-stimulating hormone (FSH). METHODS
This study was part of a larger project investigating the role of menstrual cycle phase on cardiovascular and endocrine function. For complete methodological details, refer to Ref. (8). Subjects. Fifteen women volunteers satisfied eligibility criteria because they were free from chronic diseases, unmedicated, between the ages of 18 and 30 years, nonsmoking, without a personal psychiatric history, less than 20% overweight, not pregnant or lactating in the last 12 months, and not using oral contraceptives in the last 6 months. Complete data are available for 9 of the 15eligible women. All potential study participants participated in an initial 1-hr interview session, during which extensive medical and gynecological histories were established. Participants were required to have documented menstrual cycle lengths between 26 and 34 days in each of the previous 3 months, with variability less than 4 days. No subject reported significant menstrual cycle-associated symptoms. of the women who were excluded from participating in this study, 25% were denied participation because of menstrual cycle irregularities apparent in their written records. The majority of the remaining women not invited to participate did not have complete records for the previous 3 months to document the reliability of their cycles. Procedure. During the initial interview session, we instructed eligible participants regarding the use of the ovulation test kit (Q-test). This test is a monoclonal antibody-based qualitative enzyme immunoassay test designed to visually detect luteinizing hormone in the urine, and we used it throughout the middle of each menstrual cycle as a relatively noninvasive means of identifying cycle phase. Each woman who was accepted into the study protocol came to the laboratory three separate times, corresponding to the menstrual, follicular, and luteal phases of their menstrual cycles. One-third of the women had their first testing session in each of the three phases, and the testing times for each phase were determined individually according to each woman’s previous 3-month cycle history, the results from the ovulation test kit, and the first day of her most recent menses. For example, a woman with an average 28-day cycle was tested between Days 1 and 3, Days 10 and 13, and Days 20 and 22. This same woman used the Q-test ovulation kit daily from Days 12to 16, and the luteal phase visit was scheduled about 7 days after the LH surge, as detected by the test. Testing days were adjusted for women with cycles longer or shorter than 28 days. The level of LH, FSH, progesterone, and estradiol in the blood was used to verify menstrual cycle phase at each testing session. For each session, a nurse inserted an indwelling catheter into the antecubital fossa of the arm and subjects were instructed to rest quietly and fill out questionnaires for 30 min. All testing occurred in a temperature and humidity controlled room. Thromboresistant blood withdrawal kits (Kowarski-Cormed, 1Pgauge) were pretreated with 10,000IU of heparin prior to insertion and blood withdrawal
132
BRIEF
COMMUNICATION
Blood samples for plasma were centrifuged immediately after collection in a Damon-refrigerated centrifuge (Model CRU-5000). After centrifugation and preparation for storage, all plasma and serum samples were immediately stored at - 90°C in a Revco freezer (Model ULT 790), equipped with an emergency alarm and voltage safeguard. Hematocrit tubes were centrifuged in a hematocrit centrifuge (International Equipment Co.). Progesterone, estradiol, and LH were determined from serum with a specific, solid-phase radioimmunoassay method (Diagnostic Products Corp.). A similar radioimmunoassay procedure was used for the determination of FSH (Clinical Assays Corp.), All endocrine and neuroendocrine assays were performed in duplicate. For immune assays, an additional 20 cc of blood was collected in heparinized vacutainer tubes and centrifuged over a Hypaque-Ficoll gradient for the isolation of lymphocytes. The isolated lymphocytes were washed three times in RPM1 1640 tissue culture medium by centrifugation. Cell counts were done and 50,000 lymphocytes/well in 10% autologous or normal AB male plasma were cultured in microtiter dishes. Two types of plasma were used to determine if soluble factors present in the subjects’ autologous plasma would alter mitogenic activity in comparison to male plasma which would not contain such factors. Phytohemagglutinin (Difco Laboratories, Detroit, MI) was used at dilutions of 1: 10 and 1:50. Cells were cultured for 72 hr and tritiated thymidine (6.7 Ci/mmol) was added 4 hr before harvesting. Cells were harvested with a mini-mash harvester and the amount of tritiated thymidine taken up by the cells was determined in a liquid scintillation counter. Background counts were obtained by incubating cells without added mitogen. RESULTS
Levels of estradiol, progesterone, LH, and FSH indicated that all women were correctly tested during the menstrual, follicular, and luteal phases (Table 1) (9). With one exception-a higher LH value for luteal than for menstrual or follicular for one subject-there were no individual reversals from the general pattern. There were significant increases in reactivity associated with increasing dose of PHA and with autologous plasma vs AB plasma (Fig. 1). A 3 (menstrual phase) x 2 (dose of PHA) x 2 (autologous plasma vs AB plasma) repeated measures analysis of variance for counts yielded an F(1,96 df) = 89.06, P < 0.001 for dose and TABLE CONCENTRATIONS
1
OFPLASMAHORMONESDUR~NGMENSTRUAL,FOLLICULAR,ANDLUTEAL PHASESOFTHE MENSTRUALCYCLE Menstrual
Estradiol (&ml) Progesterone (nghnl) LH (mIU/mI) FSH (mIU/ml) Note. Means t SE. N = 9.
25.1 0.45 4.9 6.8
e 2 t f
2.9 0.06 0.85 0.67
Follicular 86.6 0.33 13.6 6.4
2 f f 2
22.9 0.07 2.4 0.78
Luteal 104.8 10.58 4.6 3.8
?I 18.2 ” 1.98 k 0.78 4 0.42
BRIEF
COMMUNICATION
133
”
1:lO 150 1:lO 150 AB PLASMA AUTOLOGOUS PHYTOHEMAGGLUTININ
FIG. 1. Effects of menstrual cycle phase on the mitogenic stimulation induced by PHA, expressed as means f SE of the average triplicate counts/minute (n = 9).
F(l,% dfi = 11.5, P < 0.002 for plasma. However, there was no evidence for systematic differences in lymphocyte reactivity to PHA over the three menstrual phases sampled in the present study (F[2,% d’ = 0.06, P > 0.05). Correlational analysis indicated no relationship between counts and any of the hormones measured. The highest correlation for either AB plasma (1:lO) or autologous plasma (1: 10) was r = - 0.06. DISCUSSION
The results of this study do not support the preliminary report of Bjune (7) that responsiveness to PHA was low during ovulation and high during menstrual bleeding in two women. Possible methodological differences that might explain this discrepancy-such as differences in sampling points, hormone levels, etccannot be evaluated because of the preliminary nature of the earlier report. However, the fact that no differences were found in the present study among points deliberately selected to include peak levels of estradiol (midfollicular phase), peak levels of progesterone (midluteal phase), and the lowest level of both hormones (menstrual phase) strongly suggestslittle or no relationship between cycle-related fluctuations in ovarian hormones and this particular measure of immune function. Other studies employing hormonal assays to verify cycle phase have reported variations in some measures of immune function, but not in others. For example, cycle-related changes were found for white cell subpopulations (5) and titers of antibodies to C. albicans (6), but no relationship was found for total immunoglobulin levels or for antibodies to sheep red blood cells or Herpes virus (6). This confirms other studies indicating that reproductive hormones have different effects on different components of immune physiology (1,2). Future work on menstrual cycle-immune relationships must use multiple measures of immunocompetence and employ-as did the present study-hormone measurements to verify cycle phase.
134
BRIEF
COMMUNICATION
ACKNOWLEDGMENTS We gratefully acknowledge the expert assistance of Leslie Gedman and Richard Nelson. This work was supported by Grants HL-38712 and MH-4341.
REFERENCES 1. Ahmed, S. A., Penhale, W. J., andTala1, N., Sex hormones, immune responses, and autoimmune disease. Amer. J. Pathol. 121, 531-555, 1985. 2. Grossman, C. J., Regulation of the immune system by sex steroids. Endocr. Rev. 5, 435-455. 1985. 3. Strausser, H. R., and Fiore, R. P., Alterations in immune function with age, sex hormones, and stress. In “Stress, Immunity, and Aging” (E. L. Cooper, Ed.), pp. 157-171, Dekker, New York, 1984. 4. Stites, D. P., and Siiteri, P. K., Steroids and immunosuppressants in pregnancy. Immunol. Rev. 75, 117-138, 1983. 5. Mathur, S., Mathur, R. S., Goust, J. M., Williamson, H. O., and Fudenberg, H. H., Cyclic variations in white cell populations in the human menstrual cycle: Correlations with progesterone and estradiol. Clin. Immunol. Immunopathol. 13, 246-253, 1979. 6. Mathur, S., Mathur, R. S., Dowda, H., Williamson, H. O., Faulk, W. P., and Fundenberg, H. H., Sex steroid hormones and antibodies to Candida albicans. Clin. Exp. Immunol. 33, 79-87. 1978. 7. Bjune, G., In vitro lymphocyte responses to PHA show covariation with the menstrual cycle. Stand. J. Immunol. 10(4), 362, 1979. 8. Stoney, C. M., Owens, J. F., Matthews, K. A., Davis, M. C., and Caggiula, A., Influences of the normal menstrual cycle on physiologic functioning during behavioral stress. Psychophysiology, in press. 9. Droegemueller, W., Herbst, A. L., Michell, D. R., and Stenchever, M. A., “Comprehensive Gynecology,” Mosby, St. Louis, 1987. Received June 5, 1989; accepted with revision April 5, 1990
IMMUNOLOGY
AND
IMMUNOPATHOLOGY
56, 130-134 (I!?%)
BRIEF COMMUNICATION T-Lymphocyte
Reactivity during the Menstrual in Women
Cycle
ANTHONY R. CAGGIULA,* CATHERINE M. STONEY,t KAREN A. MA-rrHEws,t JANE F. OwtzNs,t MARY C. DAVIS,* AND BRUCE S. RABIN* Departments of Wsychiatry and SPathology, University of Pittsburgh School of Medicine and Department of *Psychology, University of Pittsburgh, Pittsburgh, Pennsylvania 15260 Reactivity of blood lymphocytes to nonspecific mitogenic stimulation with phytohemagghrtinin (PHA) was measured in nine healthy, regularly cycling women at three phases of their menstrual cycles correspondingto peak levels of estradiol (midfollicular phase), peak levels of progesterone (midluteal phase), and the lowest levels of both hormones (menstrual phase). Sampling points were verified by radioimmunoassay of estrogen, progesterone, luteinizing hormone, and follicle-stimulating hormone. There were significant increases in reactivity associated with an increasing concentration of PHA and with autologous plasma vs AB plasma. However, no differences were found in reactivity to PHA over the three menstrual cycle phases and correlational analyses indicated no relationship between counts and any of the hormones measured. 01990 Academic
Press. Inc.
INTRODUCTION The reproductive hormones estradiol, progesterone, and testosterone can influence the immune system in both animals and humans (1, 2). Differences in circulating levels of these hormones may underlie sex differences in immune function (2,3) as well as functional changes in the immune system associated with pregnancy (4). Since major fluctuations in plasma concentrations of the ovarian steroids, estrogen and progesterone, also characterize the menstrual cycle, it is reasonable to ask whether cycle-related changes in these hormones are accompanied by alterations in immunologic function. Several reports suggest cyclic variations in white cell subpopulations (5) and titers of antibodies to Can&da albicans (6). In one preliminary report of two cycling women (7), T-cell responsiveness, as measured by in vitro lymphocyte reactivity to phytohemagglutinin (PHA), was low during ovulation and high during menstrual bleeding. However, no information was provided on the method of determining cycle phase or on the levels of ovarian hormones associated with each sample point. Adequate procedures for validating cycle phase are essential to the successful execution of any research involving the menstrual cycle (8). In the present paper, nine women were followed for one complete cycle. Reactivity of blood lymphocytes to PHA was measured during the menstrual, follicular, and luteal phases. Cycle phase was determined by diary records and an 130 0090-1229/90 $1.50 Copyright 0 1990 by Academic Press, Inc. AII tights of reproduction in any form reserved.
BBIEF
COMMUNICATION
131
ovulation test kit and verified by circulating levels of estradiol, progesterone, luteinizing hormone (LH), and follicle-stimulating hormone (FSH). METHODS
This study was part of a larger project investigating the role of menstrual cycle phase on cardiovascular and endocrine function. For complete methodological details, refer to Ref. (8). Subjects. Fifteen women volunteers satisfied eligibility criteria because they were free from chronic diseases, unmedicated, between the ages of 18 and 30 years, nonsmoking, without a personal psychiatric history, less than 20% overweight, not pregnant or lactating in the last 12 months, and not using oral contraceptives in the last 6 months. Complete data are available for 9 of the 15eligible women. All potential study participants participated in an initial 1-hr interview session, during which extensive medical and gynecological histories were established. Participants were required to have documented menstrual cycle lengths between 26 and 34 days in each of the previous 3 months, with variability less than 4 days. No subject reported significant menstrual cycle-associated symptoms. of the women who were excluded from participating in this study, 25% were denied participation because of menstrual cycle irregularities apparent in their written records. The majority of the remaining women not invited to participate did not have complete records for the previous 3 months to document the reliability of their cycles. Procedure. During the initial interview session, we instructed eligible participants regarding the use of the ovulation test kit (Q-test). This test is a monoclonal antibody-based qualitative enzyme immunoassay test designed to visually detect luteinizing hormone in the urine, and we used it throughout the middle of each menstrual cycle as a relatively noninvasive means of identifying cycle phase. Each woman who was accepted into the study protocol came to the laboratory three separate times, corresponding to the menstrual, follicular, and luteal phases of their menstrual cycles. One-third of the women had their first testing session in each of the three phases, and the testing times for each phase were determined individually according to each woman’s previous 3-month cycle history, the results from the ovulation test kit, and the first day of her most recent menses. For example, a woman with an average 28-day cycle was tested between Days 1 and 3, Days 10 and 13, and Days 20 and 22. This same woman used the Q-test ovulation kit daily from Days 12to 16, and the luteal phase visit was scheduled about 7 days after the LH surge, as detected by the test. Testing days were adjusted for women with cycles longer or shorter than 28 days. The level of LH, FSH, progesterone, and estradiol in the blood was used to verify menstrual cycle phase at each testing session. For each session, a nurse inserted an indwelling catheter into the antecubital fossa of the arm and subjects were instructed to rest quietly and fill out questionnaires for 30 min. All testing occurred in a temperature and humidity controlled room. Thromboresistant blood withdrawal kits (Kowarski-Cormed, 1Pgauge) were pretreated with 10,000IU of heparin prior to insertion and blood withdrawal
132
BRIEF
COMMUNICATION
Blood samples for plasma were centrifuged immediately after collection in a Damon-refrigerated centrifuge (Model CRU-5000). After centrifugation and preparation for storage, all plasma and serum samples were immediately stored at - 90°C in a Revco freezer (Model ULT 790), equipped with an emergency alarm and voltage safeguard. Hematocrit tubes were centrifuged in a hematocrit centrifuge (International Equipment Co.). Progesterone, estradiol, and LH were determined from serum with a specific, solid-phase radioimmunoassay method (Diagnostic Products Corp.). A similar radioimmunoassay procedure was used for the determination of FSH (Clinical Assays Corp.), All endocrine and neuroendocrine assays were performed in duplicate. For immune assays, an additional 20 cc of blood was collected in heparinized vacutainer tubes and centrifuged over a Hypaque-Ficoll gradient for the isolation of lymphocytes. The isolated lymphocytes were washed three times in RPM1 1640 tissue culture medium by centrifugation. Cell counts were done and 50,000 lymphocytes/well in 10% autologous or normal AB male plasma were cultured in microtiter dishes. Two types of plasma were used to determine if soluble factors present in the subjects’ autologous plasma would alter mitogenic activity in comparison to male plasma which would not contain such factors. Phytohemagglutinin (Difco Laboratories, Detroit, MI) was used at dilutions of 1: 10 and 1:50. Cells were cultured for 72 hr and tritiated thymidine (6.7 Ci/mmol) was added 4 hr before harvesting. Cells were harvested with a mini-mash harvester and the amount of tritiated thymidine taken up by the cells was determined in a liquid scintillation counter. Background counts were obtained by incubating cells without added mitogen. RESULTS
Levels of estradiol, progesterone, LH, and FSH indicated that all women were correctly tested during the menstrual, follicular, and luteal phases (Table 1) (9). With one exception-a higher LH value for luteal than for menstrual or follicular for one subject-there were no individual reversals from the general pattern. There were significant increases in reactivity associated with increasing dose of PHA and with autologous plasma vs AB plasma (Fig. 1). A 3 (menstrual phase) x 2 (dose of PHA) x 2 (autologous plasma vs AB plasma) repeated measures analysis of variance for counts yielded an F(1,96 df) = 89.06, P < 0.001 for dose and TABLE CONCENTRATIONS
1
OFPLASMAHORMONESDUR~NGMENSTRUAL,FOLLICULAR,ANDLUTEAL PHASESOFTHE MENSTRUALCYCLE Menstrual
Estradiol (&ml) Progesterone (nghnl) LH (mIU/mI) FSH (mIU/ml) Note. Means t SE. N = 9.
25.1 0.45 4.9 6.8
e 2 t f
2.9 0.06 0.85 0.67
Follicular 86.6 0.33 13.6 6.4
2 f f 2
22.9 0.07 2.4 0.78
Luteal 104.8 10.58 4.6 3.8
?I 18.2 ” 1.98 k 0.78 4 0.42
BRIEF
COMMUNICATION
133
”
1:lO 150 1:lO 150 AB PLASMA AUTOLOGOUS PHYTOHEMAGGLUTININ
FIG. 1. Effects of menstrual cycle phase on the mitogenic stimulation induced by PHA, expressed as means f SE of the average triplicate counts/minute (n = 9).
F(l,% dfi = 11.5, P < 0.002 for plasma. However, there was no evidence for systematic differences in lymphocyte reactivity to PHA over the three menstrual phases sampled in the present study (F[2,% d’ = 0.06, P > 0.05). Correlational analysis indicated no relationship between counts and any of the hormones measured. The highest correlation for either AB plasma (1:lO) or autologous plasma (1: 10) was r = - 0.06. DISCUSSION
The results of this study do not support the preliminary report of Bjune (7) that responsiveness to PHA was low during ovulation and high during menstrual bleeding in two women. Possible methodological differences that might explain this discrepancy-such as differences in sampling points, hormone levels, etccannot be evaluated because of the preliminary nature of the earlier report. However, the fact that no differences were found in the present study among points deliberately selected to include peak levels of estradiol (midfollicular phase), peak levels of progesterone (midluteal phase), and the lowest level of both hormones (menstrual phase) strongly suggestslittle or no relationship between cycle-related fluctuations in ovarian hormones and this particular measure of immune function. Other studies employing hormonal assays to verify cycle phase have reported variations in some measures of immune function, but not in others. For example, cycle-related changes were found for white cell subpopulations (5) and titers of antibodies to C. albicans (6), but no relationship was found for total immunoglobulin levels or for antibodies to sheep red blood cells or Herpes virus (6). This confirms other studies indicating that reproductive hormones have different effects on different components of immune physiology (1,2). Future work on menstrual cycle-immune relationships must use multiple measures of immunocompetence and employ-as did the present study-hormone measurements to verify cycle phase.
134
BRIEF
COMMUNICATION
ACKNOWLEDGMENTS We gratefully acknowledge the expert assistance of Leslie Gedman and Richard Nelson. This work was supported by Grants HL-38712 and MH-4341.
REFERENCES 1. Ahmed, S. A., Penhale, W. J., andTala1, N., Sex hormones, immune responses, and autoimmune disease. Amer. J. Pathol. 121, 531-555, 1985. 2. Grossman, C. J., Regulation of the immune system by sex steroids. Endocr. Rev. 5, 435-455. 1985. 3. Strausser, H. R., and Fiore, R. P., Alterations in immune function with age, sex hormones, and stress. In “Stress, Immunity, and Aging” (E. L. Cooper, Ed.), pp. 157-171, Dekker, New York, 1984. 4. Stites, D. P., and Siiteri, P. K., Steroids and immunosuppressants in pregnancy. Immunol. Rev. 75, 117-138, 1983. 5. Mathur, S., Mathur, R. S., Goust, J. M., Williamson, H. O., and Fudenberg, H. H., Cyclic variations in white cell populations in the human menstrual cycle: Correlations with progesterone and estradiol. Clin. Immunol. Immunopathol. 13, 246-253, 1979. 6. Mathur, S., Mathur, R. S., Dowda, H., Williamson, H. O., Faulk, W. P., and Fundenberg, H. H., Sex steroid hormones and antibodies to Candida albicans. Clin. Exp. Immunol. 33, 79-87. 1978. 7. Bjune, G., In vitro lymphocyte responses to PHA show covariation with the menstrual cycle. Stand. J. Immunol. 10(4), 362, 1979. 8. Stoney, C. M., Owens, J. F., Matthews, K. A., Davis, M. C., and Caggiula, A., Influences of the normal menstrual cycle on physiologic functioning during behavioral stress. Psychophysiology, in press. 9. Droegemueller, W., Herbst, A. L., Michell, D. R., and Stenchever, M. A., “Comprehensive Gynecology,” Mosby, St. Louis, 1987. Received June 5, 1989; accepted with revision April 5, 1990