Human vaginal leukocytes and the effects of vaginal fluid on lymphocyte and macrophage defense functions Josaph A. Hill, MD, and DeborahJ. Anderson, PhD Boston, Massachusetts OBJECTIVES: The purpose of this study was to quantify, characterize, and further define the role of vaginal white blood cells in defense mechanisms and human immunodeficiency virus infection. STUDY DESIGN: Vaginal lavages were obtained from five healthy women throughout three menstrual cycles. Lymphocyte subpopulations, macrophages, and granulocytes were characterized and quantified by an immunohistologic technique. Vaginal lavage fluid was added to peripheral blood mononuclear cells, and effects on cell viability, lymphocyte proliferation, macrophage phagocytosis, and expression of various cell surface molecules critical to immunologic functions were assessed. Data were analyzed by Student's t test. RESULTS: Few lymphocytes were found at any stage of the menstrual cycle; however, granulocytes and macrophages were abundant at menstruation and present at low levels through the proliferative phase. Vaginal lavage fluid collected during menses, at midcycle, and after coitus suppressed mitogen-induced lymphocyte proliferation but had no effect on surface expression of human leukocyte antigen or CD4 antigens, or on macrophage function. Likewise, low pH «5.0) medium significantly inhibited lymphocyte proliferation but had no effect on macrophage phagocytosis. The spermicide nonoxynol 9 was toxic to both lymphocytes and macrophages. CONCLUSION: White blood cells, including lymphocytes and macrophages, are infrequently present in cervicovaginal secretions of healthy women except during menses; the vaginal environment may effect their function. (AM J OaSTEl GVNECOL 1992;166:720-6.)

Key words: Vagina, lymphocyte, macrophage, immune defense, human immunodeficiency virus type 1

Vaginal fluid is primarily derived from cervical secretions, exfoliated epithelial cells, bacterial products, and plasma transudation through the vaginal mucosa. Uterine, oviductal, follicular, and peritoneal fluid may also contribute. I Sexually communicable pathogens are transmitted through this vaginal milieu. A first line of defense is the mucosal secretory immune system of the cervix, which is capable of producing abundant amounts of secretory immunoglobulin A.2 Cellular mechanisms important in host defense against pathogenic bacteria and viruses in the vagina include cytotoxic T-cell, natural killer cell, macrophage, and granulocyte-mediated defense mechanisms." 4 This study From the Fearing Research Laboratory, Department of Obstetrics, Gynecology and Reproductive Biology, Brigham and Women's Hospital, Harvard Medical School. Supported by grants HD00815, HD23547, and A125305 United States Public Health Service, Bethesda, Maryland, and by the Fearing Laboratory Endowment. Received for publication May 3, 1991; revised August 9, 1991; accepted August 27, 1991. Reprint requests: joseph A. Hill, MD, Fearing Research Laboratory, SGMB, Room 204, 250 Longwood Ave., Boston, MA 02115. 6/1 /33381

720

was performed to characterize and quantitate lymphocytes and phagocytic cells in the vaginal environment and to determine the effects of lavage fluids on lymphocyte proliferation, macrophage phagocytosis, cell viability, and the expression of cell surface molecules on lymphocytes and macrophages, which are important in immunologic functions.

Material and methods Vaginal lavage fluid collection. Five healthy women of reproductive age (range 24 to 37 years) volunteered for the study. Three were in monogamouos sexual relationships and were taking oral contraceptives or using barrier methods of contraception. The other two women were sexually inactive and were not using any method of contraception. Vaginal lavage fluids were obtained at weekly intervals through three consecutive menstrual cycles. Women collected their own samples with an adapted 20 ml syringe (extra holes drilled at end for ease in aspiration). The syringe was filled with sterile Hanks' balanced salt solution and was inserted into the vagina with the woman in a supine position. The solution was expelled and drawn into the syringe

Volume 166 Number 2

three times, then emptied from the syringe into a 50 ml sterile test tube and delivered to the laboratory within 1 hour of collection. Samples containing < 15 ml of fluid were not included in the study. Vaginal lavage fluid preparation. Vaginal lavage fluids were centrifuged at 600 g for 10 minutes. The supernatant was filtered through a 0.22 iJ-m Millex filter (Millipore Corp., Bedford, Mass.) and aliquots were stored at - 70° C. For some experiments unfiltered aliquots of supernatant were incubated at 37° C overnight to promote growth of vaginal flora. The cell pellet was resuspended in 1 ml phosphate-buffered saline solution and 20 iJ-l was transferred to a hemocytometer where cells of white blood cell morphologic type and size (5 to 15 iJ-m) were counted. Cell viability was assessed by trypan blue exclusion. A smear of the cell suspension was made on a glass microscope slide and stained with eosin-thiazine (Hemacolor, EM Diagnostics, Gibbstown, N.J.). In those patients with >2 x 10' white blood cells per sample, cell smears were made on Teflon-coated slides for immunohistologic assessment of white blood cells sub populations as previously described.' Monoclonal antibodies. Various monoclonal antibodies to white blood cells subpopulations were used in the immunoperoxidase studies: pan-leukocyte (antiHLe-l, Becton Dickinson, Mountain View, Calif.), cytotoxic/suppressor T lymphocytes (Leu-2a, Becton Dickinson), helperlinducer T lymphocytes and some monocytes/macrophages (Leu 3a + b, Becton Dickinson), pan-T cell (Leu 4a + 5b, Becton Dickinson), macrophages (Dako-macrophage, Dako Corporation, Santa Barbara, Calif.), B lymphocytes, (Pan-B, Dako Corp.), and natural killer cells (NKH-I, Becton Dickinson). The following monoclonal antibodies to immunologically relevant lymphocyte/macrophage surface antigens were used in radioimmunoassay studies of antigen modulation by soluble vaginal fluid factors: HLAA,B,C, HLA-DR, interleukin-2 receptor (I1-2R), and anti-CD4 (Leu 3a + b); all of these were from Becton Dickinson. Immunoperoxidase technique. A streptavidin-biotin-peroxidase system (Histostain-SP kit, Zymed Laboratories, San Francisco) was used as previously described."' Briefly, acetone-fixed eight-spot Teflon slides were thawed and air-dried. Nonspecific binding sites were blocked by nonimmune rabbit serum (10 minutes). Appropriately diluted monoclonal antibodies (listed previously) or phosphate-buffered saline solution (negative control) were incubated on individual spots of eight-spot slides for 30 minutes at 3r C. Slides were washed, and a Histostain-SP kit with a biotinylated secondary antibody was used to visualize bound antibodies. Slides were counterstained with hematoxylin,

Vaginal immune defense mechanisms

721

and coverslips were mounted with glycerol-polyvinyl alcohol aqueous mounting medium. Phagocytosis. Macrophage phagocytosis of fluorescent carboxylate microspheres (Polysciences, Inc., Warrington, Pa.) was determined as previously described. 6 Briefly, human peripheral blood mononuclear cells were isolated by Ficoll-Hypaque centrifugation, washed three times, and resuspended at a concentration of I x 106 cells per milliliter in RPM I 1640 tissue culture medium buffered with sodium bicarbonate and containing 10% fetal calf serum, 10 mmollL L-glutamine, and 10 mmollL penicillin-streptomycin (GIBCO, Grand Island, N.Y.). Macrophages were obtained by incubation of a suspension of I ml of peripheral blood mononuclear cells on 12 mm round glass coverslips (Rochester Scientific Co., Inc., Rochester, N.Y.) in 24-well tissue culture plates (Falcon, Becton Dickinson, Lincoln Park, N.J.) for I hour at 37° C followed by removal of nonadherent cells by gentle pipetting and aspiration. Macrophages were cultured overnight in wells containing I ml of vaginal lavage fluid, 10% seminal plasma, 10% saliva, 1% nonoxynol 9 (C"H 6o O iO ; Ortho Pharmaceutical Corporation, Raritan, N.J.), or Hanks' balanced salt solution (control) and I ml tissue culture medium. Microspheres were added and incubation was continued for 4 hours, after which the cells were vigorously washed to remove nonadherent microspheres and recultured for an additional hour in tissue culture medium to reduce background binding of microspheres to cells. The coverslips with attached cells were rinsed in Hanks' balanced salt solution, fixed in acetone for 10 minutes, and then placed cell side up on glass slides. Coverslips (24 x 50 mm) were mounted over cell monolayers with 80% glycerol and phosphate-buffered saline solution. Macrophage phagocytosis (percent of macrophages with more than five intracellular microspheres) was assessed on a Zeiss (Carl Zeiss, Inc., Thornwood, N.Y.) phase/epifluorescence microscope. Radioimmunoassay. Human peripheral blood mononuclear cells (7 x 10 6 ) were incubated in I ml undiluted vaginal lavage fluid, in 10% or I % seminal plasma, 10% saliva, or 0.01% nonoxynol-9 diluted in Hanks' balanced salt solution, or in Hanks' balanced salt solution alone at 37° C for 1 hour. The cells were then washed three times in RPMI 1640 medium and resuspended at a concentration of 2 x 106 cells per milliliter in RPMI medium containing I % bovine serum albumin (washing solution); 0.2 ml of cell suspension was than added in triplicate to individual wells of 96well, round-bottom microtiter plates (Linbro Division, Flow Laboratories, Inc., McLean, Va.). Cells were pelleted by centrifugation (500g for 5 minutes), the supernatant was removed from each well, and 0.02 ml of appropriately diluted monoclonal antibodies was added

722

Hill and Anderson

February 1992 Am J Obstet Gynecol

Table I. Numbers and types of white blood cells in vaginal lavage fluid from healthy women (mean ± SD) Phase of cycle Menstruation (n = 15) Proliferative (n = 15) Periovulatory (n = 15) Secretory (n = 15*)

Total leukocytes 6.5 x 10 ± 2.3 x 10

T lymphocytes

Macrophages

Granulocytes

(CD45)

(CD2, CDJ)

4

3.2

X

10 ± 8.8 x 10'

3.2

X

10 ± 2.5 x 10

5 x 10' ± 3.6 x 104

2

X

103 ± 3.3 x 10'

2.5

X

10' ± 3.5 x 103

4

4

4

4

6.1 x 10' ± 2.6

X

104

ND

ND

ND

ND

ND

ND

ND

ND

ND

ND, None detected «2 x 104 white blood cells per milliliter). *Excludes the samples from two women with yeast vaginitis.

to the wells. After 1 hour of incubation at 4° C, the cells were washed three times and 200,000 counts/min of iodine 125-labeled rabbit antimouse immunoglobulin G (New England Nuclear, Boston) was added and incubated for an additional hour; then the cells were washed five times and counted in a ,,(-counter. Lymphocyte proliferation. Lymphocyte proliferation was determined as previously described.' Briefly, peripheral blood mononuclear cells were adjusted to a concentration of 2.5 x 106 cells per milliliter in tissue culture medium. Cultures were set up in quadruplicate in 96-well tissue culture plates with each well containing 2.5 x 10 5 cells in 0.1 ml tissue culture medium; either 0.09 ml undiluted vaginal lavage fluid, 10% or 1% seminal plasma, 10% saliva, 0.01 % nonoxynol-9, or Hanks' balanced salt solution; and 0.01 ml concanavalin A (Sigma Chemical ("':0., St. Louis; 30 gm/ml) or tissue culture medium. For the study of effects of pH on lymphocyte proliferation, tissue culture medium was made with RPMI 1640 that did not contain sodium bicarbonate and was pH adjusted with 0.1 moll L phosphate buffer (pH range 3 to 12). Cultures were incubated for 72 hours at 37° C in a humidified atmosphere of 5% carbon dioxide and 95% air. Lymphocyte proliferation was assessed by adding 0.5 /J-Ci of ['Hlthymidine (13.1 Ci/mmol, New England Nuclear) 8 hours before harvesting with a MASH II (Microbiological Associates, Los Angeles) automatic sample harvester. Peripheral blood mononuclear cell viability was assessed before harvesting by try pan blue exclusion. Cells on dried glass fiber filter paper disks were suspended in 2 ml Betafluor scintillation cocktail (National Diagnostics, Somerville, N .j.), and radioactivity was measured in a Beckman liquid scintillation counter (Beckman Instruments, Inc., Fullerton, Calif.) (efficacy of the system 45%). Statistical analysis. The unpaired Student t test and analysis of variance were used to calculate the significance of differences between test solutions. Statistical significance was assumed when the null hypothesis could be rejected with reasonable assurance (p < 0.05).

Results Vaginal samples collected during menses and the proliferative phase of the menstrual cycle from healthy women contained numerous macrophages and granulocytes but few T lymphocytes, B lymphocytes, or plasma cells (Table I). Fewer than 104 leukocytes were found in samples from the periovulatory and secretory phases of the menstrual cycle, except in two women during one cycle each in which a symptomatic vaginal yeast infection was noted at midcycle and during the secretory phase. In these samples 2 to 3 X 10 5 vaginal white blood cells were recovered; the majority of cells were macrophages and granulocytes, although CD4 + lymphocytes and interleukin-2 receptor+ lymphocytes were also observed. Few «10 4 ) CD8+ cytotoxic/suppressor T lymphocytes and no B lymphocytes were found. Within 1 week of therapy with fungicidal cream, hyphae and white blood cells were no longer observed in the vaginal lavage fluids. Peripheral blood mononuclear cells were used to study the effects of vaginal lavage fluids on cellular defense functions because insufficient numbers of vaginal cells could be recovered for such studies. Vaginal lavage fluids collected during various phases of the menstrual cycle did not adversely affect peripheral blood lymphocyte viability as measured by try pan blue exclusion after a 2-hour incubation. Alteration of pH in tissue culture medium by replacement of bicarbonate buffer with 0.1 moll L phosphate buffer (pH range 3 to 12) likewise did not affect leukocyte viability after 2 hours of culture. Cell viability was also unaffected by 10% seminal plasma, 10% saliva, or postcoital vaginal secretions without nonoxynol-9. The vaginal spermicide nonoxynol-9, however, when tested alone at concentrations ranging from 10% to 0.01% or when present in certain postcoital vaginal lavages, significantly affected peripheral blood leukocyte viability after 2 hours of exposure «10% viable). The function of peripheral blood-derived macrophages, as assessed by the ability of adherent cells to phagocytose fluorescent latex beads, was not signifl-

Vaginal immune defense mechanisms

Volume 166 Number 2

Helper/inducer T lymphocytes (CD4)

5.2

X

10 3 ± 9.9

X

Cytotoxic / suppressor T lymphocytes (CD8)

4.4

103

X

10' ± 1.3

X

B lymphocytes (CD22)

Interleukin-2 receptor

104

8.3

X

10' ± I

X

723

104

ND

ND

ND

ND

ND

ND

ND

ND

ND

ND

ND

ND

ND

Table II. Effects of vaginal lavage fluid on immune cell surface antigen expression (counts per minute, mean ± SD) Bovine serum albumin control

Secretion added

Control Media Saliva (l : 10) Seminal plasma (l : 10) Nonoxynol-9 (1: 10) Vaginal lavage fluid Menstrual Proliferative Periovulatory Peri ovulatory postcoital* Secretory

Class I major histocompatibility antigens

15 15 5 5

3,136 2,321 3,285 1,172

± ± ± ±

1,485 723 2,328 635

12,239 10,566 11,108 1,149

15 15 15 5 15

1,928 2,069 1,739 4,248 1,782

± ± ± ± ±

556 455 471 3,092 302

10,589 11,251 10,341 14,788 10,875

Class II major histocompatibility antigens

CD4

± 1,756

3,186 3,623 3,745 1,010

± ± ± ±

668 302 439 232t

3,561 3,227 3,332 1,058

± ± ± ±

783 780 117 576t

± 756

5,320 3,139 3,553 4,648 4,957

± ± ± ± ±

2,310 264 691 1,047 673

3,091 4,742 3,228 4,006 3,568

± ± ± ± ±

632 2,043 385 659 963

± 647 ± 1,179 ± 123** ± 653 ± 984 ± 2,341 ± 369

*P < 0.005. tp < 0.025. *Contained nonoxynol-9.

cantly affected by vaginal lavage fluid, 10% seminal plasma, or saliva. Furthermore, macro phages from vaginal lavage fluids collected at both menstrual and periovulatory phases of the cycle could adhere to plastic and phagocytose latex beads. Likewise, macrophage phagocytosis was not significantly affected by low pH conditions. However, neither adherence nor phagocytosis was observed in macrophage cultures exposed to 0.01 % nonoxynol-9 or postcoital vaginal lavage fluids containing nonoxynol-9, presumably because of the disruptive effect of this compound on cell membranes (data not shown). Expression of class I (HLA-A,B,C) and class II (HLADR) major histocompatibility antigens and the CD4 antigen (human immunodeficiency virus type 1, [HIV-l] receptor) on peripheral blood mononuclear cells was not affected by exposure to precoital or postcoital vaginallavage fluids, seminal plasma, or saliva. Nonoxynol9, when tested alone at a concentration of 1%, significantly affected surface antigen expression of all three markers; however, various postcoital vaginal lavage samples containing nonoxynol-9 had no effect on surface antigen expression (Table II). Data showing effects of vaginal lavage fluid on lymphocyte proliferation are presented in Table III. Vag-

inallavage fluids collected during the periovulatory and menstrual phases significantly inhibited lymphocyte proliferation. Vaginal lavage fluid collected and processed within 4 hours after coitus also had a very strong suppressive effect on lymphocyte proliferation, presumably because of the presence of seminal plasma, which has a strong immunosuppressive effect in these assay systems." Nonoxynol-9 at a 0.01 % concentration almost completely abolished lymphocyte proliferation, and postcoital samples containing nonoxynol-9 were more immunosuppressive than postcoital samples without nonoxynol-9 (48% vs 63% of control, p < 0.01). Saliva at a final concentration of 10% had no effect on lymphocyte proliferation. Aliquots of vaginal lavage fluids that were incubated overnight at 37° C before use had an increased inhibitory effect on lymphocyte proliferation (Table III), suggesting that vaginal flora may contribute to the immunosuppressive effect. Lymphocyte proliferation was significantly reduced in pHcontrolled medium when the pH was

Human vaginal leukocytes and the effects of vaginal fluid on lymphocyte and macrophage defense functions.

The purpose of this study was to quantify, characterize, and further define the role of vaginal white blood cells in defense mechanisms and human immu...
708KB Sizes 0 Downloads 0 Views