- Email: [email protected]
Natural killer cell cytotoxicity to herpes simplex virus-1-infected cells is not altered by pregnancy
Natural killer cell cytotoxicity to herpes simplex virus-I-infected cells is not altered by pregnancy Nancy L. Eriksen, MD, Bernard Gonik, MD, and Lian S. Loo, MS Houston, Texas There is evidence to suggest a decrease in natural killer cell cytotoxicity during pregnancy, but information regarding immune responsiveness to actual infection is limited. An in vitro study was undertaken to examine the effect of herpes simplex virus infection on natural killer cell cytotoxicity with peripheral blood mononuclear cells from pregnant (N = 8) and nonpregnant (N = 5) women. The peripheral blood mononuclear cells were separated by Ficoll-Hypaque centrifugation. Effector cells were incubated with live herpes simplex virus-1, ultraviolet-inactivated herpes simplex virus-1, or media alone for 18 hours at 37° C. K562 target cells were used in a sodium chromate release assay with an effector-to-target cell ratio of 100: 1. Baseline natural killer cell values (mean ± SE) for pregnant patients (13.4% ± 2.4%) and nonpregnant patients (19.8% ± 3.7%) were similar. Natural killer cell cytotoxicity was significantly increased by incubation with live virus for both pregnant (37.5% ± 6.2%) and nonpregnant subjects (49.8% ± 7.6%). There was no difference in mean values between media and ultraviolet-inactivated herpes simplex virus-1-exposed samples for either group. Results suggest that (1) infection with live virus, but not viral antigen alone, can augment natural killer cell response in vitro and (2) natural killer cell response to herpes simplex virus-1 infection is not altered by pregnancy. (AM J OSSTET GYNECOL 1991 ;165:965-8.)
Key words: Natural killer cell cytotoxicity, herpes simplex virus, pregnancy Natural killer cells are distinct from cytotoxic T lymphocytes in their ability to lyse target cells without prior antigenic sensitization. Therefore they playa primary role in the immune defense system against viral infection, including herpes simplex virus (HSV). Several reports suggest that pregnant women have a higher incidence of HSV infection, with a longer duration and severity of clinical symptoms. I . 4 However, epidemiologic data to support this suggestion are lacking.' Several authors have shown a decrease in natural killer cell cytotoxicity in pregnant subjects compared with nonpregnant controls. 6. 12 This has led some investigators to refer to pregnancy as an immunocompromised state. There is limited information regarding natural killer cell immune responsiveness to actual infections in humans. An in vitro study was undertaken to examine the effect of HSV infection on natural killer cell cytotoxicity with peripheral blood mononuclear cells from pregnant and nonpregnant subjects. We hypothesized that natural killer cell cytotoxicity would be augmented by HSV infection in a similar fashion for both groups. From the Department of Obstetrics, Gynecology, and Reproductive Sciences, University of Texas Health Sciences Center at Houston. Presented at the Eleventh Annual Meeting of the Society of Perinatal Obstetrics, San Francisco, California, January 31-February 2, 1991. Reprint requests: Nancy L. Eriksen, MD, University of Texas Health Sciences Center at Houston, Department of Obstetrics, Gynecology, and Reproductive Sciences, 6431 Fannin, Suite 3.204, Houston, TX 77030. 616130903
Material and methods Patients. After written informed consent was obtained, 20 ml of blood was obtained into heparinized tubes from eight pregnant subjects and five reproductive-aged nonpregnant controls. Preparation of effector cells. Dextran (3%, Sigma Chemical Co., St. Louis) was added to each tube (10%, vol/vol) and allowed to sediment for 45 minutes. The plasma was then placed over 10 ml of Ficoll-Hypaque (Pharmacia Laboratories, Inc., Piscataway, N.J.) and centrifuged at 200 g for 45 minutes. The mononuclear cell layer was washed three times in Hank's balanced salt solution (Hazelton, Lenexa, Kan.) and resuspended in 1 ml minimal essential media (Gibco Live Technologies, Grand Island, N.Y.) supplemented with 10% heat-inactivated fetal calf serum (Hazelton), penicillin (50 U Iml), and streptomycin (50 J-Lg/ml). These mononuclear cells are a heterogeneous cell population consisting of lymphocytes, macrophages, and natural killer cells. Target cells. K562 (an erythroleukemia cell line) target cells were grown in culture. A pellet was obtained by centrifuging 10 ml at 200 g for 5 minutes. The pellet was resuspended in 150 J-LI of 10% minimal essential media and incubated with 20 to 80 fJ.l sodium chromate (Dupont, Wilmington, Del.) for 1 hour at 3r C. The cells were then layered over 3 ml heat-inactivated fetal calf serum and centrifuged for 5 minutes at 200 g. The pellet was then washed with 10 ml 10% minimal essential media. The supernatant was discarded, 1 ml of 10%
965
966
Eriksen, Gonik, and Loa
••
60 50
~
0
October 199 1 Am J Obstet Gynecol
Pregnant (n=8) Non-pregnant (n=5)
Experimental release - Spontaneous release X 100 100 - Spontaneous release
40
X
0 ~ 0
!: 0 0~
and B equals counts per minute of the lysed pellet. Natural killer cell cytotoxicity equals:
Data are expressed as the mean ± SE of experiments. These data were analyzed with the arcsine transformation. Comparisons were made between the two groups by the analysis of variance, with p < 0.05 considered statistically significant.
30
20 10
Results 0+---
Fig_ 1. Baseline natural killer cell cytotoxicity (mean ± SE) in pregnant and nonpregnant subjects (p = 0.25). Effector/target cell ratio equals 100: 1.
minimal essential media was used to resuspend the pellet, and the cells were counted. The sodium chromatelabeled target cells were suspended at a final concentration of 1 X 10 5 cells per milliliter in 10% minimal essential media. Microcytotoxicity assay. Each assay was conducted in triplicate in polystyrene U-bottom-well microtiter plates (Corning, Corning, N .Y.) . The unstimulated wells, for pregnant and nonpregnant subjects, consisted of 50 ILl effector cells and 50 ILl 10% minimal essential media, yielding a final concentration of 5 x 106 cell per milliliter. Each of the HSV-stimulated wells consisted of 50 ILl effector cells and 50 ILl either live HSV -1 virus (2 X 10 5 plaque-forming units) or 50 ILl ultravioletinactivated HSV-l (2 x 105 plaque-forming units). The final concentration of effectors was 5 x 106 cells per milliliter with 1 x 10' plaque-forming units of HSV. The cells were incubated for 18 hours at 37° C. Experimental wells that contained live virus were washed, along with the unstimulated wells, three times with 10% minimal essential media. Subsequently, 50 ILl of 10% minimal essential media and 50 ILl of labeled target cells (1 x 105 cells per milliliter) were added to all of the wells. The final effector-to-target cell ratio was 100; 1. The covered plate was then centrifuged at 75 g for 5 minutes and allowed to incubate an additional 4 hours at 37° C in a humidified atmosphere of air (95%) and carbon dioxide (5%). Afterward 100 ILl was aspirated from the top of each well without disturbing the pellet. To each well, 100 ILl of 1 moll L sodium hydroxide was added, and the total volume was aspirated into a separate tube. The samples were counted in a Packard Cobra 50 10 )I-counter for 1 minute. Chromium release was calculated by the formula: Percent sodium 2A = - - x 100 chromate release A + B where A equals counts per minute in the top 100 ILl
Baseline natural killer cell cytotoxicity (mean ± SE) was lower in pregnant subjects (13.8% ± 2.4%) than in nonpregnant subjects (19.8% ± 3.7%) (Fig. 1). However, this did not reach statistical significance (p = 0.25). Incubation of effector cells with live HSV-l significantly increased natural killer cell cytotoxicity for pregnant (37.5% ± 6.2 %) and nonpregnant (49 .8% ± 7.6%) subjects, when compared with values for media alone (Fig. 2). As anticipated, the increase in cytotoxicity was similar for both groups (p = 0.25) (Fig. 2). Similar experiments with ultraviolet-inactivated HSV -1 (1 x 105 plaque forming units) showed no difference in mean natural killer cell cytotoxicity values (Fig. 3) when compared with those for media alone for either group. Additional experiments were conducted with I x 10' plaque forming units of ultraviolet-inactivated HSV-l, to determine if increasing the amount of antigen would have any effect on cytotoxicity. Again, no difference in mean natural killer cell cytotoxicity values were noted for either grou p (results not depicted).
Comment Pregnancy represents a unique immunologic state in which a balance exists between acceptance of the fetal allograft and host immunocompetence. A suppression in maternal cell-mediated immunity is assumed to be advantageous for the fetus but theoretically makes the pregnant woman more vulnerable to infections, particularly those of viral causes. Case reports 3 • 4 and animal work l3 suggest that the severity of an HSV infection is increased in pregnancy. A higher incidence of HSV infections and a longer duration of clinical symptoms have been reported in pregnancy. I. 2 Although firm epidemiologic data are lacking to confirm these observations,S these clinical reports have led to the suggestion that pregnancy is an immunocompromised state. Natural killer cells represent the first line of defense against viral pathogens and therefore play an important role in HSV infection. Data in regard to natural killer cells in pregnancy are limited and sometimes conflicting. Histochemical studies 9 and flow cytometrylO
Natural killer cell cytotoxicity to herpes in pregnancy
Volume 165 Number 4, Pan I
967
60 50
>
I::
0
><
...
30
~
20
0 0 0
#.
[3 Pregnant
40
•
(n=8)
Non-pregnant
(n=5)
*< pcO.05 vs media
10 0
Media
HSV-1
Fig. 2. Natural killer cell cytotoxicity in response to live HSV-I infection. Results (mean ± SE) show pregnant and nonpregnant subjects compared with controls (media alone, p < 0.05). Effectorltarget cell ratio is 100: 1.
have shown that the number of natural killer cells is not reduced during pregnancy. Natural killer cell function has been examined in several recent studies,7.12 which demonstrate a general decrease in natural killer cell activity during pregnancy, although this finding is not consistent. None of these previous studies have attempted to examine the effect of actual infection on natural killer cell immune responsiveness. To address this question, an in vitro model was developed in which peripheral blood mononuclear cells of pregnant and nonpregnant subjects were infected with live HSV-l virus. Natural killer cell cytotoxicity was measured in control and infected wells to determine if pregnancy altered immune responsiveness to infection. Our results in Fig. 1 show that baseline cytotoxicity tended to be lower in pregnant subjects, compared with nonpregnant controls. Although this did not reach statistical significance, the calculated power of these experiments (0 .54) to reveal a significant reduction in natural killer cells (20%) is limited. To have a power = 0.8, 14 paired experiments would need to be analyzed. Regardless, our results are consistent with data already available in the literature'·12 and are not central to the issue of HSV -induced immune responsiveness. Incubation of effector cells with live HSV-l resulted in a significant increase in natural killer cell cytotoxicity as compared with controls for both groups (Fig. 2). There was a 2.5-fold increase in cytotoxicity in the nonpregnant subjects, compared with a 2.7-fold increase in the pregnant subjects. This demonstrates that, in spite of lower baseline natural killer cell cytotoxicity values in the pregnant subjects, they were able to respond in a similar fashion as the nonpregnant subjects (Fig. 2). Viral infection leads to the production of a variety of cytokines. Several of these cytokines, in particular interferon alfa, are known to regulate natural killer cell function. Deficiencies in cytotoxic responses have been shown to result in a greater susceptibility to HSV in-
.. Pregnant (n=7) • Non-pregnant (n=7)
30
20
10
o Media
UVHSV-1
Fig. 3. Natural killer cell cytotoxicity in response to ultraviolet-inactivated HSV-l. Effector / target cell ratio IS 100 : J.
fection. 14 · 15 Furthermore, natural killer cell activity aginst HSV infection is enhanced by interferon in vitro,16. 17 and in vivo. IS Data from our laboratory confirm that interferon alfa can augment natural killer cell cytotoxicity in the pregnant host. 19 One could speculate that, although baseline natural killer cell activity during pregnancy may be decreased, cytokine production in response to infection may be normal, accounting for our results. Evidence to support this includes the ability of interferon to restore natural killer cell activity from peripheral blood of pregnant patients,1O and systemic interferon production is higher in pregnant mice than in controls. 13 Surprisingly, few data regarding cytokine production in human pregnancy exist. Studies examining the production of interleukin-Po and interleukin-22I demonstrate elevated or normal production rates, respectively, compared
968
Eriksen, Gonik, and Loa
with those of nonpregnant controls. We have shown that the monokines tumor necrosis factor and interleukin-6 also are significantly increased during pregnancy (unpublished observations). Although interferon alfa has not been directly measured during pregnancy, these data support the concept that its production during pregnancy is unchanged. Interestingly, we found that ultraviolet-inactivated HSV -1 did not augment natural killer cell activity as did infection with live virus (Fig. 3). This suggests that infection with live virus may be necessary to induce the intracellular production of interferon alfa. Kirchner et al. 22 demonstrated that interferon alfa could be detected after injection of live herpesvirus in mice but not with ultraviolet-inactivated herpes antigen. However, HSV antigen in much higher concentrations, and with extended incubation, has been shown to induce lymphocyte stimulation in previously sensitized cells!3, 24 The mechanism for this induction has not been identified, but it is most likely humoral in origin. Data from lymphocyte stimulation experiments cannot be extrapolated to explain natural killer cell function, because natural killer cells lyse target cells without prior antigenic sensitization. In conclusion, our results suggest that (1) infection with live virus, but not viral antigen, can augment natural killer cell cytotoxicity in vitro and (2) natural killer cell response to HSV -1 infection is not altered by pregnancy. We speculate that, during pregnancy, the production of interferon alfa is not attenuated in response to infection with live herpesvirus. This could explain the increase in natural killer cell cytotoxicity in vitro observed in both pregnant and nonpregnant subjects. Although our data suggest that isolated mononuclear cells respond appropriately to HSV infection during pregnancy, other factors, such as serum-mediated suppression of immune activity, may alter in vivo host responses.
REFERENCES 1. Whitley R], Nahmias A], Visintine AM, Fleming CL, Alford CA. The natural history of herpes simplex virus infection of mother and newborn. Pediatrics 1980;66:48994. 2. Brown ZA, Vontver LA, Benedetti], et al. Genital herpes in pregnancy: risk factors associated with recurrences and asymptomatic viral shedding. AM .I OBSTET GYNECOL 1985; 153:24-30. 3. Kobbermann T, Clark L, Griffen WT. Maternal death secondary to disseminated herpes virus hominis. AM ] OBSTET GYNECOL 1980;137:742-3. 4. Young E], Killam AP, Green .IF. Disseminated herpes virus infection: association with primary genital herpes in pregnancy. ]AMA 1976;235:2731-3. 5. Brabin B]. Epidemiology of infection in pregnancy. Rev Infect Dis 1985;7:579-603. 6. Luft B.J, Remington ]S. Effect of pregnancy on augmentation of natural killer cell activity by Corynebacterium
October 1991 Am J Obstet Gynecol
parvum and Toxoplasma gondii. ] Immunol 1984; 132:
2375-9. 7. Toder V, Nebel L, Gleicher N. Studies of natural killer cells in pregnancy. 1. Analysis at the single cell level. Clin Lab ImmunoI1984;14:123-7. 8. Okamura K, Furukawa K, Nakakuki M, Yamada K, Suzuki M. Natural killer cell activity during pregnancy. AM ] OBSTET GYNECOL 1984;149:396-9. 9. Baley]E, Schacter BZ. Mechanisms of diminished natural killer cell activity in pregnant women and neonates.] ImmunoI1985;134:3042-8. 10. Vaquer S, De la Hera A, ]orda.J, Martinez C, Escudero A, Alvarez-Mon M. Diminished natural killer activity in pregnancy: modulation by interleukin 2 and interferon gamma. Scand] Immunol 1987;26:691-8. 11. Toder V, Blank M, Gleicher N, Voljovich I, Mashiah S, Nebel L. Activity of natural killer cells in normal pregnancy and edema-proteinuria-hypertension gestosis. AM ] OBSTET GYNECOL 1983;145:7-10. 12. Gonik B, Loo LS, West S, Kohl S. Natural killer cell cytotoxicity and antibody-dependent cellular cytotoxicity to herpes simplex virus-infected cells in human pregnancy. Am] Reprod Immunol 1987;13:23-6. 13. Yamada K, Shimzu Y, Okamura K, Kumagai K, Suzuki M. Study of interferon production during pregnancy in mice and antiviral activity in the placenta, AM ] OBSTET GYNECOL 1985; 153:335-41. 14. Gresser MG, Tovey G, Maury C,Bandu T. Role of interferon in the pathogenesis of virus diseases in mice as demonstrated by the use of anti-interferon serum. II. Studies with herpes simplex, Maloney sarcoma, vesicular stomatitis, Newcastle disease, and influenza viruses. ] Exp Med 1976;144:1316-3. 15. Virelizier JL. Viral infections in patients with selective disorders of the interferon system. In: Dela Maza LM, Peterson EM, eds. Medical urology. Proceedings of the 1981 international symposium on medical virology, Anaheim, California, October 8-10, 1981, New York: Elsevier Biomedical, 1982. 16. Platsoucas CD. Augmentation of human natural killer cytotoxicity to alpha-interferon and inducers of gamma-interferon. Analysis of monoclonal antibodies. In: Herberman RB, ed. NK cells and other natural effector cells. New York: Academic Press, 1982:355, 17. Bishop GA, McCurry L, Schwartz SA, Glorioso ]C. Activation of human natural killer cells by herpes simplex virus type I-infected cells. Intervirology 1987;28:78-88. 18. Gidlund M, Om A, Wigzell H, Senik A, Gresser 1. Enhanced NK activity in mice infected with interferon and interferon inducers. Nature 1978;273:759-62. 19. Gonik B, Loftin K, Tan NS, Crump J. Immune modulation of natural killer cell cytotoxicity against herpes infected target cells in pregnancy. Am] Reprod Immunol 1990;24:95-8. 20. Bessler H, Sirota L, Dulitzky F, Djaldetti M. Production of interleukin-l by mononuclear cells of newborns and their mothers. Clin Exp Immunol 1987;68:655-61. 21. Hauser GJ, Lidon A, Zakuth V, et al. Immunocompetence in pregnancy: production of interleukin-2 by peripheral blood lymphocytes. Cancer Detect Prev SuppI1987;1:3942. 22. Kirchner H, Engler H, Schroder CH, Zawatzky R, Storch E. Herpes simplex virus type I-induced interferon production and activation of natural killer cells in mice. J Gen Virol 1983;64:437-41. 23. Rosenberg GL, Farber PA, Notkins AL. In vitro stimulation of sensitized lymphocytes by herpes simplex virus and vaccinia virus. Proc Natl Acad Sci USA 1972;69:75660. 24. Jacobs RP, Avrelian L, Cole GA. Cell-mediated immune response to herpes simplex virus: type-specific lymphoproliferative responses in lymph nodes draining the site of primary infection.] Immunol 1976;116;1520-5.
Key words: Natural killer cell cytotoxicity, herpes simplex virus, pregnancy Natural killer cells are distinct from cytotoxic T lymphocytes in their ability to lyse target cells without prior antigenic sensitization. Therefore they playa primary role in the immune defense system against viral infection, including herpes simplex virus (HSV). Several reports suggest that pregnant women have a higher incidence of HSV infection, with a longer duration and severity of clinical symptoms. I . 4 However, epidemiologic data to support this suggestion are lacking.' Several authors have shown a decrease in natural killer cell cytotoxicity in pregnant subjects compared with nonpregnant controls. 6. 12 This has led some investigators to refer to pregnancy as an immunocompromised state. There is limited information regarding natural killer cell immune responsiveness to actual infections in humans. An in vitro study was undertaken to examine the effect of HSV infection on natural killer cell cytotoxicity with peripheral blood mononuclear cells from pregnant and nonpregnant subjects. We hypothesized that natural killer cell cytotoxicity would be augmented by HSV infection in a similar fashion for both groups. From the Department of Obstetrics, Gynecology, and Reproductive Sciences, University of Texas Health Sciences Center at Houston. Presented at the Eleventh Annual Meeting of the Society of Perinatal Obstetrics, San Francisco, California, January 31-February 2, 1991. Reprint requests: Nancy L. Eriksen, MD, University of Texas Health Sciences Center at Houston, Department of Obstetrics, Gynecology, and Reproductive Sciences, 6431 Fannin, Suite 3.204, Houston, TX 77030. 616130903
Material and methods Patients. After written informed consent was obtained, 20 ml of blood was obtained into heparinized tubes from eight pregnant subjects and five reproductive-aged nonpregnant controls. Preparation of effector cells. Dextran (3%, Sigma Chemical Co., St. Louis) was added to each tube (10%, vol/vol) and allowed to sediment for 45 minutes. The plasma was then placed over 10 ml of Ficoll-Hypaque (Pharmacia Laboratories, Inc., Piscataway, N.J.) and centrifuged at 200 g for 45 minutes. The mononuclear cell layer was washed three times in Hank's balanced salt solution (Hazelton, Lenexa, Kan.) and resuspended in 1 ml minimal essential media (Gibco Live Technologies, Grand Island, N.Y.) supplemented with 10% heat-inactivated fetal calf serum (Hazelton), penicillin (50 U Iml), and streptomycin (50 J-Lg/ml). These mononuclear cells are a heterogeneous cell population consisting of lymphocytes, macrophages, and natural killer cells. Target cells. K562 (an erythroleukemia cell line) target cells were grown in culture. A pellet was obtained by centrifuging 10 ml at 200 g for 5 minutes. The pellet was resuspended in 150 J-LI of 10% minimal essential media and incubated with 20 to 80 fJ.l sodium chromate (Dupont, Wilmington, Del.) for 1 hour at 3r C. The cells were then layered over 3 ml heat-inactivated fetal calf serum and centrifuged for 5 minutes at 200 g. The pellet was then washed with 10 ml 10% minimal essential media. The supernatant was discarded, 1 ml of 10%
965
966
Eriksen, Gonik, and Loa
••
60 50
~
0
October 199 1 Am J Obstet Gynecol
Pregnant (n=8) Non-pregnant (n=5)
Experimental release - Spontaneous release X 100 100 - Spontaneous release
40
X
0 ~ 0
!: 0 0~
and B equals counts per minute of the lysed pellet. Natural killer cell cytotoxicity equals:
Data are expressed as the mean ± SE of experiments. These data were analyzed with the arcsine transformation. Comparisons were made between the two groups by the analysis of variance, with p < 0.05 considered statistically significant.
30
20 10
Results 0+---
Fig_ 1. Baseline natural killer cell cytotoxicity (mean ± SE) in pregnant and nonpregnant subjects (p = 0.25). Effector/target cell ratio equals 100: 1.
minimal essential media was used to resuspend the pellet, and the cells were counted. The sodium chromatelabeled target cells were suspended at a final concentration of 1 X 10 5 cells per milliliter in 10% minimal essential media. Microcytotoxicity assay. Each assay was conducted in triplicate in polystyrene U-bottom-well microtiter plates (Corning, Corning, N .Y.) . The unstimulated wells, for pregnant and nonpregnant subjects, consisted of 50 ILl effector cells and 50 ILl 10% minimal essential media, yielding a final concentration of 5 x 106 cell per milliliter. Each of the HSV-stimulated wells consisted of 50 ILl effector cells and 50 ILl either live HSV -1 virus (2 X 10 5 plaque-forming units) or 50 ILl ultravioletinactivated HSV-l (2 x 105 plaque-forming units). The final concentration of effectors was 5 x 106 cells per milliliter with 1 x 10' plaque-forming units of HSV. The cells were incubated for 18 hours at 37° C. Experimental wells that contained live virus were washed, along with the unstimulated wells, three times with 10% minimal essential media. Subsequently, 50 ILl of 10% minimal essential media and 50 ILl of labeled target cells (1 x 105 cells per milliliter) were added to all of the wells. The final effector-to-target cell ratio was 100; 1. The covered plate was then centrifuged at 75 g for 5 minutes and allowed to incubate an additional 4 hours at 37° C in a humidified atmosphere of air (95%) and carbon dioxide (5%). Afterward 100 ILl was aspirated from the top of each well without disturbing the pellet. To each well, 100 ILl of 1 moll L sodium hydroxide was added, and the total volume was aspirated into a separate tube. The samples were counted in a Packard Cobra 50 10 )I-counter for 1 minute. Chromium release was calculated by the formula: Percent sodium 2A = - - x 100 chromate release A + B where A equals counts per minute in the top 100 ILl
Baseline natural killer cell cytotoxicity (mean ± SE) was lower in pregnant subjects (13.8% ± 2.4%) than in nonpregnant subjects (19.8% ± 3.7%) (Fig. 1). However, this did not reach statistical significance (p = 0.25). Incubation of effector cells with live HSV-l significantly increased natural killer cell cytotoxicity for pregnant (37.5% ± 6.2 %) and nonpregnant (49 .8% ± 7.6%) subjects, when compared with values for media alone (Fig. 2). As anticipated, the increase in cytotoxicity was similar for both groups (p = 0.25) (Fig. 2). Similar experiments with ultraviolet-inactivated HSV -1 (1 x 105 plaque forming units) showed no difference in mean natural killer cell cytotoxicity values (Fig. 3) when compared with those for media alone for either group. Additional experiments were conducted with I x 10' plaque forming units of ultraviolet-inactivated HSV-l, to determine if increasing the amount of antigen would have any effect on cytotoxicity. Again, no difference in mean natural killer cell cytotoxicity values were noted for either grou p (results not depicted).
Comment Pregnancy represents a unique immunologic state in which a balance exists between acceptance of the fetal allograft and host immunocompetence. A suppression in maternal cell-mediated immunity is assumed to be advantageous for the fetus but theoretically makes the pregnant woman more vulnerable to infections, particularly those of viral causes. Case reports 3 • 4 and animal work l3 suggest that the severity of an HSV infection is increased in pregnancy. A higher incidence of HSV infections and a longer duration of clinical symptoms have been reported in pregnancy. I. 2 Although firm epidemiologic data are lacking to confirm these observations,S these clinical reports have led to the suggestion that pregnancy is an immunocompromised state. Natural killer cells represent the first line of defense against viral pathogens and therefore play an important role in HSV infection. Data in regard to natural killer cells in pregnancy are limited and sometimes conflicting. Histochemical studies 9 and flow cytometrylO
Natural killer cell cytotoxicity to herpes in pregnancy
Volume 165 Number 4, Pan I
967
60 50
>
I::
0
><
...
30
~
20
0 0 0
#.
[3 Pregnant
40
•
(n=8)
Non-pregnant
(n=5)
*< pcO.05 vs media
10 0
Media
HSV-1
Fig. 2. Natural killer cell cytotoxicity in response to live HSV-I infection. Results (mean ± SE) show pregnant and nonpregnant subjects compared with controls (media alone, p < 0.05). Effectorltarget cell ratio is 100: 1.
have shown that the number of natural killer cells is not reduced during pregnancy. Natural killer cell function has been examined in several recent studies,7.12 which demonstrate a general decrease in natural killer cell activity during pregnancy, although this finding is not consistent. None of these previous studies have attempted to examine the effect of actual infection on natural killer cell immune responsiveness. To address this question, an in vitro model was developed in which peripheral blood mononuclear cells of pregnant and nonpregnant subjects were infected with live HSV-l virus. Natural killer cell cytotoxicity was measured in control and infected wells to determine if pregnancy altered immune responsiveness to infection. Our results in Fig. 1 show that baseline cytotoxicity tended to be lower in pregnant subjects, compared with nonpregnant controls. Although this did not reach statistical significance, the calculated power of these experiments (0 .54) to reveal a significant reduction in natural killer cells (20%) is limited. To have a power = 0.8, 14 paired experiments would need to be analyzed. Regardless, our results are consistent with data already available in the literature'·12 and are not central to the issue of HSV -induced immune responsiveness. Incubation of effector cells with live HSV-l resulted in a significant increase in natural killer cell cytotoxicity as compared with controls for both groups (Fig. 2). There was a 2.5-fold increase in cytotoxicity in the nonpregnant subjects, compared with a 2.7-fold increase in the pregnant subjects. This demonstrates that, in spite of lower baseline natural killer cell cytotoxicity values in the pregnant subjects, they were able to respond in a similar fashion as the nonpregnant subjects (Fig. 2). Viral infection leads to the production of a variety of cytokines. Several of these cytokines, in particular interferon alfa, are known to regulate natural killer cell function. Deficiencies in cytotoxic responses have been shown to result in a greater susceptibility to HSV in-
.. Pregnant (n=7) • Non-pregnant (n=7)
30
20
10
o Media
UVHSV-1
Fig. 3. Natural killer cell cytotoxicity in response to ultraviolet-inactivated HSV-l. Effector / target cell ratio IS 100 : J.
fection. 14 · 15 Furthermore, natural killer cell activity aginst HSV infection is enhanced by interferon in vitro,16. 17 and in vivo. IS Data from our laboratory confirm that interferon alfa can augment natural killer cell cytotoxicity in the pregnant host. 19 One could speculate that, although baseline natural killer cell activity during pregnancy may be decreased, cytokine production in response to infection may be normal, accounting for our results. Evidence to support this includes the ability of interferon to restore natural killer cell activity from peripheral blood of pregnant patients,1O and systemic interferon production is higher in pregnant mice than in controls. 13 Surprisingly, few data regarding cytokine production in human pregnancy exist. Studies examining the production of interleukin-Po and interleukin-22I demonstrate elevated or normal production rates, respectively, compared
968
Eriksen, Gonik, and Loa
with those of nonpregnant controls. We have shown that the monokines tumor necrosis factor and interleukin-6 also are significantly increased during pregnancy (unpublished observations). Although interferon alfa has not been directly measured during pregnancy, these data support the concept that its production during pregnancy is unchanged. Interestingly, we found that ultraviolet-inactivated HSV -1 did not augment natural killer cell activity as did infection with live virus (Fig. 3). This suggests that infection with live virus may be necessary to induce the intracellular production of interferon alfa. Kirchner et al. 22 demonstrated that interferon alfa could be detected after injection of live herpesvirus in mice but not with ultraviolet-inactivated herpes antigen. However, HSV antigen in much higher concentrations, and with extended incubation, has been shown to induce lymphocyte stimulation in previously sensitized cells!3, 24 The mechanism for this induction has not been identified, but it is most likely humoral in origin. Data from lymphocyte stimulation experiments cannot be extrapolated to explain natural killer cell function, because natural killer cells lyse target cells without prior antigenic sensitization. In conclusion, our results suggest that (1) infection with live virus, but not viral antigen, can augment natural killer cell cytotoxicity in vitro and (2) natural killer cell response to HSV -1 infection is not altered by pregnancy. We speculate that, during pregnancy, the production of interferon alfa is not attenuated in response to infection with live herpesvirus. This could explain the increase in natural killer cell cytotoxicity in vitro observed in both pregnant and nonpregnant subjects. Although our data suggest that isolated mononuclear cells respond appropriately to HSV infection during pregnancy, other factors, such as serum-mediated suppression of immune activity, may alter in vivo host responses.
REFERENCES 1. Whitley R], Nahmias A], Visintine AM, Fleming CL, Alford CA. The natural history of herpes simplex virus infection of mother and newborn. Pediatrics 1980;66:48994. 2. Brown ZA, Vontver LA, Benedetti], et al. Genital herpes in pregnancy: risk factors associated with recurrences and asymptomatic viral shedding. AM .I OBSTET GYNECOL 1985; 153:24-30. 3. Kobbermann T, Clark L, Griffen WT. Maternal death secondary to disseminated herpes virus hominis. AM ] OBSTET GYNECOL 1980;137:742-3. 4. Young E], Killam AP, Green .IF. Disseminated herpes virus infection: association with primary genital herpes in pregnancy. ]AMA 1976;235:2731-3. 5. Brabin B]. Epidemiology of infection in pregnancy. Rev Infect Dis 1985;7:579-603. 6. Luft B.J, Remington ]S. Effect of pregnancy on augmentation of natural killer cell activity by Corynebacterium
October 1991 Am J Obstet Gynecol
parvum and Toxoplasma gondii. ] Immunol 1984; 132:
2375-9. 7. Toder V, Nebel L, Gleicher N. Studies of natural killer cells in pregnancy. 1. Analysis at the single cell level. Clin Lab ImmunoI1984;14:123-7. 8. Okamura K, Furukawa K, Nakakuki M, Yamada K, Suzuki M. Natural killer cell activity during pregnancy. AM ] OBSTET GYNECOL 1984;149:396-9. 9. Baley]E, Schacter BZ. Mechanisms of diminished natural killer cell activity in pregnant women and neonates.] ImmunoI1985;134:3042-8. 10. Vaquer S, De la Hera A, ]orda.J, Martinez C, Escudero A, Alvarez-Mon M. Diminished natural killer activity in pregnancy: modulation by interleukin 2 and interferon gamma. Scand] Immunol 1987;26:691-8. 11. Toder V, Blank M, Gleicher N, Voljovich I, Mashiah S, Nebel L. Activity of natural killer cells in normal pregnancy and edema-proteinuria-hypertension gestosis. AM ] OBSTET GYNECOL 1983;145:7-10. 12. Gonik B, Loo LS, West S, Kohl S. Natural killer cell cytotoxicity and antibody-dependent cellular cytotoxicity to herpes simplex virus-infected cells in human pregnancy. Am] Reprod Immunol 1987;13:23-6. 13. Yamada K, Shimzu Y, Okamura K, Kumagai K, Suzuki M. Study of interferon production during pregnancy in mice and antiviral activity in the placenta, AM ] OBSTET GYNECOL 1985; 153:335-41. 14. Gresser MG, Tovey G, Maury C,Bandu T. Role of interferon in the pathogenesis of virus diseases in mice as demonstrated by the use of anti-interferon serum. II. Studies with herpes simplex, Maloney sarcoma, vesicular stomatitis, Newcastle disease, and influenza viruses. ] Exp Med 1976;144:1316-3. 15. Virelizier JL. Viral infections in patients with selective disorders of the interferon system. In: Dela Maza LM, Peterson EM, eds. Medical urology. Proceedings of the 1981 international symposium on medical virology, Anaheim, California, October 8-10, 1981, New York: Elsevier Biomedical, 1982. 16. Platsoucas CD. Augmentation of human natural killer cytotoxicity to alpha-interferon and inducers of gamma-interferon. Analysis of monoclonal antibodies. In: Herberman RB, ed. NK cells and other natural effector cells. New York: Academic Press, 1982:355, 17. Bishop GA, McCurry L, Schwartz SA, Glorioso ]C. Activation of human natural killer cells by herpes simplex virus type I-infected cells. Intervirology 1987;28:78-88. 18. Gidlund M, Om A, Wigzell H, Senik A, Gresser 1. Enhanced NK activity in mice infected with interferon and interferon inducers. Nature 1978;273:759-62. 19. Gonik B, Loftin K, Tan NS, Crump J. Immune modulation of natural killer cell cytotoxicity against herpes infected target cells in pregnancy. Am] Reprod Immunol 1990;24:95-8. 20. Bessler H, Sirota L, Dulitzky F, Djaldetti M. Production of interleukin-l by mononuclear cells of newborns and their mothers. Clin Exp Immunol 1987;68:655-61. 21. Hauser GJ, Lidon A, Zakuth V, et al. Immunocompetence in pregnancy: production of interleukin-2 by peripheral blood lymphocytes. Cancer Detect Prev SuppI1987;1:3942. 22. Kirchner H, Engler H, Schroder CH, Zawatzky R, Storch E. Herpes simplex virus type I-induced interferon production and activation of natural killer cells in mice. J Gen Virol 1983;64:437-41. 23. Rosenberg GL, Farber PA, Notkins AL. In vitro stimulation of sensitized lymphocytes by herpes simplex virus and vaccinia virus. Proc Natl Acad Sci USA 1972;69:75660. 24. Jacobs RP, Avrelian L, Cole GA. Cell-mediated immune response to herpes simplex virus: type-specific lymphoproliferative responses in lymph nodes draining the site of primary infection.] Immunol 1976;116;1520-5.