OUP user menu

Pathogenesis of herpes simplex virus type 2 experimental genital infection in pregnant mice

Norberto A. Sanjuan, María Noel Zimberlin
DOI: http://dx.doi.org/10.1111/j.1574-695X.2001.tb01570.x 197-202 First published online: 1 April 2001

Abstract

The progression of herpes simplex-2 genital infection in pregnant mice was studied by detection of viral antigens using immunoperoxidase in tissue sections, electron microscopy and virus isolation. The majority of mice (66.66%) died at 8–9 days post-inoculation. Abortions were observed in 69.23% of the infected mice along with impairment of labor and delivery. Herpes antigens were detected in most of the autonomic nerves of the uterus, including those surrounding small arterioles in the myometrium and the Auerbach and Meissner plexa of the large bowel, but not in the abortions or placentas. The infection of uterine autonomic fibers and myometrial cells could explain the delivery impairment and could have provoked a decrease in blood flow leading to abortions.

Keywords
  • Herpes simplex virus
  • Pregnancy
  • Abortion
  • Delivery
  • Mouse

1 Introduction

Herpes simplex virus-2 (HSV-2) genital infection is one of the most frequent and relevant venereal diseases occurring in women. After local replication at the entrance site, HSV-2 establishes a life-long latent infection, with recurrent genital lesions and eventual systemic dissemination (reviewed in [1]).

When the HSV-2 genital infection occurs during pregnancy in women, several additional complications have been reported. Herpes antigens have been found by immunocytochemistry [2,3] and HSV DNA has been detected by PCR [4] in spontaneous and therapeutic abortions. Also, when compared with an uninfected control population, a significant increase of spontaneous abortions was described in women with latent HSV-2 genital infections (diagnosed by a microneutralization test) [5]. Preterm delivery and low infant birth weight were reported in women with HSV asymptomatic genital shedding [6]. Hydrops, cystic brain degeneration and unexplained fetal growth retardation were also suggested to be related to HSV-2 genital infection as well [3]. After birth, the HSV infection of neonates born to infected mothers is clearly the most serious and even fatal complication at hand [711].

The pathogenesis of HSV-2 genital infection during pregnancy is far from clear. Neither the tissues infected by HSV-2 in pregnant women nor the cells involved in transplacental viral dissemination to embryos have been clearly identified yet [12]. Moreover, the information obtained from many clinical trials is often limited because of ethical or technical issues. Here we present an experimental model that allows a detailed study of the progression of HSV-2 genital infection in pregnant mice.

2 Materials and methods

2.1 Virus

The ATCC VR-734 prototype strain of HSV-2 was used. Virus stock was prepared by infection of Vero cell cultures in plastic flasks, maintained with minimal essential medium (MEM, Gibco) supplemented with 5% calf serum, and incubated at 37°C in an atmosphere containing 5% CO2. When 80% of the cells showed cytopathic effect, cultures were harvested, frozen and thawed (three times) and spun down to discard cell debris. The supernatants were then stored in liquid nitrogen until used. Virus was titrated by the plaque-forming units (pfu) method in Vero cell monolayers covered with culture medium and methyl-cellulose.

2.2 Animals

Specific pathogen-free Balb/c mice were used. The animals were obtained from the bioterium of the National Academy of Medicine (Argentina), kept five to a box, and fed on pellets ad libitum. Mice were maintained at constant temperature, with natural cycles of light and darkness.

2.3 Experimental design

Twenty-one pregnant mice were infected intravaginally (IVAG) with HSV-2 and 16 were mock-infected and used as negative controls. The diagnosis of pregnancy was done by the observation of a mucous plug in the vagina, then confirmed at necropsies. Thirteen mice were infected at 10–14 days pregnant, and eight mice at 3–4 days pregnant (the normal gestation period of a mouse is 19–21 days). The control group was composed of 10 mice which were mock-infected at 10–14 days pregnant, and six animals at 3–4 days pregnant. The whole study was divided in three consecutive experiments: in the first one the eight mice at 3–4 days pregnant were infected as well as their mock-infected controls, while the animals at 10–14 days pregnant were divided in two groups, each of them composed of infected mice and their uninfected controls, and employed consecutively in the other two experiments. Before the inoculation, each and every mouse was gently swabbed IVAG with sterile, dry cotton wool. The infection was performed using 5×105 pfu of HSV-2 contained in 0.05 ml of cell culture medium. The mock-infected control group received 0.05 ml of normal cell culture medium obtained from the supernatant of uninfected Vero cell cultures. Clinical signs were then recorded daily. During the premortem stage, mice were sacrificed using an excess of ether anesthesia, following the rules regarding animal welfare protection which have been established in this institution. Immediately, heparinized blood samples were taken from each animal by cardiac puncture, then kept separately in liquid nitrogen until detection of viremia, done by adsorption on Vero cell monolayers and titration by pfu. Complete necropsies were performed as follows: the uterus was dissected first, each embryo was then grossly studied and separately processed for histology. The body of each mouse was cut into five serial transversal sections, from the vulva to the lower kidney tip. In some cases, longitudinal slices were taken. All the other organs were dissected separately including the liver, kidneys, spleen, lungs, heart, skin, brain, and bone marrow. Samples were fixed in Bouin fluid, then routinely embedded in paraffin. Sections were stained with hematoxylin and eosin, and adjacent sections were immunolabeled with the peroxidase-antiperoxidase (PAP) technique to detect the presence of HSV-2 antigens.

2.4 Immunocytochemistry

The PAP method was employed as previously described [13,14] using Dako polyclonal antibodies. Briefly, sections were consecutively treated with xylene, decreasing concentrations of ethanol and 0.05 N Tris-HCl (pH 7.6); endogenous peroxidase was removed by incubation in methanol containing 2% hydrogen peroxide, and any eventual unspecific reaction was blocked by incubation of the slides with 5% normal goat serum in 0.05 N Tris-HCl. The primary antibody was a rabbit against HSV-2, diluted 1:1000. Slides were then incubated with goat serum anti-rabbit immunoglobulins 1:50 and the PAP complex in rabbit 1:250. After every step, slides were washed with 0.05 N Tris-HCl. The developer mix used was composed of 0.03% 3,3′-diaminobenzidine and 2% hydrogen peroxide in 0.05 N Tris-HCl and slides were then slightly counterstained with hematoxylin. Paraffin-embedded slides obtained from brains of HSV-2 intracerebrally infected mice were used as positive controls, and normal uninfected mouse brains were used as negative controls in every immunolabeling.

2.5 Electron microscopy

Samples were taken from the vulva, vagina, uterus, embryos and large bowel of each animal immediately after death. Tissues were minced into 0.5-mm blocks, fixed in a solution composed of 4% formaldehyde–1% glutaraldehyde in phosphate buffer saline (PBS), pH 7.4, post-fixed in osmium tetroxide, and embedded in Vestopal. Grids were stained with uranyl acetate and lead citrate, and observed using a Zeiss EM 109-T electron microscope at 80 kV.

2.6 Viremia detection

Frozen blood samples were thawed and diluted in PBS, then adsorbed 1 h on Vero cell monolayers. Cell cultures were then washed, overlaid with MEM and methyl-cellulose containing 5% calf serum and incubated at 37°C in a 5% CO2 atmosphere for 7 days. HSV-2 stock was titrated simultaneously as a positive control, as well as mock-infected monolayers as internal controls.

3 Results

Fourteen out of 21 mice (66.66%) developed genital and neurologic signs of infection and died between 8 and 9 days post-infection (pi), (10/13 mice inoculated at 10–14 days pregnant and 4/8 mice inoculated at 3–4 days pregnant), while none of the mock-infected animals showed neurologic signs nor did they die. One mouse belonging to the group inoculated at 10–14 days pregnant was found dead during the course of the experiment and has not been included in this study.

Vulvar congestion, edema and vaginal flux were detected at 4–5 days pi. At day 6 pi, progressive loss of the hair around the external genitalia and anus was found. Hind-limb paresia developed at 7–8 days pi and all the infected animals died at 8–9 days pi, with premortem wheezing and lethargy.

At necropsy, a massive dilatation of the whole large bowel with fecal retention were observed, plus, urinary bladder distention with urine retention were also observed in all the infected animals. In the four mice that developed infection after inoculated at 3–4 days pregnant, multiple nodular, round-shaped, whitish structures with an average size of 2–3 mm were observed after opening the uterus. Also, 11–12-mm embryos were observed in the uterus of 5/9 mice which had had clinical signs of infection after inoculated at 10–14 days pregnant. The histology of these structures showed mainly necrotic embryonic tissue (not shown) and therefore they were diagnosed as abortions. The average was three abortions per uterus horn. The other four mice inoculated at 10–14 days pregnant which developed clinical signs of HSV-2 infection had full-term developed fetuses (average seven per mouse). In every case, the first fetus was trapped along the vagina, and also showed necrotic tissues by histology. None of the mock-infected mice showed any evidence of abortions and there were no failures during delivery either.

The PAP method detected HSV-2 antigens in the vulvar, vaginal and endocervical epithelia (Fig. 1A), in the autonomic Auerbach and Meissner plexa of the large bowel (Fig. 1B) and in pelvic nerves (Fig. 1C). The spinal cord showed positively labeled areas in lateral and dorsal columns and dorsal root ganglia (Fig. 1D) from the lumbar to the cervical segments. Herpes antigens were also present in periarteriolar nerves of the myometrium (Fig. 2A) and in myometrial cells (Fig. 2B) as well as in perigenital and perivesical nerves. Double-layered, 120–150-nm, particles resembling herpes virions were observed in PAP-positive tissues (Fig. 2C). There were no HSV-2 antigens in the brain, pons or medulla oblongata, and viremia was not detected in any of the infected animals. HSV-2 antigens were not observed in any of the abortions, fetuses or placentas of the infected mice. Mock-infected animals did not show any positive labeling.

Figure 1

HSV-2 progression from the vagina to the central nervous system detected by PAP technique counterstained with hematoxylin (×200). A: Vaginal epithelium showing nuclear inclusion bodies; B: the dark spots are heavily infected fibers and neurons belonging to the Auerbach and Meissner plexa of the large bowel; C: PAP-positive axons in a pelvic nerve; D: HSV-2 antigens in neurons of a dorsal root ganglion and in lateral and dorsal columns of the spinal cord not counterstained with hematoxylin (×80).

Figure 2

HSV-2 antigens in the uterus detected by PAP technique and counterstained with hematoxylin. A: The dark spots indicate heavily infected small autonomic nerves surrounding an arteriole in the myometrium (×250); B: herpes antigens in myometrial cells (×125). C: herpes particle budding from the nuclear membrane of a myometrial cell observed by transmission electron microscopy (×100 000); this finding demonstrates that herpes antigens detected by PAP method correspond with infectious virus particles.

4 Discussion

In this work we describe the progression of HSV-2 acute genital infection during pregnancy in a mouse experimental model. For this to be achieved we infected Balb/c mice IVAG at two different points of time during pregnancy: 3–4 days and 10–14 days, and detected HSV-2 antigens using the PAP method. The use of this technique on serial paraffin-embedded slides along with the observation of developing clinical signs, electron microscopy and viremia detection allowed a detailed comprehension of virus progression from the inoculation site towards organs [13]. Moreover, the PAP method enabled us to avoid using the classic procedure of virus isolation from every organ, which frequently leads to errors in the results given the risks of undesirable virus contamination of samples.

The first conclusion is that the average mortality of IVAG HSV-2-infected pregnant mice is not significantly different from that which was reported in non-pregnant IVAG infected mice [14,15]. Some authors have described pregnant or progesterone-administered mice as being more susceptible to HSV-2 IVAG infection than those which are non-pregnant [16,17], while others showed no difference in HSV-2 susceptibility between pregnant and non-pregnant mice [18]. The results obtained in the experimental model and described herein may be due to the fact that each animal was IVAG swabbed with dry cotton wool immediately before the IVAG inoculation with HSV-2. In preliminary experiments (data not shown), we had demonstrated that the microtrauma produced by this procedure in the vaginal mucosa is crucial for HSV-2 IVAG infection to occur in non-pregnant mice, the microtrauma probably being more important than the serum levels of the progesterone.

After a 4-day incubation period, HSV-2 IVAG inoculated mice developed local genital infection which led to death in 4–6 days. The most striking fact of this acute lethal infection is the high incidence of abortion in the HSV-2 IVAG infected mice. All the animals (4/4) that showed clinical infection after inoculation at 3–4 days pregnant and 5/9 mice infected at 10–14 days pregnant had abortions. According to the size and implantation in the endometrium, the mice infected at 3–4 days pregnant had had 8–9-day-old abortions, while the mice infected at 10–14 days pregnant had had 14–15-day-old abortions. The ages of the abortions were based on their sizes, and these findings were made according to the parameters reported by K.A. Rafferty [19]. This means that the death of embryos or fetuses occurred 4–5 days after infection. Interestingly, this time coincides with the incubation period of the HSV-2 IVAG infection, when the first signs of disease appear. It could be argued that the abortions were a consequence of the inoculation trauma but, in this case, an equal incidence of abortions should have been observed in the mock-infected mice. This last group of animals showed no abortions at all. It is also interesting that when infection was performed earlier during pregnancy, the incidence of abortions was higher (4/4 in animals 3–4 days pregnant and 5/9 in those which were 10–14 days pregnant). Even though the number of mice that developed acute infection and had abortions after being inoculated at 3–4 days pregnant is small (n=4), none of the mice in the mock-infected control group showed the same effect. Also, the Balb/c strain used in this experiment normally does not have a high incidence of spontaneous abortions. The presence of HSV-2 antigens was not detected in any of the placentas, fetuses or dead embryos of either group. Moreover, viremia was absent, and no PAP-positive labeling was observed in the liver, spleen, bone marrow or kidneys, casting doubts on an eventual massive dissemination of HSV-2 through the bloodstream.

How were the abortions induced? The histology of the abortions showed areas of coagulative necrosis. This kind of necrosis is particularly observed when the blood supply to tissues is diminished. The presence of HSV-2 antigens in nerves located in the large bowel and the urinary bladder after the IVAG infection demonstrated that the fibers of the autonomic nervous system were widely involved. In this connection, the detection of HSV-2-positive labeling in the nervous fibers surrounding small arterioles of the myometrium is particularly important. It is known that the mouse uterus contains cholinergic nerves which belong to a wide autonomic plexus that also innervates the large bowel and the urinary bladder, and it is able to regulate the normal flow of blood in the small arterioles [20]. We are reporting here that most of these nerves were heavily infected after the HSV-2 IVAG infection took root. So, it is possible that the abortions described in this experiment were due to the insufficient blood supply to the embryos and fetuses due to the HSV-2 infection of autonomic nerves and fibers which control the arteriolar pressure. According to the PAP labeling, virus replication occurred initially in the vulvovaginal epithelium and the perineal skin, then progressed along nerves and fibers mainly belonging to the autonomic nervous system. This is patently demonstrated by the presence of HSV-2 antigens in the neurons and fibers of the Auerbach and Meissner plexa which were observed from rectum to caecum. This massive infection led to protracted bowel paralysis and fecal retention. The autonomic nerves of the urinary bladder were also infected causing distention of the organ, thus, urinary retention. In fact, these two features probably provoked some serious internal medium disturbances that caused the animal's death. Since there was no HSV-2 antigen observed in the cerebrum, cerebellum, pons or medulla oblongata, encephalitis described by other authors [21,22] should be ruled out as being the cause of death. Muscular cells belonging to the myometrium and small nerves located in the wall of the cervix and myometrium were infected. The autonomic failure and herpes infection of myometrial fibers could also explain why the fetuses became trapped in the vagina of animals infected at 10–14 days pregnant. In this group of mice the delivery process seemed to have been started, but somehow, could not be completed causing the death of the fetuses. The reason why this happened could have been inefficiency of myometrium contractions, a consequence of HSV-2 infection of the autonomic plexus which innerves the uterus, and the heavy infection of the myometrial muscle cells. The results obtained in this work do not allow us to explain the lack of embryo or placental infection. The presence of interferons in placenta proposed by other authors [18] could probably explain this. In conclusion, the present study suggests that the herpes simplex infection of autonomic fibers that innervate the uterus and uterine arterioles as well as the myometrial cells is associated with a high incidence of abortions and also with delivery impairment. Hopefully, the findings described in this experimental model will lead to new clinical studies in pregnant women in order to elucidate the role of the autonomic nervous system after HSV-2 primary genital infection during pregnancy.

Acknowledgements

This work was supported by University of Buenos Aires Grant TM-016, and is included in the research program of the C.A.E.H. (Argentine chapter of the International Herpes Management Forum), sponsored by Glaxo-Wellcome Argentina.

References

  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
View Abstract