Vet Pathol 36:267275 (1999)

Detection of Viral Antigen in Placenta and Fetus of Cattle Acutely Infected with Bovine Viral Diarrhea Virus
B. FREDRIKSEN, C. M. PRESS, T. SANDVIK, S. A. DEGAARD,
AND

T. LKEN

Departments of Reproduction and Forensic Medicine (BF, SA), Morphology, Genetics and Aquatic Biology (CMP), and Large Animal Clinical Sciences (TL), Norwegian School of Veterinary Science, Oslo, Norway; and Department of Virology and Serodiagnostics, National Veterinary Institute, Oslo, Norway (TS) Abstract. The reproductive organs and fetuses of seven Norwegian Red heifers were investigated for the presence of bovine viral diarrhea virus (BVDV) antigen during the time of initial transplacental transmission of the virus. The heifers were inoculated with a noncytopathogenic BVDV at day 85/86 of gestation and were slaughtered at day 7, 10, 14, 18, or 22 postinoculation (pi). Cryostat sections of uterus, ovaries, placentomes, intercotyledonary fetal membranes, and fetal organs were examined using immunohistochemical techniques. A double immunouorescence technique was used to identify cells that showed staining with antibodies against the leukocyte common antigen CD45 or the intermediate lament vimentin and BVDV antigens. The earliest stage of infection at which BVDV antigen could be detected in the fetuses was 14 days pi. At this stage, BVDV antigen was detected in cells of mesenchymal origin in the lungs and in large cells that morphologically resembled immature megakaryocytes in the liver. In the intercotyledonary fetal membranes and in the placentomes, BVDV antigen was not detected until 18 and 22 days pi, respectively. BVDV antigen was not detected in maternal tissue from any of the heifers. The present results indicate that fetal infection with BVDV can take place without preceding or simultaneous high concentrations of BVDV in uterus or placenta of acutely infected heifers. Key words: uterus. Bovine viral diarrhea virus; experimental infection; fetus; immunohistochemistry; placenta;

In susceptible pregnant animals, bovine viral diarrhea virus (BVDV) readily crosses the placenta and infects the fetus.10,28 The outcome of the fetal infection depends on the properties of the virus and on the stage of gestation at which the dam is infected.30 Only the noncytopathogenic biotype of the virus has been reported to establish a persistent infection in the fetus. In cattle, a persistent infection with BVDV-specic immunotolerance can occur when a susceptible dam is infected with BVDV from approximately day 30 of gestation or even earlier23 and until the fetus establishes immunocompetence against BVDV around day 120 of gestation. Although fetal infection during this period of gestation may also result in fetal death, abortion, mummication, malformations, or stillbirth,9,10,22,24,36 apparently healthy animals immunotolerant to and persistently infected with BVDV are of great concern for cattle populations because these cattle continue to shed large amounts of virus as long as they live and, accordingly, are very important in the epidemiology of BVDV infections.19 Even if the outcome of fetal infection with BVDV is well recognized, the effect of the virus on the placenta and the rate and manner in which the virus crosses the placenta and infects the fetus have not been

established. Most researchers have concluded that there are only minor or no lesions in the placentomes after experimental infection of cattle with BVDV10,22 or after abortions where BVDV is suspected as the causative agent.31 In other studies performed in pregnant sheep, prominent lesions have been discovered in the placentomes after experimental infection.32,41 The discrepancies among the results of these studies may to some extent be explained by the use of different strains or biotypes or even mixtures of different pestiviruses. It is still unclear whether fetal death and abortions follow directly from viral invasion of the fetuses or whether they are caused by damage to the maternal placenta that disrupts the vascular supply of nutrients.7 The tropism of BVDV for lymphoid, neural, and epithelial tissues has been demonstrated;1,35,15,18 however, most studies of viral distribution have been focused on postnatal persistently infected animals, and the presence of virus in the uterus, placenta, and fetus during the time of initial transplacental infection has received less attention. The placenta has been considered the organ of choice for virus replication,41 and both maternal and fetal epithelial cells of uterus, placentomes, and intercotyledonary fetal membranes have

267

268

Fredriksen, Press, Sandvik, degaard, and Lken

Vet Pathol 36:4, 1999

been shown to be infected in persistently infected heifers.16 Thus, direct cell-to-cell transmission may represent the pathway for transplacental infection of the fetus. Alternatively, transplacental infection with BVDV may occur via a vascular pathway. To investigate the possible routes of fetal infection with BVDV, and in particular the involvement of the placentomes, immunohistochemical techniques were used to determine the time of appearance and the cellular localization of viral antigens in the uterus, placenta, fetal membranes, and some fetal organs in heifers during an acute experimental infection with a noncytopathogenic BVDV in early pregnancy. Materials and Methods
Animals and inoculation Eight pregnant heifers of the Norwegian Red breed were housed in an isolation unit at the Norwegian School of Veterinary Science during the experiment. The heifers had been inseminated with semen from a certied articial insemination center (Norwegian Cattle Association, Hamar, Norway) where all bulls have tested free from BVDV since 1985. The heifers were all shown to be free from BVDV by an antigen enzyme-linked immunosorbent assay (ELISA) performed on blood leukocytes38 and by virus isolation in cell cultures.25 They were free from serum neutralizing antibodies to BVDV at the beginning of the experiment. Seven of the animals (heifer Nos. 17) were inoculated intramuscularly with BVDV at day 85/86 of gestation and slaughtered 7, 10, 14, 14, 18, 18, or 22 days postinoculation (pi), respectively. The remaining animal served as a negative control and was given an intramuscular inoculation with 5 ml of a cell culture medium at day 85 of gestation and slaughtered 14 days later (heifer No. 8). One pregnant uterus at a comparable stage of gestation was collected from a heifer at the abattoir and used as an additional negative control (heifer No. 9). Inoculum The virus used was a noncytopathogenic BVDV isolated in 1993 from a persistently infected but clinically healthy Norwegian heifer (No. 93/4618-226).37 Partial nucleotide sequencing of the 5 untranslated region of this isolate showed greatest homology to type-1 BVDVs of European origin (unpublished data). Because eld isolates may consist of several strains, the virus was biologically cloned three times by endpoint titrations in fetal bovine kidney cell cultures from which the inoculum was harvested at a concentration of 104.44.9 CCID50/ml in Eagles minimal essential medium. Each animal was inoculated with approximately 104 CCID50 BVDV diluted with saline to a volume of 5 ml. Clinical, virologic, serologic, and hematologic examinations The animals were examined clinically twice daily from inoculation to slaughter. Rectal temperature, appetite, the presence of a nasal and/or vaginal discharge, and the consistency of the feces were recorded. Blood samples were collected daily, and serum was examined for BVDV by in-

oculation of fetal bovine kidney cell cultures, which subsequently were immunostained for noncytopathogenic BVDV.25 Serum samples obtained twice weekly were examined for antibodies to BVDV using a commercial indirect ELISA kit (Svanovir , SVANOVA, Uppsala, Sweden).21 Leukocyte and platelet counts and electrophoresis of serum proteins were also performed with samples collected twice weekly. Postmortem examination and collection of tissue specimens At slaughter of the heifers, the uterus, ovaries, placentomes, fetal membranes, and fetuses were examined grossly. Fetal blood samples were collected for virus isolation in cell cultures25 or for examination with a reverse-transcriptase nested polymerase chain reaction (RT-PCR).39 The serum samples were also examined for antibodies to BVDV using the ELISA kit.21 Tissues for immunohistochemical examination were collected from the liver, lung, spleen, brain, and intestine of the fetuses, the ovaries and uterus of the heifers, and the placentomes and intercotyledonary fetal membranes (chorioallantois and/or amniochorion) using sterile disposable equipment. Tissue samples were embedded in TissueTek OCT compound (4583, Miles, Elkhart, IN) and frozen in chlorodiuoromethane chilled in liquid nitrogen, wrapped in aluminum foil, and stored at 70 C. Sections (68 m thick) were cut with a cryostat and mounted on glass slides precoated with glycerol or poly-L-lysine (P8920, Sigma Chemical Co., St. Louis, MO). Tissue from the fetal livers, spleens, and lungs were also frozen at 70 C and tested for BVDV in cell cultures25 and with RT-PCR.39 Immunohistochemical investigations An avidinbiotinperoxidase complex method (ABC-PO, Vectastain ABC kit, Vector Laboratories, Burlingame, CA) was used to examine sections from all samples. In addition, an indirect immunouorescence technique and a double immunouorescence technique were used as previously described.16 For both the immunoperoxidase and the immunouorescence techniques, a mixture of ve commercially available murine monoclonal antibodies (WB 160, WB 162, WB 210, WB 215, and WS 363, Central Veterinary Laboratory, Weybridge, UK) was used as the primary antibody for demonstration of BVDV antigens. This mixture, which contains four antibodies of isotype IgG1 and one of isotype IgG2b, is expected to detect most isolates of BVDV and border disease virus.13 A mixture of two other monoclonal antibodies (13G4 and VPM20, Moredun Animal Health, Edinburgh, UK) of isotype IgG2a against the nonstructural p125/p80 polypeptide of BVDV was also used on some sections with the ABC-PO method.12,14 The double immunouorescence technique used isotype-specic secondary antibodies to distinguish between a murine monoclonal IgG2a antibody against the leukocyte common antigen CD45 (SBU-LCA) or a murine monoclonal IgG2a antibody against the intermediate lament vimentin (3B4, M7020 Dako, Denmark) and the mixture of BVDV antibodies. The isotypespecic secondary antibody used to detect the BVDV antibody mixture would not detect the BVDV antibody with the IgG2b isotype. For each organ examined, equivalent sections

Vet Pathol 36:4, 1999

BVDV in Placenta and Fetus of Acutely Infected Cattle

269

Table 1. Examinations for BVDV and antibodies against BVDV in 8 heifers and their fetuses after intramuscular inoculation with BVDV at day 85/86 of gestation.
Blood Heifer Heifer No. Slaughter Day (pi)* Virus Isolation Antibodies Virus Isolation Fetus PCR Antibodies Fetal Organs Uterus (virus Virus isolation) Isolation PCR

1 2 3 4 5 6 7 8 9

7 10 14 14 18 18 22 Not infected Not infected

(days 5, 6) (days 5, 6, 7) (days 6, 7) (day 6) (day 6) n.d. (day (day (day (day (day n.d. 13) 13) 16) 16) 16) n.d. n.d. n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d.

/ n.d. n.d. n.d. n.d. # # # ( )/

n.d. n.d. n.d.

* pi postinoculation. PCR reverse-transcriptase nested polymerase chain reaction. Including liver, lung, and/or spleen. n.d. not done. Lung was positive, liver was negative. # For these samples, the parallel negative control cell cultures were also positive. Lung and spleen were weakly positive; liver was negative.

from persistently infected animals16 were used as positive controls. As a negative control for the immunohistochemical procedures, the primary antibodies were excluded from the tissue sections.

Results
Clinical ndings

All experimentally infected animals except one had a moderate rise in body temperature for 1 day. Four of the animals had a slight nasal discharge for a few days, and two had a cough for 12 days. Appetite and feces remained normal throughout the experiment, and vaginal discharge was not observed from any of the animals. The control animal remained healthy throughout the study. The concentrations of serum proteins and the gamma globulin fraction remained stable throughout the experiment for all animals. Leukocyte and platelet counts were within normal range.
Virology and serology

from the fetus examined at day 7 pi (heifer No. 1), which was positive for BVDV with RT-PCR but negative with virus isolation. The organs of the fetus examined 10 days pi (heifer No. 2) and the abattoir control (No. 9) were negative for BVDV both in cell cultures and with RT-PCR. The fetal organ samples from the control animal (No. 8) were positive for BVDV by isolation in cell cultures, and the lung and spleen samples were weakly positive also with RT-PCR, but the liver sample was negative. For virus isolation, samples from the control animal were examined in a set-up where the parallel negative control cell cultures were also positive. Because of this fetal blood samples corresponding to the affected samples were examined by RT-PCR; heifer No. 8 (control) was negative and heifer Nos. 5 and 6 (both day 18 pi) were positive. All fetal blood samples tested were negative for antibodies to BVDV.
Pathologic ndings

The results are summarized in Table 1. Microscopic examination of immunostained cell cultures revealed BVDV in blood from ve of the heifers between days 5 and 7 pi. Antibodies against BVDV were not demonstrated until days 1316 pi, with increasing antibody levels in subsequent samples. The control animal (No. 8) remained both virus and antibody negative until slaughter. In the fetal samples, BVDV was demonstrated by virus isolation in one or more organs from the fetuses surviving until day 14 pi. For the organ samples examined by RT-PCR, the results were in accordance with those of virus isolation except for the lung sample

Gross morphologic changes were not detected in the uterus or its contents from any of the animals. One of the heifers (No. 6) carried twins.
Immunohistochemical investigations

The results for both mixtures of BVDV antibodies used in the immunoperoxidase technique and for the immunouorescent staining are given in Table 2; the results of the double immunouorescent staining are given in Table 3. BVDV antigens were not detected in cryosections from the ovaries or uterus from the heifers or from the maternal part of the placentomes (the caruncles) from any of the animals. None of the sec-

270

Fredriksen, Press, Sandvik, degaard, and Lken ABC Pb

Vet Pathol 36:4, 1999

Table 2. Results of immunohistochemical staining* of organs from heifers and their fetuses after intramuscularly inoculation of the heifers with a noncytopathogenic BVDV at day 85/86 of gestation.

Heifer Uterus Ovary Placentomes Fetal membranes

Fetus Intestine Liver Lung Spleen Brain

* Pa ABCperoxidase with a mixture of ve monoclonal antibodies against BVDV; IF indirect immunouorescence with the same mixture of antibodies as for Pa; Pb peroxidase with a mixture of two monoclonal antibodies against nonstructural proteins of BVDV.

Organ

tions from the control animals or the animals slaughtered 7 or 10 days pi showed any specic staining. The two mixtures of BVDV antibodies both gave specic staining in all organs when used on samples from persistently infected animals. In the samples from the two animals slaughtered 14 days pi, BVDV-positive cells were detected in the lungs of both fetuses and in the liver of one fetus (heifer No. 3). In this liver, only large cells negative for both CD45 and vimentin and that morphologically resembled immature megakaryocytes were positive for BVDV. In the lungs, most virus-positive cells were positive for vimentin and negative for CD45 (Fig. 1). For the two fetuses collected 18 days pi, reactivity for BVDV was detected in the spleen, brain, and intestine and in the intercotyledonary fetal membranes. There was a large increase in the proportion of BVDVpositive cells in the fetal liver and lungs compared with the proportion in these organs in fetuses collected 14 days pi. At 18 and 22 days pi, epithelial cells in the lungs also showed reactivity for BVDV (Fig. 2), as did smooth muscle cells in the tunica media of blood vessels. Both epithelial and nonepithelial cells in the fetal intestine showed reactivity for BVDV. The nonepithelial BVDV-positive cells were mainly localized in the lamina propria and were positive for vimentin but negative for CD45. In the liver, lymphocytelike cells that also showed reactivity for CD45 were positive for BVDV at this stage. The hepatocytes did not show reactivity for BVDV. In the spleen, many cells showed reactivity for BVDV at 18 and 22 days pi. These cells were negative both for CD45 and for vimentin. Many platelets showed reactivity for BVDV. Large BVDV-positive cells that resembled immature megakaryocytes were also present in the spleen (Fig. 3). In the fetal membranes, mostly small vessels and subepithelial cells stained for BVDV (Fig. 4). These cells also showed reactivity for vimentin but not for CD45. A few epithelial cells, both in the chorion and in the amnion/allantois, also showed staining for BVDV. BVDV antigen was demonstrated in the placentomes only in the animal slaughtered 22 days pi (Fig. 5). Small vimentin-positive cells on the fetal side of the fetomaternal interface were the main cell population that stained for BVDV. Occasionally, CD45-positive fetal cells in the placentomes showed reactivity for BVDV antigen. In the other organs, the results at 22 days pi were similar to the results seen at 18 days pi, but in most organs the presence of cells showing staining for BVDV had increased. Discussion In the present study, the earliest stage of infection at which BVDV antigen was detected in fetal or uter-

Heifer Nos.

Pa

IF

Pb

Pa

IF

Pb

Pa

IF

Pb

Pa

IF

Pb

Pa

IF

Pb

Pa

IF

Pb

Pa

IF

Pb

Pa

IF

Pb

Pa

IF

Vet Pathol 36:4, 1999

BVDV in Placenta and Fetus of Acutely Infected Cattle * Cells positive for BVDV were either positive ( ) or negative ( ) for the respective antigen; ( ) the main cell population positive for BVDV was negative for vimentin/CD45, but a few cells showed staining for both antigens; / of the BVDV positive cells, both cells positive and negative for CD45 occurred frequently. pi postinoculation.

271

Table 3. Results of double immunouorescence* with antibodies against BVDV and vimentin/leukocyte common antigen CD45 performed on samples from placenta and fetuses of heifers intramuscularly inoculated with a noncytopathogenic BVDV at day 85/86 of gestation.

Small cells of mesenchymal origin, capillaries Cells of mesenchymal origin, blood vessels

Epithelial cells, cells of mesenchymal origin in lamina propria Immature megakaryocytes, lymphocytes Cells of mesenchymal origin, epithelium Immature megakaryocytes, platelets

ine tissues using immunohistochemical techniques was 14 days after inoculation of the heifers. At this early stage of infection, scattered clusters of BVDV-infected cells were detected in fetal liver and lung only. In the intercotyledonary fetal membranes and the placentomes, BVDV was not detected until days 18 and 22 pi, respectively. In most studies where fetal infection has been investigated, early transfer of virus has not been addressed. However, in one study, BVDV was detected in ovine fetal tissues 7 days after inoculation of pregnant ewes,32 and in another the time of fetal death in cattle was determined to be from 12 days pi with BVDV. The time of virus transfer from the mother to the fetus is probably inuenced by virus strain, host species, and individual variation. In the present study, intramuscular inoculation was used to reduce possible individual variation in the time from virus exposure to viremia so that the time from inoculation of the heifers to detection of virus in the placenta and fetus in the different animals could be compared. The detection of a fetal infection with BVDV at 1014 days pi in the present study was supported by the results of conventional virologic examination of fetal organs, which did not reveal virus before 14 days pi. The positive RT-PCR results for the fetus euthanatized at 7 days pi were obtained with the lung sample, but the liver sample was negative. RT-PCR is a sensitive assay for viral RNA6 and accordingly is susceptible to contamination, resulting in false-positive results. Thus, the signicance of the positive RT-PCR result on day 7 pi for the lung sample remains to be determined. Contamination would appear to be the most reasonable explanation for the positive results of the virus examinations of some of the fetal organs of the negative control (heifer No. 8). In the RT-PCR, the fetal liver was negative and the fetal spleen and lung were only weakly positive as compared with the other positive samples. Blood samples from the heifer (heifer No. 8) and her fetus were both negative when tested with RTPCR. In addition, a cell culture line was established from turbinate cells of the fetus and has proven to be negative for BVDV (data not shown). The validity of the immunohistochemical approaches was supported by the general consistency in results achieved with different immunohistochemical techniques. In addition, both mixtures of BVDV antibodies did not identify viral antigens in uterus, placentomes, or intercotyledonary fetal membranes at 14 days pi or in the liver of heifer No. 3. Further, the immunohistochemical and virologic methods showed a similar sensitivity in fetal and maternal tissues. However, although viral antigen was not detected immunohistochemically in tissue from the heifers during an acute experimental infection in the present study, other investigators have demonstrated BVDV antigen in the

22 Days pi

Main Localization of Antigen

( )

18 Days pi

( )

CD45

14 Days pi

22 Days pi

Vimentin

18 Days pi

14 Days pi

Fetal membranes

Organ

Placentomes

Fetus Intestine

Liver Lung Spleen

( )

( )

( )

/ ( )

/ ( )

272

Fredriksen, Press, Sandvik, degaard, and Lken

Vet Pathol 36:4, 1999

Fig. 1. Fetal lung; heifer No. 3, 14 days pi. L lumen. Double immunouorescence for BVDV antigen (uorescein isothiocyanate) and CD45 (Texas red). Bar 22 m. Fig. 1A. Cells showing staining for BVDV antigen (green) surround an unstained bronchiole (L). Fig. 1B. Same section showing cells staining for BVDV (yellow) and for CD45 (red). One CD45 cell identied in Fig. 1B is also weakly stained for BVDV (Fig. 1A, arrow), whereas the other CD45 cells show no reactivity for virus antigen. Fig. 2. Fetal lung; heifer No. 7, 22 days pi. A small bronchiole contains many epithelial cells that are stained for BVDV (brown; arrows), but the epithelium of a larger bronchiole is not stained. Many of the cells surrounding the bronchioles show reactivity for BVDV antigen. L lumen. ABC-PO method, Mayers hematoxylin counterstain. Bar 22 m.

Vet Pathol 36:4, 1999

BVDV in Placenta and Fetus of Acutely Infected Cattle

273

tissues of animals experimentally infected with noncytopathogenic BVDV.27,42,44 In these other studies, BVDV antigens were detected predominantly in lymphoid tissues while uterus, placenta, and ovaries were not examined. Thus, a direct comparison with the present results is not possible. The observed lack of staining in the maternal tissues is in accordance with the negative results of virus isolation from uterine tissues and indicates that the passage of virus to the fetus occurs without high concentrations of virus in uterine tissue. Alternatively, high virus concentration in uterus may have been of very short duration and therefore not detected in the present study. Epithelial cells and certain leukocyte populations have been shown to be predilection sites for BVDV replication.3,5,29 In the present study, the expression of the leukocyte common antigen CD45 was used to identify cells of a hemopoietic origin.2,26 However, virus-infected CD45 cells were only consistently demonstrated in the fetal liver. Another phenotypic marker of cell populations investigated in the present study was the expression of the intermediate lament vimentin. The presence of vimentin was used to provide an indication of the mesenchymal origin of virus-infected cells. The expression of intermediate laments in embryonal and fetal cells changes during development and is less stable than in mature cells.17 The functions of intermediate lament proteins have been considered related to the specic functions of differentiated cells rather than to cell growth and division, and these proteins were not present in rapidly dividing cells.11 The present study did not address the cell cycle status of virus-infected cells. In all the organs studied, some nonepithelial BVDV-positive cells were negative for both vimentin and CD45. The absence of these markers suggests that the viral antigen-containing cells were probably undifferentiated cells. Previous studies in calves and heifers have found differentiated leukocyte populations such as B-cells, T-cells, and null cells to be major virus-infected cell populations.4,27,44 At the earliest stages of fetal infection, 14 days pi, viral antigen was readily detected in fetal organs but was not detected in the placentomes. At 22 days pi, viral antigen was weakly present in the cotyledons.

The sparsity of viral antigen suggests that the fetal placenta is not a predilection site for virus replication early in fetal infection. Some studies have reported high titers of BVDV in placental tissue,32,40,41 but none of these studies were performed on acutely infected cattle. Based on the presence of high virus titers in ovine placentomes, the placentome appeared to be the organ of choice for virus multiplication.41 However, the pregnant ewes were inoculated with BVDV isolated from outbreaks of mucosal disease, but it is unclear whether the cytopathogenic virus used was free from contamination with noncytopathogenic BVDV.41 In addition, the virus strain used was known to be more pathogenic than several other strains. Similarly, a high virus titer was reported in placental tissue after infection with a virus isolated from a fatal case of mucosal disease.40 In a previous study, the maternal crypt epithelium in placentomes from persistently infected animals was conrmed to be heavily infected, whereas only scattered cells in the cotyledons showed staining for BVDV antigen.16 In the placentomes of persistently infected animals, however, the main cell population showing reactivity for BVDV was a binuclear trophoblast cell, whereas in the acutely infected animal (No. 7) in the present study, mainly small cells of mesenchymal origin were infected. In the acutely infected animals, the distinct staining of cells in chorioallantois/amniochorion 18 days pi is notable in relation to the absence of staining in the placentomes at this stage and the difference in the distribution of BVDV-positive cells compared with that shown in persistently infected heifers.16 In the fetal membranes of persistently infected animals, only scattered cells of mesenchymal origin showed reactivity for BVDV, whereas a high proportion of the epithelial cells in the chorion was affected. In the present study with acutely infected animals, the main cell population showing reactivity for BVDV was of mesenchymal origin, and epithelial cells only occasionally showed reactivity for BVDV. The observed differences between the two studies may be due to the BVDV infection status of the dams, although the possibility that the differences are a consequence of the time interval from the initial fetal infection seems more likely. This pos-

Fig. 3. Fetal spleen; heifer No. 6, 18 days pi. A large, immature multinucleated megakaryocyte (large arrow), a small lymphocyte-like cell (small arrow), and numerous platelets show staining for BVDV antigen. Immunouorescence for BVDV (uorescein isothiocyanate). Bar 22 m. Fig. 4. Intercotyledonary fetal membranes; heifer No. 5, 18 days pi. Many subepithelial cells are stained (brown) for BVDV. e epithelium of amnion/allantois. ABC-PO method, Mayers hematoxylin counterstain. Bar 22 m. Fig. 5. Placentome; heifer No. 7, 22 days pi. Small fetal cells show reactivity for BVDV antigen (brown, arrows). Binuclear trophoblast cells (t) and maternal epithelium (me) are negative. ABC-PO method, Mayers hematoxylin counterstain. Bar 22 m.

274

Fredriksen, Press, Sandvik, degaard, and Lken

Vet Pathol 36:4, 1999

sibility is supported by the results of the immunohistochemical staining of fetal lungs, in which the distribution of BVDV antigen in samples collected 22 days pi resembled that in similar sections from the persistently infected animals but was different in the earlier stages of acute infection. Further studies of placentomes and intercotyledonary fetal membranes from acutely infected dams are needed to determine the basis for these observed differences. A possible route of passage for the virus from dam to fetus could be via uterine secretions to the chorion. However, most BVDV-positive cells seen at the earliest stages of infection were subepithelial cells of mesenchymal origin and not epithelial cells, which does not indicate the transfer of infection via the uterine secretions. BVDV has been thought to have a predilection for vascular endothelial cells, and virus replication in these cells could result in vascular damage facilitating the spread of the virus across the placenta.43 The initial detection of BVDV-positive cells in fetal liver and lung 14 days pi without simultaneous detection of virus in the placenta supports the proposition that the passage of virus has taken place via the vasculature and not via local cell-to-cell spread. The pathogenic effect of BVDV infection on the placenta is a point of contention. Most authors have concluded that only minor or no lesions are found in the placenta after infection of the bovine fetus,10,20,22 whereas others have reported lesions in the ovine placenta.32,41 Lesions in placenta have not been reported after infection with noncytopathogenic BVDV. Although histologic examination of the placenta was not included in the present study, the presence of a heavy infection of fetal organs without simultaneous detection of BVDV antigen in the placenta indicates that a fetal infection with noncytopathogenic BVDV can occur in the absence of signicant levels of virus in the placenta. The fetal cells that were shown to be infected with BVDV at the earliest stage of infection were probably immature megakaryocytes in the liver and cells of mesenchymal origin in the lungs. Many platelets in the spleen were also infected with BVDV at an early stage of infection. BVDV has previously been demonstrated in megakaryocytes of acutely infected calves,27,42 which could be of importance for the pathogenesis of thrombocytopenia in calves acutely infected with virulent noncytopathogenic strains of BVDV.35 These strains are characterized as atypical or type-2 BVDV,33,34 which are antigenically different from the classic or type-1 BVDV strains. The present results indicate that the megakaryocytes may have an important role also in the initial stages of intrauterine infection in acutely infected dams and the production of persistently infected calves. However, further investigations are needed to clarify the role of megakaryocytes in the pathogenesis of BVDV infection of the fetus.

Acknowledgements
We thank Dr. Vibeke Dantzer, Department of Anatomy and Physiology, The Royal Veterinary and Agricultural University, Denmark, for assistance and advice in the evaluation of the placenta sections and Dr. Hans Gamlem, Department of Pathology, National Veterinary Institute, Oslo, for assistance and advice in the evaluation of cell morphology. We also thank Laila Aune, Department of Morphology, Genetics and Aquatic Biology, Norwegian College of Veterinary Medicine, and Wanda Gajowniczek, Department of Virology and Serodiagnostics, National Veterinary Institute, Oslo, for their skillful technical assistance.

References
1 Alcigir G: Pathology of persistent bovine diarrhoea pestivirus infection in sheep [Beitrag zur pathologie der persistierenden bovinen virusdiarrhoe (BVD)-infection der schafe]. PhD Thesis, Tierarztliche Hochschule, Hanno ver, Germany, 1983 2 Barclay AN, Birkeland ML, Brown MH, Beyers AD, Davis SJ, Somoza C, Williams AF: The Leucocyte Antigen Factsbook. Academic Press, London, England, 1997 3 Bielefeldt Ohmann H: Pathogenesis of bovine viral diarrhoeamucosal disease: distribution and signicance of BVDV antigen in diseased calves. Res Vet Sci 34:510, 1983 4 Bielefeldt Ohmann H: Double-immunolabeling systems for phenotyping of immune cells harboring bovine viral diarrhea virus. J Histochem Cytochem 35:627633, 1987 5 Bielefeldt Ohmann H: BVD virus antigens in tissues of persistently viraemic, clinically normal cattle: Implications for the pathogenesis of clinically fatal disease. Acta Vet Scand 29:7784, 1988 6 Brock KV: Diagnosis of bovine viral diarrhea virus infections. Vet Clin North Am Food Anim Pract 11:549 561, 1995 7 Brownlie J: The pathways for bovine virus diarrhoea virus biotypes in the pathogenesis of disease. Arch Virol Suppl 3:7996, 1991 8 Carlsson U, Fredriksson G, Alenius S, Kindahl H: Bovine virus diarrhoea virus, a cause of early pregnancy failure in the cow. J Vet Med Ser A 36:1523, 1989 9 Casaro APE, Kendrick JW, Kennedy PC: Response of the bovine fetus to bovine viral diarrheamucosal disease virus. Am J Vet Res 32:15431562, 1971 10 Done JT, Terlecki S, Richardson C, Harkness JW, Sands JJ, Patterson DSP, Sweasey D, Shaw IG, Winkler CE, Duffell SJ: Bovine virus diarrhoeamucosal disease virus: pathogenicity for the fetal calf following maternal infection. Vet Rec 106:473479, 1980 11 Dornell J, Lodish H, Baltimore D: Molecular Cell Biology. Scientic American, New York, NY, 1986 12 Dutia BM, Entrican G, Nettleton PF: Cytopatic and noncytopathic biotypes of border disease virus induce polypeptides of different molecular weight with common antigenic determinants. J Gen Virol 71:12271232, 1990 13 Edwards S, Sands JJ, Harkness JW: The application of monoclonal antibody panels to characterize pestivirus

Vet Pathol 36:4, 1999

BVDV in Placenta and Fetus of Acutely Infected Cattle

275

14

15

16

17

18

19 20

21

22

23

24

25 26

27

28

29

isolates from ruminants in Great Britain. Arch Virol 102: 197206, 1988 Entrican G, Dand A, Nettleton PF: A double monoclonal antibody ELISA for detecting pestivirus antigen in the blood of viraemic cattle and sheep. Vet Microbiol 43: 6574, 1995 Fernandez A, Hewicker M, Trautwein G, Pohlenz J, Liess B: Viral antigen distribution in the central nervous system of cattle persistently infected with bovine viral diarrhea virus. Vet Pathol 26:2632, 1989 Fredriksen B, Press CM, Lken T, degaard SA: Distribution of viral antigen in uterus, placenta and foetus of cattle persistently infected with bovine virus diarrhoea virus. Vet Microbiol 64:109122, 1999 Goel A, Gupta I, Joshi K: Immunohistochemical analysis of human embryos and fetuses: an insight into the mechanism of subversion of antigenic differentiation in neoplasia. Arch Pathol Lab Med 121:719723, 1997 Hewicker M, Wohrmann T, Fernandez A, Trautwein G, Liess B, Moennig V: Immunohistological detection of bovine viral diarrhoea virus antigen in the central nervous system of persistently infected cattle using monoclonal antibodies. Vet Microbiol 23:203210, 1990 Houe H: Epidemiology of bovine viral diarrhea virus. Vet Clin North Am Food Anim Pract 11:521547, 1995 Hubbert WT, Fernelius AL, Van Der Maaten MJ, Estes PC, Bryner JH: Concurrent dual viral infection of a bovine foetus. Vet Rec 93:4546, 1973 Juntti N, Larsson B, Fossum C: The use of monoclonal antibodies in enzyme linked immunosorbent assay for detection of antibodies to bovine viral diarrhoea virus. J Vet Med Ser B 34:356363, 1987 Kendrick JW: Bovine viral diarrheamucosal disease virus infection in pregnant cows. Am J Vet Res 32:533 544, 1971 Kirkland PD, McGowan MR, Mackintosh SG: Factors inuencing the development of persistent infection of cattle with pestiviruses. Proc Symp Pestiviruses Eur Soc Vet Virol 2:117121, 1992 Larsson B, Niskanen R, Alenius S: Natural infection with bovine virus diarrhoea virus in a dairy herd: A spectrum of symptoms including early reproductive failure and retained placenta. Anim Reprod Sci 36:3748, 1994 Lken T: Pestivirus infections in Norway. Epidemiological studies in goats. J Comp Pathol 103:110, 1990 Maddox JF, Mackay CR, Brandon MR: The sheep analogue of leucocyte common antigen (LCA). Immunology 55:347353, 1985 Marshall DJ, Moxley RA, Kelling CL: Distribution of virus and viral antigen in specic pathogen-free calves following inoculation with noncytopathic bovine viral diarrhea virus. Vet Pathol 33:311318, 1996 McClurkin AW, Littledike ET, Cutlip RC, Frank GH, Coria MF, Bolin SR: Production of cattle immunotolerant to bovine viral diarrhea virus. Can J Comp Med 48:156 161, 1984 Meyling A: Demonstration of VD-virus by the uores-

30

31

32

33

34

35

36

37

38

39

40

41

42

43 44

cent antibody technique in tissues of cattle affected with bovine viral diarrhea (mucosal disease). Acta Vet Scand 11:5972, 1970 Moennig V, Liess B: Pathogenesis of intrauterine infections with bovine viral diarrhea virus. Vet Clin North Am Food Anim Pract 11:477487, 1995 Murray RD: Lesions in aborted bovine fetuses and placenta associated with bovine viral diarrhoea virus infection. Arch Virol Suppl 3:217224, 1991 Parsonson IM, OHalloran ML, Zee YC, Snowdon WA: The effects of bovine viral diarrhoeamucosal disease (BVD) virus on the ovine fetus. Vet Microbiol 4:279 292, 1979 Paton DJ, Sands JJ, Lowings JP, Smith JE, Ibata G, Edwards S: A proposed division of the pestivirus genus using monoclonal antibodies, supported by cross-neutralisation assays and genetic sequencing. Vet Res 26: 92109, 1995 Pellerin C, Van Den Hurk J, Lecomte J, Tijssen P: Identication of a new group of bovine viral diarrhea virus strains associated with severe outbreaks and high mortalities. Virology 203:260268, 1994 Rebhun WC, French TW, Perdrizet JA, Dubovi EJ, Dill SG, Karcher LF: Thrombocytopenia associated with acute bovine viral diarrhea infection in cattle. J Vet Intern Med 3:4246, 1989 Roeder PL, Jeffrey M, Cranwell MP: Pestivirus fetopathogenicity in cattle: changing sequelae with fetal maturation. Vet Rec 118:4448, 1986 Sandvik T, Fredriksen B, Lken T: Level of viral antigen in blood leucocytes from cattle acutely infected with bovine viral diarrhoea virus. J Vet Med Ser B 44:583590, 1997 Sandvik T, Krogsrud J: Evaluation of an antigen capture ELISA for detection of bovine viral diarrhea virus in cattle blood samples. J Vet Diagn Invest 7:6571, 1995 Sandvik T, Paton D, Lowings P: Detection and identication of ruminant and porcine pestiviruses by nested amplication 5 untranslated cDNA regions. J Virol Methods 64:4356, 1997 Scott FW, Kahrs RF, Parsonson IM: A cytopathogenic strain of bovine viral diarrheamucosal disease virus isolated from a bovine fetus. Cornell Vet 62:7484, 1972 Snowdon WA, Parsonson IM, Broun ML: The reaction of pregnant ewes to inoculation with mucosal disease virus of bovine origin. J Comp Pathol 85:241251, 1975 Spagnuolo M, Kennedy S, Foster JC, Moffett DA, Adair BM: Bovine viral diarrhoea virus infection in bone marrow of experimentally infected calves. J Comp Pathol 116:97100, 1997 Van Oirschot JT: Congenital infections with nonarbo togaviruses. Vet Microbiol 8:321361, 1983 Wilhelmsen CL, Bolin SR, Ridpath JF, Cheville NF, Kluge JP: Experimental primary postnatal bovine viral diarrhea viral infections in six-month-old calves. Vet Pathol 27:235243, 1990

Request reprints from Dr. B. Fredriksen, National Veterinary Institute, P.O. Box 8156, Dep, 0033 Oslo (Norway). E-mail: bente.fredriksen@vetinst.no.