- Email: [email protected]
Diagnosis of fetal infection with porcine parvovirus by in situ hybridization
veterinary microbiology ELSEVIER
Veterinary Microbiology 47 (1995) 377-385
Diagnosis of fetal infection with porcine parvovirus by in situ hybridization Andreas S. Waldvogel a,*, Susanne Broll a, Max Rosskopf b, Martin Schwyzer b, Andreas Pospischil a aInstitutfir Veteritirpathologie der Universitiit Ziirich, Winterthurerstrasse 268, 8057 Ziirich, Switzerland b Virologisches Institut der Veten’&--medizinischen Fakultiit der Universitiit Ziin’ch, Winterthurerstrasse 266, 8057 Ziirich, Switzerland Received 22 November 1994; accepted 26 April 1995
Abstract In situ hybridization (ISH) for the diagnosis of fetal infection with porcine parvovirus (PPV) was compared with immune electron microscopy (IEM) and serology by immunofluorescence (IF) for its sensitivity and its applicability in a routine diagnostic laboratory. The technique was applied to the examination of sections of formalin-fixed paraffin-embedded tissues from 68 fetuses. Fifty-three of these fetuses were diagnosed serologically since they had a crown rump length of more than 17 cm, i.e. they were mature enough to mount a humoral immune response; 38 were positive and 15 negative. Eleven out of 15 smaller fetuses examined for the presence of viral antigen by immune electron microscopy (IEM) were positive and 4 were negative. Heart and lung were found to he the most suitable organs for in situ hybridization. In situ hybridization yielded a positive result in 8 of the 11 IEM positive fetuses and in 33 of the 38 serologically positive fetuses. No signal was detected in any of the 4 IEM or the 13 serologically negative fetuses. Expenses for IEM were estimated to be 179% of the expenses for ISH. Expenses for serology by IF on the other hand were 67% of the expenses for ISH. From this it was concluded that the most efficient way to diagnose a fetal infection with PPV was serology by IF, if possible with samples from several fetuses and that the other techniques, IEM or ISH, ought to be reserved for those cases where no immunocompetent fetuses were available for diagnosis. Keywords:
Porcine parvovirus; pig; Fetal infection
1. Introduction Porcine or aborted
parvovirus fetuses
(PPV)
is the infectious
in Switzerland
(Broll
agent diagnosed
most frequently
in stillborn
et al., 1993). The diagnosis of fetal infection with
* Corresponding author, Tel: +41 1 365 1204, Fax: +41 1 313 0130. 0378-l 135/95/$09.50 0 1995 Elsevier Science B.V. All rights reserved SSD10378-1135(95)00120-4
378
AS. Waldvogel et al. /Veterinary
Microbiology
47 (1995) 377-385
PPV depends on the demonstration of the virus in fetuses of less than 70 days of age. This can be accomplished either by immunofluorescence on frozen sections or hemagglutination or immune electron microscopy (IEM). Immunofluorescence on frozen sections of lung is more sensitive than hemagglutination (Leslie-Steen and Kirkbride, 1983; Zanoni et al., 1984). IEM is markedly more sensitive than immunofluorescence. When samples of 56 fetuses were examined by both methods, a PPV infection was diagnosed in 34 fetuses by IEM, but only in 18 by immunofluorescence on frozen sections (Zanoni et al., 1984). However IEM is more labor intensive and depends on the availability of an electron microscope. This may explain why immunofluorescence on frozen sections is still used. Both techniques, IEM or immunofluorescence would yield false negative results if the relevant viral epitopes were masked by fetal antibody. Diagnosis of the infection in older fetuses is based on the demonstration of specific antibodies since porcine fetuses acquire the ability to mount a humoral immune response against PPV at approximately 70 days of gestation (Hogg et al., 1977; Molitor et al., 1986; Gradil et al., 1990; Nielsen et al., 1991; Lager et al., 1992). This is conveniently done by an indirect immunofluorescence assay (IFA), where fetal serum is incubated with PPV infected cultured cells fixed on microscopic slides and specifically bound fetal antibodies are detected with fluorescein labelled antibodies directed against porcine immunoglobulins. In situ hybridization (ISH) is an increasingly popular technique to localize and identify nucleic acids in paraffin-embedded, formalin-fixed tissues. In contrast to IEM it allows to recognize specific staining by its consistent association with particular cellular structures and to distinguish it from nonspecific staining scattered randomly throughout a tissue section. Because it does not depend on antigen, the target cannot be masked by the host’s antibodies. Furthermore it can be performed in any laboratory equipped for histology and immunohistochemistry. Since parvovirus DNA could still be detected by ISH in severely autolytic canine and feline intestines (Waldvogel et al., 1992)) this technique might be a valuable alternative to the currently used methods to demonstrate the virus. For this reason we decided to investigate whether this technique could also be used to demonstrate PPV in porcine fetuses, which are often severely autolytic or even mummified, and whether the ISH could be used as an alternative method to the IEM and serology by IFA to diagnose a fetal PPV infection in stillborn and aborted porcine fetuses.
2. Materials and methods The tissues used in this study (see results) originated from individuals which had also been examined for parvoviral infection according to the protocol published by Zanoni et al. ( 1984). Supematant of organ suspension of fetuses of less than 17 cm crown rump length (CBL) , i.e. of less than 74 days of age (Marrable and Ashdown, 1967), was examined for the presence of PPV by IEM. Infection of older fetuses was diagnosed by binding specific antibodies from body cavity or organ fluid to PPV infected cells fixed on microscopic slides and demonstrating this binding with fluorescein labelled rabbit antibodies (Nordic immunological laboratories, 5038 SC Tilburg, the Netherlands) directed against the Fc and Fab fragments of porcine IgG (Leslie-Steen and Kirkbride, 1983; Zanoni et al., 1984). A titer of 1: 10 or more was considered to be specific.
AS. Waldvogel et al. /Veterinary Microbiology 47 (1995) 377-385
319
The probe was a kind gift of Dr. Krell. It consisted of a 3 kb fragment of genomic DNA of the NADL-2 strain of PPV cloned into the vector pUC-19 (Krell et al., 1988)) amplified in E. coli (JM109) according to standard protocols (Sambrook et al., 1989)) and labelled with Biotin-16-dUTP overnight using a commercially available random primed labelling kit (Boehringer Mannheim GmbH, Mannheim, Germany). The whole plasmid was digested with RsaI before labelling to control the length of the probe and keep it small enough for easy penetration to the target DNA. The volume of probe needed for one slide was determined by spotting serial dilutions of a biotinylated control DNA and of the freshly labelled PPV specific probe on nylon membranes according to the recommendations of the supplier (Boehringer). The DNA was detected by incubating the filters for 2 h with streptavidin peroxidase (Boehringer Mannheim GmbH, Mannheim, Germany) diluted 1:3000 in 100 mM Tris-HCl, 150 mM NaCl pH 7.5, washing twice for 15 min in 100 mM Tris-HCl, 150 mM NaCl pH 7.5 and using 3amino-9-ethylcarbazol (AEC) as a chromogen (Biomeda Corp., Foster City, California, USA). ISH was done on formalin fixed paraffin-embedded tissue. Two to 3 pm thick sections were mounted on organosilan-treated glass slides (Rentrop et al., 1986), dried at 60°C overnight and deparaffinized. Pretreatment included blocking of endogenous peroxidase, digestion with proteinase K, and post-fixation in 4% pamformaldehyde ( WaIdvogel et al., 199 1) . The slides were dehydrated in graded ethanol, air-dried and the probe was applied to the section in an aqueous solution of 50% (v/v) formamide, 2 X SSC, 10% (w/v) dextran sulfate, 0.5% SDS, 1 X (v/v) Denhardt’s solution, 1 mM EDTA, and 250 nglpl salmon sperm DNA. Target and probe DNA were denatured simultaneously on the section at 85°C for 10 min, and hybridization was performed overnight at 42°C. Post-hybridization washing included a brief rinse in 1 X PBS at 42”C, three consecutive washes on a shaker platform in 1 X PBS, 0.1 X PBS, and 0.01 X PBS at 42°C for 10 min each, a further wash in 0.1 %Triton- X -lOO/ 1 X PBS at 42°C for 5 min and a brief rinse in 1 X PBS at room temperature. The hybrid was detected as described above for the determination of the amount of probe needed, the sections were briefly counterstained in hematoxylin, and coverslips were mounted with Crystal Mount (Biomeda Corp., Foster City, California, USA). The result of the ISH was evaluated by light microscopy at a magnification of at least 200 X , and if the result was equivocal the assay was repeated. Positive controls included pellets of ST cells infected with the NADL-2 strain of PPV. These pellets were fixed in formahn, slightly stained with hematoxylin and inserted into a block of boiled chicken egg white (Jaspers, 1982). The egg white with the pellets was embedded into paraffin and 2 to 4 pm sections were cut and mounted on organosilan-treated slides. Furthermore, 2 sections of fetal tissue with only few infected cells were included in each assay and served as weak positive control. Negative controls included sections of uninfected ST cells as well as tissue sections of 93 day old porcine fetuses negative for PPV by serology. These sections were hybridized with the biotinylated probe. Furthermore, at least one section of each paraffin block was hybridized with pUC19 digested with Ah1 prior to labelling and labelled in the same way as the pPPV probe. A signal was considered to be specific if it was associated with cellular
AS. Waldvogel et al. /Veterinary
380
Microbiology 47 (1995) 377-385
structures and no staining of cellular structures was observed on the sections hybridized with labelled pUC19. Expenses for labour and reagents were estimated for processing batches of samples from 10 fetuses. Expenses of equipment were calculated under the assumption that the assays were done in a diagnostic laboratory, where standard equipment for histology, cytology and immunohistology was available. Expenses for the electron microscope were calculated under the assumption that we could join a group of users and would contribute to its maintenance, depreciation, and water and electricity bills on an hourly basis.
3. Results Tissue of 68 fetuses was available for ISH. Eleven out of 15 mummified fetuses were positive for PPV by IEM. Thirty-eight of the remaining 53 fetuses with a CRL of more than 17 cm had seroconverted (Table 1) . These fetuses were at various stages of autolysis. PPV DNA could be detected by ISH in 8 out of 11 PPV positive mummies and 33 of 38 fetuses with antibodies specific for PPV. No PPV specific DNA was detected in any of the fetuses negative for a PPV infection by IEM or IF (Table 1) . PPV DNA was detected in heart and lung in 19 fetuses, in the lung only in 14 fetuses, and in the heart only in 3 fetuses (Table 2). The positive reaction consisted of a nuclear staining in myocardial cells (Fig. 1) . Background staining in the form of granular precipitates was occasionally a serious problem, particularly in mummified fetuses. This was aggravated when cellular structures could no longer be clearly recognized. The positive cells in the lung were always located within the alveoli. The staining reaction was confined to the nucleus in some cells. On the other hand there were always numerous cells with cytoplasmic staining, occasionally even without any nuclear staining (Fig. 2). These two staining patterns were observed in lungs of both, fresh and mummified fetuses. Table 1 Correlation
between ISH results and previous PPV diagnosis by IEM and IF, respectively: Numbers fetuses previously
tested by serology by IF
IEM
Total Thereof ISH positive ISH negative Sensitivity’ Specificity’ ‘Percentage ‘Percentage
Positive
Negative
Positive
Negative
11
4
38
15
8
0
33
0
3
4
5
15
87%
13% 100%
of ISH positive relative to IEM positive or IF positive, respectively. of ISH negative relative to IEM negative or IF negative, respectively
100%
AS. Waldvogel et al. /Veterinary
Microbiology 47 (1995) 377-385
381
Table 2 ISH results in lung and heart of fetuses tested positive for PPV infection by IEM or IF, respectively: ISH result by organ
Numbers of fetuses positive by ISH and previously
Lung
Heart
IEM
IF
+ + + _ n.a. _
+ n.a. + + na.
2 4 1 0 1 2 0 1 11
17 10 2 3 1 4 0
*.a. total
positive by
1 38
No signal was detected in sections, where pUC19 had been used as negative control probe, even if a strong signal was generated with the specific probe (Fig. 3 and Fig. 4). False positive signals were observed in bronchial epithelial cells and occasionally in bacteria in fetuses infected or contaminated by these organisms. The fact that the signal could be assigned to anatomical structures allowed in most cases background or nonspecific staining to be distinguished from true signal. Tissue sections of various organs in addition to lung and heart were assayed for the presence of PPV by ISH in 9 reasonably fresh, close to term and 8 mummified fetuses. The most consistent results, however, were obtained in heart and lung. Liver and kidney were
Fig. 1. Myocardium of a mummifted porcine fetus. The positive reaction consisted of a nuclear staining in myocardial cells. In situ hybridization on sections of formalin-fixed paraffin-embedded tissues. Although these sections were counterstained. hematoxylin did not stain the severely autolytic tissues. Bars: 10 pm.
382
AS. Waldvogel et al. /Veterinary
Microbiology 47 (1995) 377-385
Fig . 2. Lung of an autolytic fetus: Positive cells within alveoli. Nuclear (arrow) stai ning. See Fig. 1 for legend.
and cytoplasmic
(open am DW)
Fig :. 3. ISH with the probe specific for PPV. While numerous cells reacted positively when the specific probe was aPI Aied, no staining was observed with the non-specific probe. See Fig. 1 for legend.
A.S. Waldvogel et al. / Veterinary Microbiology
383
47 (1995) 377-385
Fig. 4. ISH with the probe specific for pUC-19. While numerous cells reacted positively when the probe spec:ific for PPV was applied, no staining was observed with the non-specific probe. See Fig. 1 for legend.
Table 3 Comparison
of expenses (in SFr) for IEM, ISH, and IF for the diagnosis of PPV infection in fetal pigs
Labour for 10 samples Price of reagents for 10 samples Total expenses “Additional costs for use of electron microscope “Salary of technician only.
IEM
ISH
IF
7 h, 50.-/h” 25.375.-
5 h, 25.-/hb 84.209.-
3 h, 25.-/hb 65.140.-
in salary of technician
included.
not suitable bedause of nonspecific staining of the cytoplasm of hepatocytes and tubular epithelial cells. Other organs such as spleen, brain, lymph nodes, or ileum with Peyer’s patches were negative. Furthermore placentae were available from 30 litters. However, this tissue was not suitable for routine examination because advanced autolysis and dehydration artifacts made it impossible to assign a signal to cellular structures. Furthermore background due to autolysis, bacterial contamination, and dirt was severe and this tissue had a tendency to slough off the slide during hybridization. Expenses for labour and reagents were highest for the IEM (Table 3). Three squares of a grid were scanned for the presence of clumped viruses in the electron microscope. This took 20 to 30 min for one fetus when no virus was present. On the other hand reading the ISH slides of 10 fetuses on the light microscope could easily be achieved in 1.5 h.
384
AS. Waldvogel et al. / Veterinary Microbiology 47 (1995) 377-385
4. Discussion
We were able to show in this study that PPV DNA can be detected even in mummified porcine fetuses by ISH. The ISH was less expensive than the IEM. The fact that the ISH could be done in a laboratory with standard equipment for histology and immunohistochemistry and that the signal could be appreciated in its anatomical context were further advantages of this technique. It was, however, less sensitive than the IEM. Substitution of the peroxidase with alkaline phosphatase and detection with nitroblue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate was reported to render the detection more sensitive (Nuovo and Richart, 1989). And indeed, alkaline phosphatase was up to 100 times more sensitive on filters. However this effect was not appreciated on tissue sections (data not shown). Background was, on the other hand, a serious problem, especially on tissues from autolytic fetuses. It consisted of fine black granules which formed clusters of variable size, sometimes as large as nuclei. When such background formed on cellular structures, it could not be distinguished from the black precipitate of true signal and it was not possible to recognize any difference between positive and negative controls in many cases for this reason. The red precipitate of AEC could easily be distinguished from such background. Therefore peroxidase labelled avidin and AEC were used in the subsequent experiments for the detection of the hybrids. Although PPV antigen could be demonstrated in various organs in experimentally infected fetuses (Cartwright et al., 1971; Mengeling and Cutlip, 1976; Joo et al., 1977) lung and heart turned out to be most suitable for diagnosing a PPV infection by in situ hybridization. Liver and kidney were not suitable because of considerable problems with nonspecific binding of the biotinylated probe. Brain was only obtained from reasonably fresh fetuses, which were also immunocompetent. It contained only very few positive endothelial cells. This may have to do with the limited sensitivity of our assay, but may as well be the result of destruction of the virus by autolysis or the elimination of the virus by the fetal immune response. The viral DNA could be demonstrated most abundantly and frequently in cells within the alveoli of the fetal lungs. The fact that the signal was either intranuclear or intracytoplasmic is in agreement with the findings of Mengeling and Cutlip ( 1976). Intranuclear staining would be the staining pattern expected in cells with viral replication (Bems, 1991). Since RNase A treatment prior to hybridization did not diminish the cytoplasmic staining (not shown), this signal was not the result of hybridization to viral mRNA. Rather these cells might be phagocytic cells which accumulated viral DNA in their cytoplasm. IP and IEM turned out to be more sensitive than ISH. Since serology on fetal body cavity fluids by IF was the cheapest way to diagnose the fetal infection with PPV, producers and veterinarians ought to be advised to submit fresh fetuses of more than 17 cm CRL whenever possible. The sensitivity of this approach could be further enhanced if several, at least three fetuses, were assayed (Broll et al., 1993). If only fetuses of less than 17 cm CRL were available, IEM, and if no electron microscope was available, ISH would be valuable tools to diagnose a fetal infection with PPV. Acknowledgements The authors thank Dr. P.J. Krell for the DNA probe and Mrs. Roseline Weilenmann and Mrs. Verena Henn for their excellent technical support. This study was supported by the Bundesamt fur Veterin&rwesen der Schweiz (Projekt-Nr. 012.91.5).
A.S. Waldvogel et al. /Veterinary
Microbiology
47 (1995) 377-385
385
References Bems, K.I., 1991. Parvoviridae and their replication, In: B.N. Fields, D.M. Knipe (Eds.), Fundamental Virology, 2nd edition. Raven Press Ltd., N.Y., pp. 817-837. Broll, S., Waldvogel, AS., Rosskopf, M., Corboz, L., Pospischil, A. 1993. Untersuchungen zu infektiosen Ursachen ftir Aborte und Totgeburten beim Schwein in der Schweiz. J. Vet. Med. B, 40: 641-653. Cartwright, S.F., Lucas, M., Huck, R.A. 1971. A small haemagglutinating porcine DNA-virus. II. Biological and serological studies. J. Comp. Pathol., 81: 145-155. Gradil, C.M., Joo, H.S., Molitor. T.W. 1990. Persistence of porcine parvovirus in swine infected in utero and followed through maturity. J. Vet. Med. B, 37: 309-316. Hogg, G.G., Lenghaus, C., Forman, A.J. 1977. Experimental porcine parvovims infection of foetal pigs resulting in abortion, histological lesions and antibody formation. J. Comp. Pathol., 87: 539-549. Jaspers, B. 1982. Htihnereiweisseinbettung. Histochem. J. 1: 32-33. Joo, H.S., Donaldson-Wood, C.R., Johnson, R.H., Campbell, R.S.F. 1977. Pathogenesis of porcine parvovirus infection: Pathology and immunfluorescence in the foetus. J. Comp. Pathol., 87: 383-391. Krell, P.J., Salas, T., Johnson, R.P. 1988. Mapping of porcine parvovirus DNA and development of a diagnostic DNA probe. Vet. Microbial., 17: 2943. Lager, K.M., Mengeling, W.L., Liu, W. 1992. Comparison of the virulence of two isolates of porcine parvovirus in 72-day-old porcine fetuses. J. Vet. Diagn. Invest., 4: 245-248. Leslie-Steen, P., Kirkbride, C. 1983. Comparison of diagnostic methods for in utero parvovirus infection of swine fetuses. Proc. 3rd Intl. Symposium of World Ass. of Vet. Lab. Diagnosticians, Ames, Iowa, USA: 11 l-l 17. Marrable, A.W., Ashdown, R.R. 1967. Quantitative observations on pig embryos of known ages. J. Agric. Sci. Camb., 69: 443-447. Mengeling, W.L., Cutlip, R.C. 1976. Reproductive disease experimentally induced by exposing pregnant gilts to porcine parvovirus. Am. J. Vet. Res., 37: 1393-1400. Molitor, T.W., Choi, C.S., Harding, M.J., Joo, H.S. 1986. In utero inoculation of swine fetuses to study viral pathogenesis and immunity, In: Tumbleson M.E. (Ed.), Swine in biomedical research, Vol. III, Plenum Press, N.Y., pp. 1945-1954. Nielsen, J., Ronsholt, L., Sorensen, K.J. 1991. Experimental in utero infection of pig foetuses with porcine parvovinrs (PPV). Vet. Microbial., 28: l-l 1. Nuovo, G.J., Richart, R.M. 1989. A comparison of biotin-and %i-based in situ hybridization methodologies for detection of human papillomavirus DNA. Lab. Invest., 61: 471-476. Rentrop, M., Knapp, B., Winter, H., Schweizer, J. 1986. Aminoalkylsilane-treated glass slides as support for in situ hybridization of keratin cDNA to frozen tissue sections under varying fixation and pretreatment conditions. Histochem. J., 18: 271-276. Sambrook, J., Fritsch, E.F., Maniatis, T. 1989. Molecular Cloning. 2nd edition, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Waldvogel, A.S., Hassam, S., Weilenmann, R., Tratschin, I.D., Siegl, G., H&nichen, T., Briner, J., Pospischil, A. 1991. Retrospective study of myocardial canine parvovirus infection by in situ hybridization. J. Vet. Med. B, 38: 353-357. Waldvogel, A.S., Hassam, S., Stoerckle, N., Weilenmann, R., Tratschin, J.D., Siegl, G., Pospischil, A. 1992. Specific diagnosis of parvovirus enteritis in dogs and cats by in situ hybridization. J. Comp. Pathol., 107: 141146. Zanoni, R.G., Henn, V., Rutishauser, U.P., Wyler, R. 1984. Haufigkeit der porcinen Parvovirusinfektion in der Schweiz und ein neuer Vimsnachweis mittels Immunelektronenmikroskopie. Zbl. Vet. Med. B, 31: 729-742.
Veterinary Microbiology 47 (1995) 377-385
Diagnosis of fetal infection with porcine parvovirus by in situ hybridization Andreas S. Waldvogel a,*, Susanne Broll a, Max Rosskopf b, Martin Schwyzer b, Andreas Pospischil a aInstitutfir Veteritirpathologie der Universitiit Ziirich, Winterthurerstrasse 268, 8057 Ziirich, Switzerland b Virologisches Institut der Veten’&--medizinischen Fakultiit der Universitiit Ziin’ch, Winterthurerstrasse 266, 8057 Ziirich, Switzerland Received 22 November 1994; accepted 26 April 1995
Abstract In situ hybridization (ISH) for the diagnosis of fetal infection with porcine parvovirus (PPV) was compared with immune electron microscopy (IEM) and serology by immunofluorescence (IF) for its sensitivity and its applicability in a routine diagnostic laboratory. The technique was applied to the examination of sections of formalin-fixed paraffin-embedded tissues from 68 fetuses. Fifty-three of these fetuses were diagnosed serologically since they had a crown rump length of more than 17 cm, i.e. they were mature enough to mount a humoral immune response; 38 were positive and 15 negative. Eleven out of 15 smaller fetuses examined for the presence of viral antigen by immune electron microscopy (IEM) were positive and 4 were negative. Heart and lung were found to he the most suitable organs for in situ hybridization. In situ hybridization yielded a positive result in 8 of the 11 IEM positive fetuses and in 33 of the 38 serologically positive fetuses. No signal was detected in any of the 4 IEM or the 13 serologically negative fetuses. Expenses for IEM were estimated to be 179% of the expenses for ISH. Expenses for serology by IF on the other hand were 67% of the expenses for ISH. From this it was concluded that the most efficient way to diagnose a fetal infection with PPV was serology by IF, if possible with samples from several fetuses and that the other techniques, IEM or ISH, ought to be reserved for those cases where no immunocompetent fetuses were available for diagnosis. Keywords:
Porcine parvovirus; pig; Fetal infection
1. Introduction Porcine or aborted
parvovirus fetuses
(PPV)
is the infectious
in Switzerland
(Broll
agent diagnosed
most frequently
in stillborn
et al., 1993). The diagnosis of fetal infection with
* Corresponding author, Tel: +41 1 365 1204, Fax: +41 1 313 0130. 0378-l 135/95/$09.50 0 1995 Elsevier Science B.V. All rights reserved SSD10378-1135(95)00120-4
378
AS. Waldvogel et al. /Veterinary
Microbiology
47 (1995) 377-385
PPV depends on the demonstration of the virus in fetuses of less than 70 days of age. This can be accomplished either by immunofluorescence on frozen sections or hemagglutination or immune electron microscopy (IEM). Immunofluorescence on frozen sections of lung is more sensitive than hemagglutination (Leslie-Steen and Kirkbride, 1983; Zanoni et al., 1984). IEM is markedly more sensitive than immunofluorescence. When samples of 56 fetuses were examined by both methods, a PPV infection was diagnosed in 34 fetuses by IEM, but only in 18 by immunofluorescence on frozen sections (Zanoni et al., 1984). However IEM is more labor intensive and depends on the availability of an electron microscope. This may explain why immunofluorescence on frozen sections is still used. Both techniques, IEM or immunofluorescence would yield false negative results if the relevant viral epitopes were masked by fetal antibody. Diagnosis of the infection in older fetuses is based on the demonstration of specific antibodies since porcine fetuses acquire the ability to mount a humoral immune response against PPV at approximately 70 days of gestation (Hogg et al., 1977; Molitor et al., 1986; Gradil et al., 1990; Nielsen et al., 1991; Lager et al., 1992). This is conveniently done by an indirect immunofluorescence assay (IFA), where fetal serum is incubated with PPV infected cultured cells fixed on microscopic slides and specifically bound fetal antibodies are detected with fluorescein labelled antibodies directed against porcine immunoglobulins. In situ hybridization (ISH) is an increasingly popular technique to localize and identify nucleic acids in paraffin-embedded, formalin-fixed tissues. In contrast to IEM it allows to recognize specific staining by its consistent association with particular cellular structures and to distinguish it from nonspecific staining scattered randomly throughout a tissue section. Because it does not depend on antigen, the target cannot be masked by the host’s antibodies. Furthermore it can be performed in any laboratory equipped for histology and immunohistochemistry. Since parvovirus DNA could still be detected by ISH in severely autolytic canine and feline intestines (Waldvogel et al., 1992)) this technique might be a valuable alternative to the currently used methods to demonstrate the virus. For this reason we decided to investigate whether this technique could also be used to demonstrate PPV in porcine fetuses, which are often severely autolytic or even mummified, and whether the ISH could be used as an alternative method to the IEM and serology by IFA to diagnose a fetal PPV infection in stillborn and aborted porcine fetuses.
2. Materials and methods The tissues used in this study (see results) originated from individuals which had also been examined for parvoviral infection according to the protocol published by Zanoni et al. ( 1984). Supematant of organ suspension of fetuses of less than 17 cm crown rump length (CBL) , i.e. of less than 74 days of age (Marrable and Ashdown, 1967), was examined for the presence of PPV by IEM. Infection of older fetuses was diagnosed by binding specific antibodies from body cavity or organ fluid to PPV infected cells fixed on microscopic slides and demonstrating this binding with fluorescein labelled rabbit antibodies (Nordic immunological laboratories, 5038 SC Tilburg, the Netherlands) directed against the Fc and Fab fragments of porcine IgG (Leslie-Steen and Kirkbride, 1983; Zanoni et al., 1984). A titer of 1: 10 or more was considered to be specific.
AS. Waldvogel et al. /Veterinary Microbiology 47 (1995) 377-385
319
The probe was a kind gift of Dr. Krell. It consisted of a 3 kb fragment of genomic DNA of the NADL-2 strain of PPV cloned into the vector pUC-19 (Krell et al., 1988)) amplified in E. coli (JM109) according to standard protocols (Sambrook et al., 1989)) and labelled with Biotin-16-dUTP overnight using a commercially available random primed labelling kit (Boehringer Mannheim GmbH, Mannheim, Germany). The whole plasmid was digested with RsaI before labelling to control the length of the probe and keep it small enough for easy penetration to the target DNA. The volume of probe needed for one slide was determined by spotting serial dilutions of a biotinylated control DNA and of the freshly labelled PPV specific probe on nylon membranes according to the recommendations of the supplier (Boehringer). The DNA was detected by incubating the filters for 2 h with streptavidin peroxidase (Boehringer Mannheim GmbH, Mannheim, Germany) diluted 1:3000 in 100 mM Tris-HCl, 150 mM NaCl pH 7.5, washing twice for 15 min in 100 mM Tris-HCl, 150 mM NaCl pH 7.5 and using 3amino-9-ethylcarbazol (AEC) as a chromogen (Biomeda Corp., Foster City, California, USA). ISH was done on formalin fixed paraffin-embedded tissue. Two to 3 pm thick sections were mounted on organosilan-treated glass slides (Rentrop et al., 1986), dried at 60°C overnight and deparaffinized. Pretreatment included blocking of endogenous peroxidase, digestion with proteinase K, and post-fixation in 4% pamformaldehyde ( WaIdvogel et al., 199 1) . The slides were dehydrated in graded ethanol, air-dried and the probe was applied to the section in an aqueous solution of 50% (v/v) formamide, 2 X SSC, 10% (w/v) dextran sulfate, 0.5% SDS, 1 X (v/v) Denhardt’s solution, 1 mM EDTA, and 250 nglpl salmon sperm DNA. Target and probe DNA were denatured simultaneously on the section at 85°C for 10 min, and hybridization was performed overnight at 42°C. Post-hybridization washing included a brief rinse in 1 X PBS at 42”C, three consecutive washes on a shaker platform in 1 X PBS, 0.1 X PBS, and 0.01 X PBS at 42°C for 10 min each, a further wash in 0.1 %Triton- X -lOO/ 1 X PBS at 42°C for 5 min and a brief rinse in 1 X PBS at room temperature. The hybrid was detected as described above for the determination of the amount of probe needed, the sections were briefly counterstained in hematoxylin, and coverslips were mounted with Crystal Mount (Biomeda Corp., Foster City, California, USA). The result of the ISH was evaluated by light microscopy at a magnification of at least 200 X , and if the result was equivocal the assay was repeated. Positive controls included pellets of ST cells infected with the NADL-2 strain of PPV. These pellets were fixed in formahn, slightly stained with hematoxylin and inserted into a block of boiled chicken egg white (Jaspers, 1982). The egg white with the pellets was embedded into paraffin and 2 to 4 pm sections were cut and mounted on organosilan-treated slides. Furthermore, 2 sections of fetal tissue with only few infected cells were included in each assay and served as weak positive control. Negative controls included sections of uninfected ST cells as well as tissue sections of 93 day old porcine fetuses negative for PPV by serology. These sections were hybridized with the biotinylated probe. Furthermore, at least one section of each paraffin block was hybridized with pUC19 digested with Ah1 prior to labelling and labelled in the same way as the pPPV probe. A signal was considered to be specific if it was associated with cellular
AS. Waldvogel et al. /Veterinary
380
Microbiology 47 (1995) 377-385
structures and no staining of cellular structures was observed on the sections hybridized with labelled pUC19. Expenses for labour and reagents were estimated for processing batches of samples from 10 fetuses. Expenses of equipment were calculated under the assumption that the assays were done in a diagnostic laboratory, where standard equipment for histology, cytology and immunohistology was available. Expenses for the electron microscope were calculated under the assumption that we could join a group of users and would contribute to its maintenance, depreciation, and water and electricity bills on an hourly basis.
3. Results Tissue of 68 fetuses was available for ISH. Eleven out of 15 mummified fetuses were positive for PPV by IEM. Thirty-eight of the remaining 53 fetuses with a CRL of more than 17 cm had seroconverted (Table 1) . These fetuses were at various stages of autolysis. PPV DNA could be detected by ISH in 8 out of 11 PPV positive mummies and 33 of 38 fetuses with antibodies specific for PPV. No PPV specific DNA was detected in any of the fetuses negative for a PPV infection by IEM or IF (Table 1) . PPV DNA was detected in heart and lung in 19 fetuses, in the lung only in 14 fetuses, and in the heart only in 3 fetuses (Table 2). The positive reaction consisted of a nuclear staining in myocardial cells (Fig. 1) . Background staining in the form of granular precipitates was occasionally a serious problem, particularly in mummified fetuses. This was aggravated when cellular structures could no longer be clearly recognized. The positive cells in the lung were always located within the alveoli. The staining reaction was confined to the nucleus in some cells. On the other hand there were always numerous cells with cytoplasmic staining, occasionally even without any nuclear staining (Fig. 2). These two staining patterns were observed in lungs of both, fresh and mummified fetuses. Table 1 Correlation
between ISH results and previous PPV diagnosis by IEM and IF, respectively: Numbers fetuses previously
tested by serology by IF
IEM
Total Thereof ISH positive ISH negative Sensitivity’ Specificity’ ‘Percentage ‘Percentage
Positive
Negative
Positive
Negative
11
4
38
15
8
0
33
0
3
4
5
15
87%
13% 100%
of ISH positive relative to IEM positive or IF positive, respectively. of ISH negative relative to IEM negative or IF negative, respectively
100%
AS. Waldvogel et al. /Veterinary
Microbiology 47 (1995) 377-385
381
Table 2 ISH results in lung and heart of fetuses tested positive for PPV infection by IEM or IF, respectively: ISH result by organ
Numbers of fetuses positive by ISH and previously
Lung
Heart
IEM
IF
+ + + _ n.a. _
+ n.a. + + na.
2 4 1 0 1 2 0 1 11
17 10 2 3 1 4 0
*.a. total
positive by
1 38
No signal was detected in sections, where pUC19 had been used as negative control probe, even if a strong signal was generated with the specific probe (Fig. 3 and Fig. 4). False positive signals were observed in bronchial epithelial cells and occasionally in bacteria in fetuses infected or contaminated by these organisms. The fact that the signal could be assigned to anatomical structures allowed in most cases background or nonspecific staining to be distinguished from true signal. Tissue sections of various organs in addition to lung and heart were assayed for the presence of PPV by ISH in 9 reasonably fresh, close to term and 8 mummified fetuses. The most consistent results, however, were obtained in heart and lung. Liver and kidney were
Fig. 1. Myocardium of a mummifted porcine fetus. The positive reaction consisted of a nuclear staining in myocardial cells. In situ hybridization on sections of formalin-fixed paraffin-embedded tissues. Although these sections were counterstained. hematoxylin did not stain the severely autolytic tissues. Bars: 10 pm.
382
AS. Waldvogel et al. /Veterinary
Microbiology 47 (1995) 377-385
Fig . 2. Lung of an autolytic fetus: Positive cells within alveoli. Nuclear (arrow) stai ning. See Fig. 1 for legend.
and cytoplasmic
(open am DW)
Fig :. 3. ISH with the probe specific for PPV. While numerous cells reacted positively when the specific probe was aPI Aied, no staining was observed with the non-specific probe. See Fig. 1 for legend.
A.S. Waldvogel et al. / Veterinary Microbiology
383
47 (1995) 377-385
Fig. 4. ISH with the probe specific for pUC-19. While numerous cells reacted positively when the probe spec:ific for PPV was applied, no staining was observed with the non-specific probe. See Fig. 1 for legend.
Table 3 Comparison
of expenses (in SFr) for IEM, ISH, and IF for the diagnosis of PPV infection in fetal pigs
Labour for 10 samples Price of reagents for 10 samples Total expenses “Additional costs for use of electron microscope “Salary of technician only.
IEM
ISH
IF
7 h, 50.-/h” 25.375.-
5 h, 25.-/hb 84.209.-
3 h, 25.-/hb 65.140.-
in salary of technician
included.
not suitable bedause of nonspecific staining of the cytoplasm of hepatocytes and tubular epithelial cells. Other organs such as spleen, brain, lymph nodes, or ileum with Peyer’s patches were negative. Furthermore placentae were available from 30 litters. However, this tissue was not suitable for routine examination because advanced autolysis and dehydration artifacts made it impossible to assign a signal to cellular structures. Furthermore background due to autolysis, bacterial contamination, and dirt was severe and this tissue had a tendency to slough off the slide during hybridization. Expenses for labour and reagents were highest for the IEM (Table 3). Three squares of a grid were scanned for the presence of clumped viruses in the electron microscope. This took 20 to 30 min for one fetus when no virus was present. On the other hand reading the ISH slides of 10 fetuses on the light microscope could easily be achieved in 1.5 h.
384
AS. Waldvogel et al. / Veterinary Microbiology 47 (1995) 377-385
4. Discussion
We were able to show in this study that PPV DNA can be detected even in mummified porcine fetuses by ISH. The ISH was less expensive than the IEM. The fact that the ISH could be done in a laboratory with standard equipment for histology and immunohistochemistry and that the signal could be appreciated in its anatomical context were further advantages of this technique. It was, however, less sensitive than the IEM. Substitution of the peroxidase with alkaline phosphatase and detection with nitroblue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate was reported to render the detection more sensitive (Nuovo and Richart, 1989). And indeed, alkaline phosphatase was up to 100 times more sensitive on filters. However this effect was not appreciated on tissue sections (data not shown). Background was, on the other hand, a serious problem, especially on tissues from autolytic fetuses. It consisted of fine black granules which formed clusters of variable size, sometimes as large as nuclei. When such background formed on cellular structures, it could not be distinguished from the black precipitate of true signal and it was not possible to recognize any difference between positive and negative controls in many cases for this reason. The red precipitate of AEC could easily be distinguished from such background. Therefore peroxidase labelled avidin and AEC were used in the subsequent experiments for the detection of the hybrids. Although PPV antigen could be demonstrated in various organs in experimentally infected fetuses (Cartwright et al., 1971; Mengeling and Cutlip, 1976; Joo et al., 1977) lung and heart turned out to be most suitable for diagnosing a PPV infection by in situ hybridization. Liver and kidney were not suitable because of considerable problems with nonspecific binding of the biotinylated probe. Brain was only obtained from reasonably fresh fetuses, which were also immunocompetent. It contained only very few positive endothelial cells. This may have to do with the limited sensitivity of our assay, but may as well be the result of destruction of the virus by autolysis or the elimination of the virus by the fetal immune response. The viral DNA could be demonstrated most abundantly and frequently in cells within the alveoli of the fetal lungs. The fact that the signal was either intranuclear or intracytoplasmic is in agreement with the findings of Mengeling and Cutlip ( 1976). Intranuclear staining would be the staining pattern expected in cells with viral replication (Bems, 1991). Since RNase A treatment prior to hybridization did not diminish the cytoplasmic staining (not shown), this signal was not the result of hybridization to viral mRNA. Rather these cells might be phagocytic cells which accumulated viral DNA in their cytoplasm. IP and IEM turned out to be more sensitive than ISH. Since serology on fetal body cavity fluids by IF was the cheapest way to diagnose the fetal infection with PPV, producers and veterinarians ought to be advised to submit fresh fetuses of more than 17 cm CRL whenever possible. The sensitivity of this approach could be further enhanced if several, at least three fetuses, were assayed (Broll et al., 1993). If only fetuses of less than 17 cm CRL were available, IEM, and if no electron microscope was available, ISH would be valuable tools to diagnose a fetal infection with PPV. Acknowledgements The authors thank Dr. P.J. Krell for the DNA probe and Mrs. Roseline Weilenmann and Mrs. Verena Henn for their excellent technical support. This study was supported by the Bundesamt fur Veterin&rwesen der Schweiz (Projekt-Nr. 012.91.5).
A.S. Waldvogel et al. /Veterinary
Microbiology
47 (1995) 377-385
385
References Bems, K.I., 1991. Parvoviridae and their replication, In: B.N. Fields, D.M. Knipe (Eds.), Fundamental Virology, 2nd edition. Raven Press Ltd., N.Y., pp. 817-837. Broll, S., Waldvogel, AS., Rosskopf, M., Corboz, L., Pospischil, A. 1993. Untersuchungen zu infektiosen Ursachen ftir Aborte und Totgeburten beim Schwein in der Schweiz. J. Vet. Med. B, 40: 641-653. Cartwright, S.F., Lucas, M., Huck, R.A. 1971. A small haemagglutinating porcine DNA-virus. II. Biological and serological studies. J. Comp. Pathol., 81: 145-155. Gradil, C.M., Joo, H.S., Molitor. T.W. 1990. Persistence of porcine parvovirus in swine infected in utero and followed through maturity. J. Vet. Med. B, 37: 309-316. Hogg, G.G., Lenghaus, C., Forman, A.J. 1977. Experimental porcine parvovims infection of foetal pigs resulting in abortion, histological lesions and antibody formation. J. Comp. Pathol., 87: 539-549. Jaspers, B. 1982. Htihnereiweisseinbettung. Histochem. J. 1: 32-33. Joo, H.S., Donaldson-Wood, C.R., Johnson, R.H., Campbell, R.S.F. 1977. Pathogenesis of porcine parvovirus infection: Pathology and immunfluorescence in the foetus. J. Comp. Pathol., 87: 383-391. Krell, P.J., Salas, T., Johnson, R.P. 1988. Mapping of porcine parvovirus DNA and development of a diagnostic DNA probe. Vet. Microbial., 17: 2943. Lager, K.M., Mengeling, W.L., Liu, W. 1992. Comparison of the virulence of two isolates of porcine parvovirus in 72-day-old porcine fetuses. J. Vet. Diagn. Invest., 4: 245-248. Leslie-Steen, P., Kirkbride, C. 1983. Comparison of diagnostic methods for in utero parvovirus infection of swine fetuses. Proc. 3rd Intl. Symposium of World Ass. of Vet. Lab. Diagnosticians, Ames, Iowa, USA: 11 l-l 17. Marrable, A.W., Ashdown, R.R. 1967. Quantitative observations on pig embryos of known ages. J. Agric. Sci. Camb., 69: 443-447. Mengeling, W.L., Cutlip, R.C. 1976. Reproductive disease experimentally induced by exposing pregnant gilts to porcine parvovirus. Am. J. Vet. Res., 37: 1393-1400. Molitor, T.W., Choi, C.S., Harding, M.J., Joo, H.S. 1986. In utero inoculation of swine fetuses to study viral pathogenesis and immunity, In: Tumbleson M.E. (Ed.), Swine in biomedical research, Vol. III, Plenum Press, N.Y., pp. 1945-1954. Nielsen, J., Ronsholt, L., Sorensen, K.J. 1991. Experimental in utero infection of pig foetuses with porcine parvovinrs (PPV). Vet. Microbial., 28: l-l 1. Nuovo, G.J., Richart, R.M. 1989. A comparison of biotin-and %i-based in situ hybridization methodologies for detection of human papillomavirus DNA. Lab. Invest., 61: 471-476. Rentrop, M., Knapp, B., Winter, H., Schweizer, J. 1986. Aminoalkylsilane-treated glass slides as support for in situ hybridization of keratin cDNA to frozen tissue sections under varying fixation and pretreatment conditions. Histochem. J., 18: 271-276. Sambrook, J., Fritsch, E.F., Maniatis, T. 1989. Molecular Cloning. 2nd edition, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Waldvogel, A.S., Hassam, S., Weilenmann, R., Tratschin, I.D., Siegl, G., H&nichen, T., Briner, J., Pospischil, A. 1991. Retrospective study of myocardial canine parvovirus infection by in situ hybridization. J. Vet. Med. B, 38: 353-357. Waldvogel, A.S., Hassam, S., Stoerckle, N., Weilenmann, R., Tratschin, J.D., Siegl, G., Pospischil, A. 1992. Specific diagnosis of parvovirus enteritis in dogs and cats by in situ hybridization. J. Comp. Pathol., 107: 141146. Zanoni, R.G., Henn, V., Rutishauser, U.P., Wyler, R. 1984. Haufigkeit der porcinen Parvovirusinfektion in der Schweiz und ein neuer Vimsnachweis mittels Immunelektronenmikroskopie. Zbl. Vet. Med. B, 31: 729-742.