Persistent Rat Virus Infection in Smooth Muscle of Euthymic and Athymic Rats

Abstract

Rat virus (RV) infection can cause disease or disrupt responses that rely on cell proliferation. Therefore, persistent infection has the potential to amplify RV interference with research. As a step toward determining underlying mechanisms of persistence, we compared acute and persistent RV infections in infant euthymic and athymic rats inoculated oronasally with the University of Massachusetts strain of RV. Rats were assessed by virus isolation, in situ hybridization, and serology. Selected tissues also were analyzed by Southern blotting or immunohistochemistry. Virus was widely disseminated during acute infection in rats of both phenotypes, whereas vascular smooth muscle cells (SMC) were the primary targets during persistent infection. The prevalence of virus-positive cells remained moderate to high in athymic rats through 8 weeks but decreased in euthymic rats by 2 weeks, coincident with seroconversion and perivascular infiltration of mononuclear cells. Virus-positive pneumocytes and renal tubular epithelial cells also were detected through 8 weeks, implying that kidney and lung excrete virus during persistent infection. Viral mRNA was detected in SMC of both phenotypes through 8 weeks, indicating that persistent infection includes virus replication. However, only half of the SMC containing viral mRNA at 4 weeks stained for proliferating cell nuclear antigen, a protein expressed in cycling cells. The results demonstrate that vasculotropism is a significant feature of persistent infection, that virus replication continues during persistent infection, and that host immunity reduces, but does not eliminate, infection.

Rat virus (RV) is a common virus of laboratory rats and the prototype virus for the family Parvoviridae (29, 35). It is one of three parvovirus serotypes which infect rats; the others are H-1 virus (54) and rat parvovirus (4). RV can disrupt research by causing disease or distorting biological responses in laboratory rats (53). These effects have been attributed to the proclivity of autonomous parvoviruses for mitotically active cells (18). RV infection in fetal and infant rats, which have numerous cycling cells, can lead to severe tissue necrosis and clinical morbidity (30, 34). The preference of RV for mitotically active cells is also thought to account for its ability to distort responses dependent on cell proliferation, including suppression of tumor growth (7) and immune responses to transplantable neoplasms (13) and tissue alloantigens (40).

The risks to biomedical research from RV are heightened by the persistence of infection after the onset of antiviral immunity. A capacity for persistence was suspected from a early study of RV which demonstrated infectious virus in immune rats (50). We subsequently found that some rats inoculated with RV by the oronasal route at 2 days of age harbored infection for at least 6 months and excreted virus for up to 11 weeks, well after the onset of antiviral immunity (32). Susceptibility to persistent infection appeared to be age dependent, since randomly bred rats inoculated as young adults with the Yale strain of RV (RV-Y) rarely remained infected for more than 4 weeks unless they were immunodeficient (25, 32).

The adverse implications of persistent RV infection prompted a search for causative factors. Preliminary studies of athymic rats demonstrated that T-cell-mediated immunity is essential to eliminate infection (25) and that humoral immunity alone suppresses but does not eliminate preexisting infection (23). These and other results (32) also indicated that mature tissues contain cells susceptible to infection, but they did not confirm the distribution or replication status of virus during persistent infection or explore further the role of host immunity. Additionally, they emphasized that investigation of persistent infection required a more reliable induction strategy. Although inoculation of 2-day-old euthymic rats with RV-Y resulted in persistent infection, the prevalence varied from 0 to 50% during 6 months of periodic sampling (32). Furthermore, approximately one-third of the infants developed severe clinical illness or died during acute infection. Inoculation of older euthymic infants with RV-Y or decreasing the virus dose reduced clinical morbidity but lowered the prevalence of persistent infection. These drawbacks were overcome by inoculating 6-day-old infants with the University of Massachusetts strain of RV (RV-UMass), a more virulent strain (26). This regimen induced infection in 19 of 20 euthymic rats through 8 weeks postinoculation without producing clinical signs and induced asymptomatic persistent infection consistently in athymic rats. Further, the results implied that the influence of host immunity on the distribution and replication status of virus could be investigated by comparing levels of persistent infection in rats of the two phenotypes. This paper reports initial results of that comparison for an 8-week interval after inoculation of infant rats by a natural (oronasal) route. Smooth muscle cells (SMC) were the most conspicuous sites of viral replication during persistent infection in both phenotypes. The onset of immunity in euthymic rats reduced but did not eliminate infection.

MATERIALS AND METHODS

Virus and virus isolation.

RV-UMass was obtained from Arthur Like, University of Massachusetts School of Medicine, Worcester). Virus stocks were prepared and quantified in NRK cells as previously described (26).

Rats.

Pregnant Rowett rats, heterozygous at the rnu locus (rnu/+), which had been mated with athymic (rnu/rnu) males, were obtained from the Animal Genetics and Production Branch, National Cancer Institute, Bethesda, Md. Litters consisted of approximately equal numbers of rnu/+ (euthymic) and rnu/rnu (athymic) pups. All rats were housed in microisolette cages under barrier conditions (25). Dams tested prior to inoculation did not have serum antibodies to RV, rat parvovirus, rat coronaviruses, Sendai virus, pneumonia virus of mice, Theiler's murine encephalomyelitis virus, or Mycoplasma pulmonis. Sera collected at necropsy from rats experimentally inoculated with RV were tested for antiviral antibodies as described below. All positive sera had antibodies to RV but not to the other rodent viruses. No clinical signs were detected during daily observation of inoculated rats.

Unanesthetized 6-day-old rats were restrained manually and inoculated oronasally by placing virus suspension containing a total of 100 50% tissue culture infective doses in the nares (10 μl) and the mouth (10 μl). Suckling rats were weaned at approximately 3 weeks of age and group housed by sex and genotype. Thirty-five euthymic rats and 33 athymic rats were inoculated with virus, and 5 rats of each genotype were sham inoculated to serve as controls.

Tissue collection.

Tissues were collected from virus-inoculated rats at 4, 6, 8, and 10 days and at 2, 4, and 8 weeks after inoculation. The interval through week 2 was designated the acute infection period because it included the preimmune and early immune phases of infection. Results at weeks 4 and 8 represented persistent infection, during which established antiviral immunity was detected by serological testing in euthymic rats. Four to six rats of each genotype were selected randomly at each time point and euthanized with carbon dioxide gas. Tissues collected for microscopic examination included lung, trachea, heart, great vessels, thymus, spleen, lymph nodes, salivary glands, liver, small intestine, pancreas, mesenteric and genital vessels, kidney, testis, epididymus, ovaries, and uterus. Tissues were immersed in periodate-lysine-paraformaldehyde (PLP) (41) for 16 h, transferred briefly to phosphate-buffered saline (PBS), embedded in paraffin wax, and sectioned at a thickness of 5 μm preceding in situ hybridization (ISH) and immunohistochemistry. The lungs were inflated with fixative prior to immersion. Fresh pieces of lung, liver, spleen, and kidney were flash frozen in liquid nitrogen and stored at −80°C prior to extraction of nucleic acid. Fresh pieces of lung, kidney, and spleen collected at 2, 4, and 8 weeks were explanted to detect infectious virus. Blood samples were collected by cardiac puncture, and sera were stored individually at −80°C prior to assay.

Explant culture.

Infectious virus was detected by explant cultures that were prepared and evaluated as described previously (45). Briefly, pieces of lung, spleen, and kidney were collected aseptically at necropsy and minced into 2- to 3-mm fragments. Seven to nine fragments of each tissue were cultured, by tissue, in 25-cm² flasks for 3 weeks to permit significant cell outgrowth. Cultures were then lysed by freezing and thawing and inoculated into 324K cell monolayers for detection of cytopathic effect and/or viral antigen.

Serology.

Sera from euthymic and athymic rats were tested initially for antibodies to RV by an immunofluorescence assay (51). Sera obtained from euthymic rats was tested subsequently by enzyme-linked immunosorbent assay for RV antibodies among immunoglobulin (Ig) classes M, G1, and G2a using bacterially expressed RV VP2 (L. J. Ball-Goodrich, E. A. Johnson, and R. O. Jacoby, submitted for publication) as the antigen. VP2 was purified from the insoluble fraction using His-bind resin (Novagen, Madison, Wis.). Ninety-six-well flat-bottom polystyrene microtiter plates (Nunc MaxiSorp, Roskilde, Denmark) were coated with 150 ng of RV VP2 or bacterially expressed β-galactosidase (β-Gal) protein purified from the soluble fraction of bacteria as a negative control. Plates were incubated at room temperature for 2 h and then overnight at 4°C. Plates were washed three times with PBS containing 0.5% Tween 20 and blocked with 250 μl of 3% gelatin in PBS for 1 h at 37°C. Plates were washed again, and 100-μl samples of serial dilutions of sera, in 0.5% bovine serum albumin (BSA)–PBS, were added to antigen- and β-Gal-coated wells. Plates were incubated at 37°C for 2 h and washed. Horseradish peroxidase (HRP)-conjugated secondary antibodies, goat anti-rat Ig (IgG, IgA, and IgM) used at a 1:10,000 dilution in PBS–0.5% BSA (ICN Pharmaceuticals, Inc., Aurora, Ohio), or mouse anti-rat IgM, IgG1, or IgG2a used at a 1:2,000 dilution in PBS-BSA (Serotec, Inc., Raleigh, N.C.) was added to the appropriate wells and incubated at 37°C for 1 h. After the plates were washed three times, 3,3′,5,5′-tetramethylbenzidine peroxidase substrate (Kirkergaard & Perry, Gaithersburg, Md.) was added to the wells for 5 min, followed by addition of 1 N HCl to halt the reaction. Absorbance was measured in a plate reader at 450 nm. Titer endpoints were defined as the inverse dilution of the last absorbance value greater than the mean β-Gal absorbance value plus 2 standard deviations for that dilution.

DNA extraction.

Small-molecular-weight DNA was extracted from liver, kidney, spleen, and lung by a modified Hirt protocol (27). Frozen tissue was pulverized in cold Teflon chambers with stainless steel balls using a Dismembranator (Braun Instruments, Allentown, Pa.). Powdered tissue was added to lysis buffer containing 0.7% sodium dodecyl sulfate (SDS), 1.25 M NaCl, 20 mM Tris (pH 8.0), 10 mM EDTA, and 250 μg of proteinase K per ml. Suspensions were incubated at 37°C for 2 h, held overnight at 4°C, and then centrifuged at 22,000 × g for 30 min to remove precipitated chromatin. Low-molecular-weight DNA present in the supernate was precipitated with ethanol, pelleted by centrifugation, and dried. Samples were resuspended in Tris-EDTA containing 100 μg of RNase A per ml and incubated at 37°C for 1 h, followed by phenol-chloroform extraction and a second ethanol precipitation.

Southern analysis.

DNA samples were electrophoresed in 1% agarose (SeaKem LE, Rockland, Maine), denatured, and transferred to a Hybond N+ membrane (Amersham, Piscataway, N.J.) using standard protocols. Membranes were baked for 2 h at 80°C. Blots were hybridized at 42°C as described elsewhere (24) using a randomly primed, ³²P-labeled probe with a specific activity of 1.2 × 10⁹ cpm per μg of DNA. Blots were washed twice for 30 min in 2× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate)–0.2% SDS and then twice for 30 min in 0.2× SSC–0.2% SDS and exposed to Kodak X-Omat AR film (Eastman Kodak, Rochester, N.Y.).

Molecular probes.

Randomly primed ³²P-labeled DNA probes were prepared using a commercial kit (New England Biolabs, Beverly, Mass.) and purified RV-UMass DNA (nucleotides 1086 to 4300) as a template. The same template was used to prepare biotinylated randomly primed DNA probes to detect virus by immunoperoxidase staining augmented by tyramine-based amplification (catalyzed reporter deposition [CARD] probe) (NEN Life Science Products, Boston, Mass.) (28).

Strand-specific ³⁵S-labeled riboprobes (Promega, Madison, Wis.) were used to differentiate cells containing viral mRNA, an indicator of viral replication, from cells containing virion and RF DNA. RV-UMass DNA from nucleotides 2655 to 4277 was cloned into Bluescript II KS and SK vectors (Stratagene, La Jolla, Calif.). RNA probes, detecting either plus-sense or minus-sense virus strand, were transcribed in equivalent amounts using the T7 promoter. The threshold for detection of RV virion DNA in PLP-fixed cells with the plus-sense probe was established previously as 2.3 × 10⁴ copies (Ball-Goodrich et al., submitted).

ISH.

Hybridization of tissue sections with randomly primed or strand-specific radiolabeled probes was performed and assessed as previously described (24). Tissues were hybridized with the same probe batch and same exposure times to minimize animal-to-animal and tissue-to-tissue variation in autoradiography. Assessment of radiolabeled hybridizations by light microscopy was based on a semiquantitative scale. Signal prevalence was defined as high (greater than 100 positive cells per tissue section), moderate (50 to 100 cells positive per tissue section), low (5 to 50 positive cells per tissue section), trace (1 to 5 positive cells per tissue section), or negative. A cell was scored as positive if the overlying grain count was at least 8 grains (twice background).

Tissue preparation for the CARD probe was identical to that for the radiolabeled probe. The probe was denatured at 99°C for 4 min and cooled, and 2 ng was added to 50 μl of hybridization mix containing 60% formamide, 10% dextran sulfate, 2× SSC, and 50 μg of salmon sperm DNA per ml. Slides were coverslipped, immersed in mineral oil, and incubated for 48 h at 42°C. Mineral oil was removed with three washes in chloroform, and coverslips were removed during two 4× SSC washes. Tissue sections were then washed twice in 60% formamide–2× SSC at 42°C, twice with 2× SSC at 42°C, and once at room temperature with PBS containing 0.05% Tween 20. Detection reactions were performed according to the manufacturer's protocols (Renaissance, TSA-Indirect; NEN Life Science Products). Streptavidin-HRP was used at a dilution of 1:100 for both reactions, and tissue was exposed to biotinylated tyramide for 10 min and to 4,4-diaminobenzidine for 15 min. Tissue sections were counterstained with hematoxylin.

Immunohistochemistry for PCNA.

Paraffin sections were dewaxed and hydrated. Endogeneous peroxidase activity was quenched by incubation in 3% hydrogen peroxide. Tissue was incubated for 45 min at 37°C with a mouse monoclonal antibody against proliferating-cell nuclear antigen (PCNA) linked to HRP by a flexible polymer backbone which facilitates attachment of numerous HRP molecules (Dako Corporation, Carpinteria, Calif.). After being washed in PBS, sections were exposed to 4,4-diaminobenzidine for 4 min and counterstained with hematoxylin. The sensitivity of PCNA staining was determined by counting the fractions of PCNA-labeled cells in sections of liver from two 4-week-old uninfected rats. Color photographs were taken of randomly selected fields at a ×400 magnification, and 1,000 to 1,500 hepatocytes per animal were evaluated. Approximately 60% of hepatocytes were scored as PCNA positive.

Colabeling for PCNA and viral mRNA.

Sections were first labeled for PCNA by immunohistochemistry as described above. Then they were hybridized for detection of viral mRNA using an ³⁵S-labeled, minus-sense RNA probe containing 5 × 10⁵ cpm/50 μl of hybridization solution. Slides were washed as previously described and coated with autoradiographic emulsion (24). Emulsion was developed after a 27- to 30-h exposure to labeled tissue sections. One hundred thirteen cells containing viral mRNA in tissues of persistently infected rats were examined for coexpression of PCNA. Cells were scored as having strong, weak, or no PCNA staining.

Immunohistochemistry for RV NS and VP2.

Immune serum for the RV capsid protein (VP2) and the nonstructural protein (NS) was prepared during a prior study of RV infection (Ball-Goodrich et al., submitted). Selected tissues were stained by the avidin-biotin complex immunoperoxidase method (30).

RESULTS

Clinical signs and gross lesions.

No clinical signs or gross lesions were found with either phenotype.

Virus isolation.

Explant culture amplifies infectious RV in tissue samples, so it is highly sensitive for detecting small quantities of infectious virus encountered during persistent RV infection (45). Lung, spleen, and kidney were assayed by explant culture at weeks 2, 4, and 8. All rats of both phenotypes had infectious virus through week 4, and all but one (a euthymic rat) were virus positive at week 8 (Table 1). Twenty-seven of 29 tissues (93%) from euthymic rats tested through week 4 yielded infectious virus, but the prevalence decreased to 7 of 18 tissues (39%) by week 8. All tissue samples from athymic rats were positive through week 8.

TABLE 1.

Prevalence of RV-UMass in lung, kidney, and spleen as determined by explant culture^a

Week	Phenotype	Virus-positive tissues			Virus-positive rats
		Lung	Spleen	Kidney
2	Euthymic	4/4	4/4	4/4	4/4
	Athymic	4/4	4/4	4/4	4/4
4	Euthymic	5/5	4/6	6/6	6/6
	Athymic	5/5	5/5	5/5	5/5
8	Euthymic	0/6	3/6	4/6	5/6
	Athymic	5/5	5/5	5/5	5/5

^aOne sample per tissue was tested. Results are expressed as numbers of positive tissues (rats)/numbers of tissues (rats) tested. 

ISH with randomly primed probes.

A randomly primed, ³²P-labeled DNA probe which detects a segment of the RV genome encoding the NS and VP genes was used to estimate the distribution and frequency of infected tissues and cells. It also served to detect virus in tissues that were not conducive to explant culture. A biotinylated CARD probe (28) was used to confirm infected cell types.

The radiolabeled, randomly primed probe revealed that euthymic and athymic rats developed widespread infection during the first 10 days after inoculation, consistent with previous results (24). Viral DNA was detected in thoracic and abdominal viscera by day 4, and the prevalence in positive tissues, including lymph nodes and spleen, was 100% on day 6 through week 2 (Table 2). Infected tissues contained few necrotic cells, in contrast to the severe necrosis which typifies acute infection in rats inoculated at 2 days of age. The frequency of virus-positive tissues declined by week 8 and to a greater extent among euthymic rats than athymic rats (Table 2).

TABLE 2.

Tissues containing RV detected by ISH using a randomly primed ³²P-labeled probe^a

Tissue	Euthymic rats at week:			Athymic rats at week:
	2	4	8	2	4	8
Heart	4/4	4/6	1/6	4/4	5/5	3/5
Lung	4/4	6/6	4/6	4/4	5/5	5/5
Liver	4/4	6/6	0/6	4/4	5/5	5/5
Kidney	4/4	6/6	5/6	4/4	5/5	5/5
Spleen	4/4	5/6	2/6	4/4	5/5	5/5
Lymph nodes	4/4	5/5	6/6	4/4	5/5	5/5
Small intestine	4/4	5/6	0/6	4/4	5/5	5/5
Great vessels	4/4	6/6	5/6	4/4	4/4	5/5
Mesenteric vessels	4/4	6/6	2/4	4/4	5/5	5/5
Gonadal vessels	ND	3/3	3/3	ND	3/3	2/2
Total	4/4	6/6	6/6	4/4	5/5	5/5

^aResults are expressed as numbers of positive rats/numbers of rats examined. ND, not determined. 

The frequency of virus-positive cells in infected tissues was high in euthymic and athymic rats through day 10 (Table 3). It decreased progressively in both phenotypes from week 2 onward, but the decline was more pronounced in euthymic rats. By week 8, tissues of euthymic rats still contained positive cells, but in small numbers. All tissues examined in athymic rats at week 8 had positive cells at low-to-moderate prevalence.

TABLE 3.

Mean prevalence of RV-positive cells detected by ISH using a randomly primed ³²P-labeled probe^a

Tissue	Euthymic rats				Athymic rats
	Day 10	Week			Day 10	Week
		2	4	8		2	4	8
Heart	+++	+	+/−	+/−	+++	++	+/−	+/−
Lung	+++	+++	+	+/−	+++	+++	+++	++
Liver	+++	++	+/−	−	+++	+++	+++	+
Kidney	+++	++	+	+/−	+++	+++	++	++
Spleen	+++	+	+/−	+/−	+++	++	+/−	+
Lymph nodes	+++	++	+	+/−	+++	+++	++	++
Small intestine	+++	++	+	−	+++	++	++	+
Great vessels	+++	ND	+/−	+/−	+++	ND	++	+
Mesenteric vessels	+++	+	+	+/−	+++	+++	+++	++
Gonadal vessels	+++	++	++	+/−	+++	+++	+++	++

^aSignal prevalence was defined as high (+++) (greater than 100 positive cells per tissue section), moderate (++) (50 to 100 cells positive per tissue section), low (+) (5 to 50 positive cells per tissue section), trace (+/−) (1 to 5 positive cells per tissue section), or negative (−). A cell was scored as positive if the overlying grain count was at least 8 (twice background). The results are expressed as mean scores for the corresponding tissues from three to six rats per time point. ND, not determined. 

Blood vessels were common sites of acute and persistent infection in both phenotypes. The aorta and pulmonary, mesenteric, renal, and gonadal arteries, in addition to smaller arteries and arterioles within lung, liver, kidney, and other tissues, were involved. During acute infection, signal was prominent among endothelial cells; however, it also was found in SMC within vessel walls. During persistent infection, the number of positive endothelial cells decreased, leaving SMC as the primary targets. This finding was confirmed using a biotinylated probe to examine tissues from two rats of each genotype at weeks 4 and 8. Infected SMC occurred singly or in groups and were distributed randomly or adjacent to the intima (Fig. 1A). Most SMC in affected vessels were histologically normal; however, some had swollen, vacuolated, angular, or pyknotic nuclei consistent with cell injury or impending death. Intramural hemorrhage and inflammation was not detected, however, in affected vessels. Virus-positive SMC also were found in the muscle tunics of the small intestine, especially in athymic rats.

FIG. 1.

Tissues from rats inoculated oronasally with RV-UMass at 6 days of age and necropsied 4 weeks after inoculation. (A to C) Tissues subjected to CARD ISH using a randomly primed RV probe; (D) SMCs from an infected athymic rat stained for PCNA by immunohistochemistry followed by ISH for RV using a minus-sense ³⁵S-labeled RV riboprobe. (A) Branch of a mesenteric artery from an athymic rat demonstrating RV DNA in SMC nuclei. Bar = 40 μm. (B) (Left) Lung from a euthymic rat with an infected pneumocyte. Bar = 40 μm. (Right) Kidney from a euthymic rat with a infected tubule epithelial cell. Bar = 40 μm. (C) Kidney from a euthymic rat with an infected arteriole and perivascular infiltration by mononuclear cells. Bar = 40 μm. (D) (Left) SMC in a gonadal artery that is positive for RV mRNA and for PCNA. The plane of focus was chosen to enhance visualization of overlying grains, so the cell is slightly out of focus. Bar = 16 μm. (Right) Intestinal SMC that is positive for RV mRNA and negative for PCNA. As with the left image, the plane of focus was chosen to enhance visualization of the overlying grains. Bar = 16 μm.

Lung was a site of persistent infection in rats of both phenotypes, and pneumocytes were the most commonly affected cells (Fig. 1B [left]). Respiratory epithelium also was infected in athymic rats. Renal tubular epithelium was a common site of acute infection, and positive cells were seen occasionally in rats of both phenotypes during persistent infection (Fig. 1B [right]). A few positive biliary epithelial cells and hepatocytes were found in the livers of persistently infected athymic rats.

ISH with strand-specific riboprobes.

³⁵S-labeled, strand-specific riboprobes were used to detect the frequency and distribution of cells containing RV DNA and RV mRNA. The plus-sense probe was used to detect virion and replicative-form DNA, and the minus-sense probe was used to detect viral mRNA, indicative of active or recently completed replication (8). Tissues from two rats of each phenotype were examined at days 6, 8, and 10 and at week 2, and tissues from five to six rats of each phenotype were examined at weeks 4 and 8. Hybridization with the plus-sense riboprobe detected a slightly lower prevalence of positive cells compared to results using the randomly primed probes, indicating some reduction in sensitivity (Tables 4 and 5). This reduction was attributed largely to the lower energy and extent of labeling of the single-stranded ³⁵S riboprobe than was observed with the double-stranded, randomly primed ³²P probe. The distribution of virus-positive cells was identical, however, to that detected by the randomly primed probes.

TABLE 4.

Prevalence of RV-positive tissues during persistent infection detected by ISH using strand-specific ³⁵S-labeled riboprobes^a

Tissue	Euthymic rats				Athymic rats
	Week 4		Week 8		Week 4		Week 8
	DNA	mRNA	DNA	mRNA	DNA	mRNA	DNA	mRNA
Heart	3/6	1/6	1/6	0/6	3/5	1/5	4/5	2/5
Lung	6/6	4/6	1/6	0/6	5/5	5/5	5/5	4/5
Liver	5/5	2/6	0/6	0/6	5/5	5/5	5/5	4/5
Kidney	6/6	4/6	3/6	1/6	5/5	5/5	5/5	5/5
Spleen	0/6	0/6	0/6	0/6	5/5	1/5	2/5	0/5
Lymph nodes	5/5	0/6	0/6	0/6	5/5	1/5	5/5	0/5
Small intestine	3/6	2/6	0/6	0/6	5/5	3/5	4/5	3/5
Great vessels	2/5	0/5	1/5	0/6	5/5	5/5	4/5	2/5
Mesenteric vessels	5/6	4/6	5/6	0/6	5/5	5/5	5/5	5/5
Gonadal vessels	3/3	3/3	4/5	1/5	3/3	3/3	2/3	2/3

^aResults are expressed as numbers of positive rats/numbers of rats examined. 

TABLE 5.

Mean prevalence of RV-positive cells during persistent infection detected by ISH using strand-specific ³⁵S-labeled riboprobes^a

Tissue	Euthymic rats				Athymic rats
	Week 4		Week 8		Week 4		Week 8
	DNA	mRNA	DNA	mRNA	DNA	mRNA	DNA	mRNA
Heart	+/−	+/−	+/−	−	+/−	+/−	+/−	+/−
Lung	+/−	+/−	+/−	−	+++	+	+	+/−
Liver	+/−	+/−	−	−	++	+	+	+/−
Kidney	+/−	+/−	+/−	+/−	+	+	+	+/−
Spleen	−	−	−	−	+/−	+/−	+/−	−
Lymph nodes	+/−	−	−	−	+	+/−	+	−
Small intestine	+/−	+/−	−	−	+	+/−	+/−	+/−
Great vessels	+/−	−	+/−	−	+	+/−	+/−	+/−
Mesenteric vessels	+	+/−	+/−	−	+++	+	+	+/−
Gonadal vessels	+	+/−	+	+/−	++	+	+	+/−

^aResults are expressed as mean scores for the corresponding tissues from three to six rats per time point. The scoring system is summarized in the footnote for Table 3 and in Materials and Methods. 

The prevalence of tissues and cells positive for viral DNA was consistently higher in rats of both phenotypes than that for viral mRNA. However, the prevalence of tissues and cells positive for viral mRNA during persistent infection was higher in athymic rats than in euthymic rats, except with peripheral (mesenteric) lymph nodes. In athymic rats, mesenteric lymph nodes contained ample RV DNA, especially in germinal centers, but no mRNA was detected. Nevertheless, viral mRNA was found in other tissues of both phenotypes through week 8. SMC in vessels and intestinal muscle tunics were the most commonly affected cell types identified by both probes. SMC also contained RV NS and VP antigens by immunoperoxidase staining (data not shown). RV mRNA-positive cells in euthymic rats were limited to renal and gonadal vessels by week 8 (Fig. 2), whereas vessels in many tissues of athymic rats were positive at this time point.

FIG. 2.

Renal artery from a euthymic rat 8 weeks after inoculation with RV-UMass. An SMC positive for RV mRNA is identified (arrow). ISH with a minus-sense ³⁵S-labeled RV riboprobe is shown. Bar = 40 μm.

Colabeling for and PCNA and RV mRNA.

The replication requirements of autonomous rodent parvoviruses, including utilization of host cell DNA polymerase, imply that RV mRNA should be detected primarily or solely in cycling cells. To test this expectation, tissues were stained for PCNA by immunohistochemistry followed by ISH for RV mRNA. PCNA is a highly stable protein which is first synthesized during late G₁, attains peak concentrations during S phase of the cell cycle, and is not expressed during G₀ (52). Staining for PCNA alone was performed in preliminary experiments to determine whether infected rats had more PCNA-positive cells than uninfected, age-matched control rats. One thousand to 1,500 hepatocytes were counted in each of four rats (one infected and one uninfected of each phenotype) 4 weeks after inoculation. The fractions of PCNA-positive cells were approximately equal in all livers: 54 and 59% in infected rats and 62 and 63% in uninfected rats.

Mesenteric and gonadal vessels and small intestine from two infected athymic rats at 4 weeks after inoculation were colabeled for PCNA and viral mRNA. A total of 113 mRNA-positive cells were scored for PCNA staining (strong, weak, or negative) by two independent observers using light microscopy. The mean counts were negative (48%), weak (34%), and strong (18%). Therefore, approximately half of the mRNA-positive cells were scored as negative for PCNA (Fig. 1D).

Southern analysis.

DNA was prepared from athymic and euthymic rat tissues (spleen, kidney, liver, and lung) to enrich for small-molecular-weight DNA. The band pattern observed during acute infection (day 8) was similar to that established for RV-UMass replication in synchronized tissue culture cells (Ball-Goodrich et al., submitted). It included minus-sense (virion), single-strand DNA which migrates at approximately 2.5 kb and two double-stranded, replicative forms which are approximately 5 kb (monomer) and 10 kb (dimer) (Ball-Goodrich et al., submitted). All three forms also were identified in athymic and euthymic tissues at weeks 4 and 8. However, signal strength correlated with ISH results. Thus, bands were easily visualized in athymic rats (Fig. 3), but bands obtained from euthymic rats, although at the same location, were too faint to photograph. The results provide further evidence, however, that RV replicates during persistent infection.

FIG. 3.

Southern blot demonstrating replicative forms (RF) of RV (monomer RF [mRF], dimer RF [dRF]) and single-strand DNA (ssDNA) in representative tissues from persistently infected athymic rats. Lanes 1 to 6 contain RV DNAs from two rats at week 4, and lanes 7 to 9 contain RV DNA from one athymic rat at week 8. Lanes 1, 4, and 7, kidneys; lanes 2, 5, and 8, livers; lanes 3, 6, and 9, lungs.

Antiviral immunity.

The decrease in virus-positive cells in euthymic rats began with the onset of seroconversion. IgM antibody against RV VP2 was detected by day 10, and IgG antibodies appeared by day 14. IgM titers were not detected by week 4, whereas IgG titers increased slowly (Table 6). The IgG response consisted primarily of IgG2a, the titers for which were consistently higher than for IgG1. IgG antibodies against RV NS proteins also were detected by 14 days and were present through week 8 (data not shown). Athymic rats developed weak anti-RV IgM responses beginning at day 14 but did not develop anti-RV IgG.

TABLE 6.

Titers of RV antibody in euthymic rats determined by enzyme-linked immunosorbent assay using RV VP2 as the antigen

Week (no. of rats)	Antibody	Titer in an indiviudal rat (103)a						Mean titer ± SD
2 (4)	Ig	6.25	6.25	6.25	12.5			7.81 ± 3.13
	IgM	1.25	2.5	6.25	6.25			3.5 ± 3.2
	IgG1	1.25	1.25	1.25	6.25			2.5 ± 2.5
	IgG2a	1.25	6.25	1.25	6.25			3.75 ± 2.89
4 (6)	Ig	6.25	6.25	12.5	6.25	6.25	6.25	7.29 ± 2.56
	IgG1	0.25	0.25	6.25	1.25	0.25	0.25	1.42 ± 2.4
	IgG2a	6.25	1.25	6.25	6.25	6.25	6.25	5.42 ± 2.04
8 (6)	Ig	25.0	12.25	25.0	25.0	25.0	50.0	27.08 ± 12.29
	IgG1	1.25	1.25	1.25	1.25	0.25	12.5	2.96 ± 4.69
	IgG2a	25.0	12.5	25.0	25.0	12.5	50.0	25.0 ± 13.69

^aEach column indicates titers for an individual rat. 

Perivascular accumulations of mononuclear cells appeared in tissues of euthymic rats by week 2, suggesting activation of cell-mediated immunity. They occurred in hepatic portal triads, adjacent to pulmonary vessels and mesenteric vessels, but were most conspicuous surrounding renal cortical vessels, where they continued to intensify through week 4 (Fig. 1C). Mononuclear cell infiltrates were not a feature of early infection in athymic rats, but mild perivasculitis developed among some renal and gonadal vessels and in a few hepatic portal triads by week 4. Several athymic rats also had developed mild bile duct hyperplasia.

DISCUSSION

Vasculotropism is an established feature of RV infection (5, 17, 20, 30, 38) and one that it shares with numerous other viruses (22). For most of these viruses endothelium is the prominent target. However, some vasculotropic viruses, such as encephalomyocarditis virus (12), human and murine herpesviruses (6, 42, 49, 56), and Seoul virus (R. O. Jacoby, S. R. Compton, F. X. Paturzo, and E. A. Johnson, unpublished data), have the capacity to infect SMC. A few early reports noted acute RV infection of SMC using routine histopathology (37, 38), and our prior ISH studies suggested that SMC support RV during acute and persistent infection (24, 26). This paper confirms and extends those results by examining persistent vascular infection with strand-specific probes and immunohistochemistry. For both phenotypes, infection was more prominent in SMC than in endothelium 4 and 8 weeks after inoculation of virus. Infection included virus replication, since RV DNA, mRNA, and NS and VP proteins were present in vascular (and intestinal) SMC. The demonstration of RV replicative forms by Southern analysis provided additional evidence that viral replication occurred during persistent infection, and explant cultures confirmed the presence of infectious virus in persistently infected kidney, lung, and spleen. Detection of virus by ISH compared favorably with that by explant culture. Minor discrepancies between the methods indicate, however, the value of using complementary approaches to detect small quantities of virus during persistent infection.

Virus-positive SMC were localized to subendothelial SMC or dispersed throughout the muscle tunic. These patterns suggest that SMC infection occurred by cell-to-cell extension from overlying endothelium or through capillaries (vasa vasorum) which supply blood to the muscle tunics. Although RV also may have reached the SMC by penetrating between virus-damaged endothelial cells, possibly adherent to red blood cells (47), the absence of intravascular hemorrhage makes this pathway less likely.

The results of PCNA staining support previous evidence that SMC cycle in young adult rats (16). Thus, these cells meet an important criterion for enabling RV replication. Furthermore, vascular SMC in rats can proliferate in response to direct or endothelial injury (14). Therefore, virus-induced endothelial injury and subsequent infection of SMC may potentiate SMC turnover and promote local RV infection. However, the pace of cell-to-cell infection implicit to this possibility may be slow enough to mask increased cell turnover in affected tissues.

The prevalence of viral-DNA-positive cells was consistently higher than for viral-mRNA-positive cells during all stages of infection. Since comparatively small amounts of mRNA are needed to initiate parvoviral replication, this difference may indicate that RV mRNA concentrations—which vary with the stage of virus replication (Ball-Goodrich et al., submitted)—were below the level of ISH detection in some infected cells. Additionally, one expects RV mRNA to be more dispersed intracellularly than RV DNA, which is concentrated in the nucleus and easier to detect by ISH. The lability of mRNA compared to that of DNA also may have reduced detectable levels of RV mRNA during the brief interval between euthanasia and tissue fixation. Further, the comparatively higher frequency of RV DNA-infected cells may reflect intracellular accumulation of nonreplicating virus (i.e., virus sequestration), as has been hypothesized for persistent parvovirus infection of mink. Mori and coworkers (44) found, in this regard, that viral DNA in lymph nodes was prevalent among germinal center cells resembling follicular dendritic cells or macrophages. The presence of RV DNA, but not mRNA, in lymph node follicles of persistently infected athymic rats resembles the results obtained with mink. This distribution may represent sequestered RV and/or concentration of scavenged RV DNA from persistent infection in other tissues.

Viral mRNA was detected in PCNA-positive and PCNA-negative SMC, suggesting that RV replication can proceed in cells that are in the G₀ or early G₁ phase of the cell cycle. A report by Lenghaus and coworkers offers some precedent for this possibility (36). They demonstrated replication of feline parvovirus in cultured cells in which DNA synthesis was blocked by 6 mM thymidine. Although the mechanism was not determined, they speculated that a cell function blocked by thymidine may have been assumed by a viral protein or that part of an infected cell's DNA-replicating machinery was sufficient and available to support parvoviral replication. Nevertheless, other explanations must be explored before concluding that active RV replication occurs in cells that are not in S phase. For example, cell death may not be the sole outcome of RV replication during a single pass of infected SMC through the cell cycle. If viral transcription or replication does not attain peak levels in SMC prior to the end of S phase, it may be present when the cell returns to G₀, an interval when PCNA is not expressed.

Technical deficiencies also must be considered to explain the variable presence of PCNA in SMC containing RV mRNA. Inadequate sensitivity of the immunostaining method for PCNA is an unlikely factor, since the results were at least as sensitive as those reported previously for rat tissues (19, 21). However, the intensity of PCNA staining varies during the cell cycle (21), with weaker reactions expected during G₁ or early S phase. Therefore, small amounts of PCNA may not have been detected by standard light microscopy. In this context, hybridization with the radiolabeled riboprobe may have quenched detection of PCNA, producing false-negative results. However, the fraction of PCNA-positive cells was approximately the same in tissues that were labeled for PCNA and mRNA as in those stained only for PCNA. Furthermore, the ratios of cells which stained strongly versus weakly for PCNA in the two groups of tissues were similar. The lack of a definitive explanation for RV mRNA signal in PCNA-negative SMC justifies closer examination of RV replication in such cells. Infection of synchronized SMC in vitro, including staining for cell cycle proteins less generic than PCNA, may clarify whether RV replication can progress in noncycling cells.

The comparison between euthymic and athymic rats confirmed that the intensity of persistent infection is strongly influenced by host immunity. The reduction in the number of virus-infected cells after seroconversion in euthymic rats is consistent with a role for humoral immunity. However, the development of mononuclear cell infiltrates in infected tissues in our study and in prior investigations (24, 30) and the prominence in this study of IgG2a responses, consistent with a Th1 response in the rat (33), justify exploration for cell-mediated responses. Weak IgM and mononuclear responses occurred in the athymic rats. This result was not surprising given the proclivity of athymic rats to develop some T cells as they age (57). Additionally, athymic rats are capable of producing IgM responses to thymus-independent antigens (60), which may mean that RV has at least one epitope of this type.

The onset of immunity did not eliminate infection in euthymic rats but may promote the sequestration of RV and/or reduce virus replication. These possibilities are consistent with the results of passive immunization, wherein RV immune serum, administered to infected juvenile athymic rats, transiently suppressed detection of infectious virus (23). Antiviral immunity also may reduce cell-associated expression of viral proteins, impeding effective immune recognition of infected cells (1, 46). Alternatively, initial exposure to RV prior to immunologic maturity may result in delayed elimination of virus due to suboptimal immune responses. Immunologic maturation in rats is not complete until at least 1 month after birth (3, 43, 59, 61). Therefore, the 6-day-old rats used in this study were probably immunologically immature when they were inoculated. Additionally, RV may retard anti-RV immunity because it is at least transiently immunosuppressive (40). We are currently pursuing the influence of anti-RV immunity further by determining responses to virus and viral proteins in adult rats.

Viral persistence is a feature of infection caused by other autonomous parvoviruses (15, 48, 55, 58). Among these, the most detailed pathogenesis studies have been performed with Aleutian disease virus (ADV) (2, 9, 10, 39). Persistent RV and ADV infection share the properties of low-level infection after the onset of host immunity and a tropism for lymphoid tissue, but they differ in several significant ways. For example, persistent ADV infection can be induced in adult mink and encompasses viremia, plasmacytosis, hypergammaglobulinemia, and immune complex disease, which reflect elevated but functionally inadequate immune responsiveness. It is also associated with reduced viral replication in individual cells. Additionally, ADV does not target SMC during persistent infection. Therefore, host and viral factors contributing to persistent RV and ADV infection, while similar in some respects, also display potentially important differences.