SJL/J Mice Are Highly Susceptible to Infection by Mouse Adenovirus Type 1

Abstract

Mouse adenovirus type 1 (MAV-1) targets endothelial and monocyte/macrophage cells throughout the mouse. Depending on the strain of mouse and dose or strain of virus, infected mice may survive, become persistently infected, or die. We surveyed inbred mouse strains and found that for the majority tested the 50% lethal doses (LD₅₀s) were >10^4.4 PFU. However, SJL/J mice were highly susceptible to MAV-1, with a mean LD₅₀ of 10^−0.32 PFU. Infected C3H/HeJ (resistant) and SJL/J (susceptible) mice showed only modest differences in histopathology. Susceptible mice had significantly higher viral loads in the brain and spleen at 8 days postinfection than resistant mice. Infection of primary macrophages or mouse embryo fibroblasts from SJL/J and C3H/HeJ mice gave equivalent yields of virus, suggesting that a receptor difference between strains is not responsible for the susceptibility difference. When C3H/HeJ mice were subjected to sublethal doses of gamma irradiation, they became susceptible to MAV-1, with an LD₅₀ like that of SJL/J mice. Antiviral immunoglobulin G (IgG) levels were measured in susceptible and resistant mice infected by an early region 1A null mutant virus that is less virulent that wild-type virus. The antiviral IgG levels were high and similar in the two strains of mice. Taken together, these results suggest that immune response differences may in part account for differences in susceptibility to MAV-1 infection.

For a complete understanding of viral disease, knowledge of the host factors involved is crucial. Differential host susceptibility to a variety of animal viruses has been observed (reviewed in reference 9), including retroviruses (6), poxviruses (14), papovaviruses (28), rhabdoviruses (23), and herpesviruses (15, 38). Human immunodeficiency virus type 1 (HIV-1) is a well-studied example of a human virus in which host genetics plays a role in the outcome of infection. At least eight human loci in which allelic differences affect HIV-1 infection and AIDS progression have been identified (reviewed in references 16, 31, and 34). Genes affected include those for HIV coreceptors and chemokines and genes of the major histocompatibility complex (MHC).

Mouse adenovirus type 1 (MAV-1), which causes acute and persistent infections in mice, has different disease outcomes depending on the dose of virus administered and strain of mice infected. At sufficient doses, MAV-1 causes an acute fatal disease in both newborn and adult mice (19, 22, 27). In infected outbred and inbred mice, brains and spinal cords exhibit encephalomyelitis. MAV-1 infects endothelial cells and cells of the monocyte/macrophage lineage and is disseminated throughout many organs (12, 24).

We identified outbred mice with different susceptibilities to MAV-1 infection (27). Except for the quantity of virus needed to induce infection, infected susceptible NIHS and resistant CD-1 mice were similar in all criteria tested, including outward signs of disease, histology, presence and quantity of viral DNA in various organs, and presence of anti-MAV-1 serum antibodies. At the same time, Guida et al. (19) reported differences in susceptibility in inbred strains: C57BL/6 (B6) mice were at least 100-fold more susceptible to MAV-1 than BALB/cJ mice. Clinical signs were not seen in the resistant mice, and virus was found at different levels in the brains and spinal cords of infected mice of the susceptible and resistant strains.

We report here a survey of MAV-1 infection of a number of inbred strains of mice. For the majority, the 50% lethal doses (LD₅₀s) were very high. However two strains, SWR/J and SJL/J, were considerably more susceptible. Brains of SJL/J mice showed more degenerative vascular changes and higher levels of infectious virus than those of resistant C3H/HeJ mice. Infection of primary cells from susceptible and resistant mice yielded equivalent levels of virus. Sublethal gamma irradiation of resistant C3H/HeJ mice resulted in their having a much lower LD₅₀, consistent with a role for the immune system in control of the viral infection. The implications of these results are discussed, particularly with respect to known genetic defects of SJL/J and C3H/HeJ mice that affect interactions with infectious agents via innate and adaptive immunity.

MATERIALS AND METHODS

Cells and viruses.

Cell culture media and components were obtained from Life Technologies unless otherwise indicated. Mouse NIH 3T6 cells were maintained in Dulbecco modified Eagle medium plus 5% heat-inactivated calf serum. All media were supplemented with 100 U of penicillin, 100 μg of streptomycin, and 20 U of nystatin per ml. Primary mouse embryo fibroblasts were prepared from 14- to 15-day embryos and maintained in Dulbecco modified Eagle medium containing 10% heat-inactivated fetal bovine serum. Cells were passaged once prior to infection. Primary mouse intraperitoneal (i.p.) macrophages were prepared by i.p. lavage from 4- to 6-week-old male mice 3 days after injection with 3% thioglycolate medium (Difco). Cells were plated in RPMI 1640 medium supplemented with 50 μg of gentamicin sulfate (Sigma) per ml and 10% fetal bovine serum. Cells were washed at 24 h postplating to remove nonadherent cells, and the adherent cells were then infected. Wild-type (wt) MAV-1 was the standard MAV-1 stock originally obtained from S. Larsen (1). pmE109 is an early region 1A (E1A) null mutant of MAV-1 in which the initiator methionine has been mutated, and the virus does not express detectable levels of E1A protein (40). Plaque assays were carried out in 3T6 cells as described previously (11).

Mice.

All animal work complied with all relevant federal guidelines and institutional policies. Mice were obtained from Jackson Laboratories and maintained in microisolator cages. Mice were infected with the indicated virus doses by i.p. injection in a volume of 0.1 ml of conditioned medium or phosphate-buffered saline (PBS). Infected mice were monitored twice daily for signs of disease. Mice were euthanized by inhalation of CO₂ at 3, 5, or 8 days postinfection (p.i.) or if moribund. The LD₅₀ experiments were carried out as described previously with 4- to 6-week-old mice (10). Briefly, 10-fold serial dilutions of virus were prepared, and groups of four to six mice were infected for each dose; six to eight doses were used in each LD₅₀ determination. LD₅₀s were determined using the method of Reed and Muench (37). Organs were harvested from mice and processed for histopathology and in situ hybridization using an antisense digoxigenin-labeled probe corresponding to MAV-1 early region 3 (E3) as described previously (24). This probe hybridizes to both viral DNA and mRNA. When a sense probe was used, the signals were slightly decreased, indicating that the majority of the signal detected is from viral DNA (data not shown).

Five- to 9-week old male C3H/HeJ mice were irradiated with 700 rads from a ⁶⁰Co source, a dose that was determined empirically to be sublethal for C3H/HeJ mice: 20 of 20 irradiated uninfected mice survived for 22 days following irradiation. At 24 h after irradiation, mice were inoculated with MAV-1 for an LD₅₀ determination, with five mice per virus dose. Irradiated mice were maintained in autoclaved microisolator cages and provided autoclaved food and water ad libitum.

Determination of virus loads in organs.

Whole brains or spleens were collected aseptically from euthanized mice. Suspensions (5 to 10%, wt/vol) of 20 to 100 mg of tissue in PBS were homogenized with 100 to 200 mg of sterile sand in 1.5-ml microcentrifuge tubes with plastic pestles (VWR Scientific Products). Sand and tissue debris were removed by centrifugation (5 min, 700 × g) at room temperature. Aliquots and 10- and 100-fold dilutions of the aliquots were assayed by plaque assay on 3T6 cells (11).

Virus titers in organs were considered to be log-normally distributed. The means of the log titers were compared by a two-tailed t test, assuming equal variance. Samples with no plaques were omitted from the statistical analyses. Counts of fewer than 20 plaques per 60-mm-diameter plate were considered to be unreliable; thus, we calculated a detection limit for the organ homogenates of 2 × 10³ PFU/g of tissue.

Analysis of viral mRNAs in organs.

Brain or spleen samples (≈100 mg) were mechanically disrupted in 3.5 ml of TRI reagent (Molecular Research Center, Inc.) using a Kinematica Polytron homogenizer. RNAs were isolated and expected yields were obtained as per the manufacturer's instructions (Molecular Research Center, Inc.). Positive control RNAs were prepared as described previously (5) from 3T6 cells infected with MAV-1 at a multiplicity of infection (MOI) of 0.5. Yeast RNA (Ambion) was used as a negative control in RNase protection assays. An [α-³²P]UTP-labeled MAV-1 hexon probe was prepared by T7 polymerase transcription of a genomic hexon plasmid, pHEX, that had been digested at nucleotide (nt) 16432 in the viral sequence at a BamHI site introduced by PCR cloning (MAV-1 numbers are according to GenBank accession no. NC_000942). The full-length probe was 395 nt (MAV-1 nt 16432 to 16769, plus 58 nt of vector), and the protected size after RNase digestion was 337 nt. A mouse actin ³²P-labeled probe was prepared by T7 polymerase transcription of pTRI-actin-mouse (Ambion); full-length and protected fragment sizes were 304 and 245 nt, respectively. RNase protection assays were carried out by hybridizing 50 μg of RNA and the indicated viral probe plus the actin probe overnight at 42°C. RNase and digestion buffer were obtained from Ambion and used according to the standard conditions. Samples were ethanol precipitated and electrophoresed on 5% polyacrylamide–8 M urea gels that were fixed for 1 h in 45% methanol–10% acetic acid, dried, and analyzed with a phosphorimager.

For reverse transcription-PCR (RT-PCR) analysis, DNase-treated mouse brain RNA (10 μg) was reverse transcribed with avian myeloblastosis virus reverse transcriptase as described previously (2). MAV-1 E3 primers for PCR were MAVR24718 (5′TTC CTG TGC CTG CTT CTA CTC GTA TT3′) and MAVR25148 (5′AAA CAG GGC AGC AGC CAC GCT GCT GTT A3′), which span an E3 intron and therefore can be used to distinguish cDNA from any contaminating viral genomic DNA. PCRs (40, 45, or 50 cycles) were carried out with annealing at 65°C. Products were analyzed on 7% polyacrylamide gels. We conducted reconstruction experiments in which we mixed an E3 plasmid template with cDNA made from uninfected cells prior to PCR amplification. Assuming that 3 to 5% of total RNA is mRNA and that there are 500,000 transcripts per cell, our detection level in a PCR amplification was 1.8 E3 transcripts per cell. As a control for cDNA synthesis and PCR, we assayed Kitl, which is expressed in mouse brains (3). All cDNA samples tested positive using Kitl mRNA primers (SL5 [5′CGG TGC GTT TTC TTC CAT GCA3′] and SL21 [5′CTA TCT GCA GCC GCT GCT3′]).

ELISA.

SJL/J or C3H/HeJ mice were infected with pmE109 virus at various doses in an LD₅₀ experiment. At 21 days p.i., surviving mice were euthanized and serum samples were collected. Threefold serum dilutions ranging from 1:10 to 1:3,000 were tested for antiviral immunoglobulin G (IgG) using commercial MAV-1 enzyme-linked immunosorbent assay (ELISA) plates (Charles River) or ELISA plates prepared as follows. Mock- or MAV-1-infected 3T6 cells were trypsinized at 12 h p.i., and 10⁵ cells were plated in alternate rows of a 96-well plate. After 24 h of incubation, cells were washed three times with PBS, fixed at room temperature in 50% acetone–50% methanol for 2 to 5 min, and air dried. Plates were stored at −20 or −80°C until use. Mouse anti-MAV-1 antisera were detected with secondary peroxidase-conjugated goat anti-mouse IgG serum (Charles River). Net ELISA scores were calculated according to the manufacturer's instructions. Net scores of ≥3 for 1:30 serum dilutions are considered highly positive for MAV-1-specific IgG.

RESULTS

Mouse strain differences in infection by MAV-1.

Two outbred strains of Swiss mice differ in their susceptibility to MAV-1 by more than 3 orders of magnitude, as determined by LD₅₀ experiments (27). Two inbred strains, B6 and BALB/cJ, were reported to have LD₅₀s of 10³ and ≥10⁵ PFU, respectively (19). We surveyed additional inbred strains of mice to identify resistant and susceptible mouse strains. Adult male mice (4 to 6 weeks old) were inoculated i.p. with wt MAV-1 at various doses, and LD₅₀s were determined (Table 1). The majority of strains tested were resistant to MAV-1 (LD₅₀ of >10^4.4 PFU). However, two strains, SWR/J and SJL/J, were more susceptible to MAV-1; the geometric mean of the LD₅₀s for the SJL/J strain was 10^−0.32. For all strains the time of death was dose dependent. Mice given the highest doses died at 3 to 4 days p.i., and mice succumbing to the lowest doses died by 12 to 14 days p.i.

TABLE 1.

Susceptibilities of inbred mouse strains to MAV-1 infection

Strain	LD50 (PFU)a in expt:
	1	2	3	4	5	Mean
129/J	>104.4	—b	—	—	—
A/J	>104.4	—	—	—	—
BALB/cJ	>104.4	—	—	—	—
C3H/HeJ	>104.4	>104.4	—	—	>104.4	>104.4
C57BL/6J	>104.4	>104.4	—	—	—
DBA/J	>104.4	—	—	—	—
SWR/J	102.6	—	—	—	—
SJL/J	<100.5	<10−1.6	100.15	—	—	<10−0.32
SJL/J females	—	—	—	<100.20	—
C3H/HeJ, irradiatedc	—	—	—	—	<10−1.5

^aLD₅₀ for 4- to 6-week-old mice inoculated i.p. All were males except as noted. Virus titers were determined in mouse 3T6 cells, in which the particle/PFU ratio is ∼1,000 (Spindler, unpublished data). Values with “>” are for determinations in which >50% of the mice survived the indicated (maximum possible) dose. Values with “<” are for determinations in which <50% of the mice survived the indicated dose. 

^b—, not determined. 

^cMice were gamma irradiated 24 h prior to infection. 

Because SJL/J mice exhibit sex and age dependence for susceptibility to certain diseases (41, 47), we examined whether the sex or age of SJL/J mice could account for their susceptibility to MAV-1. There was no significant difference in LD₅₀ between male and female SJL/J mice (Table 1) or in mortality between male and female mice given a single low i.p. dose of MAV-1 (30 PFU) (Table 2). There was no significant difference in mortality between mice infected at 4 to 6 or at 12 weeks of age.

TABLE 2.

Mortalities of SJL/J mice at different ages

Age (wk)	Sex	No. of deaths/totala	Day of death
4–6	Male	15/18	6–17
	Female	6/6	7–13
12	Male	2/2	10–11
	Female	6/6	8–13

^aMice were inoculated i.p. with 30 PFU of MAV-1. 

A histopathological assessment of C3H/HeJ (resistant) and SJL/J (susceptible) mice infected i.p. with 10² to 10⁴ PFU MAV-1 was conducted. In general, the morphological changes in both strains were similar to those reported previously for MAV-1 infection of outbred mice (10, 24, 27). Some mice had early germinal center development in spleen or lymph nodes, but there were no differences between the two strains. At the doses of 10³ and 10⁴ PFU at 3 days p.i., the C3H/HeJ mice showed somewhat more perivascular edema in the brain than the SJL/J mice. At 8 days p.i., C3H/HeJ mice infected with 10² PFU showed modest diffuse perivascular edema with some focal neuronal degeneration (Fig. 1A). In contrast, similarly infected SJL/J mice showed more distinct degenerative vascular changes, including infiltration of inflammatory cells into the vessel wall and fibrinoid changes (Fig. 1B). By in situ hybridization with a probe that can detect both viral DNA and mRNA, the only virus-positive cells occurred in the SJL/J mice infected with 10² PFU and euthanized at 8 days p.i. (Fig. 1C and D). In all three mice there was evidence of replicating virus in scattered endothelial cells of brain, with fewer positive endothelial cells in spleen and lymph nodes. In contrast, in the C3H/HeJ mice there was evidence of replicating virus in only one endothelial cell in the brain of only one mouse at the same dose and time p.i. There was no positivity in any other organs.

FIG. 1.

Infection of brains of susceptible and resistant mice. Sections of cerebrum from mice infected with 10² PFU of virus at 8 days p.i. were stained with hematoxylin and eosin (A and B) or by in situ hybridization with an E3 riboprobe (C and D). (A) C3H/HeJ mouse. Note the distinct perivascular edema and lack of endothelial cell reactivity. (B) SJL/J mouse. Note the mild perivascular edema and transmural inflammation of the vascular wall. (C) C3H/HeJ mouse. There is no positive staining by in situ hybridization. (D) SJL/J mouse. Note the positive endothelial cell staining. Bar, 50 μm.

Quantitation of virus infection in mice.

To determine whether virus loads differed between resistant and susceptible strains, spleens and brains from C3H/HeJ and SJL/J mice at various doses and times p.i. were homogenized and assayed for infectious virus by plaque assay. As shown in Fig. 2, at 8 days p.i. susceptible SJL/J mice infected with 10² PFU of MAV-1 showed significantly higher levels of virus in both the brains and spleens compared to infected C3H/HeJ mice (two-tailed t test, P < 0.0001 and P = 0.005, respectively, for brains and spleens). Although at this dose virus was not detected in brains in either mouse strain until 8 days p.i. (Fig. 3A), at higher doses virus was detected in brains at 3 days p.i. and was found at higher levels in SJL/J mice than in C3H/HeJ mice (Fig. 3B). Brain and spleen DNAs from the same mice analyzed in Fig. 2 were prepared, and viral DNA levels were quantitated in a dot blot analysis (27). The results confirmed the quantitative differences observed by assay of infectious virus (data not shown).

FIG. 2.

Virus in organs of susceptible and resistant mice at 8 days p.i. Brains or spleens were obtained at 8 days p.i. from SJL/J (○) and C3H/HeJ (×) mice infected with 10² PFU of MAV-1. Organs were homogenized, and virus yields were determined by plaque assay. Each symbol represents an individual mouse. The short horizontal lines represent the means of the log-transformed titers. For the brains, duplicate brain homogenates were prepared and assayed at different times (experiments 1 and 2). The dotted line at 2 × 10³ represents the lower limit of detection of virus in the organs.

FIG. 3.

Time and dose dependence of virus in infected mouse brains. (A) Homogenates from brains obtained from infection of SJL/J or C3H/HeJ mice with 10² PFU of MAV-1 were assayed at the indicated times p.i. (B) Homogenates from brains obtained at 3 days p.i. from mice infected with the indicated doses of MAV-1.

We examined whether viral mRNA levels differed among brains of infected susceptible and resistant mice. mRNAs isolated at 4, 5, or 8 days p.i. from brains of mice infected with 10² PFU were prepared and analyzed by RNase protection assay. We obtained a positive RNase protection assay signal at 8 days p.i. with a late structural gene probe (hexon) in two of four susceptible SJL/J mice and zero of four resistant C3H/HeJ mice (data not shown). Hexon mRNA was not detected in any mice at 4 or 5 days p.i. We examined these same RNA samples by RT-PCR, because it can be a more sensitive method for detecting low-abundance mRNAs. When we performed PCR using high-specificity primers and 40, 45, and 50 cycles of amplification, we were able to detect low levels of amplified E3 product from all susceptible and resistant mice (Fig. 4). This finding of viral mRNAs in all mice correlates with the finding of infectious virus in all susceptible and resistant mouse brains at 8 days p.i. (Fig. 2). Although these PCRs were not quantitative, positive amplification was consistently detected in samples from susceptible mice at fewer amplification cycles than in samples from resistant mice (Fig. 4 and data not shown). This suggests that higher levels of viral mRNA were present in susceptible mice than in resistant mice, consistent with the significantly higher levels of infectious virus (Fig. 2).

FIG. 4.

RT-PCR of RNAs from brains of susceptible and resistant mice. RNAs from brains of individual mice of the indicated strain that were mock infected (M) (lanes 1, 6, and 7) or infected with 10² PFU of MAV-1 (INF) (lanes 2 to 5 and 8 to 11) were isolated at 8 days p.i. RNAs were analyzed by RT-PCR with MAV-1 E3-specific primers MAVR24718 and MAVR25148 for 40, 45, or 50 cycles (top, middle, and bottom panels, respectively). cDNA templates were as follows: lane 1, mock-infected SJL/J brain; lanes 2 to 5, infected SJL/J brains; lanes 6 and 7, mock-infected C3H/HeJ brain; lanes 8 to 11, infected C3H/HeJ brains. Lanes 12 to 14 show reconstruction experiments in which plasmid DNA (E3 cDNA, pZU14 [2]) was mixed with cDNA prepared from mock-infected brains: top panel, 30, 100, and 300 fg of pZU14, respectively; middle and bottom panels: 3, 10, and 30 fg of pZU14, respectively. We have observed that in these reconstruction experiments, with >30 fg of pZU14 PCR, product yields were reproducibly inversely correlated with pZU14 template concentration (e.g., see top panel, lanes 12 to 14). Similar results obtained by others have been interpreted to occur because when target DNA is present in high concentrations, rehybridization of the amplified fragments occurs more readily than their hybridization to primer molecules (32). Lane 15, 1 pg of MAV-1 virion DNA. Lane 16, water (no added template). Lane 17, DNA marker fragments; sizes (in base pairs) are indicated on the right. V, 430-bp PCR product from genomic DNA. C, 273-bp PCR product from E3 cDNA. + and −, reactions positive and negative for E3 cDNA, respectively.

Investigation of possible mechanisms of resistance to MAV-1.

We determined whether primary cells isolated from susceptible and resistant mice differed in their ability to support virus growth. If viral growth differences were found, they might be accounted for by differences in factors at the individual cellular level, such as virus receptors, or factors involved in virus replication. We prepared primary embryo fibroblasts from resistant and susceptible mice and tested their ability to support growth of MAV-1. As shown in Fig. 5A, there was no difference in the yields of MAV-1 grown on cells from the two mouse strains. One target of MAV-1 infection is cells of the mononuclear phagocytic system (24). Primary macrophages from C3H/HeJ and SJL/J mice infected with MAV-1 yielded equivalent levels of virus (Fig. 5B). These results indicate that a difference in susceptibility in the strains is not reflected in virus growth in the isolated primary cells tested. This suggests that there may be a strain difference at the systemic level that is not evident in cell culture, such as immune system differences.

FIG. 5.

Growth of MAV-1 in primary cells from susceptible and resistant mice. (A) Primary mouse embryo fibroblasts were prepared from SJL/J and C3H/HeJ mice and infected at an MOI of 1.5. Yields were determined by plaque assay on mouse 3T6 cells. Results from three independent experiments are shown. Error bars indicate standard deviations. (B) Primary i.p. macrophages were prepared from SJL/J and C3H/HeJ mice and infected at an MOI of 10.

We tested whether immunosuppression of resistant C3H/HeJ mice by irradiation would increase their susceptibility to MAV-1 infection. Adult mice were irradiated with a sublethal dose of gamma irradiation; 100% of irradiated (uninfected) mice survived for 22 days postirradiation. Twenty-four hours after irradiation, mice were infected with MAV-1. The LD₅₀ for the irradiated C3H/HeJ mice was calculated to be <10^−1.5 PFU (Table 1). This value was much lower than the LD₅₀ for unirradiated C3H/HeJ mice and was comparable to that for the susceptible SJL/J mice. The irradiated C3H/HeJ mice died at the same times as SJL/J mice at each corresponding virus dose (data not shown).

One assay of B-cell function is to measure serum IgG levels in mice by ELISA at various times after infection with wt virus. However, we were unsuccessful in these attempts, because at doses of wt virus low enough to allow survival of the mice until 21 days p.i., the antiviral IgG levels were not significantly different from those of mock-infected mice. At higher wt virus doses, mice died too early (3 to 9 days p.i.) to develop detectable antiviral IgG levels. We were able to overcome this difficulty by using the mutant virus pmE109, which is null for E1A due to a site-directed mutation altering the initiator codon of the protein (39, 40). pmE109 is less virulent than wt virus (39). In susceptible outbred mice the LD₅₀ of pmE109 is higher than that of wt virus (10^3.5 and <10^−1.0 PFU, respectively) (Table 3) (39). In SJL/J mice the LD₅₀ of pmE109 was also higher than that of wt virus (LD₅₀s of 10^2.0 and 10^−0.32 PFU, respectively). Sera from C3H/HeJ and SJL/J mice surviving for 21 days after infection with pmE109 at doses of 10¹, 10³, and 10⁵ PFU/mouse were tested for antiviral response by ELISA. All mice had high IgG levels, as indicated by highly positive ELISA scores. There were no significant differences between IgG levels in pmE109-infected C3H/HeJ and SJL/J mice, and there was no correlation of initial virus dose with antibody response (data not shown). These data, together with the irradiation data, suggest that differences in innate immunity between SJL/J and C3H/HeJ mice may contribute to differences in susceptibility.

TABLE 3.

Susceptibilities of mouse strains to mutant MAV-1 infection

Virus	LD50 (PFU)a for strain:
	Susceptible outbred NIHSb	SJL/J	C3H/HeJ
wt	<10−1.0	<10−0.32	>104.4
pmE109	103.5	102.0	>105.0

^aLD₅₀ for 4- to 6-week-old male mice inoculated i.p. 

^bData are from reference 39. 

DISCUSSION

We have expanded the number of inbred mouse strains that have been tested for susceptibility to MAV-1 infection. The LD₅₀s for SWR/J and SJL/J mice are lower than those for the majority of strains, including 129/J, A/J, BALB/cJ, C3H/HeJ, B6, and DBA/J mice. Our data for B6 mice differ from those of Guida et al. (19), who reported that the LD₅₀ for B6 mice was ≥2 log units lower than that for BALB/c mice. We do not have an explanation for this difference. The age, source, pathogen status, and maintenance of mice prior to infection were the same in both laboratories. The virus strains and routes of infection were also the same. The sex of the mice may account for some of the difference; female B6 mice (predominantly used by Guida et al.) are very slightly more susceptible to MAV-1 (M. Horwitz, personal communication).

We investigated the almost 5-log-unit difference in LD₅₀s between SJL/J and C3H/HeJ mice by examining various aspects of MAV-1 infection in adult mice. We noted two differences in histopathology between these mouse strains. At high virus doses at 3 days p.i., the C3H/HeJ (resistant) mice had more perivascular edema in the brain than did SJL/J (susceptible) mice. This increased edema is consistent with a vigorous innate immune response in C3H/HeJ mice that could aid either in viral clearance or in reducing viral replication. Accordingly, we found lower levels of infectious MAV-1 in brains of the C3H/HeJ mice compared to SJL/J mice regardless of viral dose (Fig. 2 and 3). The second histopathological difference between the two mouse strains was the greater degenerative vascular changes at 8 days p.i. in the susceptible SJL/J mice given a dose of 10² PFU compared to resistant mice. At this dose, which is lethal for this strain, some of the mice had ruffled fur and a hunched posture, indicating severe depression. Susceptible mice given this virus dose were generally moribund by 9 or 10 days p.i. In contrast, at the same dose and time after infection the resistant mice had only modest histopathological changes compared to mock-infected mice, and they never showed clinical signs of disease. The increased vascular permissiveness in susceptible mice correlated with their significantly higher levels of infectious virus in the brain and spleen relative to resistant mice (Fig. 2). This suggests that massive and acute vascular damage due to this permissiveness may be a major factor contributing to the susceptibility of SJL/J mice.

The levels of infectious virus and viral DNA in brains and spleens were significantly higher in susceptible mice than in resistant mice. The amounts of infectious virus in the brain were both dose and time dependent. Mice infected with higher initial doses had more infectious virus, and mice had more virus at 8 days p.i. than at 3 to 5 days p.i. Because detection of MAV-1 mRNAs in brains and spleens of inbred mice at 4 days p.i. had been reported earlier (19), we were surprised that viral mRNAs from brains were difficult to detect by RNase protection assay and RT-PCR analysis at 8 days p.i. from mice given 10-fold-lower virus doses in our experiments. In reconstruction experiments we estimated that we could detect approximately two early mRNA transcripts per cell. Assuming that there are 500 E3 mRNA transcripts per infected cell, the levels of mRNA that we detected would thus correspond to one infected cell per 250 to 500 brain cells. This is consistent with the levels of virus-positive brain cells we detected in outbred Swiss mice by in situ hybridization (24).

It is possible that immune system differences account for some of the observed differences in susceptibility to MAV-1 infections. Little is known about the immune response to MAV-1 infections. Mice infected with MAV-1 elicit virus-specific cytotoxic T-lymphocyte responses (21) and produce neutralizing and complement-fixing antibodies (43). We tested whether immunosuppression by gamma irradiation would alter resistance to MAV-1 infection in resistant C3H/HeJ mice. C3H/HeJ mice were highly susceptible to MAV-1 infection after sublethal gamma irradiation (Table 1), suggesting that resistance to MAV-1 infection has an immunological basis. This is consistent with results of infecting SCID mice, which lack B and T lymphocytes and are highly susceptible to MAV-1 infection (12). We have similar preliminary evidence with Rag1-deficient mice, which also lack B and T cells. Rag1-deficient mice on a B6 background had increased susceptibility to MAV-1 infection relative to B6 controls (M. Moore and K. Spindler, unpublished data). Our result with RAG-1 B6 mice contrasts with a report that Rag1 mice on a B6 background are as susceptible as control B6 mice to MAV-1 infection (12). The difference in response to infection in the two studies may be due to the higher virus dose used by those authors and by the susceptibility to MAV-1 that they observed in B6 control mice and that we do not see. However, taken together, these results indicate that a functioning immune system is critical for survival of a MAV-1 infection.

We demonstrated that susceptible and resistant mice infected by an E1A mutant had similar high anti-MAV-1 IgG responses. This indicates that a difference in antiviral IgG is not responsible for the different courses of MAV-1 infection in susceptible and resistant mice. However, we have preliminary evidence that B cells play an important role in early control of acute MAV-1 infection (Moore and Spindler, unpublished data). It will be interesting to determine whether there is a difference between susceptible and resistant mice in IgM responses or in other aspects of B-cell function. We have preliminary evidence that mice lacking T cells do not differ from control (resistant B6) mice in their susceptibility to acute MAV-1 infections (Moore and Spindler, unpublished data). The fact that T-cell-deficient mice survive acute MAV-1 infection argues against the importance of T cells for resistance to MAV-1 infection.

Charles et al. concluded that the difference in MAV-1 disease outcome in resistant (BALB/c) and susceptible (B6) mice was determined by the ability of the virus to replicate in the vascular endothelium, rather than by differences in immune response (12). They presented evidence for a strain-dependent replication of MAV-1 in the central nervous system: they detected essentially no replication of MAV-1 in brains of BALB/c mice. In contrast, we found evidence of MAV-1 replication in BALB/c brains (G. Hirsch, K. Dokubo, and K. Spindler, unpublished data). One possibility that Charles et al. suggested for the susceptibility difference they observed between BALB/cJ and B6 mice was that there is a receptor difference between the two mouse strains. We do not believe that a difference in target cell replication or receptors accounts for the difference in susceptibility between SJL/J and C3H/HeJ mice. We found distinct evidence of viral replication in brain endothelia of both susceptible and resistant mice, with some evidence of increased vascular damage in susceptible SJL/J mice. Infectious MAV-1 was found in brains of both strains of mice, albeit at higher levels in the susceptible mice. We found no difference in levels of virus yields from infected primary fibroblast and macrophage cells from the two strains of mice (Fig. 5), although we cannot exclude the possibility that there may be differences in other cell types. We do not know whether differences in susceptibility occur before, at, or after entry of virus into the central nervous system. Furthermore, strain differences in replication in the vascular endothelium and strain differences in the immune response are not mutually exclusive. Among other things, cytokine expression, natural killer (NK) cell activity, and infection of and dissemination by macrophages may affect virus replication in endothelial cells. Experiments to address these possibilities are in progress.

SJL/J and C3H/HeJ mice exhibit different susceptibilities to infectious agents. For example, in addition to MAV-1, SJL/J mice are susceptible to Theiler's murine encephalomyelitis virus, street rabies virus, and Plasmodium chabaudi, while they are resistant to coronavirus, measles virus, Listeria monocytogenes, Cryptococcus neoformans, and Trichonella spiralis (reviewed in reference 30). The resistance of SJL mice to mouse hepatitis virus A-59 infection is due to lack of a viral receptor on target tissues; the receptor is found in susceptible and semisusceptible mice (7, 48). C3H/HeJ mice, while resistant to MAV-1, are susceptible to varicella-zoster virus (44) and are semisusceptible to coronavirus infection (7). C3H/HeJ mice have a defect in responsiveness to lipopolysaccharide, encoded by the Tlr4^Lps-d allele (35, 36, 44–46). This defect makes C3H/HeJ mice highly resistant to endotoxic shock, in contrast to most inbred mouse strains that express the normal Tlr4^lps-n allele. It does not seem likely that this known defect of C3H/HeJ mice is responsible for their resistance to MAV-1, since B6, BALB/c, and other mice (Table 1) are also resistant to MAV-1 and yet do not have the Tlr4^Lps-d allele (45, 46).

SJL/J mice have an unusually high resistance to X irradiation, and mice >8 months old have a high incidence of spontaneous tumors, originally described as reticulum cell neoplasms (33) and now considered to be B-cell tumors (reviewed in reference 30). These tumors express endogenous mouse mammary tumor virus superantigens (42) that stimulate autoreactive T cells to secrete cytokines that promote tumor growth (13). The difference in susceptibility to MAV-1 between SJL/J and other inbred strains of mice is seen at 4 to 9 days p.i. of 4- to 6-week-old mice, well before the time when the B-cell tumors appear in SJL/J mice. It thus seems unlikely that this tendency to develop B-cell tumors is related to SJL/J susceptibility to MAV-1. A potential lack of immune diversity in T cells could contribute to the susceptibility of SJL/J mice to MAV-1, either because of the mouse mammary tumor virus superantigen expression or because SJL/J mice have a germ line deletion of 50% of T-cell receptor V_β genes (4). BALB/c, B6, and C3H inbred strains that were resistant to MAV-1 (Table 1) have the full complement of V_β genes. We have preliminary evidence that mice lacking α/β T cells and mice lacking α/β and γ/δ T cells do not differ from control (resistant B6) mice in their response to acute MAV-1 infections (Moore and Spindler, unpublished data). The fact that T-cell-deficient mice survive acute MAV-1 infection argues against the importance of T cells as mediators of resistance. Thus, a lack of T-cell diversity may not be a major contributor to susceptibility of SJL/J mice.

Development of genome analysis methods has made the genetics of susceptibility to infectious diseases more amenable to study (29). We are interested in determining the gene(s) involved in susceptibility of SJL mice to MAV-1. Some specific abnormalities in the adaptive (both humoral and cell-mediated) and innate immune responses have been noted in SJL mice, but none have been specifically correlated with susceptibility to infectious agents (reviewed in reference 30). It is possible that some of these play a role in susceptibility of SJL mice to MAV-1 infection. These include altered levels of serum IgG isotypes and certain T-cell-receptor-expressing T cells relative to other strains of mice. SJL NK cells (a component of innate immunity) have low endogenous activity against lymphoma targets, whereas two other mouse strains with the same MHC class I haplotype (s), A.SW and B10.S, have inducible or high levels of NK activity, respectively (25). F1 mice resulting from crosses between SJL and the other strains exhibited the high-NK phenotype, indicating that the low-NK phenotype of SJL mice is inherited as a recessive trait (26). Additional mapping experiments indicated that a difference in at least three genes accounts for the low NK activity in SJL mice. Experiments to map susceptibility of SJL mice to MAV-1 infection should help address whether any of these known defects of SJL mice play a role in the susceptibility.

It is possible that susceptibility of SJL mice to MAV-1 infection is a polygenic trait. Susceptibility to infectious agents often involves both MHC class I (H-2) and non-H-2 components. For example, susceptibility to intracellular parasites or polyomavirus-induced tumors is dependent on H-2 genes (8, 17, 20) and non-H-2 genes (17, 18). The outcome of HIV-1 infection is also dependent on both HLA and non-HLA genes (16, 31, 34). Experiments with mouse strains that, like SJL, carry the MHC class I H-2^s haplotype are in progress to determine whether this contributes to susceptibility to MAV-1. Identification of the host gene(s) involved in susceptibility by genetic mapping, combined with biological experiments with wt and mutant MAV-1, will provide important information about virus-host interactions involved in infectious disease.