Infectious RNA Isolated from the Spinal Cords of Mice Chronically Infected with Theiler's Murine Encephalomyelitis Virus

Abstract

The DA strain of Theiler's murine encephalomyelitis virus (TMEV) persistently infects cells of the spinal cord during the chronic phase of infection. Although in situ hybridization and reverse transcription-PCR have demonstrated the presence of viral RNA in the spinal cord, it has not been determined whether this RNA is infectious and, if so, how many PFU equivalents of virus the RNA can yield. In this study, we demonstrated that TMEV RNA isolated from the spinal cords of chronically infected mice is infectious and that there is at least 30-fold more infectious RNA than infectious virus in the spinal cords of these mice.

Theiler's murine encephalomyelitis viruses (TMEV) are positive, single-stranded (ss) RNA viruses that occur naturally in mice as enteric pathogens and are spread by the fecal-oral route (18, 19). They belong to the family Picornaviridae and the genus Cardiovirus. Two groups of TMEV have been defined based on neurovirulence: TO and GDVII. The TO group, including the Daniels (DA) and BeAn strains, is the less-neurovirulent group, causing a biphasic disease in susceptible mouse strains. During the acute stage, 1 to 2 weeks after infection, TMEV infects neurons in the gray matter of the brain, causing polioencephalomyelitis. During the chronic stage, which starts around 4 weeks after infection, TMEV induces an inflammatory demyelinating disease, similar to multiple sclerosis (MS) (8, 22). During the chronic stage of TMEV infection, TMEV RNA and antigen have been found in macrophages/microglia, oligodendrocytes, and astrocytes in the white matter of the spinal cord by in situ hybridization and immunohistochemistry (10, 14, 24). Infectious TMEV has been detected in the spinal cord by plaque assays (3, 9, 23).

Restricted viral gene expression may not only reduce the lysis of virus-infected host cells (1, 2) but may also provide an effective mechanism by which virus escapes detection by the host immune responses in TMEV infection (22). In DA virus infection, the number of viral RNA-containing cells has been shown to be larger than that of viral antigen-positive cells (1, 24). Using in situ hybridization, Cash et al. (1, 2) showed that central nervous system (CNS) cells infected with DA virus contain 100 to 500 copies of viral RNA. Using Northern hybridization and real-time reverse transcription (RT)-PCR, Trottier et al. (20, 21) reported a large number of viral genomes (mean, 3.0 × 10⁹ per spinal cord) in the spinal cords of mice infected with BeAn virus. BeAn viral genomes were full-length in size, and no subgenomic species were detected.

mRNA, derived from the cDNA clones of TMEV by linearization and in vitro transcription, has been shown to be infectious if it is transfected into cells (15, 17). While TMEV genome detected by in situ hybridization or isolated from infected mice has been shown to be full-length in size (20), it is unknown whether the TMEV RNA detected in the spinal cord from TMEV-infected mice is infectious and, if so, how many PFU of virus or infectious genomes the TMEV RNA can yield, based on plaque assays.

We infected 6-week-old female SJL/J mice (Jackson Laboratory, Bar Harbor, ME) intracerebrally with 2 × 10⁵ PFU of the DA strain of TMEV (23). The spinal cords were flushed from the spinal canals of mice with phosphate-buffered saline at 6 months postinfection; the spinal cords were weighed and stored frozen until use. The spinal cords (average mass, 0.123 g per spinal cord) from five mice were pooled, 1 ml of phosphate-buffered saline was added, and the spinal cords were homogenized. The homogenate was divided into two portions. Plaque assays (7) were performed using one portion (42%) of the homogenate in order to determine the titer of intact, infectious viral particles present in the spinal cords. The titer of DA virus present in the spinal cords of chronically infected mice was determined to be 1.25 × 10⁴ PFU/g of spinal cord (Table 1). This is in close agreement with what has been determined by us and others: <10⁴ PFU/CNS from BeAn-infected SJL/J mice at 1, 3, and 6 months postinfection (9); <1.5 × 10⁵ PFU/CNS from DA-infected SJL/J mice at 2, 2.5, 3, and 4 months postinfection (3); and < 1.7 × 10⁵ PFU/g of spinal cord from DA-infected SJL/J mice at 1 and 4 months postinfection (23). The total RNA (88.54 μg) was isolated, using TRIzol reagent (Invitrogen, San Diego, CA) according to the manufacturer's recommendation, from the other portion (58%) of the homogenate for use in transfection experiments to determine the infectivity of the viral RNA present in the spinal cords. The total RNA was transfected into BHK-21 cells (American Type Culture Collection, Manassas, VA) via DMRIE-C reagent (Invitrogen). After a 4-h incubation, the lipid-RNA solution was replaced with growth medium containing 2% Cosmic calf serum (HyClone, Logan, UT), and after a 24-h incubation in growth medium, cell monolayers were overlaid with agar and later stained with crystal violet, as per the plaque assay protocol (7). Plaques were generated from the RNA isolation/transfection procedure so the TMEV RNA isolated from the spinal cords of chronically infected mice was infectious, and the titer was determined, after adjustment for transfection efficiency (as described below), to be 3.8 × 10⁵ PFU/g of spinal cord (Table 1). This is a 30-fold increase over the titer of DA virus present in the spinal cords; therefore, there is 30-fold more infectious RNA than intact, infectious virus in the spinal cords of these mice. This disparity between infectious RNA and infectious virus may reflect a restriction of infectious virus production and/or neutralization of infectious virus by TMEV-specific antibody that could be present in the CNS homogenate (20). Another factor contributing to the disparity is nonencapsidated viral RNA. Our RNA isolation procedure should result in the isolation of not only viral RNA from the virion (encapsidated RNA) but also nonencapsidated RNA from the cytoplasm (either after decapsidation or before virion assembly). The infectivity of nonencapsidated versus encapsidated RNA in vivo can be estimated by (i) a comparison between total viral RNA levels (both encapsidated and nonencapsidated) from the spinal cord quantified by RT-PCR and the viral RNA level from virions purified from infected spinal cord (encapsidated only) and (ii) a transfection experiment using viral RNA from purified virions from the spinal cord.

TABLE 1.

Quantitation of infectious virus and infectious RNA from purified DA virus and spinal cord homogenate from mice chronically infected with TMEV

Source	Plaque assay titer (PFU)a	Transfection titer (PFU)a	Efficiency (%)	Fold increase
Purified DA virus	1.21 × 107	7.86 × 106	65	NAb
Spinal cord homogenate	1.25 × 104	3.80 × 105	NA	30.4

^aTiters are expressed per milliliter for DA virus and per gram for spinal cord homogenate.

^bNA, not applicable

In order to determine the efficiency of the RNA isolation and transfection procedure, purified DA virus was isolated from infected BHK-21 cells through sonication, detergent lysis, low-speed centrifugation, and CsCl gradient centrifugation (11). The titer of the purified DA virus was determined by plaque assay, and RNA isolation and transfection were carried out as described above. From a preparation of purified DA virus with a titer of 1.21 × 10⁷ PFU/ml, we isolated 5.85 μg of RNA, which yielded 7.86 × 10⁶ PFU/ml (Table 1). Thus, RNA isolation and transfection resulted in a 35% decrease in virus titer. Therefore, the 30-fold increase in virus titer in the spinal cord RNA transfection experiment cannot be due to the RNA isolation or transfection procedure. In addition, it is possible that the infectivity determined by transfection of RNA extracted from the spinal cord is lower than the actual infectivity, since different kinds of inhibitors might exist in the RNA preparation from the spinal cord and in the RNA preparation from purified virions (although TRIzol isolates RNA from both DNA and proteins).

This efficient transfection rate (65%) in cell culture suggests that the discrepancy between viral RNA and infectivity in the spinal cord homogenate is not an intrinsic inefficiency of TMEV infection. Using the BeAn strain of TMEV for infection, Trottier et al. (20) showed that the ratio of viral genomes to PFU was on the order of 10⁶:1 to 10⁷:1 during the chronic phase, compared to a ratio of 10³:1 during the acute phase, suggesting a restriction in the production of infectious virus during the chronic phase. This is in accord with our present results.

One of many factors that may contribute to the difference is the host cell types that harbor TMEV. During the acute phase, TMEV predominantly infects neurons in the gray matter of the brain, while during the chronic phase, virus infects only glial cells and macrophages in the white matter of the spinal cord. Thus, the influence of cell-type-specific responses against TMEV will be worth investigating, as discussed below.

In this study, we demonstrated that TMEV RNA isolated from the spinal cords of chronically infected mice is infectious. Although we do not know the precise mechanism by which host cells control replication of infectious virus from viral genomes in vivo, cell-type-specific and virus-specific immune responses could be the main factors responsible for the restriction of the synthesis of infectious virus during persistent infection in the murine CNS. One such host factor that influences proliferation of ssRNA virus is Toll-like receptors (TLRs), by which cells recognize invariant molecular structures of pathogens termed pathogen-associated molecular patterns or PAMPs (6). ssRNA is recognized by murine TLR7 and human TLR8, while double-stranded (ds) RNA is recognized by TLR3 (CD283) (4, 13). Stimulation of both TLR3 and TLR7 causes induction of a type I interferon, which is important in controlling virus replication, including TMEV (5). TMEV carries a positive ssRNA genome and can form dsRNA in the infection cycle. During the chronic stage of TMEV infection, virus has been shown to persistently infect oligodendrocytes, astrocytes, and microglia/macrophages (22). Thus, in TMEV infection, it will be worth studying the role of TLR3 and TLR7 in infected glial cells and macrophages in controlling infectious viral RNA genome during the chronic stage in vivo, while in vitro studies suggest a role of TLR3 in microglia (12) and astrocyte cultures (16).