Endoplasmic reticulum stress is a determinant of retrovirus-induced spongiform neurodegeneration

Abstract

FrCas(E) is a mouse retrovirus that causes a fatal noninflammatory spongiform neurodegenerative disease with pathological features strikingly similar to those induced by transmissible spongiform encephalopathy (TSE) agents. Neurovirulence is determined by the sequence of the viral envelope protein, though the specific role of this protein in disease pathogenesis is not known. In the present study, we compared host gene expression in the brain stems of mice infected with either FrCas(E) or the avirulent virus F43, differing from FrCas(E) in the sequence of the envelope gene. Four of the 12 disease-specific transcripts up-regulated during the preclinical period represent responses linked to the accumulation of unfolded proteins in the endoplasmic reticulum (ER). Among these genes was CHOP/GADD153, which is induced in response to conditions that perturb endoplasmic reticulum function. In vitro studies with NIH 3T3 cells revealed up-regulation of CHOP as well as BiP, calreticulin, and Grp58/ERp57 in cells infected with FrCas(E) but not with F43. Immunoblot analysis of infected NIH 3T3 cells demonstrated the accumulation of uncleaved envelope precursor protein in FrCas(E)- but not F43-infected cells, consistent with ER retention. These results suggest that retrovirus-induced spongiform neurodegeneration represents a protein-folding disease and thus may provide a useful tool for exploring the causal link between protein misfolding and the cytopathology that it causes.

FIG. 1.

The region of the brain analyzed in this study (brain stem) is illustrated above. Panel A shows sections of brain stem from mice either mock inoculated or inoculated as neonates intraperitoneally with F43 or FrCas^E. Sections are stained with a polyclonal antiserum to MLV surface glycoprotein (SU). Immunoreactivity was detected with the substrate AEC (red), and the sections were counterstained with hematoxylin (blue). Both viruses infected cells associated with the microvasculature (yellow arrows), as well as cells in the parenchyma consisting of primarily microglial cells (black arrows). Microglial cells are distinguished by highly arborized processes and small ovoid nuclei. Neurons in these sections (red arrows) are identified by the largeness of their nuclei and their prominent nucleoli. At 10 dpi both viruses have spread beyond the microvasculature and infected parenchymal microglial cells. At 17 dpi infection of microglial cells by both viruses is extensive. Focal spongiosis can be detected at 10 dpi only in the FrCas^E-infected brain stem (black arrowheads), and by 17 dpi the spongiosis is extensive. No neuropathology is detectable in the F43-infected brain stem at either time point. Magnification before reduction, ×125. Panel B shows viral RNA levels in the brain stem as measured by real-time RT-PCR with a probe specific for sequences within the gag gene shared by both viruses (n = 8). The difference between FrCas^E and F43 was not statistically significant at 10 dpi (P > 0.05) but was significant at 17 dpi (P < 0.001). Note the inverse relationship between viral RNA levels and neurovirulence. Data were normalized with GAPDH and are shown as mean ± standard error of the mean. In panel A there appears to be a dramatic increase in the level of staining of the SU protein from 10 to 17 dpi, which is not apparent in respective levels of viral RNA shown in panel B. RNA levels correlate with previous quantitative Western blot analyses for FrCas^E that indicate no difference in viral protein levels between 10 and 17 dpi (11). Immunohistochemistry is not a quantitative technique and is subject to substantial sampling error, so it is not used here to indicate levels of virus infection.

FIG. 2.

Quantification of mRNA levels of select ER stress genes in brain stems of FrCas^E-infected mice. Real-time RT-PCR was used to quantify transcript levels in groups of eight mice sacrificed at 10 and 17 dpi. At 10 dpi there was a small increase in FrCas^E-infected brain stem in CHOP, BiP, Grp58/ERp57, and calreticulin, though statistical significance was reached only for CHOP and BiP. At 17 dpi CHOP was up-regulated, but BiP, GrP58/ERp57, and calreticulin each appeared to be down-regulated. PERK was up-regulated as well at 17 dpi (Fig. 3). Data are shown as mean ± standard error of the mean. P for FrCas^E versus: F43: *, <0.05; **, <0.01; and ***, <0.001.

FIG. 3.

Expression of ER stress genes in FrCas^E-infected and tunicamycin-treated NIH 3T3 cells. NIH 3T3 cells were infected with either FrCas^E or F43 or were mock infected. RNA was extracted at 48 (A) or 96 (B) h after infection and was analyzed by real-time RT-PCR with probes for BiP, CHOP/GADD153, calreticulin, Grp58/ERp57, and PERK. Each data set represents infections in triplicate. (C) NIH 3T3 cells were treated with 2.0 μg of tunicamycin or control (DMSO) per ml for 6 h and with RNA subject to real-time RT-PCR with the probes listed above (n = 4). Virus infection data were analyzed with ANOVA with Tukey's multiple-comparison test and are shown as average relative expression values normalized to GAPDH. Data from tunicamycin-treated cells were analyzed with unpaired t tests. For comparisons of FrCas^E, F43, or tunicamycin with DMSO, *, P < 0.05; **, P < 0.01; and ***, P < 0.001. Data are shown as mean ± standard error of the mean.

FIG. 4.

Steady-state levels of viral envelope protein in infected NIH 3T3 cells. Panel A schematically shows the proteolytic processing of the envelope protein of MLV (see text). The viral genome is shown at the top. MSD, membrane-spanning domain. Panel B shows an immunoblot of lysates of FrCas^E-, F43-, and mock-infected cells probed with a polyclonal antiserum to MLV SU protein. This antiserum recognizes pr85^env as well as SU protein but does not react with the TM protein (p15E). The blot was stripped and was reprobed with anti-GAPDH to control for loading differences. Note that the ratio of pr85^env to SU is different for these two viruses, indicating that in FrCas^E-infected cells there is accumulation of the envelope precursor protein, suggestive of ER retention.