Main

Moloney murine leukemia virus-TB (MoMuLV) is a type C retrovirus that induces T-cell lymphomas in susceptible strains of mice.1 Its temperature-sensitive mutant ts1 has a single point mutation in the envelope gene, which confers upon the virus the ability to cause loss of T cells and motor neurons in susceptible host strains.2, 3, 4 ts1 infection of the CNS causes a progressive neurodegenerative disease that morphologically manifests as spongiform polioencephalomyelopathy.5 ts1-mediated neuronal degeneration is likely due to loss of glial support and release of proinflammatory cytokines and neurotoxins such as tumor necrosis factor-α (TNF-α), interleukin-1 (IL-1), and nitric oxide (NO) from surrounding ts1-infected glial cells.5, 6, 7

In brainstems of ts1-infected mice, neurons display apoptosis, vacuolization and inclusion bodies.5, 8 These histopathological hallmarks have been associated with animal and human neuronal diseases such as human immunodeficiency virus (HIV)-induced neurodegenerative disease,9 Parkinson's disease,10 amyotrophic lateral sclerosis,11 motor neuronopathy,12 and transmissible spongiform encephalopathy.13 However, there have been no detailed in vivo investigations regarding the molecular mechanism(s) involved in neuronal loss in ts1-induced encephalomyelopathy.

The ER is a principal site for protein synthesis and folding, and also serves as a cellular storage site for Ca2+. ER homeostasis is sensitive to a variety of different stimuli, such as glucose deprivation, perturbation of intracellular and extracellular Ca2+ concentration, and exposure to free radicals. Such conditions can trigger ER stress signals that provoke three functionally distinct responses.14, 15, 16 The first one is upregulation of genes encoding ER chaperone proteins such as glucose-regulated protein 78 (Bip/GRP78). The second is translational attenuation by phosphorylating the double-stranded RNA-dependent protein kinase-like ER kinase (PERK) and eukaryotic initiation factor 2-α(eIF2-α). Unless those two previous adaptive responses are sufficient, a cell death program is activated by transcriptional activation of the gene for GADD153/CHOP (CCAAT/enhancer-binding protein (C/EBP)-homology protein) and by cleavage activation of the ER-associated caspase-12.17, 18, 19, 20 Prolonged ER stress is a primary component in the neuropathology of a wide variety of neurological disorders.16, 21

Recent evidence suggests that either TNF-α or NO causes the ER to release Ca2+, leading to activation of ER-stress signaling pathways and cell death in various types of cells.22, 23, 24 ER stress activates the mitochondrial apoptosis pathway.25, 26 The release of proapoptotic factors from mitochondria is triggered by the proapoptotic Bcl-2 family members such as Bax.26, 27 ts1 infection of glial cells secretes neurotoxic products such as TNF-α and NO, and impair proteasome activity in neighboring neurons.6, 7, 8 We have shown that the expression level of Bax is increased in neuronal cells in brainstems of ts1-infected mice.28 In addition, our preliminary studies show that mitochondria of degenerating neurons have damaged cristae and Ca2+ deposits. All of these connected events may cause ER stress in neurons, leading to neuronal death.

The aim of this study is to elucidate the causes of neuronal death in the ts1-infected brainstems, with emphasis on ER stress signaling pathways. We report here that ts1 infection causes ER stress, leading to neuronal death, presumably initiated by interference with Ca2+ homeostasis. These findings highlight the involvement of the ER stress signaling pathway in retroviral infection-induced neuronal cell degeneration and death in vivo.

Materials and methods

Reagents and Antibodies

Sodium orthovanadate (Na3VO4), sodim β-glycerophosphate, sodium fluoride (NaF), dithiothreitol (DTT), and phenylmethylsulfonyl fluoride (PMSF) were purchased from Sigma Chemical Company (St Louis, MO, USA). Leupeptin, pepstatin, and aprotinin were purchased from Boehringer Mannheim (Indianapolis, IN, USA). Iodoacetamide (an isopeptidase inhibitor) was obtained from Fluka Chemical Corp. (Milwaukee, WI, USA).

Goat IgG antibody against gPr80env has been described previously.4, 5 Rabbit IgG antibodies against phosphospecific eIF2-α and PERK, caspase-3, cleaved caspase-3, and caspase-12 were purchased from Cell Signaling Technology (Beverly, MA, USA). Rabbit IgG antibodies against GADD153/CHOP, phosphorylated calmodulin kinase II-α (CaMKII-α), and goat IgG antibody against Bip/GRP78 were from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Rabbit IgG antibody against glial fibrillary acidic protein (GFAP) was purchased from DAKO (Carpinteria, CA, USA). Monoclonal mouse IgG antibody against β-actin was purchased from Sigma.

Virus

ts1 virus was propagated in the thymus-bone marrow cell line TB. Virus titers were determined using a modified direct focus assay in 15F cells, which are a murine sarcoma-positive, leukemia-negative cell line, as described.29

Animals and Virus Infection

FVB/N mice were obtained from Taconic Farms (Germantown, NY, USA). Mice were maintained in sterilized microisolators and supplied with autoclaved food and water ad libitum. Newborn FVB/N mice were inoculated intraperitoneally with 0.1 ml of ts1 viral suspension containing 106–107 infectious units/ml, as described previously,8 while control mice were inoculated with medium only. The mice were observed daily for signs of paralysis, and were killed at 20 and 25 days post-infection (dpi). We have determined previously that histologically detectable neuronal loss is not significant at 20 dpi, but is evident at 25 dpi.6, 8 This experimental protocol was approved by the Texas A&M University Institutional Animal Care and Use Committee.

Tissue Processing

For histopathology and immunohistochemistry, ts1-infected (n=5) and control (n=5) mice were anesthetized using an intraperitoneal injection of pentobarbital (150 mg/kg) and perfused with 10% buffered formalin, using a peristaltic pump. After a 12 h fixation, the brain segments were separated for further processing. For RT-PCR and Western blotting analyses of mRNAs and proteins in the brainstem, control (n=4) and ts1-infected (n=4) mice were killed, their brains removed, and the brainstem tissues snap-frozen in liquid nitrogen and stored at −80°C.

Semiquantitative RT-PCR Analyses

Total RNA was isolated using the Trizol reagent (Invitrogen, Carlsbad, CA, USA), as described previously.30 Reverse transcription was performed using the ultra-HF RT-PCR system kit (Strategene, La Jolla, CA, USA). Briefly, 1 μg of DNAase-treated total RNA was mixed with oligo (dT) primers and dNTPs. Synthesized cDNA was amplified by a standard PCR protocol using Taq polymerase and mouse specific primers for gPr80env (forward: 5′-ATA ACA ATC TCA CCT CTG ACC AGG-3′; reverse: 5′-GAG CAG AGG TAT GGT TGG AGT AGG-3′), GADD153/CHOP (forward: 5′-TGC CTT TCA CCT TGG AGA CG-3′; reverse: 5′-CCA TAG AAC TCT GAC TGG AAT CTG G-3′), Bip/GRP78 (forward: 5′-CCA CTA ATG GAG ATA CTC ACC TGG G-3′; reverse: 5′-GTA AGG GGA CAC ACA TCA AGC AG-3′), caspase-12 (forward: 5′-GGA GGT AAA TGC TGG ATT GGC-3′; reverse: 5′-GCT TTT TGT TTC GGT GCT TCA C-3′), and caspase-3 (forward: 5′-GAG CAC TGG AAT GTC ATC TCG C-3′; reverse: 5′-AAG CAT ACA GGA AGT CAG CCT CC-3′). Parallel amplifications with primers for β-actin cDNA (forward: 5′-ATG TAC GTA AGC CAT CCA GGC-3′; reverse: 5′-AAG GAA GGC TGG AAA AGA GC-3′) were performed. The numbers of cycles (between 21 and 24) giving the linear range and annealing temperatures (between 56.0 and 58.3°C) were adjusted depending on the genes amplified. Reaction products were electrophoresed on 1.5% agarose gels. Band intensities were analyzed using a densitometer (Model GS-690; Bio-Rad, Hercules, CA, USA) equipped with the Multi-Analyst software program (Version 1.01; Bio-Rad). For quantification, the densities of the signals were normalized to that for β-actin.

Tissue and Cell Extracts

Brainstem tissue lysates were prepared by homogenization of frozen tissues in 10 volumes of lysis buffer containing 50 mM Tris-HCl (pH 7.9), 150 mM NaCl, 1 mM EDTA, 1 mM Na3VO4, 30 mM sodium β-glycerophosphate, 50 mM NaF, 10 mM iodoacetamide, 1 mM DTT, 1% Nonidet P-40, 1 mM PMSF, 10 μg/ml pepstatin, 10 μg/ml leupeptin, and 5 μg/ml aprotinin. The lysates were cleared by centrifugation at 13,000 g at 4°C for 20 min, and the supernatants kept frozen at –80°C. The protein content of the lysates was determined using the Bradford Assay (Bio-Rad), with bovine serum albumin as the standard.

Western Blotting

Proteins (20–50 μg) were separated by 6–12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and then transferred to nitrocellulose membranes (Schleicher and Schuell, Keene, NH, USA). The membranes were incubated for 1 h in blocking buffer (20 mM Tris-HCl buffered saline containing 5% nonfat milk powder and 0.1% Tween 20) at room temperature, and then probed with appropriate antibodies in blocking buffer overnight at 4°C. Normal mouse, rabbit, or goat IgGs at the same dilutions were used as controls. The blots were incubated with appropriate secondary IgG-peroxidase conjugates (1:10,000 dilution; Kirkegaard Perry Laboratories, Gaithersburg, MD, USA), and developed using the enhanced chemiluminescence (ECL) method (Amersham Life Science, Arlington Heights, IL, USA). After imaging, the blots were stripped and reincubated with a mouse monoclonal anti-β-actin antibody to confirm equal protein loading. Densitometric analysis of autoradiographs was performed using a densitometer as described above.

Immunohistochemistry

Paraffin-embedded sections (5 μm) were first deparaffinized and were then subjected to an antigen-retrieval protocol, in which sections were heated in 10 mM citrate buffer (pH 6.0). for 10 min, blocked with 5% normal serum in Tris-buffered saline (TBS; 100 mM Tris, 150 mM NaCl, pH 7.4), and then incubated overnight at 4°C with primary antibodies at a dilution of 1:100. After washing, sections were incubated with biotin-conjugated secondary IgG, and then treated with reagents from a Vecta-Elite streptavidin–peroxidase kit with a benzidine substrate for color development. For double immunostaining, sections were pretreated with rabbit anti-GADD153/CHOP IgG as described above. Sections were then blocked with 5% normal goat or rabbit serum in TBS, after which the tissue sections were incubated with either rabbit anti-GFAP, cleaved caspase-3, or phosphorylated-CaMKII-α antibody at a dilution of 1:100 overnight at 4°C. After incubation with secondary antibodies, sections were treated with an alkaline phosphatase–avidin–biotin substrate and with red chromagen (Vector red; Vector Laboratories) for color development. Sections were counterstained with diluted hematoxylin. Sections without primary antibodies were similarly processed to control for binding of the secondary antibodies. On control sections, no specific immunoreactivity was detected.

Ultrastructural Localization of Ca2+

The oxalate–pyroantimonate procedure was used to visualize the subcellular distribution of Ca2+ in cells from mouse brainstems, as previously described.31 At 20 dpi, control (n=3) and ts1-infected mice (n=3) were anesthetized with pentobarbital (150 mg/kg, intraperitoneal injection), and perfused with a solution containing 2% paraformaldehyde, 2% glutaraldehyde and 90 mM potasium oxalate (pH 7.4) and kept overnight at 4°C in the same solution. Brainstems were removed and washed for 15 min in 7.5% sucrose containing 90 mM potassium oxalate and then postfixed in a mixture of 1% osmium tetroxide and 2% potassium pyroantimonate in 0.01 N acetic acid (pH 7.4) for 2 h. Ultrathin sections were stained with 0.5% uranyl acetate and 0.4% lead citrate and examined with a Zeiss EM 10 electron microscope. The specificity of the cytochemical reaction was tested by treating mounted ultrathin sections with 3 mM ethylene glycol bis(b-aminoethylether)-N,N′-tetraacetic acid (EGTA) at 60°C for 1 h. EGTA treatment, in turn, was controlled by incubation of sections for 1 h at 60°C in distilled water.

Statistical Analysis

Data obtained from densitometric analyses of RT-PCR and Western blots were analyzed by Student's t-test. P-values of <0.05 were considered as statistically significant.

Results

Activation of Upstream Regulators of ER Stress Signaling Pathway in Brainstems of ts1-infected Mice

To determine whether ts1 infection causes ER stress in the brainstem, we first examined whether PERK and eIF2-α, two major upstream transducers of ER stress signaling pathways, undergo phosphorylation.14 Phosphorylated eIF2-α in ts1-infected brainstems was increased 2.0- and 3.3-fold at 20 and 25 dpi, respectively, compared to controls (Figure1a and b; P<0.01). Phosphorylated PERK in ts1-infected brainstems was increased 2.3- and 3.8-fold at 20 and 25 dpi, respectively, compared to controls (Figure1a and b; P<0.01). To ascertain whether phosphorylation of PERK and eIF2-α is correlated with viral infection in brainstems, we determined the amounts of ts1 viral envelope gPr80env mRNA and protein in brainstems. The expression of gPr80env mRNA and protein was increased in ts1-infected mice in a time-dependent manner (Figure 1c). This result is consistent with those reported previously.4, 32

Figure 1
figure 1

Increased phosphorylation of eIF2-α and PERK in ts1-infected brainstems. (a) Representative Western blots showing levels of phospho-eIF2-α and -PERK and in control and ts1-infected brainstems at 20 and 25 dpi. (b) Quantitative analyses of levels of phospho-eIF2-α and -PERK in control and ts1-infected brainstems at 20 and 25 dpi. The results shown in the histogram are the mean±s.d. for tissues from four control and four ts1-infected mice. *Significantly different from controls (P<0.01). (c) Representative photographs of semiquantitative RT-PCR and Western blots showing levels of viral envelope gPr80env mRNA and protein in control and ts1-infected brainstems at 20 and 25 dpi.

Full size image

Induction of ER Stress-associated Genes in Brainstems of ts1-infected Mice

To determine whether phosphorylation of PERK and eIF2-α is accompanied by upregulation of ER stress-associated genes, GADD153/CHOP and GRP78, semiquantitative RT-PCR was used to compare amounts of GADD153/CHOP and Bip/GRP78 mRNAs in brainstems of control and ts1-infected mice (Figure 2a). Figure 2b shows that ts1 infection induced 2.4- and 3.5-fold increases in GADD153/CHOP mRNA at 20 and 25 dpi, respectively, compared to controls (P<0.01). Bip/GRP78 mRNA increased 2.0- and 2.3-fold at 20 and 25 dpi (P<0.01), respectively, compared to controls. The pattern of increase in GADD153/CHOP and Bip/GRP78 proteins was consistent with the increases in GADD153/CHOP and Bip/GRP78 mRNAs (Figure 2c).

Figure 2
figure 2

Induction of GADD153/CHOP and Bip/GRP78 in brainstems of ts1-infected mice. (a) Representative photographs of semiquantitative RT-PCR analyses for GADD153/CHOP and Bip/GRP78 mRNAs in control and ts1-infected brainstems at 20 and 25 dpi. (b) Quantitative analysis of GADD153/CHOP and Bip/GRP78 mRNA levels in control and ts1-infected brainstems. (c) Representative Western blots showing levels of GADD153/CHOP and Bip/GRP78 proteins in control and ts1-infected brainstems. β-Actin served as a loading control. The results shown in the histogram are the mean±s.d. for tissues from four control and four ts1-infected mice. *Significantly different from controls (P< 0.01).

Full size image

Increased Expression of GADD153/CHOP in Neurons Showing Degenerative Changes

Immunostaining for GADD153/CHOP was performed to localize GADD153/CHOP in infected brainstem. Figure 3 shows that GADD153/CHOP were virtually absent in neurons and glial cells of control mice (Figure 3a and c), but abundant in neurons in ts1-infected brainstems (Figure 3b and d). Notably, immunoreactivity was evident in both intact and visibly damaged neurons showing swelling, cytoskeletal collapse, vacuolization, and disintegration of nuclear and cytoplasmic membrane structures in the early and late stages of disease progression (Figure 3e and f). Double immunostaining for GADD153/CHOP and GFAP showed astrocytes in affected areas to be negative for GADD153/CHOP (Figure 3g).

Figure 3
figure 3

Increased expression of GADD153/CHOP protein in neurons of brainstems of ts1-infected mice. (a) GADD153/CHOP immunoreactivity (brown) in the brainstem from a control mouse. (b) GADD153/CHOP immunoreactivity in neurons of ts1-infected brainstem at 25 dpi. (c) Higher magnification view of the glial and neuronal cells in panel (a). Note the absent or very weak staining in neurons (arrows). (d) Higher magnification views of the boxed area in panel (b). Note the strong staining in neurons undergoing degenerative changes such as vacuolization (arrowheads) and the absence of immunoreactivity in neurons with normal appearance (arrows). (e) Higher magnification views of the boxed area in panel (d). (f) GADD153/CHOP immunoreactivity in neurons of ts1-infected brainstem in early stage (at 20 dpi). Note the strong cytoplasmic and nuclear staining in neurons without morphologic changes (arrows). (g) Double immunostaining for GADD153/CHOP (brown) and GFAP (red) in ts1-infected brainstem (at 20 dpi). Note the absence of GADD153/CHOP immunoreactivity in astrocytes showing GFAP immunoreactivity (arrowheads), alongside strong GADD153/CHOP staining in neurons (arrows). Bars indicate 200 μm.

Full size image

Activation of Caspase-12 and -3 Pathways in ts1-infected Brainstems

In response to ER stress, caspase-12 specifically cleaves and activates effector caspases such as caspase-3.20, 33, 34 Figure 4a and b show that brainstem tissues from ts1-infected mice had significantly higher levels of cleaved forms of caspase-12 (approximately 35–39 kDa) and caspase-12 than those from uninfected controls at 20 and 25 dpi (P<0.01). Figure 4a and b show that ts1-infected brainstem tissues also contained more cleaved caspase-3 at 20 and 25 dpi (P<0.01), being associated with decrease in amounts of caspase-3 at 25 dpi (P<0.01). As shown in Figure 4c, levels of caspase-12 mRNA were consistent with changes in amounts of caspase-12 protein (immunoblotting results in Figure 4a and b) at the corresponding time points. In contrast, levels of caspase-3 mRNA in ts1-infected brainstems were not significantly different from ones of uninfected controls.

Figure 4
figure 4

Expression of caspase-12 and -3 in brainstems of control and ts1-infected mice. (a) Representative Western blots showing levels of caspase-12 and cleaved caspase-12, caspase-3, and cleaved caspase-3 proteins in control and ts1-infected brainstems at 20 and 25 dpi. β-Actin acted as a loading control. (b) Quantitative analyses of levels of caspase-12, cleaved caspase-12, caspase-3, and cleaved caspase-3 proteins in control and ts1-infected brainstems. The results shown in the histogram are the mean±s.d. for tissues from four control and four ts1-infected mice. *Significantly different from controls (P<0.01). (c) Representative photographs of semiuantitative RT-PCR analyses for caspase-12 and -3 mRNAs in control and ts1-infected brainstems.

Full size image

Coexpression of Cleaved Caspase-3 with GADD153/CHOP in Neurons Undergoing Degenerative Changes

To determine whether increased GADD153/CHOP protein levels occur together with cleavage of caspase-3, double immunostaining for GADD153/CHOP and cleaved caspase-3 was performed (Figure 5). While no immunoreactivity for either GADD153/CHOP or cleaved caspase-3 was evident in brainstem neurons of control mice (Figure 5a), most of the neurons positive for cleaved caspase-3 immunostaining had moderate to strong nuclear GADD153/CHOP immunoreactivity at 20 and 25 dpi (Figure 5b–d).

Figure 5
figure 5

Colocalization of GADD153/CHOP and cleaved caspase-3 in neurons in brainstem section from ts1-infected mice. (a) Double immunostaining for GADD153/CHOP and cleaved caspase-3 in brainstem from a control mouse. (b) Double immunostaining to GADD153/CHOP (brown, nuclei) and cleaved caspase-3 (red, cytoplasm) in neurons undergoing degenerative changes in ts1-infected brainstem at 25 dpi. (c) Higher magnification views of the boxed area in panel (b). (d) Double immunostaining for GADD153/CHOP (brown, nuclei) and cleaved caspase-3 (red, cytoplasm) in neurons with normal appearance in ts1-infected brainstem at 20 dpi. Bars indicate 200 μm.

Full size image

Phosphorylation of CaMKII-α is Increased in Brainstems of ts1-infected Mice and Correlated with Induction of GADD153/CHOP in Neurons

Perturbation of Ca2+ homeostasis can elicit ER stress.35, 36, 37 One of the target kinases activated by intracellular Ca2+ increase is CaMKII.38, 39 Therefore, we examined whether phosphorylation of neuron-specific CaMKII-α is increased in brainstems of ts1-infected mice. As shown in Figure 6a and b, phosphorylated CaMKII-α was increased 2.9- and 5.1-fold in ts1-infected brainstems at 20 and 25 dpi, respectively, compared to controls (P<0.01). Immunostaining for phosphorylated CaMKII-α showed that neurons with degenerative changes in ts1-infected brainstems showed moderate to strong cytoplasmic staining for phosphorylated CaMKII-α (Figure 7b and d), compared to controls (Figure 7a and c). Most (>95%) of neurons positive for immunoreactive GADD153/CHOP were also positive for phosphorylated CaMKII-α (Figure 7e and f). We used transmission electron microscopy to determine whether intracellular Ca2+ increase is involved in early development of neuronal degeneration. As shown in Figure 7h, increased intracellular Ca2+ was illustrated by cytochemical staining for Ca2+ deposits in damaged mitochondria of neurons at early stage, compared to control (Figure 7g).

Figure 6
figure 6

Increased expression of phosphorylated CaMKII-α in brainstems of ts1-infected mice. (a) Representative Western blots showing levels of phosphorylated CaMKII-α (p-CaMKII-α) in control and ts1-infected brainstems at 20 and 25 dpi. β-actin served as a loading control. (b) Quantitative analysis of levels of p-CaMKII-α in control and ts1-infected brainstems. The results shown in the histogram are the mean±s.d. for tissues from four control and four ts1-infected mice. *Significantly different from controls (P < 0.01).

Full size image
Figure 7
figure 7

Increased expression of p-CaMKII-α protein and Ca2+ deposits in neurons of brainstems of ts1-infected mice. (a) p-CaMKII-α immunoreactivity (brown) in brainstem from a control mouse. (b) p-CaMKII-α immunoreactivity in ts1-infected brainstem at 25 dpi. (c) Higher magnification view of the neuronal cells (arrows) in panel (a). Note the absent or very weak cytoplasmic staining in neurons (arrows). (d) Higher magnification views of the neuronal cells in panel (b). Note the strong staining in neurons undergoing degenerative changes such as vacuolization (arrowheads) and the weak immunoreactivity in neurons with normal appearance (arrows). (e, f) Coexpression of GADD153/CHOP (brown, nuclei) and p-CaMKIIα (red, cytoplasm) in the neurons (arrowheads) in ts1-infected brainstem at 25 dpi (e) and 20 dpi (f). Bars indicate 200 μm. (g, h) Electron micrograph showing dense granular Ca2+ pyroantimonate deposits in mitochondria of neurons (arrows) from ts1-infected brainstem at 20 dpi (h). Note the absence of granular deposits in mitochondria of neurons (arrows) from control (g). Bars indicate 3 μm.

Full size image

Discussion

The present study shows that ts1 infection triggers activation of ER stress signaling pathways in the brainstem, as evidenced by increases in phosphorylation of PERK and eIF2-α, induction of GADD153/CHOP and GRP78, and activation of the ER-associated caspase-12. The concomitant increases in GADD153/CHOP, cleaved caspase-3, phosphorylated CaMKII-α, and intracellular Ca2+ in neurons implicate ER stress following intracellular Ca2+ increase in mechanisms leading to neurodegeneration in ts1-infected brainstems. Induction of these ER stress markers is correlated temporally with increases in viral burden and noxious cytokines, neuronal loss, and the development of neuroclinical signs in infected animals.4, 6, 8, 32

Particularly relevant to the present work are two recent studies describing roles of ER stress in neurodegeneration. Hetz et al40 demonstrated that ER stress and caspase-12 mediate neurotoxicity of pathological prion protein. Dimcheff et al41 also showed that another murine neurotropic virus, FrCasE, triggers ER stress in the brainstems leading to spongiform degeneration. Together with these recent publications, our results strongly support the idea that spongiform neurodegeneration is likely to be mediated by ER stress.

GADD153/CHOP has emerged recently as a ubiquitous regulator of neuronal death.42 Induction of GADD153/CHOP appears to be an early event contributing to neuronal dysfunction and death in ts1-infected brainstems. This conclusion is based on the observation that increased expression and nuclear translocation of GADD153/CHOP is evident early in the disease course in neurons without morphological changes and becomes more pronounced in neurons showing advanced degenerative changes in late stages.

What is the role of GADD153/CHOP in neuronal degeneration in ts1-infected brainstems? A previous study showed that GADD153/CHOP is a negative regulator of Bcl-2 expression.43 Also, GADD153/CHOP may sensitize cells to noxious stimuli by inducing the expression of pro-apoptotic genes such as Bax and caspases.44, 45 It is noteworthy that the Bcl-2 protein levels in brainstems of ts1-infected mice are decreased at late stages of disease (data not shown). However, the precise pro-cell-death targets of GADD153/CHOP transcription factor during ts1 infection of the brainstem remain to be identified.

In contrast to GADD153/CHOP, the chaperone GRP78 promotes protein folding and thus attenuates cellular death under stress conditions.46, 47 While GRP78 was found to be increased in early and late stages of disease in ts1-infected brainstems, there was no progressive increase in GRP78 mRNA and protein from 20 to 25 dpi. Similarly, previous studies with FrCasE- or scrapie-infected mice have shown that there is no clear correlation between the expression of GRP78 and the development of ER stress-mediated neurodegeneration.40, 41 These combined results indicate that GRP78 may be increased early after infection, but, under the conditions of prolonged ER stress in late stages of diseases, the induction of GRP78 may not be as efficient as those of other ER stress markers. Such levels of GRP78 induction may not be sufficient to protect neurons against ER stress in ts1-, FrCasE-, or scrapie-infected mice, since GRP78 overexpression protects cellular death under ER stress conditions by stabilizing Ca2+ homeostasis.47, 48

The activation of caspase-12 is a crucial event in ER stress-mediated cell death.20 In ts1-infected brainstem tissue, expression and activation of caspase-12 are increased, and this is accompanied by increases in cleavage activation of its downstream target, caspase-3. This cleaved caspase-3 is the caspase most frequently reported to participate in neuronal cell death. Interestingly, increase in caspase-12 protein was correlated with increased amounts of caspase-12 mRNA. This induction may be due to expression of interferon-γ induced during ts1 infection, as demonstrated by recent publications that either interferon-γ alone or in concert with TNF-α, is capable of activating the transcription of caspase-12 gene as well as the processing of caspase-12.49, 50

Previous studies have shown that the ER Ca2+ ATPase inhibitor Thapsigarin, or Ca2+ ionopores induce transcriptional activation of GADD153/CHOP by elevating intracellular Ca2+.35, 51 We therefore attempted to dissect the mechanism of induction of ER stress signaling pathway in ts1-infected brainstems by investigating Ca2+ signaling pathways. In this study, we found that phosphorylation of CaMKII-α is increased and that phosphorylated CaMKII-α is present, together with GADD153/CHOP and cleaved caspase-3, in most neurons in brainstems from both early and advanced stages of the disease (Figures 6 and 7). This result suggests that perturbation of Ca2+ homeostasis is a potential mechanism responsible for induction of ER stress in neurons. This notion is supported by electron microscopic evidence of increased intracellular Ca2+, manifested as Ca2+ precipitates in the mitochondria in neurons during early stages (Figure 7).

One key issue that should be addressed is what types of cells in brainstems are specially affected by ER stress. We have obtained immunohistological data showing the induction of ER stress markers in neurons, but not in glial cells (Figures 3, 5, and 7). Involvement of ER stress in neurons is further supported by observation of increase in intracellular Ca2+ (Figure 7). These results suggest that activation of ER stress signaling pathways in neurons are likely due to an indirect mechanism(s), since our previous work has revealed that ts1 virus infects glial cells, but not neurons.5 Therefore, the present study strongly indicates that ER stress-mediated spongiform neurodegeneration is likely to be caused by neurotoxic products secreted from glial cells by ts1 infection.

Among the elements present within the lesions of the ts1-infected brainstem, TNF-α and NO are probably prime candidates as triggers for ER stress in neurons, since both are upregulated in astrocytes by ts1 infection.6, 7 These agents may cause ER Ca2+ release in surrounding neurons, leading to activation of ER stress signaling pathways, as demonstrated by previous reports.22, 23, 24 Also, TNF-α may cause glutamate accumulation in the vicinity of neurons, which activates N-methyl-D-aspartate (NMDA) receptor-operated channels on the neuronal surface, causing Ca2+ influx and activation of CaMKII,52, 53 as shown in Figure 7. This notion is supported by recent reports that excitotoxicity associated with NMDA receptor induces GADD153/CHOP production, and activation and phosphorylation of CaMKII-α, leading to neuronal death.54, 55, 56

In conclusion, we show here that activation of ER stress signaling pathways constitutes a sequential and inter-related cytotoxic signaling cascade that causes neuronal injury in the ts1-infected brainstem. The potential importance of the Ca2+–ER stress link in the pathogenesis of spongiform encephalopathy by ts1 infection, particularly neuronal degeneration, is now evident. Furthermore, the data presented here support the idea that ER stress-mediated neurodegeneration is likely to be caused by neurotoxic products secreted from glial cells by ts1 infection.