|Home | About | Journals | Submit | Contact Us | Français|
Viral infections in the central nervous system (CNS) can lead to neurological disease either directly by infection of neurons or indirectly through activation of glial cells and production of neurotoxic molecules. Understanding the effects of virus-mediated insults on neuronal responses and neurotrophic support is important in elucidating the underlying mechanisms of viral diseases of the CNS. In the current study, we examined the expression of neurotrophin- and neurotransmitter-related genes during infection of mice with neurovirulent polytropic retrovirus. In this model, virus-induced neuropathogenesis is indirect, as the virus predominantly infects macrophages and microglia and does not productively infect neurons or astrocytes. Virus infection is associated with glial cell activation and the production of proinflammatory cytokines in the CNS. In the current study, we identified increased expression of neuropeptide Y (NPY), a pleiotropic growth factor which can regulate both immune cells and neuronal cells, as a correlate with neurovirulent virus infection. Increased levels of Npy mRNA were consistently associated with neurological disease in multiple strains of mice and were induced only by neurovirulent, not avirulent, virus infection. NPY protein expression was primarily detected in neurons near areas of virus-infected cells. Interestingly, mice deficient in NPY developed neurological disease at a faster rate than wild-type mice, indicating a protective role for NPY. Analysis of NPY-deficient mice indicated that NPY may have multiple mechanisms by which it influences virus-induced neurological disease, including regulating the entry of virus-infected cells into the CNS.
The early innate immune response to virus infection in the central nervous system (CNS) plays an important regulatory role in controlling both viral infection and pathogenesis. The neuroinflammatory response can limit virus replication through production of type I interferons and recruitment of virus-specific T cells to the CNS (5, 9, 12, 15, 19). However, the neuroinflammatory response can also lead to chronic gliosis, the production of cytokines that are toxic to neurons, and the recruitment of virus-infected cells to the CNS (6, 8, 18, 35). Understanding the relationship between the innate immune response and viral disease is essential in order to manipulate this response to control virus infection in the CNS.
To better understand the role of the innate immune responses in viral pathogenesis in the CNS, we have utilized a mouse model of polytropic retrovirus infection. In this model, neuropathogenesis is indirect, since the polytropic retroviruses do not productively infect neurons. Instead, the viruses predominantly infect macrophages and microglia in the CNS (32). Despite severe neurological disease development following polytropic retrovirus infection, the only histologic changes observed in the brain are the activations of microglia and astrocytes (31). In addition, we have found high levels of proinflammatory cytokines and chemokines in brain tissue from infected mice, including tumor necrosis factor (TNF); interleukin 1 alpha (IL-1α), IL-1β, and IL-6; and the chemokines chemokine ligand 2 (CCL2/MCP-1), CCL3 (MIP-1α), CCL4 (MIP-1β), CCL5 (RANTES), and chemokine (C-X-C motif) ligand 10 (CXCL10/IP-10) (28). Studies with different chemokine receptors and cytokine-deficient mice demonstrated that at least two of these proinflammatory cytokines, CCL2 and TNF, can contribute to retrovirus-induced neurological disease (26, 27). However, neither of these molecules was necessary for neurological disease for all of the neurovirulent polytropic retroviruses studied, suggesting that other host factors contribute to retroviral pathogenesis.
Analysis of the envelope protein of the neurovirulent polytropic retrovirus identified key residues in the envelope protein that influence the ability of the virus to induce neurological disease (28). In this study, we utilized neurovirulent and nonneurovirulent chimeric viruses that differ by only a few amino acid residues in these envelope regions to identify host response factors whose expression correlated with neurovirulence. We also utilized two different mouse strains, Inbred Rocky Mountain White (IRW) and 129S6, to confirm that expression of these host response factors is consistently induced or suppressed during neurovirulent virus infection. We determined that, although a number of host response genes are induced by polytropic retrovirus infection of the CNS, the expression of several of these factors correlated only with neuroinvasion and was not strongly correlative of neurovirulence. However, we have identified a neurotrophin, neuropeptide Y (NPY), whose expression strongly correlates with neurovirulence. We found that NPY had a protective influence on retroviral neuropathogenesis and examined the mechanisms by which NPY influences retrovirus infection of the CNS.
Inbred Rocky Mountain White mice were maintained at Rocky Mountain Laboratories or Louisiana State University. 129S6 mice were purchased from Taconic. 129S1 (129SvImJ) and 129S1 Npy−/− mice were purchased from Jackson Laboratories. Single nucleotide polymorphism (SNP) analysis of 1,449 SNPs was completed on DNA from 129S1 and 129S1 Npy−/− mice by Taconic (Hudson, NY) and confirmed at least 99.8% homology of backgrounds (equivalent to 10 backcrossings) between 129S1 and 129S1 Npy−/− mice. All of the animal procedures were conducted in accordance with the Louisiana State University Animal Care and Use Committee guidelines or the Rocky Mountain Laboratories Animal Care and Use Committee guidelines.
The construction of chimeric virus clones for BE has been previously published (16, 28). Virus stocks were prepared from the supernatants of confluent cultures of infected Mus dunni fibroblasts. Virus titers were determined by focus-forming assays using the envelope-specific monoclonal antibodies 514 and 720 (32). Mice were infected within 24 h of birth by intraperitoneal injection with 100 μl of cell culture supernatant containing 104 focus-forming units (FFU) of virus. Mice were observed daily for clinical signs, i.e., hyperexcitability, followed by the development of multiple severe seizures and/or ataxia, which precedes death by 1 or 2 days. When mice developed multiple severe seizures or ataxia, they were scored as clinical and euthanized and tissues were removed. No severe seizures or ataxia was observed in uninfected wild-type or uninfected Npy−/− mice.
Infected and uninfected mice were anesthetized by deep-inhalation anesthesia, followed by axillary incision and cervical dislocation. Brains were removed and cut into two halves by midsagittal dissection. One half was immediately frozen in liquid nitrogen and stored at −80°C for molecular analysis, and the other half was fixed in 10% neutral buffered formalin for 48 h prior to processing for histological analysis.
Total RNA was extracted from frozen tissue using Trizol reagent (Invitrogen) by following the manufacturer's instructions. RNA was treated with DNase I and repurified using an RNA cleanup kit (Zymo Research). cDNA was generated from RNA samples using an iScript reverse transcription (RT) kit (Bio-Rad). Primers for real-time PCR analysis are shown in Table Table1.1. All primers used for real-time analysis were designed using Primer3 software (33). A BLAST search of primer sequences against the National Center for Biotechnology Information (NCBI) database was performed to confirm that all primer pairs were specific for the gene of interest and that no homology to other genes was present. PCRs were prepared using SYBR green mix with Rox (Bio-Rad) in a 10-μl volume with approximately 10 ng of cDNA and 1.8 μM forward and reverse primers. Samples were run in triplicate on an ABI Prism 7900 sequence detection system (Applied Biosystems). Analysis of dissociation curves was used to confirm the amplification of a single product for each sample. Confirmation of a lack of DNA contamination was achieved by analyzing samples that had not undergone reverse transcription. Gene expression was quantified by the cycle number at which each sample reached a fixed fluorescence threshold (CT). To control for variations in RNA amounts among samples, data were calculated as the difference between the CT value (log2) for the housekeeping gene Gapdh and that for the gene of interest for each sample (ΔCT = CT for Gapdh − CT for the gene of interest). Data are presented as percentages of Gapdh expression for each gene of interest.
RNA samples were isolated and treated with DNase I as described above. First-strand cDNA was synthesized using 100 ng of cleaned up RNA and analyzed with a mouse neurotrophin and receptor PCR array or a mouse neurotransmitter receptor and regulator PCR array according to the manufacturer's instructions (SABiosciences) on an ABI Prism 7900 sequence detection system (Applied Biosystems). A total of 84 genes related to the PCR array were analyzed in a 384-well format. Genes for the PCR array are shown in Table Table2.2. The CT values were analyzed using the RT2 Profiler PCR array data analysis template (SABiosciences). The samples were analyzed only if the test passed all the quality controls, including RT efficiency and lack of DNA contamination. Data were calculated as fold differences between values for the treatment groups and the mock-treated groups.
The half of the brain tissue that was formalin fixed was further divided into four coronal sections, embedded in paraffin, and cut in 4-μm sections. Slides were rehydrated, and antigen retrieval was performed by incubating slides in 0.018 M citric acid and 8.2 μM sodium citrate dihydrate, pH 6.0, for 20 min at 120°C. Slides were then incubated in 0.2% fish skin gelatin (FSG) (Sigma-Aldrich, St. Louis, MO) in phosphate-buffered saline (PBS). Sections were blocked for a minimum of 30 min with blocking solution consisting of 2% donkey serum (Sigma), 1% bovine serum albumin (Sigma), 0.05% FSG, 0.1% Triton X-100 (Sigma), and 0.05% Tween 20 (Bio-Rad) in PBS. Slides were incubated in primary antibodies at 4°C overnight. For the detection of virus-infected cells, a 500-fold dilution of goat polyclonal anti-gp70 was used (10). Other primary antibodies included rabbit polyclonal anti-neuropeptide Y IgG (Chemicon), mouse monoclonal anti-NeuN IgG (Millipore), rabbit anti-Iba1 IgG (Wako), rabbit anti-active caspase 3 IgG (Promega), and rabbit anti-Olig2 IgG (ProteinTech). Primary antibodies were detected using Alexa Fluor 594-conjugated donkey anti-rabbit IgG, AlexaFluor488-conjugated goat anti-mouse IgG, Alexa Fluor 488-conjugated chicken anti-goat IgG, biotin-labeled goat anti-rabbit IgG, and Alexa Fluor 594 conjugated to streptavidin (Invitrogen). All sections were counterstained with 100 ng/ml DAPI (4′,6-diamidino-2-phenylindole; Molecular Probes) for 20 min. Slides were mounted with ProLong Gold antifade reagent (Molecular Probes). Additional slides were incubated with individual primary antibodies or no primary antibodies prior to incubation with secondary antibodies to confirm the lack of nonspecific staining or cross-reactivity. Slides from NPY-deficient mice were used as a negative control for anti-NPY staining to confirm specificity. Digital images were captured using NIS Elements software (Nikon) and compiled using Canvas XI software (ACD Systems). For counting positive cells in each field, images were captured, using NIS Elements software, of the appropriate regions of tissue in a blind fashion. The number of positive cells per image per 200× field of view was recorded. Multiple images from at least three mice per group were used for analysis.
Colorimetric immunohistochemistry was utilized for the detection of virus envelope protein and glial fibrillary acidic protein (GFAP) expression. Primary antibodies were diluted 5,000-fold for goat polyclonal anti-gp70 or 1,000-fold for rabbit polyclonal anti-GFAP (Dako). Slides were then incubated with biotinylated anti-goat or anti-rabbit secondary antibody (Vector Laboratories) and horseradish peroxidase (HRP)-conjugated streptavidin with AEC (3-amino-9-ethylcarbazole; Biogenex). Tissue sections from 6 to 10 mice per group were analyzed for virus envelope and GFAP, as well as being used for hematoxylin and eosin (H&E) staining and analysis. Additional slides were incubated with individual primary antibodies or no primary antibodies prior to incubation with secondary antibodies to confirm the lack of nonspecific staining or cross-reactivity. For detection of terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end labeling (TUNEL)-positive cells, a NeuroTacs II detection kit (R&D Systems) was used by following the manufacturer's instructions.
Tissue sections were scored for vacuoles (lesion score) and level of virus infection (virus score) in a blind study. Scales to rank severity were determined prior to analysis. Tissue samples from uninfected and BE-infected 129S1 and 129S1 Npy−/− mice were scored as to the severity of vacuoles in the internal capsule. A scale of 0 to 5 was used: 0, no vacuoles; 1, 1 to 5 undefined or small vacuoles; 2, multiple undefined or small vacuoles; 3, 1 to 10 large vacuoles with clearly defined edges; 4, 10 to 20 large defined vacuoles; 5, over 50 large defined vacuoles. Tissue samples from BE-infected 129S1 and 129S1 Npy−/− mice were scored as to the presence of virus in the internal capsule. A scale of 0 to 5 was used: 0, no virus-positive cells; 1, 1 to 10 positive cells; 2, 10 to 20 positive cells; 3, 20 to 50 positive cells; 4, 50 to 100 positive cells; 5, over 100 positive cells.
Since polytropic retrovirus infection can lead to the development of clinical neurological disease, we analyzed whether polytropic retrovirus infection led to the alteration of expression of neurotrophin genes, neurotransmitter genes, and related genes. We utilized two real-time PCR arrays that analyzed 84 genes per array (Table (Table2).2). For these initial studies, we used the neurovirulent polytropic retrovirus BE, which induces strong clinical signs of seizures and ataxia in infected mice starting approximately 20 days postinfection (dpi). Total RNA was isolated from whole brain tissue of uninfected or virus-infected 129S6 mice at 20 to 22 dpi, a time point when mice first start showing signs of severe ataxia and seizures. Surprisingly, of the 168 genes analyzed, there were only 7 genes whose levels of expression differed by more than 2-fold between the uninfected and BE-infected mice and were significantly (P < 0.05) different between the two groups (Fig. (Fig.1,1, gray areas). These genes were the corticotropin-releasing hormone binding protein gene (Crhbp), the melanocortin 2 receptor gene (Mc2r), the neuropeptide Y gene (Npy), the prokineticin receptor 2 gene (Prokr2), the transcription factor gene Myc, and the genes encoding proinflammatory cytokine interleukin 6 (Il-6) and leukemia inhibitory factor (Lif).
To confirm that the mRNA levels of the genes identified by array analysis were affected by BE infection, quantitative real-time RT-PCR analysis was completed. Analysis of the appropriate gene was performed using two different sets of primer pairs for each gene (data not shown). mRNA expression of the above-identified genes confirmed mRNA upregulation following BE infection for Crhbp, Il6, Lif, Mc2r, Myc, and Npy (data not shown). Prokr2 mRNA, which was downregulated in BE-infected mice as determined by array analysis, was not significantly downregulated in BE-infected wild-type mice compared to uninfected mice, as determined by real-time RT-PCR analysis (data not shown).
To determine if mRNA expression was consistently upregulated during neurovirulence, we analyzed the mRNA expression of the above genes in another BE-susceptible mouse strain, IRW. BE infection of IRW mice results in clinical disease similar to that in 129S6 mice. Of the identified genes, only Crhbp, Il6, Lif, and Npy had mRNA expression levels significantly upregulated following BE infection in IRW mice (Fig. (Fig.2).2). Thus, the upregulation of Mc2r and Myc mRNA did not correlate with neurovirulence in multiple mouse strains, while Crhbp, Il6, Lif, and Npy mRNA upregulation was correlative.
The upregulation of gene expression may be a response to virus infection per se and not a component of disease pathogenesis. To determine which of the identified genes correlated with neurovirulence rather than just virus infection, we utilized chimeric viruses that were neuroinvasive but differed in neurovirulence. Three chimeric viruses were analyzed, one that induces disease comparable to BE infection (BE-2) and two viruses that do not induce substantial clinical disease (BE-6 and BE-8) (28). Of the four genes analyzed, Npy mRNA expression correlated the most strongly with neurovirulence, as it was upregulated at significant levels in IRW mice infected with the BE-2 virus, but not with BE-6 or BE-8 virus (Fig. (Fig.33 D). In contrast, neither Il-6 nor Lif mRNA expression correlated with virulence, as there was not clear separation between virulent and nonvirulent virus strains (Fig. 3B and C). Crhbp mRNA levels were induced by two of the three virus infections, with the highest increase observed during BE-2 infection (Fig. (Fig.3A).3A). Thus, increased Npy and Crhbp mRNA expression was the primary correlate of neurological disease in two different strains of mice (IRW and 129S6) infected with two different neurovirulent viruses (BE and BE-2).
Neuropeptide Y is a 36-amino-acid peptide that is produced at high levels in the nervous system and also produced by a number of immune cells (29, 34). To determine the cellular source of NPY during BE infection, we examined tissue sections from uninfected and BE-infected mice by immunohistochemistry. NPY-positive cells were occasionally found in the cortices of uninfected mice (Fig. (Fig.44 A). In BE-infected mice, an increase in NPY-positive cells was observed (Fig. (Fig.55 A), with positive cells located primarily in the cortex and hippocampal areas (Fig. (Fig.4B;4B; data not shown). Dual staining for virus and NPY demonstrated that NPY-positive cells were located in areas of virus infection but were not positive for virus (Fig. (Fig.4C).4C). Dual staining for NPY and the neuronal nuclear marker NeuN indicated that the NPY-positive cells were neurons (Fig. (Fig.4D).4D). Thus, BE infection appears to induce NPY expression by neurons.
Two of the main clinical signs exhibited by BE-infected mice are severe seizures and ataxia. Interestingly, NPY-deficient mice spontaneously develop mild seizures at 6 to 8 weeks of age, indicating a role for NPY in protection from seizure development (3, 14). To examine the role of NPY in retroviral pathogenesis, we utilized mice deficient in NPY. These mice are on a 129S1 background, so they are susceptible to BE-induced pathogenesis, but at a much reduced rate compared to either 129S6 or IRW mice (our unpublished observations). Comparison of the development of clinical signs of multiple severe seizures, ataxia, or death in 129S1 wild-type mice with that in 129S1 NPY-deficient mice demonstrated a substantial difference in the development of clinical disease, with only 49% of wild-type 129S1 mice developing disease compared to 100% of the 129S1 NPY-deficient mice (Fig. (Fig.5B).5B). Thus, NPY appears to play a protective role, limiting the development of seizures and ataxia following BE infection. No clinical signs of multiple severe seizures or ataxia were noted in uninfected 129S1 or 129S1 NPY-deficient mice.
Histologic examination of brain tissue from wild-type and NPY-deficient mice demonstrated no substantial differences in pathology. The only notable pathology associated with BE infection was small regions of mild myelin sheath vacuolation, located primarily in the white matter tracts of the internal capsule at 21 dpi in mice with or without clinical disease (Fig. (Fig.66 A and B). The vacuoles were associated with areas of virus infection (Fig. 6C and D). Interestingly, these vacuoles were detected by 14 dpi in NPY-deficient mice, but not in wild-type mice (Fig. 6E and F). Analysis of the severity of the lesions in the internal capsule region showed a significant difference in the severity of vacuolation at 14 dpi, but not 21 dpi, between wild-type and NPY-deficient mice (Fig. (Fig.77 A). Thus, the increased development of neurological disease in NPY-deficient mice correlated with an early onset of histological changes in NPY-deficient mice. The presence of these vacuoles was generally associated with areas of high virus infection in the internal capsule (Fig. 6G and H).
To examine if the difference in spongiform lesions between wild-type and NPY-deficient mice was influenced by cell tropism, sections were costained for virus envelope protein and Iba1, a marker for microglia and macrophages. In both wild-type and NPY-deficient mice, virus-infected macrophages were detected near areas of vacuolation (Fig. 7C and D, white arrowheads), although these cells did not appear to have vacuoles. Additional virus-infected cells that were not positive for Iba1 were detected (Fig. 7C and D, yellow arrows). Some of these virus-infected cells were blood capillary endothelia (Fig. 7E and F, white arrowheads), while others were oligodendrocytes, as detected by Olig2 staining (Fig. 7E and F, yellow arrows). Analysis of the number of virus-infected oligodendrocytes in the internal capsule showed a slight, but not significant, increase in the number of infected oligodendrocytes in NPY-deficient mice compared to wild-type controls at 14 dpi (Fig. (Fig.7B).7B). These data indicate that microglia, endothelia, and oligodendrocytes in the internal capsule were infected with virus, with no significant difference in cell tropism between 129S1 and 129S1 NPY-deficient mice. Cells containing vacuoles were occasionally positive for virus envelope (Fig. (Fig.7D,7D, pink arrow) and generally positive for Olig2 (Fig. (Fig.7E,7E, pink arrows), suggesting that oligodendrocytes were the primary cells with vacuolation.
One of the strong correlates with polytropic retrovirus infection in the CNS is the production of proinflammatory cytokines and the activation of glial cells (25, 28). Analysis of RNA from brain tissue at 14 and 21 dpi demonstrated similar increases in mRNA expression of the astrocyte activation marker gene Gfap, as well as the cytokine genes Tnf and Il6, in both wild-type and NPY-deficient BE-infected mice (Fig. (Fig.8).8). The lack of difference in astrocyte activation was confirmed by immunohistochemical analysis of brain tissue from wild-type and NPY-deficient mice (data not shown). A significant difference in Cxcl10 mRNA expression between wild-type and NPY-deficient mice at 14 dpi was noted but was not observed at 21 dpi. Thus, NPY deficiency did not appear to have a pronounced effect on the neuroinflammatory response in the CNS, with the exception of inducing higher levels of Cxcl10 mRNA at an early time point postinfection. This suggests that NPY does not inhibit disease induction by suppressing the proinflammatory response to virus infection.
Analysis of active caspase 3 expression indicated that NPY deficiency did not influence neuronal apoptosis at either time point, with only a few random active caspase 3-positive cells detected in the cortices of BE-infected 129S1 or 129S1 NPY-deficient mice at either 14 dpi (data not shown) or 21 dpi (Fig. 4E and F). Thus, the lack of NPY did not lead to the onset of substantial neuronal apoptosis. Only a few proapoptotic active caspase 3-positive cells were detected in the internal capsules of 129S1 or 129S1 NPY-deficient mice at 14 (data not shown) and 21 dpi (Fig. 4G and H). These cells were not positive for active caspase 3 in the nucleus and did not appear to contain vacuoles. This suggests that the vacuolated cells were not undergoing apoptosis. Similar results were observed using a TUNEL assay to detect apoptotic cells (data not shown).
The primary histological difference between wild-type and NPY-deficient mice was the presence of mild vacuolation in NPY-deficient mice a week earlier than that in wild-type mice (Fig. 6E and F). This was associated with stronger virus staining in the internal capsule area at 14 dpi (Fig. 6G and H), although relative scoring of virus levels in this region by blind analysis did not reveal a statistically significant difference (Fig. (Fig.99 A). No observable difference in virus infection was detectable in other areas of virus infection within the brain (Fig. (Fig.10).10). To quantitatively examine whether NPY deficiency influenced virus entry or replication in the CNS, we analyzed viral RNA levels using quantitative real-time PCR. A significant difference in virus levels at 14 dpi was observed, with higher levels of virus gag RNA in NPY-deficient mice brain tissue than in wild-type controls (Fig. (Fig.9B).9B). This difference was not due to a generalized difference in virus levels between wild-type and NPY-deficient mice since splenic virus levels were comparable (Fig. (Fig.9C).9C). No differences were observed at 21 dpi. Thus, deficiency in NPY appears to influence virus levels early during infection, which correlates with increased susceptibility of NPY-deficient mice to BE-induced neurological disease.
In the current study, increased mRNA expression of Npy was found as a strong correlate to retroviral virulence in the CNS. Increased Npy mRNA expression was found in two separate strains of mice following retrovirus infection and was induced by two neurovirulent viruses, but not by two nonneurovirulent viruses. Mice deficient in NPY had increased incidence and onset of neurological disease. The mechanism by which NPY affects retroviral pathogenesis in the CNS appears to be 2-fold. NPY deficiency altered virus levels in the CNS during the early stages of virus infection, indicating that NPY influences virus infection in the CNS. Additionally, NPY was expressed by neurons during the later stages of virus infection, suggesting that NPY may also influence the ability of neurons to respond to virus-mediated insults.
NPY deficiency enhanced early virus infection in the CNS, particularly in the internal capsule (Fig. (Fig.99 and and10).10). One mechanism by which NPY may affect virus levels is by limiting the recruitment and/or infiltration of virus-infected macrophages into the brain. NPY has multiple reported effects on macrophages, including inhibiting chemotaxis and modulating phagocytosis, oxidative burst, and tumor growth factor beta production (11, 13, 38). It is possible that NPY expression in the periphery affects macrophage activation and limits the migration of infected cells to the brain. Alternatively, NPY may play a role in the infection of brain capillary endothelial cells or in the response of endothelial cells to virus-infected cells from the periphery. NPY can influence cerebral blood flow, and brain capillary endothelial cells express at least one of the NPY receptors (17). Since upregulation of NPY in the CNS is not observed until the late stages of infection, basal levels of NPY in the periphery or the CNS may be responsible for the effect of NPY on virus recruitment to the CNS.
NPY may also influence disease by affecting oligodendrocytes. There was a detectable, but not significant, difference in virus-infected oligodendrocytes between NPY-deficient mice and wild-type controls (Fig. (Fig.7B).7B). Oligodendrocytes have previously been shown to be one of the cell types infected by polytropic retroviruses, although the infection of these cells is relatively low compared to that of microglia or endothelia (32). Infected or uninfected oligodendrocytes may be the primary cell type responsible for the lesions in the internal capsule since several of cells with vacuoles could be identified as oligodendrocytes (Fig. (Fig.7E,7E, pink arrows). Although vacuolation of oligodendrocytes is not a hallmark of polytropic retrovirus infection, it is one of the hallmarks of ecotropic retrovirus infection, with widespread lesions throughout the brain, including the brain stem (2, 30). Recent studies with ecotropic viruses indicate that vacuolation of oligodendrocytes may be due to viral-envelope-induced endoplasmic reticulum (ER) stress response (7). It is possible that a similar stress response is induced by polytropic retrovirus of the CNS, but only in specific areas of intense viral replication. It remains unclear what potential influence these vacuoles have on BE pathogenesis since the vacuolation was also observed in multiple mice that did not show signs of clinical disease, including wild-type 129S1 mice (data not shown).
NPY was upregulated by neurons at 21 dpi, suggesting that NPY may also influence the later stages of disease development in the CNS. NPY plays an important role in inhibiting excitatory neurotransmission and can suppress kainic acid-induced seizures (21, 36). Seizures are one of the clinical signs of BE-induced neurological disease. The increased expression of NPY by neurons during BE infection suggests that NPY production may be a protective response by neurons during virus infection of the CNS. Since NPY expression is induced following neurovirulent virus infection, but not avirulent virus infection (Fig. (Fig.3),3), it is probable that the upregulation of NPY by neurons is a direct response to the insult induced by neurovirulent virus infection. In the absence of NPY, neurons may be more susceptible or unable to respond to virus-induced insults, leading to increased susceptibility of neurons and earlier onset of clinical disease. It cannot be ruled out that seizures induced by NPY deficiency and seizures induced by retrovirus are two separate neurological conditions that combine to result in increased incidence and kinetics of retrovirus-induced disease. However, the upregulation of NPY during virus infection and its correlation with neurological disease suggest that NPY is involved in mediating retroviral pathogenesis rather than being a completely separate pathway.
NPY binds six distinct receptors in the mouse, Y1 through Y6. Neurons express receptors Y1, Y2, and Y5, all of which have been associated with a protective role in kainic acid-induced seizures (21, 23, 24, 36, 37). The Y1, Y2, and Y5 receptor are also expressed by macrophages, with different functions for these receptors in the macrophage response to NPY (11). Only one receptor, Y1, is expressed by capillary endothelia (1, 17). Surprisingly, mRNA expression levels of Y1, Y2, and Y5 were not altered during BE infection (Table (Table2).2). The Y6 receptor, which is functional in mice despite a lack of function in humans or rats, was inconsistently upregulated in some BE-infected mice but not others (Fig. (Fig.1;1; Table Table2),2), suggesting that its expression may not be important in regulating disease pathogenesis. The overlapping functions of Y1, Y2, and Y5 receptors in regulating excitatory neurotransmission as well as regulating macrophage function may make it difficult to fully understand the mechanisms by which NPY influences viral pathogenesis.
Changes in NPY expression have been associated with a number of maladies ranging from high blood pressure to seizure induction. While NPY has been found to be upregulated during viral infections of the CNS, including HIV and Borna disease virus (4, 22), the impact of NPY expression on viral pathogenesis had not previously been explored. Although the full mechanisms by which NPY influences virus-induced neurological disease remain unclear, the current study clearly demonstrates that NPY plays a protective role during retroviral pathogenesis in the CNS and suggests that future studies of NPY as a therapeutic in treatment of virus-induced neurological orders may be warranted.
We thank Sue Priola, John Portis, Jim Striebel, Leonard Evans, and Susan Pourciau for critical reading of the manuscript. We also thank Dan Long and Rebecca Rosenke for technical assistance with immunohistochemistry.
This work was supported in part by the Intramural Research Program of the NIH and in part by National Center for Research Resources grant IP20RR020159.
Published ahead of print on 11 August 2010.