Overall, our findings imply a beneficial immunomodulatory effect of MCMV infection in the TMEV infection MS model. MCMV prior to TMEV infection may therefore contribute to our understanding to the clinical observation that underlying CMV infection contributes to protection from MS. The overall reduction in the proportion of CD3+ cells and the observed increase in the proportion of Mac-1+ macrophages may contribute to this effect.
The role and significance of HCMV infection from the standpoint of MS development remains controversial. Similarly to most putative viral causes of MS, one key issue is the almost ubiquitous positivity of the average population to markers of these infections. A potential causative role for HCMV was suggested over 30 years ago based on primate experiments in which a strain of CMV was isolated from the brain and lymph nodes of a chimpanzee that developed paralysis after intracerebral inoculation with brain cell cultures derived from an MS case 
. However, most studies have been unable to confirm this purported HCMV association with MS. Meanwhile, a potential disease initiating role for EBV was suggested as far back as 1983 in a study where the HCMV infection rate and HCMV complement fixing antibody production was found to be lower in MS cases 
. However, a study on post-mortem brain tissue using PCR based detection techniques failed to find any statistically significant association between MS and common viral infections including HCMV and EBV 
. Meanwhile, a study in Norway found elevated titers to EBV but not to HCMV in MS cases 
. Recently, Ascherio reported a pathogenic role of EBV in MS whereas no similar association was found regarding HCMV 
. More recently, EBV-specific CD8+ T-cell responses were shown to be decreased in patients with clinically isolated syndrome (CIS). In contrast, there was no difference between categories for EBV-specific CD4+ T cell, or for HCMV-specific CD4+ and CD8+ T-cell responses 
. Intrathecal enrichment in EBV-, but not HCMV-specific CD8+ CTL was also reported in early MS patients by another group 
. EBV but not HCMV IgG antibody indexes were also increased in the CSF in this study. These studies overwhelmingly demonstrate that while EBV may be an important “trigger” of MS development, HCMV doesn't contribute to MS pathogenesis.
A surprising finding about HCMV was its potential protective role in MS, both from the standpoint of MRI and functional outcome measures, suggesting that HCMV infection in MS patients results in a beneficial modulation of the immune response 
. This previous study demonstrated that recent HCMV infection either by 1) primary infection or 2) secondary infection (since humans can be infected with multiple CMV strains), or 3) reactivation of latent virus leading to recent replication has occurred in those patients, as reflected by the increased HCMV antibody titers. Therefore, it is possible CMV-specific T cells were activated or cytokine induction occurred, which has had an immunomodulatory effect resulting in attenuated MS phenotype. It is clear that both HCMV and MCMV infection affect the responsiveness of T cells since CMV infected dendritic cells can modulate naive and antigen specific T cell responses 
. There is also evidence that HCMV may cause activation of immunomodulating and immune evasion mechanisms 
which could alter the adaptive immune processes involved in the pathogenesis of MS 
. The role of HCMV as an immunomodulating virus has been recognized in solid organ transplantation where it appears to contribute to the immunosuppressed state 
Infection of mice with MCMV reflects the infection of humans with HCMV in many respects. MCMV infects numerous tissues and cell types during the acute phase of infection and establishes a lifelong persistent or latent infection similar to HCMV 
. Similar to HCMV, MCMV encodes viral genes which function to evade or alter the host immune response 
. Some of these appear to contribute to the observed lifelong viral persistence, while others exploit immune cells that contribute to antiviral immunity 
. Both HCMV and MCMV inhibit MHC class I expression on infected cells 
, and MCMV has been shown to impair IFN-gamma induced MHC class II-dependent antigen presentation by macrophages 
. HCMV was also demonstrated to inhibit the induction of HLA class II antigens by IFN-beta dependent and independent molecules including ICAM-1 and VCAM-1, which are thought to be involved in MS pathogenesis 
. In addition, CMV-infected endothelial cells have the capability to induce IFN-beta production 
. Interferon beta represents the most commonly used disease modifying agent for relapsing forms of MS 
. Another possible mechanism to explain the effects of prior MCMV infection on TMEV-mediated MS development could result from modulation of the immune response to TMEV infection itself. In mouse infection studies with either MCMV or other viruses, prior infection of mice with one virus influences the immune response to other heterologous infections 
. Thus, prior MCMV infection in the TMEV model of MS could lead to attenuation of the MS-like disease as shown in our studies by modulation of the immune response to TMEV infection. In addition, cytokines released as part of an antiviral immune response can activate the hypothalamo-hypophyseal axis, resulting adrenal glucocorticoid release, which in turn provides strong negative feedback on the further synthesis and release of cytokines, and exerts an overall protective effect from the detrimental consequences of an overactive immune response 
. Lastly, as another potential, but at this stage purely speculative explanation to our observations of increased proportion of Mac-1+ cells, these cells may contribute to a more efficient elimination of TMEV infected CNS cells, and as such led to a better overall outcome. In addition, macrophages may exert immunosuppressive effects on T-cells, as commonly demonstrated in cancer models.
Given that our study was designed as a pilot project paving the way to future larger scale proposals, there are clear limitations to our data. These include the relatively low number of mice per group, which resulted in being underpowered from the standpoint of demonstrating significant MRI-based differences. In addition, due to the same limitation, only one time point was studied with FACS, and we only studied brain and not spinal cord samples - ideally both should be assessed given the prevalence of TMEV in the spinal cord at late time points. We utilized the most commonly studied immune cell markers in the FACS based studies; however, additional immune subset markers, cytokine assay, microarray studies would enable us to study this phenomenon in more details, and all the above are planned in future extensions of this study. The 2-week lag between MCMV and TMEV infections was chosen because MCMV virus titers peak in the salivary gland at 2 weeks and all other tissues are infected and some are even starting to be cleared from the tissue, indicating virus-specific immune response. However, it is possible that a different lag time may have resulted in enhanced immunomodulatory effects. We also did not demonstrate the effects of MCMV alone on the observed outcome measures, including there was no control group where SJL/J mice would be infected with MCMV alone. Histology time course analysis was also not done, but given that we clearly documented significant differences in motor performance, and motor performance in this model is determined by the extent of both demyelination and axonal loss, we anticipate that quantitative measures of the aforementioned would also have demonstrated significant differences.
In conclusion, in the studied chronic-progressive model of MS, MCMV infection prior to demyelinating disease induction resulted in an attenuated clinical phenotype. Overall, the mechanisms by which human or mouse CMV effectively results in reduced disease activity in MS or in the studied MS model remain unclear. However, this study recapitulates in a rodent model the clinical observation that underlying HCMV infection is protective in human MS. We plan to study the observed phenomenon in additional details in future extensions to this study, which will address all the limitations of the current study, and include additional time points and outcome measures. In our view, the presented new model system will enable us to gain insights into the beneficial immunomodulatory mechanism(s) associated with this common viral infection, and may pave the way to novel therapeutic strategies in MS.