**4. Reactivation mechanisms of persistent MV**

It is known that persistent MV infection is asymptomatic but can eventually result in SSPE [2]. The latent MV should be reactivated at the onset of disease, resulting in clinical signs of SSPE (Figure 1C). However, the molecular mechanisms of MV persistence and reactivation are yet to be elucidated.

**Heat shock protein 72 (hsp72).** One potential molecule involved in MV reactivation is hsp72. Hsp72 binds to two conserved motifs in the variable tail of the N protein, known as box 2 (amino acids 489–506) and box 3 (amino acids 517–525) [40]. The tail of the N protein is within the same area where the XD domain of the P protein (amino acids 459– 507) binds to the N protein [41]. *In vivo* models using mice expressing hsp72, or hyperthermal preconditioned mice, have revealed that hsp72 levels can serve as a host deter‐ minant of viral neurovirulence in mice. This indicates the direct influence of hsp72 on viral gene expression [42, 43]. Hsp72 induction by some type of reactivation event might enhance the replication of persistent MV in the CNS, resulting in the onset of clinical symptoms. Accumulation of the H protein inside the cell during persistent MV infection might be such a reactivation event, as antibodies against the MV can decrease cell sur‐ face expression of viral glycoproteins, which has been suggested to contribute to the es‐ tablishment of MV persistence [44, 45]. Indeed, overexpression of the H protein leads to induction of hsp72 (Figure 2).

**Figure 1.** A model for the pathogenesis of persistent MV infection. (A) Acute infection. MV enters the CNS and infects neurons and oligodendrocytes. (B) Persistent infection. MV establishes a persistent infection in the CNS. MV replica‐ tion is attuned to the host cells, with minor or reversible modifications of the cells. Minor or reversible modifications, such as alterations in lipid metabolism, in MV-infected cells might be involved in a progressive infection. (C) Reactiva‐ tion. Some reactivation events stimulate the latent MV, leading to rapid replication in the CNS. (D) Demyelination. Re‐ activated MV destroys host cells, including oligodendrocytes, and drives damaging inflammatory responses, resulting in demyelination. Damaging resulting from MV infection can lead to a spreading of epitopes that generate autoim‐ mune responses. The oligoclonal IgG found in the SSPE brain and the CSF, which is directed against MV, possibly crossreacts with myelin proteins. Activated autoreactive T cells, or T cells activated by viral antigens can cross the bloodbrain barrier and enter the brain parenchyma. These infiltrating inflammatory cells induce extensive lesions in the CNS.

Pathogenesis of Encephalitis Caused by Persistent Measles Virus Infection

http://dx.doi.org/10.5772/54434

255

doplasmic reticulum (ER) stress and apoptosis in PBMCs and human lung epithelial cells [30, 32]. All these molecules affected in MV-infected cells might be involved in SSPE pathogenesis. As an example, long-term administration of IFNs is one type of SSPE ther‐ apy [33]. NF-κB may be a determinant of multiple sclerosis (MS) susceptibility, a chronic demyelinating disease of the CNS in humans [34]. As glial cells appear to be vulnerable to ER stress, altered expression of the molecules involved in ER stress can perturb myeli‐ nation by oligodendrocytes [35]. Apoptotic processes have also been suggested to con‐ tribute to MS, where local tissue damage involves apoptosis of oligodendrocytes and

**Lipid metabolism in cells persistently infected with MV.** Most studies examining gene ex‐ pression in MV-infected cells have been performed in non-neuronal cells. Because modula‐ tion in gene expression is cell-type dependent [37], studies using neuronal cells are more informative. The molecules affected during persistent infection might be different from those in the acute infection. Two studies using neuronal cells persistently infected with MV revealed alterations in lipid metabolism, such as decreased cholesterol synthesis and im‐ paired β-oxidation, that were associated with MV persistence [38, 39]. Myelination is a com‐ plex process that requires a precise stoichiometry for gene dosage, along with protein and lipid synthesis. An alteration in lipid metabolism during persistent MV infection would af‐

It is known that persistent MV infection is asymptomatic but can eventually result in SSPE [2]. The latent MV should be reactivated at the onset of disease, resulting in clinical signs of SSPE (Figure 1C). However, the molecular mechanisms of MV persistence and reactivation

**Heat shock protein 72 (hsp72).** One potential molecule involved in MV reactivation is hsp72. Hsp72 binds to two conserved motifs in the variable tail of the N protein, known as box 2 (amino acids 489–506) and box 3 (amino acids 517–525) [40]. The tail of the N protein is within the same area where the XD domain of the P protein (amino acids 459– 507) binds to the N protein [41]. *In vivo* models using mice expressing hsp72, or hyperthermal preconditioned mice, have revealed that hsp72 levels can serve as a host deter‐ minant of viral neurovirulence in mice. This indicates the direct influence of hsp72 on viral gene expression [42, 43]. Hsp72 induction by some type of reactivation event might enhance the replication of persistent MV in the CNS, resulting in the onset of clinical symptoms. Accumulation of the H protein inside the cell during persistent MV infection might be such a reactivation event, as antibodies against the MV can decrease cell sur‐ face expression of viral glycoproteins, which has been suggested to contribute to the es‐ tablishment of MV persistence [44, 45]. Indeed, overexpression of the H protein leads to

neurons [36].

254 Encephalitis

fect the maintenance of myelin in the CNS.

are yet to be elucidated.

induction of hsp72 (Figure 2).

**4. Reactivation mechanisms of persistent MV**

**Figure 1.** A model for the pathogenesis of persistent MV infection. (A) Acute infection. MV enters the CNS and infects neurons and oligodendrocytes. (B) Persistent infection. MV establishes a persistent infection in the CNS. MV replica‐ tion is attuned to the host cells, with minor or reversible modifications of the cells. Minor or reversible modifications, such as alterations in lipid metabolism, in MV-infected cells might be involved in a progressive infection. (C) Reactiva‐ tion. Some reactivation events stimulate the latent MV, leading to rapid replication in the CNS. (D) Demyelination. Re‐ activated MV destroys host cells, including oligodendrocytes, and drives damaging inflammatory responses, resulting in demyelination. Damaging resulting from MV infection can lead to a spreading of epitopes that generate autoim‐ mune responses. The oligoclonal IgG found in the SSPE brain and the CSF, which is directed against MV, possibly crossreacts with myelin proteins. Activated autoreactive T cells, or T cells activated by viral antigens can cross the bloodbrain barrier and enter the brain parenchyma. These infiltrating inflammatory cells induce extensive lesions in the CNS.

ces and direct effects on the proliferation of lymphocytes are reportedly implicated with the immunosuppression. In contrast, a persistent brain infection leads to a hyperimmune anti‐ body response, a pathogenic feature of SSPE [10, 11]. For example, there are extremely high titers of neutralizing antibodies in the serum and CSF against viral structural proteins. The

Pathogenesis of Encephalitis Caused by Persistent Measles Virus Infection

http://dx.doi.org/10.5772/54434

257

**Direct cytopathic effects.** Persistent MV infection might destroy infected cells, including oli‐ godendrocytes, and damage inflammatory responses, thereby resulting in demyelination. Consistent with this idea, there is a strong correlation among the extent of viral fusion activi‐ ty, cytopathic effects of MV, and severity of neurovirulence in a hamster model [23]. More commonly, T and B cells may directly attack viral antigens expressed on persistently infect‐ ed glial cells and destroy these cells. Damage resulting from MV infection can lead to a spreading of epitopes that may result in the generation of autoimmune responses [52]. In

**Autoantigen.** Autoimmune responses to myelin proteins are considered to be possible caus‐ es of some demyelinating diseases including SSPE. The level of antibodies against CD9, a glycoprotein that is abundant at the surface of myelin, is elevated and reaches a peak that coincides with the appearance of brain atrophy in SSPE patients [54]. It has also been sug‐ gested that autoimmunity could arise as a result of cross-reactivity between viral and myelin antigens [55, 56]. Myelin basic protein (MBP)-homologous sequences in the N and C pro‐ teins may account not only for encephalomyelitis in humans, but also for cross-reactions as

**Superantigen.** Another mechanism has been proposed that implicates superantigens in the etiology of autoimmune demyelinating diseases [58]. Superantigens activate T cells through the variable domain of the T cell receptor β chain. This distinctive mode of T cell activation, together with the ability of superantigens to bind to a wide variety of ma‐ jor histocompatibility complex molecules outside the antigen groove, leads to one super‐ antigen activating a whole class of T cells irrespective of antigen specificity. Activated T cells can cross the blood-brain barrier and enter the brain parenchyma. A few cells hom‐ ing to the brain have been shown to be enough to induce extensive lesions in the CNS [58]. Once activated, autoreactive T cells enter the brain and initiate inflammatory le‐ sions. The permeability of the blood-brain barrier increases, leading to an influx of solu‐ ble factors, such as tumor necrosis factor, into the CNS. All these events will result in extensive CNS lesions. Exogenous superantigens can be produced by bacteria, mycoplas‐ ma or viruses [59], and therefore the existence of superantigens during persistent MV in‐

Many previous studies have demonstrated that changes in host cell homeostasis contrib‐ ute to the pathogenesis of persistent MV infections. Rapid replication of MV that has been quiescent for years is triggered by some reactivation event(s) and results in hyper-

immune system would appear to be involved in SSPE pathogenesis (Figure 1D).

SSPE patients, brain-antigen-reactive T cells are found in the periphery [53].

detected by delayed skin tests with MBP in measles-sensitized guinea pigs [57].

fection should be investigated in future studies.

**6. Conclusion**

**Figure 2.** Hsp72 induction by the H protein. 293T cells were mock-transfected, or transfected with the H protein. At 24 h post-transfection, cells were harvested, and quantitative analysis of hsp72 was performed using quantitative realtime RT-PCR. Values are expressed as mean plus S.E. and compared with those from mock-transfected cells. \* *p* < 0.05.

**Peroxiredoxin 1 (Prdx1).** Prdx1, another potential molecule involved in SSPE, has recently been identified as a critical component during MV replication and transcription [46]. It was shown to bind to the same area of the N protein as the P protein (box 2), and competes with binding of the P protein. A reduction in Prdx1 expression appears to result in a steeper MV transcription gradient, as it has less of an effect on the N protein expression compared with the L protein expression. The binding affinity of Prdx1 to the N protein is approximately 40 fold lower than that for the P protein. This would suggest that Prdx1 may only play a role in MV RNA synthesis during the early stages of infection, when the amount of cellular Prdx1 is much greater than that of the viral P protein [46]. Likewise, Prdx1 might play a role in the reactivation of latent MVs that are attuned to host cells. Recent studies have implicated Prdx as a target of age-related modifications [47]. Age-related modifications, such as hyperoxidi‐ zation, likely affect Prdx1 thereby influencing MV transcription, and may explain why it takes several years after an acute MV infection for the first symptoms of SSPE to appear.

**Post-translational modifications.** Generally, infectious virus cannot be recovered from the CNS at autopsy, or from a biopsy of SSPE cases. In SSPE, MV-specific inclusions are present in the cytoplasm and nuclei of infected cells, and the incidence of certain types of inclusion bodies decline with prolonged duration of the disease [12, 13]. The N protein is most abun‐ dantly expressed in infected cells, and a major component of MV-specific inclusions. The N protein has been shown to be modified post-translationally by phosphorylation [48, 49]. The phosphorylation at serine residues 479 and 510 in the tail of the N protein has been shown to play an important role in viral replication and transcription [48]. Some reactivation events might stimulate host cell kinases responsible for these phosphorylations. Other post-transla‐ tional modifications could possibly be involved in the reactivation of latent MV.

#### **5. Pathogenesis of persistent MV infection**

MV infection induces clinically significant immunosuppression, which can continue for many weeks after an apparent recovery from measles [50, 51]. Long-lived cytokine imbalan‐ ces and direct effects on the proliferation of lymphocytes are reportedly implicated with the immunosuppression. In contrast, a persistent brain infection leads to a hyperimmune anti‐ body response, a pathogenic feature of SSPE [10, 11]. For example, there are extremely high titers of neutralizing antibodies in the serum and CSF against viral structural proteins. The immune system would appear to be involved in SSPE pathogenesis (Figure 1D).

**Direct cytopathic effects.** Persistent MV infection might destroy infected cells, including oli‐ godendrocytes, and damage inflammatory responses, thereby resulting in demyelination. Consistent with this idea, there is a strong correlation among the extent of viral fusion activi‐ ty, cytopathic effects of MV, and severity of neurovirulence in a hamster model [23]. More commonly, T and B cells may directly attack viral antigens expressed on persistently infect‐ ed glial cells and destroy these cells. Damage resulting from MV infection can lead to a spreading of epitopes that may result in the generation of autoimmune responses [52]. In SSPE patients, brain-antigen-reactive T cells are found in the periphery [53].

**Autoantigen.** Autoimmune responses to myelin proteins are considered to be possible caus‐ es of some demyelinating diseases including SSPE. The level of antibodies against CD9, a glycoprotein that is abundant at the surface of myelin, is elevated and reaches a peak that coincides with the appearance of brain atrophy in SSPE patients [54]. It has also been sug‐ gested that autoimmunity could arise as a result of cross-reactivity between viral and myelin antigens [55, 56]. Myelin basic protein (MBP)-homologous sequences in the N and C pro‐ teins may account not only for encephalomyelitis in humans, but also for cross-reactions as detected by delayed skin tests with MBP in measles-sensitized guinea pigs [57].

**Superantigen.** Another mechanism has been proposed that implicates superantigens in the etiology of autoimmune demyelinating diseases [58]. Superantigens activate T cells through the variable domain of the T cell receptor β chain. This distinctive mode of T cell activation, together with the ability of superantigens to bind to a wide variety of ma‐ jor histocompatibility complex molecules outside the antigen groove, leads to one super‐ antigen activating a whole class of T cells irrespective of antigen specificity. Activated T cells can cross the blood-brain barrier and enter the brain parenchyma. A few cells hom‐ ing to the brain have been shown to be enough to induce extensive lesions in the CNS [58]. Once activated, autoreactive T cells enter the brain and initiate inflammatory le‐ sions. The permeability of the blood-brain barrier increases, leading to an influx of solu‐ ble factors, such as tumor necrosis factor, into the CNS. All these events will result in extensive CNS lesions. Exogenous superantigens can be produced by bacteria, mycoplas‐ ma or viruses [59], and therefore the existence of superantigens during persistent MV in‐ fection should be investigated in future studies.
