*3.2. Chronological Changes in the Interpretations of EBV Infection in SS*

With respect to the relationship between SS and EBV infection, our PubMed search identified 149 reports, with many conflicting results and opinions regarding SS per se and EBV-associated hematological tumors. Representative studies about the detection of EBV in SS are listed in Table 1.



EA-D: EBV-encoded early antigen, ELS: ectopic lymphoid structure, FCM: flow cytometry, IF: immunofluorescence, IB: immunoblot, ISH: in situ hybridization, PBMC: peripheral blood mononuclear cell, PCR: polymerase chain reaction, PC: plasma cell, RA: rheumatoid arthritis, RFLP: restriction fragment length polymorphism, RT: reverse transcription, SG: salivary gland.

A first definitive study [40] regarding EBV in Sjögren's syndrome was conducted in 41 patients with RA; IgG antibody against EBV capsid antigen (EBVCA) was detected in all 41 RA patients, 26 patients with sicca syndrome, and 26 healthy subjects. Subsequently, Fox et al. [41] reported that EBV DNA revealed by slot hybridization was detected in 8 of 20 parotid saliva samples from SS patients, whereas EBV DNA was not detected in saliva from control subjects. Fox et al. also reported positive cytoplasmic staining of EBV-encoded early antigen (EA-D) in 8 of 14 salivary gland samples from SS patients. Whittingham et al. [54] hypothesized that EBV is an etiological agent of SS (especially for early-stage SS) in light of their observation of the binding of anti-La/SS-B antibody to viral RNAs or molecules encoded by EBV.

It was reported that, compared to control subjects, the titer of anti-ENBA antibodies but not that of anti-cytomegalovirus (CMV) antibodies was higher in sera from SS patients, and anti-EBNA antibodies were dominantly detected in anti-La/SS-B antibody-positive patients with SS [55]. Venables et al. subsequently reported that EBV DNA was present in the salivary gland biopsy specimen of 6 of 10 control subjects (60%) compared to 17% in the salivary gland biopsy specimens of SS patients [44]. They also detected EBV DNA in the salivary glands of five control subjects without HLA-class II antigen, suggesting that EBV infection was not specific for SS. With respect to the tumorigenicity of EBV, Fox et al. used a restriction fragment length polymorphism technique and reported that EBV or HHV-6 was associated with the onset of lymphomas [56]. In a study applying probes for both EBV and HHV-6 in cultured salivary gland epithelial cells (SGECs), only HHV-6 was detected, but HHV-6 was not related to SS [46].

Because benign lymphoepithelial lesions (BLELs) are unique pathological features in SS, Di Giuseppe et al. examined the involvement of EBV in BLELs [57]: they detected no EBER1 expression in salivary glands from SS patients and suggested that there is no involvement of latent EBV infection in SS. Regarding the relationship between CMV and EBV, their DNA in labial salivary glands (LSGs) from SS patients and subjects with non-specific sialadenitis was examined by polymerase chain reaction (PCR), which revealed no significant difference between the two patient groups [58]. Di Giuseppe et al. also contended that the persistence of these two viruses after the initial infection in LSGs was reasonable, but a direct involvement of EBV in the pathogenesis of SS was not suggested by their findings. Other investigators detected EBV DNA by in situ hybridization (ISH) and PCR in 19% of their SS patients versus 3% of the control subjects [49]. In contrast, EBV LMP was identified in 17% of their SS patients and 22% of the control subjects, indicating that the involvement of EBV can be influenced by the presence of undetermined factors.

Regarding the involvement of EBV in the pathogenesis of lymphomas, Royer et al. examined the occurrence of non-Hodgkin's lymphoma (NHL) and they identified cases with low-grade marginal zone lymphoma (MZL) and mucosa-associated lymphoid tissue (MALT) lymphoma; however, no obvious association between these SS and EBV infection was observed based on the presence of LMP proteins and EBER RNA [59]. In addition, the clinical implications of EBV in relation to the frequency of the detection of EBV DNA and related proteins have not been established. Regarding the association between the disease activity of SS and EBV infection, it was first demonstrated that significant levels of anti-EA-D antibodies were present in SS patients with articular manifestations described by items of the EULAR Sjögren's Syndrome Disease Activity Index (ESSDAI) [52]. Croia et al. took ectopic lymphoid structures (ELSs) into account, considering the selectivity of active EBV infection in SS as described below in the Pathogenesis section (Section 3.4) [53]. Regarding lymphoma observed in SS, it was demonstrated that among 382 patients with SS, 2.6% (*n* = 10) were diagnosed with EBV-associated follicular lymphoma, and 8 of these 10 patients showed a positive expression of LMP1 [60].

#### *3.3. The Reactivation and Detection of EBV in SS*

The word 'reactivation' in SS was first used in the 1980s by researchers examining the reactivity of monoclonal antibodies against EBV in salivary glands of individuals with SS [61,62]. Since then, the interpretation of the concept of reactivation has changed, as have the techniques for detecting EBV DNA and proteins. Saito et al. [55] demonstrated the usefulness of a PCR method to detect and monitor EBV DNA in salivary gland epithelial cells and peripheral blood. They also highlighted the importance of the rapid detection of EBV reactivation under immunosuppressive conditions and in lymphoproliferative disorders. Mariette et al. [56] introduced a combination of ISH using BamH1-W fragment and PCR reaction to detect EBV DNA in salivary glands from SS patients and control subjects. They observed positivity predominantly in the specimens from the SS patients, but they reported

that there was no evidence to show that EBV infection was directly involved in the destruction of the glandular structure.

ISH was later used to detect EBV-specific DNA in patients with secondary SS in which epithelial cells positive for EBV DNA were observed around areas with salivary gland destruction or lymphoepithelial lesions [63]. By using an enzyme-linked immunosorbent assay (ELISA) and a western blot analysis, Inoue et al. then observed high IgG antibody titers against EBNA antigens in sera from SS patients compared to sera from normal subjects [46]. With respect to the reactivation of EBV, Saito et al. used reverse transcriptase PCR coupled with PCR and immunohistochemistry (IHC), which revealed a strong expression of thioredoxin (TRX) in infiltrating B cells and epithelial cells in salivary glands from most of their SS patients [64]. In addition, an anatomical association between TRX and EBV as well as the co-expression of TRX message and EBV DNA were confirmed. In an in vitro experiment by those authors, B-cell lines that were infected with EBV frequently expressed TRX. These findings were the first to suggest the usefulness of detecting factors of EBV reactivation.

Regarding a link between tumorigenesis and EBV reactivation, increased reactivity was demonstrated toward EBV EA proteins such as BHRF1 (which is a viral homologue of Bcl-2 in rheumatic diseases including SS) [48]. A clear and frequent expression of interleukin (IL)-12 in infiltrating B cells and salivary gland tissue of SS patients was shown to correspond to EBV DNA [65], suggesting a relationship between EBV reactivation and Th1 cytokine. The involvement of the aryl hydrocarbon receptor (AhR, which binds to 2, 3, 7, 8-tetrachlorodibenzo-p-dioxin, or TCDD) in the reactivation of EBV was also reported [51]; that research group demonstrated an enhancement of BZLF1 transcription that mediated the switch to the lytic form, as well as a BZLF1 message and EBV DNA due to TCDD. These findings suggest that the ligation of AhR had the potential to induce the reactivation of EBV in both B cells and salivary epithelial cells.
