We investigated the hypothesis that salivary gland inoculation stimulates formation of ectopic germinal centers (GCs), transforming the gland into a mucosal inductive site. Intraglandular infection of mice with murine cytomegalovirus (MCMV; control: UV-inactivated MCMV) induces salivary gland ectopic follicles comprising cognate interactions between CD4(+) and B220(+) lymphocytes, IgM(+) and isotype-switched IgG(+) and IgA(+) B cells, antigen presenting cells, and follicular dendritic cells. B cells coexpressed the GC markers GCT (57%) and GL7 (52%), and bound the lectin peanut agglutinin. Lymphoid follicles were characterized by a 2- to 3-fold increase in mRNA for CXCL13 (lymphoid neogenesis), syndecan-1 (plasma cells), Blimp-1 (plasma cell development/differentiation), and a 2- to 6-fold increase for activation-induced cytidine deaminase, PAX5, and the nonexcised rearranged DNA of an IgA class-switch event, supporting somatic hypermutation and class-switch recombination within the salivary follicles. Intraglandular inoculation also provided protection against a systemic MCMV challenge, as evidenced by decreased viral titers (10(5) plaque-forming units to undetectable), and restoration of normal salivary flow rates from a 6-fold decrease. Therefore, these features suggest that the salivary gland participates in oral mucosal immunity via generation of ectopic GCs, which function as ectopic mucosal inductive sites.

While the salivary gland has been recognized as an important effector site of the common mucosal immune system, a useful model for studying anti-viral salivary gland immune responses in vivo and for exploring the role of the salivary gland within the common mucosal system has been lacking. Murine cytomegalovirus (MCMV) is a beta-herpesvirus that displays a strong tropism for the salivary gland and produces significant morbidity in susceptible mice when introduced by intraperitoneal (i.p.) inoculation. This study tested the hypothesis that MCMV morbidity and pathology could be reduced by injecting the virus directly the submandibular salivary gland (intraglandular (i.g.)), using either in vivo derived MCMV or the less virulent, tissue-culture-derived MCMV (tcMCMV). Peak salivary gland viral titers were completely unaffected by infection route (i.p vs. i.g.) after inoculation with either MCMV or tcMCMV. However, i.g. tcMCMV inoculation reduced viremia in all systemic tissues tested compared to i.p. inoculation. Furthermore, systemic organ pathology observed in the liver and spleen after i.p. inoculation with either MCMV or tcMCMV was completely eliminated by i.g. inoculation with tcMCMV. Cellular infiltrates in the salivary glands, after i.p. or i.g. inoculation were composed of both B and T cells, indicating the potential for a local immune response to occur in the salivary gland. These results demonstrate that a focused MCMV infection of the salivary gland without systemic organ pathology is possible using i.g. delivery of tcMCMV.

Intraduodenal priming of mice with reovirus serotype 1/strain Lang (reovirus 1/L) stimulates gut lymphocytes and generates precursor and effector CTLs. Our earlier studies demonstrated that germinal center and T cell Ag (GCT) is a marker which identifies reovirus 1/L-specific precursor CTL and effector CTL in Peyer’s patches (PP) of reovirus 1/L-inoculated mice. In this study, we characterized the expression of the activation markers, GCT and CD11c, on reovirus 1/L-stimulated gut lymphocytes and the effector mechanisms involved in reovirus 1/L-specific cytotoxicity. We found that intraduodenal reovirus 1/L inoculation of mice induced the expression of both GCT and CD11c on PP lymphocytes (PPL), intraepithelial lymphocytes (IEL), and lamina propria lymphocytes (LPL), and these activated cells expressed Fas ligand (FasL). The majority of the GCT+CD11c+ IEL and LPL expressed a phenotype, TCRαβ+Thy-1+CD8+ similar to that expressed on reovirus 1/L-stimulated PPL. However, splenic lymphocytes expressed GCT but not CD11c after stimulation with reovirus 1/L. Perforin, Fas-FasL, and TRAIL pathways were found to be involved in PPL, IEL, and LPL cytotoxic activity against reovirus 1/L-infected targets. In PPL, perforin and Fas-FasL pathways were more effective than TRAIL. In IEL, all three cytotoxic mechanisms were equally as effective. However, LPL prefer Fas-FasL and TRAIL over perforin. Further, we demonstrated the preferential migration of GCT+ PPL to the intraepithelial compartment and the lamina propria. These results suggest that GCT and CD11c can be used as activation markers for gut lymphocytes and CD11c can also be used to differentiate between activated gut and systemic lymphocytes.