Volume 9 Supplement 3
CXCL13, CCL21 and CXCL12 are upregulated in mucosal-associated lymphoid tissue lymphomas in patients with Sjorgen's syndrome and cooperate in the maintenance of the immune response and malignant cell survival
© BioMed Central Ltd 2007
Published: 19 October 2007
We previously demonstrated in minor salivary glands of patients with Sjogren's syndrome (SS-mSGs) that a true phenomenon of ectopic lymphoneogenesis takes places, with the formation of germinal centres (GCs), in association with the ectopic expression of the lymphoid chemokines (CKs) CXCL13 and CCL21. In Sjogren's syndrome major salivary glands (SS-MSGs) these structures are known to support clonal immunoglobulin heavy chain gene rearrangements and somatic hypermutation, favouring expansion and selection of autoreactive B-cell clones, clonally related to lymphomatous B-cell clones (responsible for lymphoma development in Sjogren's syndrome (SS)). The involvement of lymphoid CKs and CXCL12 in the development of lymphoid malignancies has been demonstrated. At present, however, no data on the expression of these CKs nor their cellular source within SS-MSGs and mucosal-associated lymphoid tissue lymphomas (MALT-Ls) from patients with SS have been described.
To assess the specific contribution of lymphoid CKs in the organization of lymphoid proliferation within SS-MSGs and SS-MALT-Ls.
We studied 12 SS-mSGs, four SS-MSGs with lympho-epithelial lesion and 20 SS-MSGs with non-Hodgkin B-cell MALT-Ls. In order to define the histological organization of the lymphoid infiltrate (reactive versus malignant areas) and identify the B-cell subpopulations infiltrating the glands, immunohistochemistry was carried out for cellular (CD21, bcl-2, bcl-6, IgD, CD20, CD3, CD68) and vascular markers (CD31, PNAd). On sequential sections, digital images for CXCL13, CCL21 and CXCL12 were analysed by thresholding positive staining in detected areas of interest. Digital image analysis estimated CK's volume fraction areas as a ratio between CXCL13, CCL21 and CXCL12 positive areas over malignant or reactive areas. CK's producing cells were identified by double staining on sequential sections. Mononuclear cells were isolated from two SS-MALT-L parotids following enzymatic digestion and stained for CD19, CD38, CD27, CD24, IgM, IgD, CD10, CD5, CD3, CD4, CD69, CD45RO, CXCR5, CXCR4 and CCR7. Finally, mRNA from total SS-MSGs and MALT-Ls and isolated B cells and T cells from MALT-Ls were analysed by RT-PCR for the transcript levels of CXCL13, CCL21 and CXCL12.
Reactive areas, characterized by T-cell/B-cell segregation, CD20+IgD+bcl-2+ follicular B cells, presence of follicular dendritic cell networks in GCs and high endothelial venule formation were detected in 100% of MALT-Ls. A B-cell population (CD20+IgD-bcl-2+) characterized by nuclear abnormalities and monocytoid appearance was consistently observed within the proliferating ducts and identified as the B-cell malignant component. FACS analysis on isolated MALT-L mononuclear cells showed B cells in diverse maturative stages (transitional, mature and memory B cells) and a B-cell population CD19posIgDlowCD24negCD27low/negCD5negCD10neg not detectable in the controls. Ectopic expression of CXCL13 and CCL21 was observed in 100% of SS-MSGs and MALT-Ls. A significant increase in CXCL13's volume fraction analysis was observed both in SS-MSGs and SS-MALT-L reactive areas, as compared with SS-mSG follicular areas (P < 0.05 and P < 0.05, respectively). A strong difference in CXCL13 within the MALT-L reactive area compared with malignant areas was detected (P < 0.001). CCL21 was significantly increased in MALT-Ls compared with both SS-mSGs and SS-MSGs and was mainly confined to the T-cell area. CXCL12 was strongly expressed by ductal epithelial cells, vessels and reactive areas in SS-mSGs, SS-MSGs and MALT-Ls. Interestingly a significant increase in CXCL12 expression on MALT-L malignant areas as compared with reactive areas in SS-mSGs and SS-MSGs was detected (P < 0.01 and P < 0.05). RT-PCR analysis showed increased CXCL13 and CCL21 level in SS-MSGs compared with MALT-Ls and SS-mSGs, while strong upregulation in CXCL12 transcript in MALT-Ls as compared with SS-mSGs and SS-MSGs was detected.
CXCL13, CCL21 and CXCL12 were detected on CD68+ cells by immunohistochemistry, while CD20 and CD3, CXCL13 and CCL21 double staining and mRNA analysis on MALT-extracted T and B lymphocytes showed negligible expression of the two CKs in MALT-L-extracted lymphocytes. Interestingly we identified strong CXCL12 expression on CD19+ MALT-L isolated cells both at the protein and mRNA level and on ductal epithelial cells in close contact with the malignant B-cell infiltration. In agreement with this increase, we demonstrated a significant downregulation of CXCR4 on MALT-L-isolated B cells, likely to be the result of CXCR4 internalization upon CXCL12 ligation.
In SS-MALT-Ls a strong upregulation of the lymphoid CKs CXCL13 and CCL21 takes place and is associated with the organization of the reactive areas involved in the maintenance of the autoimmune process within the malignancy. These findings support a contributory role for CXCL13 and CCL21 in the pathogenesis of MALT lymphomas in SS. These results imply that the CK/CK- receptor axis is functional in MALT-L malignant cells, and suggest that the glandular microenvironment, B-cell receptor signalling upon antigen engagement and autocrine signals from the same malignant population concur in MALT lymphomagenesis, triggering local activation of malignant B cells and favouring their survival and expansion.