CCR6+ Th cells in the cerebrospinal fluid of persons with multiple sclerosis are dominated by pathogenic non-classic Th1 cells and GM-CSF-only-secreting Th cells.

Considerable attention has been given to CCR6+ IL-17-secreting CD4+ T cells (Th17) in the pathology of a number of autoimmune diseases including multiple sclerosis (MS). However, other Th subsets also play important pathogenic roles, including those that secrete IFNγ and GM-CSF. CCR6 expression by Th17 cells allows their migration across the choroid plexus into the cerebrospinal fluid (CSF), where they are involved in the early phase of experimental autoimmune encephalomyelitis (EAE), and in MS these cells are elevated in the CSF during relapses and contain high frequencies of autoreactive cells. However, the relatively low frequency of Th17 cells suggests they cannot by themselves account for the high percentage of CCR6+ cells in MS CSF. Here we identify the dominant CCR6+ T cell subsets in both the blood and CSF as non-classic Th1 cells, including many that secrete GM-CSF, a key encephalitogenic cytokine. In addition, we show that Th cells secreting GM-CSF but not IFNγ or IL-17, a subset termed GM-CSF-only-secreting Th cells, also accumulate in the CSF. Importantly, in MS the proportion of IFNγ- and GM-CSF-secreting T cells expressing CCR6 was significantly enriched in the CSF, and was elevated in MS, suggesting these cells play a pathogenic role in this disease.

Dynamic regulation of spinal pro-inflammatory cytokine release in the rat in vivo following peripheral nerve injury.

Spinal release of cytokines may play a critical role in the maladapted nociceptive signaling underlying chronic pain states. In order to investigate this biology, we have developed a novel 'high flux' intrathecal microdialysis approach in combination with multiplex bead-based immunoassay technology to concurrently monitor the spinal release of interleukin (IL)-1beta, IL-6 and tumour necrosis factor (TNF)alpha in rats with unilateral sciatic nerve chronic constriction injury (CCI). Intrathecal microdialysis was performed under isoflurane/N(2)O anaesthesia in rats with confirmed mechanical hypersensitivity. In a first study, C-fiber strength electrical stimulation of the operated nerve in neuropathic rats was found to evoke a dramatic increase in IL-1beta efflux ( approximately 15-fold) that was significantly greater than that observed in the sham-operated group. Spinal IL-6 efflux was also responsive to primary afferent stimulation, whereas TNFalpha was not. In a second study, treatment with the glial inhibitor propentofylline for 7days normalized CCI-induced mechanical hypersensitivity. In the same animals, this treatment also significantly reduced intrathecal IL-1beta, IL-6 and TNFalpha and prevented afferent stimulation-evoked cytokine release of both IL-1beta and IL-6. These results provide support for glia as the source of the majority of intrathecal IL-1beta, IL-6 and TNFalpha that accompanies mechanical hypersensitivity in the CCI rat. Moreover, our studies demonstrate the ability of a neurone-glia signaling mechanism to dynamically modulate this release and support a role of spinal IL-1beta in the phasic transmission of abnormal pain signals.

Elevated cerebrospinal fluid (CSF) leptin in idiopathic intracranial hypertension (IIH): evidence for hypothalamic leptin resistance?

OBJECTIVE: The aetiology of idiopathic intracranial hypertension (IIH) is not known, but its association with obesity is well-recognized. Recent studies have linked obesity with abnormalities in circulating inflammatory and adiposity related cytokines. The aim of this study was to characterize adipokine and inflammatory cytokine profiles in IIH. DESIGN: Paired serum and cerebrospinal fluid (CSF) specimens were collected from 26 patients with IIH and compared to 62 control subjects. Samples were analysed for leptin, resistin, adiponectin, insulin, IL-1beta, IL-6, IL-8 (CXCL8), TNFalpha, MCP-1 (CCL2), hepatocyte growth factor, nerve growth factor and PAI-1 using multiplex bead immunoassays. RESULTS: CSF leptin was significantly higher in patients with IIH (P = 0.001) compared to controls after correction for age, gender and body mass index (BMI). In the control population, BMI correlated with serum leptin (r = 0.34; P = 0.007) and CSF leptin (r = 0.51; P < 0.0001), but this was not the case for the IIH population. Profiles of other inflammatory cytokines and adipokines did not differ between IIH patients and controls once anthropometric factors had been accounted for. CONCLUSIONS: IIH was characterized by significantly elevated CSF leptin levels which did not correlate with BMI. We suggest that CSF leptin may be important in the pathophysiology of IIH and that obesity in IIH may occur as a result of hypothalamic leptin resistance.

Serum cytokine profiles in Behçet's disease: is there a role for IL-15 in pathogenesis?

OBJECTIVES: Behçet's disease (BD) is a multisystem inflammatory disease characterised by recurrent orogenital ulceration, ocular inflammation and skin lesions whose aetiology is currently unknown. We hypothesized that levels of cytokines in the serum might provide either diagnostic or activity markers for the disease. METHODS: Levels of 10 cytokines were analysed in a multiplex bead analysis system as well as IL-15 by ELISA, in 79 serum samples from 52 patients with BD. The same cytokines were also measured in serum samples from 20 patients with recurrent aphthous stomatitis (RAS), as disease controls, and 15 healthy volunteers. The results were correlated with disease activity and current drug therapy. RESULTS: CXCL8 and TNF were the most abundant cytokines and were significantly raised compared to both patients with RAS and healthy controls. IL-15 was present in all samples and was significantly raised in both patients with BD and RAS compared to healthy controls. By comparison, cytokines associated with an adaptive immune response such as IFNgamma and IL-2 were found in few samples, while IL-4 and IL-10 were not detected in any sample. Levels of cytokines correlated with each other suggesting a response to the same stimulus, however, there was no association with either disease activity or treatment. CONCLUSION: Cytokines related to activity of the innate immune response were most prominent in this study and showed good correlation with each other. In particular, it was shown that IL-15 was raised in BD. However, there was no pattern of cytokine expression relating to disease activity or treatment.

Phenotypic and functional differentiation of KG-1 into dendritic-like cells.

The cell line KG-1 has been used as an in vitro model for human dendritic cell (DC) differentiation. We have investigated the response of KG-1 cells to stimulation with a number of factors known to induce differentiation and/or maturation of DCs in vitro. KG-1 cells showed no differentiation in response to LPS, CpG oligodeoxynucleotide or CD40 ligation. Culture in the presence of TNF-alpha induced some differentiation, but only treatment with PMA and ionomycin (with or without prior culture in GM-CSF and IL-4) induced morphological and phenotypic changes consistent with DC-like maturation, and even these maximally differentiated KG-1 cells showed lower levels of surface marker expression, macromolecular endocytosis, and ability to stimulate in allogeneic MLR compared with in vitro monocyte-derived DCs. Our data show that KG-1 cells differentiate in vitro into cells with DC-like functional characteristics under the influence of strong inducers of cellular activation, but lack the potency of mature DCs in key aspects of professional antigen presenting cells.

Induction of primary immune responses by allogeneic human myoblasts: dissection of the cell types required for proliferation, IFNgamma secretion and cytotoxicity.

Non-professional antigen-presenting cells (APC) have a limited ability to activate T lymphocytes during normal and auto-immune responses. Myoblasts could play an important role as APC in the etiology of autoimmune myasthenia gravis and polymyositis, as well as during muscle graft rejection. We examined the role of different component cell subsets in the response of human peripheral blood mononuclear cells (PBMC) to allogeneic myoblasts. Treatment of myoblasts with TNFalpha or IFNgamma led to the expression of a range of immunostimulatory molecules including MHC class I and II, and CD95 (Fas), but not B7 family molecules. Whole PBMC, cultured with allogeneic myoblasts, proliferated, secreted IFNgamma, and were cytotoxic. Proliferation and IFNgamma secretion were largely dependent on the presence of CD4+ lymphocytes, but neither CD4+ nor CD8+ T cells were responsible for cytotoxicity, which was mediated by MHC class II+ non-T mononuclear cells. However, purified CD4+ lymphocytes co-cultured with allogeneic myoblasts required co-stimulation with anti-CD28 antibodies for proliferation and IFNgamma secretion, which only induced a low level of IFNgamma secretion by CD8+ lymphocytes and did not induce cytotoxic function. These results suggest that human myoblasts can act as antigen-presenting cells for naive T lymphocytes, but only with additional co-stimulation.

Diverse patterns of unresponsiveness in an acetylcholine receptor-specific T-cell clone from a myasthenia gravis patient after engaging the T-cell receptor with three different ligands.

Using an acetylcholine receptor-specific T-cell clone (TCC) from a myasthenia gravis patient, we have compared the induction of unresponsiveness by three agents: an anti-V beta monoclonal antibody (mAb), complexes of MHC class II with specific peptide (DR4:peptide) and free peptide. Pretreatment with each agent without antigen-presenting cells specifically rendered the TCC unresponsive to a subsequent optimal stimulus. A substantial proportion of the DR4:peptide pretreated cells underwent apoptosis and the remainder were profoundly unresponsive. Apoptosis was less prominent after pretreatment with free peptide and was not significant with anti-V beta mab; with both, the unresponsiveness was transient in the survivors. These diverse effects of engaging the T-cell receptor in the absence of costimulation suggest that these agents act via different mechanisms, and give insights to the potential for specific immunotherapy.

High sensitivity and specificity of elevated cerebrospinal fluid kappa free light chains in suspected multiple sclerosis.

Cerebrospinal fluid (CSF) analysis is routinely used in the diagnostic work-up of multiple sclerosis (MS), by detecting CSF-specific oligoclonal bands (OCB). More recently, several studies have reported CSF free light chains (FLC) as an alternative. We show that absolute CSF κFLC concentrations were highly sensitive - more than OCB testing - and specific for clinically isolated syndrome, relapsing remitting and primary progressive MS. Measurement of κFLC alone was sufficient. Our results suggest that CSF κFLC levels measured by nephelometry, if validated in a larger series, are a preferred test to OCB analysis in the diagnostic work-up of patients suspected of having MS.

Multiple cytokines in human tear specimens in seasonal and chronic allergic eye disease and in conjunctival fibroblast cultures.

BACKGROUND: Several cytokines are involved in the recruitment and activation of inflammatory cells in ocular allergic diseases. The purpose of the study was to assay multiple cytokines and chemokines in tears, to compare subgroups of allergic conjunctivitis (AC) with controls, and in culture supernatants to determine whether conjunctival fibroblasts produce some of these cytokines. METHODS: Fifty to one hundred microlitre tears were obtained from patients with active seasonal allergic conjunctivitis (SAC; n=12), vernal keratoconjunctivitis (VKC; n=18), atopic keratoconjunctivitis (AKC; n=6) and non-atopic controls (n=14). Primary conjunctival fibroblasts grown in vitro were stimulated with IL-4, IL-13 or TNF-alpha for 24 h. Cell-free tear and culture supernatants were assayed for IL-1beta, IL-2, IL-4, IL-5, IL-6, IL-8, IL-10, IL-12, IL-13, IFN-gamma, TNF-alpha, eotaxin, MCP-1 and RANTES using multiplex bead analysis. Induction of chemokine gene expression was determined by PCR. RESULTS: IL-1beta, IL-2, IL-5, IL-6, IL-12, IL-13, MCP-1 were increased in all tears groups compared with controls, with highly significant correlations between many of these molecules. In addition IL-4, IFN-gamma, and IL-10 were elevated in SAC and VKC, while eotaxin and TNF-alpha were only increased in VKC. IL-6, IL-8, MCP-1, RANTES and eotaxin were detected from fibroblasts cultures, and were all up-regulated by TNF-alpha. By PCR, fibroblasts expressed MCP-1 transcripts constitutively, whereas IP-10 and Mig were up-regulated by TNF-alpha. CONCLUSIONS: Differential cytokine levels support tears as a useful indicator of immune mechanisms occurring during AC. The striking similarities in chemokine profiles between tears and fibroblasts suggest these cells as likely sources of chemokines in tears.

CD8 requirements for negative selection events are directly related to the TCR-antigen interaction.

Positive selection of class I-restricted T cells has been suggested to always require the surface expression of CD8 molecules on CD4+8+ thymocytes, whilst negative selection was found to be differentially dependent, relating to the antigen being engaged by the T cell receptor (TCR). We have studied the CD8-dependency of positive and negative selection using two TCR-transgenic (Tg) mice models, which both react against the same allo-antigen, H-2kb, back crossed with mice which are deficient for the expression of CD8. Whilst CD8 expression was always required for positive selection of cells expressing high levels of the Tg-TCR, events of negative selection were differentially dependent upon CD8, reflecting the CD8-dependency of the original CTL clones. For one TCR-Tg model (derived from a CD8-dependent CTL clone), deletion of CD4+8+ thymocytes was partially dependent upon the expression of CD8, though there was still selection against Tg-TCR positive CD4+ cells, which normally exit to the periphery expressing low levels of Tg-TCR. For the other model (derived from a CD8-independent CTL clone) negative selection was unaffected by the absence of CD8. For both models the CD4-8- Tg-TCR positive population was unaffected by the absence of CD8, including, for the CD8-independent model, reduced expression of the Tg-TCR/CD3 complex. These results suggest that CD8 expression may be a prerequisite for positive selection, whilst negative selection events can occur in the absence of CD8, depending directly upon the TCR-antigen interaction.

The balance between deletion and activation of CD4+8+ thymocytes is controlled by T cell receptor-antigen interactions and is affected by cyclosporin A.

The sensitivity of immature thymocytes to antigen-induced deletion has been shown to correlate with their differentiation status. By using an in vitro approach we have investigated whether parameters of antigenic stimulation may also affect the response of thymocytes. Two T cell receptor (TcR)-transgenic (Tg) mouse models have been compared, both of which recognize the allo-antigen H-2Kb but are functionally CD8"-dependent" (KB5.C20-Tg) and "-independent" (BM3.3-Tg). Presentation of the antigen H-2Kb on the surface of fibroblasts; to thymocytes in vitro, resulted in the apoptosis of CD4+8+ thymocytes. In contrast to in vivo deletion, in vitro deletion was much greater for KB5.C20-Tg than for BM3.3-Tg. In the absence of engagement of CD8 (using an altered H-2Kb-alpha 3 domain or CD8-specific antibodies), the H-2Kb-induced deletion of CD4+8+ thymocytes was decreased for KB5.C20-Tg, but no change in the pattern of deletion for BM3.3-Tg occurred. CD4+8+ thymocytes which remained viable after in vitro exposure to antigen, were shown to have been activated. Cyclosporin A (CsA), which has been reported to inhibit activation-induced cell death, did not affect antigen-induced deletion of CD4+8+ thymocytes from KB5.C20-Tg. More strikingly, deletion of CD4+8+ thymocytes from BM3.3-Tg increased, whilst activation was partially inhibited by CsA. These results provide direct evidence that presentation of antigen to thymocytes can result in deletion or activation, depending on not only the differentiation status of the cell, but also parameters of TcR-antigen interaction. Additionally, the effects of CsA suggest that activation can prevent the induction of deletion.

Cytoplasmic calcium fluxes induced in cytotoxic effector cells by engagement of Fc gamma receptors I, II, and III.

Changes in the intracellular calcium ion concentration ([Ca2+]i) of monocytes, granulocytes, and NK cells have been studied following either (1) independent cross-linking of Fc gamma receptors (Fc gamma R) I, II, or III, with F(ab gamma')2 fragments of monoclonal antibodies; or (2) linking of a selected Fc gamma R to a tumour cell target with bispecific F(ab' gamma)2 antibodies. Upon cross-linking each Fc gamma R with antibody an increase in the [Ca2+]i was observed, although all receptors apart from Fc gamma RIII on NK cells required additional cross-linking with an anti-mouse Fab' gamma. These results indicate that each type of receptor can transduce signals to the cell independently. Bispecific antibodies (anti-Fc gamma R x anti-target) linking cytotoxic Fc gamma R-bearing effector cells to tumour target cells also mediated increases in [Ca2+]i for all Fc gamma R tested except for Fc gamma RIII on granulocytes. The failure to transduce a signal via this receptor may be related to the GPI link, which is in contrast to the transmembrane link of Fc gamma RIII on NK cells. Significant lysis of tumour cell targets occurred when bispecific antibodies recruited NK cells or monocytes, but not granulocytes, via Fc gamma R. Chelation of intracellular Ca2+ in the effector cells reduced the observed lysis, suggesting a role for Ca2+ in the pathways leading to cytotoxicity.

The role of apoptosis in antibody-dependent cellular cytotoxicity.

Apoptosis in three lymphoma cell lines has been studied following cytotoxicity induced in vitro by normal human blood lymphocytes utilizing either natural killer (NK) or antibody-dependent cellular cytotoxic (ADCC) mechanisms. Guinea-pig L2C leukaemic lymphocytes, but not the human cell lines Daudi and Jurkat, revealed a degree of time- and temperature-dependent apoptotic death upon simple culture in vitro. NK cytotoxicity at low effector:target ratios (E:T) induced both release of 51Cr and apoptosis. However NK cytotoxicity at higher E:T, and ADCC at all E:T, increased the level of 51Cr release while reducing the level of apoptosis. The findings were consistent with the apoptotic process being cut short by intervention of necrotic death. The same characteristics accompanied ADCC whether the effectors were recruited by Fc gamma regions of antibody coating the targets, or by bispecific antibodies attaching one arm to the targets and the other to Fc gamma receptors type III on effectors. This finding, and the high level of cytotoxicity elicited by the bispecific method, confirm the belief that NK cells, in addition to exerting NK cytotoxicity, represent the principal effectors for ADCC among blood mononuclear cells. Our results suggest that NK cells have both apoptotic and necrotic mechanisms available for killing their targets, but use only the latter for ADCC.

Flow-cytometric analysis of apoptotic and nonapoptotic T-cell receptor-transgenic thymocytes following in vitro presentation of antigen.

The use of flow cytometry to detect apoptotic thymocytes is now well established. We have further developed the technique of Hoechst 33342 vital staining to identify discrete stages of murine thymocyte apoptosis (induced by 37 degrees C culture), in conjunction with propidium iodide (PI), cell scatter profile, and surface marker analysis. The first detectable stage was an increase in Hoechst fluorescence without any change in plasma membrane permeability (measured by PI staining). At this early stage thymocytes had already reduced in size, fragmented their DNA, and for the predominant CD4+ CD8+ double positive population, reduced expression of CD4 and CD8. Subsequent to this stage thymocytes continued to reduce in size and decrease expression of CD4 and CD8, though this was accompanied by an increase in membrane permeability. This technique was applied to an in vitro antigen-specific deletion system, where apoptosis of T cell-receptor-transgenic thymocytes was induced upon presentation of self-antigen. Although self-antigen-induced apoptotic thymocytes showed similar characteristics to those undergoing spontaneous apoptosis, there was a significant population of nonapoptotic CD4+ 8+ thymocytes that also had reduced expression of CD4 and CD8. Therefore, we have been able to show that the reduced expression of CD4 and CD8 is not limited to apoptotic thymocytes.

TCR-associated zeta-Fc epsilon RI gamma heterodimers on CD4-CD8- NK1.1+ T cells selected by specific class I MHC antigen.

The origin of autoreactive CD4-CD8- T cells is largely unknown. In TCR transgenic (Tg) mice expressing the cognate class I MHC antigen, CD4-CD8- T cells differed depending on characteristics of Tg-TCR/antigen interaction. Tg-TCR/CD3lo CD4-CD8- T cells expressing the NK1.1 marker were observed only for a Tg-TCR whose stimulation by antigen was independent of CD8. Unlike normal T cells, which have essentially TCR-associated zeta homodimers, these cells had a high proportion of TCR-associated zeta-Fc epsilon RI gamma heterodimers. They were also characterized by an unusually high content of Fc epsilon RI gamma mRNA and low content of mRNA encoding CD3 epsilon, CD3 gamma, CD3 delta, and zeta. Based on their phenotype and selection requirements, it is proposed that CD4-CD8- thymic precursor cells can be driven along the CD4-CD8-NK1.1+ pathway following coreceptor-independent TCR signaling at an intrathymic stage when Fc epsilon RI gamma and CD3 components are coexpressed.

The death-inducing signalling complex is recruited to lipid rafts in Fas-induced apoptosis.

Membrane microdomains known as lipid rafts have been shown recently to be involved in Fas signalling and apoptosis in T and B cell lines. Here, we have investigated further the role of lipid rafts in Fas-induced apoptosis in non-transformed human CD4 T cells. We show that Fas-induced apoptosis in CD4 T cells was inhibited by the lipid raft disrupter methyl-beta-cyclodextrin. When lipid rafts were isolated from control and Fas ligand treated cells, we found that a small proportion of Fas was present in the raft fraction in untreated cells and that this was greatly increased upon Fas ligation. The other components of the Death Inducing Signalling Complex (DISC), FADD, and procaspase 8, were also present at higher levels in the raft fraction isolated from Fas ligand treated cells. We conclude that formation of the DISC occurs in lipid rafts and that these membrane microdomains are required for efficient Fas signalling and apoptosis.

Persistent induction of the chemokine receptor CXCR4 by TGF-beta 1 on synovial T cells contributes to their accumulation within the rheumatoid synovium.

Chemokines and their receptors determine the distribution of leukocytes within tissues in health and disease. We have studied the role of the constitutive chemokine receptor CXCR4 and its ligand, stromal-derived factor-1 (SDF-1) in the perivascular accumulation of T cells in rheumatoid arthritis. We show that synovial T cells, which are primed CD45RO+CD45RBdull cells and consequently not expected to express constitutive chemokine receptors, have high levels of the chemokine receptor CXCR4. Sustained expression of CXCR4 was maintained on synovial T cells by specific factors present within the synovial microenvironment. Extensive screening revealed that TGF-beta isoforms induce the expression of CXCR4 on CD4 T cells in vitro. Depletion studies using synovial fluid confirmed an important role for TGF-beta1 in the induction of CXCR4 expression in vivo. The only known ligand for CXCR4 is SDF-1. We found SDF-1 on synovial endothelial cells and showed that SDF-1 was able to induce strong integrin-mediated adhesion of synovial fluid T cells to fibronectin and ICAM-1, confirming that CXCR4 expressed on synovial T cells was functional. These results suggest that the persistent induction of CXCR4 on synovial T cells by TGF-beta1 leads to their active, SDF-1-mediated retention in a perivascular distribution within the rheumatoid synovium.

Memory T cells constitute a subset of the human CD8+CD45RA+ pool with distinct phenotypic and migratory characteristics.

Using HLA class I-viral epitope tetramers to monitor herpes virus-specific CD8(+) T cell responses in humans, we have shown that a significant fraction of responding cells revert from a CD45RO(+) to a CD45RA(+) state after priming. All tetramer-binding CD45RA(+) cells, regardless of epitope specificity, expressed a phenotype LFA-1(high)CCR7(low) that was stable for at least 10 years in infectious mononucleosis patients and indefinitely in asymptomatic carriers. CD8(+)CD45RA(+)LFA-1(high) cells were not present in cord blood but in adults account for up to 50% of CD8(+)CD45RA(+) cells. These CD45RA(+)LFA-1(high) cells have significantly shorter telomeres than CD45RA(+)LFA-1(low) cells, suggesting that the latter represent a naive population, while the former are memory cells. CD45RA(+) memory cells are a stable population of noncycling cells, but on stimulation they are potent producers of IFN-gamma, while naive CD8(+) cells produce only IL-2. The chemokine receptor profile and migratory potential of CD45RA(+) memory cells is very similar to CD45RO(+) cells but different to naive CD8 cells. In accord with this, CD45RA(+) memory cells were significantly underrepresented in lymph nodes, but account for virtually all CD8(+)CD45RA(+) T cells in peripheral tissues of the same individuals.

Ectopic expression of the B cell-attracting chemokine BCA-1 (CXCL13) on endothelial cells and within lymphoid follicles contributes to the establishment of germinal center-like structures in Sjögren's syndrome.

OBJECTIVE: To test the hypothesis that the formation of ectopic germinal center (GC)-like structures in Sjögren's syndrome (SS) is associated with the ectopic expression of the constitutive lymphoid tissue-homing chemokines B cell-attracting chemokine 1 (BCA-1; or, CXCL13) and stromal cell-derived factor 1 (SDF-1; or, CXCL12). METHODS: Immunohistochemical and immunofluorescence analysis was used to determine the expression of the constitutive chemokines BCA-1 (CXCL13) and SDF-1 (CXCL12) in salivary glands from 5 SS patients and 3 non-SS patients. In addition, the expression of their respective receptors (CXCR5 and CXCR4) was examined on infiltrating lymphocytes. Human tonsil was used as a positive control for secondary lymphoid tissue. RESULTS: BCA-1 (CXCL13) was expressed within lymphoid aggregates in SS, which shared many structural features with GCs in tonsil. BCA-1 (CXCL13) was completely absent in control biopsy samples from patients who did not have SS. High levels of BCA-1 (CXCL13) were also found on endothelial cells in salivary glands from SS patients. Diseased SS tissue was infiltrated by CXCR5-expressing B cells which organized into GC-like clusters. In complete contrast, SDF-1 (CXCL12), a constitutive chemokine involved in leukocyte retention within lymphoid tissue, was expressed by epithelial cells in both diseased and control samples. The chemokine receptor for SDF-1, CXCR4, was expressed on T cells that accumulated in a periductal distribution in diseased tissue. CONCLUSION: The ectopic expression of BCA-1 (CXCL13) on endothelial cells and within GC-like structures, together with the strong expression of SDF-1 (CXCL12) on ductal epithelial cells, is a unique feature of inflamed glands in SS. By creating a local microenvironment supportive of focal B cell aggregation and differentiation, with structural features that are remarkably similar to GCs, BCA-1 (CXCL13) and SDF-1 (CXCL12) may contribute to the excessive production of high-affinity, class-switched autoantibodies and to the high incidence of B cell lymphomas classically associated with SS.