Follicles, germinal centers, and immune mechanisms in primary biliary cirrhosis.

Abstract

Among the inter-related family of classical autoimmune diseases, primary biliary cirrhosis (PBC) is archetypal. It has one of the strongest associations with a specific autoantibody, with an overwhelming majority of patients producing high-titer, high-affinity antimitochondrial antibodies (Abs).1 PBC is also associated with perihepatic lymphadenopathy, marked elevations in serum immunoglobulin (Ig)M, and the development of tertiary lymphoid structures within the liver.1, 2 Therefore, activation of the lymphoid organs is implicated in PBC, and a complex role for both T and B cells is already recognized in pathogenesis. In this issue of Hepatology, Wang et al. turn the spotlight in PBC on the role of T follicular helper cells (Tfh), a subset of T cells that promote the production of high-affinity Ab by B cells.3 This work contributes, in a timely way, to the emerging increased understanding of disease mechanisms in PBC and is important to patients because it identifies potential targets to address the ongoing absence of specific biologically focused therapy. A synergy of genetic and basic immunological insights has continued to identify individual pathways and cell lineages implicated in PBC and associated autoimmune diseases and key among these are CD4+ T cells.1 Genetic risk for development of PBC is most strongly associated with major histocompatibility complex (MHC) class II variants, which are involved in the presentation of antigen to helper CD4+ T cells. Variants in loci that strongly implicate immunoregulatory molecules and pathways involved in CD4+ T-cell function, such as interleukin (IL)-12, signal transducer and activator of transcription 4, tyrosine kinase 2, and CD80, strengthen this link.4 A key role of CD4+ T cells is in providing help to maturing B cells in lymphoid follicles and producing the adaptive humoral immune response. This requirement for T cells providing “help” signals to B cells was established in the late 1960s through in vitro studies. However, the nature of these help signals was not fully recognized until several years ago. It was then that the entry into lymphoid follicles of a discrete subset of CD4+ antigen-reactive T cells, called Tfh, was established.5 The maturation of a germinal center (GC) involves the somatic hypermutation and selective clonal expansion of the most antigen-specific of variable B-cell clones; Tfh are the specialized subset of effector T cells that provide selective help to proliferation, and thereby positive selection, of high-affinity B cells in GC. Ultimately, those B cells that compete most effectively for Tfh signals are positively selected to develop into high-affinity Ab-secreting plasma cells or long-lived memory B cells (Fig. 1).5 Schematic representation of Tfh promoting B-cell maturation. Tfh are positioned at the T-zone/B-zone border according to positive interactions between CXCR5 and CXCL13 and reduced expression of CCR7. Tfh recognize antigen presented by activated B cells, provide costimulation through multiple surface molecules, and secrete interleukins IL-4, IL-10, and IL-21. BCL-6 is up-regulated in both B cells and Tfh. There is somatic hypermutation of B-cell receptors and class switching of Ab. Those B cells that develop high-affinity receptors are positively selected to proliferate; those that do not enter apoptosis. B cells subsequently mature to become high-affinity Ab-secreting plasma cells or memory B cells. Abbreviations: BCR, B-cell receptor; ICOSL, ICOS ligand; PD-L1, PD-1 ligand 1; SLAM, signaling lymphocytic activation molecule; TCR, T-cell receptor. Dashed lines denote interactions. In this context, Tfh dysfunction within GCs, and aberrant positive selection, has been previously suggested as important in the development of systemic autoimmunity.6 Functionally, Tfh are characterized by their ability to support B-cell differentiation and their tendency to migrate to follicles. Their migratory function relates to their expression of the chemokine receptor, chemokine (C-X-C motif) receptor 5 (CXCR5), and their loss of CC chemokine receptor 7 (CCR7). CXCR5's cognate ligand is chemokine (C-X-C) motif ligand 13 (CXCL13) and is strongly expressed on endothelial and reticular cells in B-cell follicles,7 whereas CCR7's ligands are most abundant in the T zones.8 Intriguingly, the CXCR5 gene locus is associated with PBC4 and of recent complimentary interest is an early report from Japan that has described up-regulated CXCL13 within the spleens of PBC, but not hepatitis C, patients at autopsy.9 Once Tfh have migrated to the T-zone/B-zone interface, they interact with B cells through three key mechanisms: paracrine through interleukins; continued antigen presentation through MHC class II induction on B cells; and direct activity of costimulatory molecules through multiple receptor-ligand pairs, especially CD40 and inducible T-cell costimulator (ICOS).5 Of relevance to this pathway, recent murine work has shown a protective effect of CD40-ligand blockade in a murine model of autoimmune cholangitis10; CD40L promoter methylation levels have also been correlated to serum IgM in PBC patients.11 The major interleukins are IL-21, IL-4, and IL-10, which collectively promote somatic B-cell hypermutation, Ab class switching, and cell proliferation. Innovative work employing intravital microscopy in mice has highlighted the dynamic nature of this help process, with Tfh moving into and around the T-zone/B-zone interface and forming multiple brief interactions with B cells. Costimulatory molecules lengthen these interactions and mediate signaling through intracellular calcium fluctuations.12 A further key feature of Tfh is the presence of the transcription factor, B-cell lymphoma 6 protein (BCL-6). BCL-6 suppresses competing transcription factors associated with development into other CD4+ subsets while promoting Tfh characteristics. Indeed, elegant knockout and transfection experiments have shown that BCL-6 is necessary and sufficient for Tfh development.13 There is also increasing interest in another transcription factor that appears central for the production of Tfh and which appears to act upstream of BCL-6: interferon-regulatory factor 4 (IRF-4). In the absence of IRF-4, there is an absence of Tfh, CD4+ cells expressing BCL-6 are absent, and there is widespread cell death in lymph nodes on pathogenic stimulation.14 Obtaining lymphoid tissue to permit analysis of GC Tfh is invasive and so has limited initial research on humans with active immune dysfunction. However, the identification of a subset of circulating CD4+ lymphocytes that express CXCR5, programmed cell death protein 1 (PD1), ICOS, and CD40L, and which appear related to true Tfh, has provided a surrogate, and one that Wang et al. have now capitalized on in their studies of PBC. Elevations in the frequency of these peripheral Tfh (pTfh) exist in a number of autoimmune conditions, including Sjögren syndrome, rheumatoid arthritis, juvenile dermatomyositis, autoimmune thyroiditis, myasthenia gravis, and systemic lupus erythematosus,5 all of which are more common in PBC patients. In lupus, Tfh numbers appear to reduce in response to appropriate immunosuppression, and in several, there is a correlation with disease severity scores. It is also interesting to note that, in the context of influenza immunization, there is a correlation between neutralizing Ab responses with frequencies of pTfh in young, but not elderly, adults.15 However, whether pTfh are representative of their GC counterparts is a source of some controversy. Much of this stems from their low levels of BCL-6 expression, the master transcription factor in true GC Tfh. Several pieces of recent work have now suggested that there is a direct correlation and that pTfh do indeed represent a useful proxy of their lymphoid counterparts: Transferred CXCR5+ cells migrate preferentially to the T-zone/B-zone border in mice16; identified human conditions associated with a failure of GC formation through TFH deficiency, including CD40L and ICOS deficiencies as well as autosomal-dominant hyper IgE syndrome, have correspondingly reduced numbers of pTfh7; evidence of low-level BCL-6 expression has been demonstrated in peripheral murine CD4+CXCR5+ cells; BCL-6, albeit at modest levels, is induced among pTfh on antigenic stimulation16; and, critically, peripheral CD4+PD-1+ CXCR3–CXCR5+ T cells are significantly more able to drive B-cell proliferation and Ab generation than other populations in vitro.17 Nevertheless, it has been noted that the level of CXCR5 expression of pTfh is lower than their TFH counterparts and that pTfh cells retain CCR7—which Tfh lose—leaving debate between immunologists as to how related these two populations are.18 It is on this background that Wang et al.3 present their first-in-field work interrogating pTfh number and function in patients with PBC, as well as histochemical analyses of liver and spleen tissue. Here, the group investigated both PD1+CXCR5+ and ICOS+ CXCR5+ subsets of CD4+ T lymphocytes in peripheral blood, both definitions having previously been used to define pTfh. The frequency of both was increased in PBC against control, and PD1+CXCR5+ cells were also more frequent than in autoimmune hepatitis patients of similar age and gender. These cells were highly functional in promoting B-cell proliferation, demonstrated increased production of IL-21 when stimulated, increased Ab production in coculture (with a greater effect on IgM than IgG classes), and correlated in frequency with plasma cells. Previously, the Gershwin group has linked CD38+ plasma cells with destructive peribiliary lesions observed in PBC, but not other liver diseases.19 Intriguingly, some measures of pTfh correlated with disease status: More frequent pTfh correlated with later disease stage, with greater elevations of alkaline phosphatase and with a lack of biochemical response to ursodeoxycholic acid (UDCA). Major current prognostic measures for PBC relate to changes in biochemical values after 1 year of UDCA therapy. However, in young patients—who may have aggressive disease—a year is a long time to wait. Although prospective validation is required, biomarkers of poor disease responsiveness potentially such as this are needed to guide introductions of future second-line therapies and therefore offer potential future value.20 Wang et al. also demonstrated that PD1+CXCR5+ Tfh are present in PBC liver tissue: They appear to concentrate around damaged bile ducts and tertiary lymphoid tissue. In PBC spleen, Tfh were demonstrated in the GCs, suggesting a role in driving the Ab response, whereas in control tissue they remained at the T-zone/B-zone interface. An added complexity to this field is that 10%-15% of Tfh cells express forkhead box protein P3, the master transcription factor of regulatory T cells, alongside the expected BCL-6 and CXCR5. These act to dampen Tfh activity and thereby the GC response: so-called T-follicular regulatory cells.6 Might perturbations in this effector/regulatory balance contribute to the changes noted by Wang et al. and the autoantibody production characteristic of PBC? A further speculation is that circulating pTFH might be relevant to PBC recurrence after transplantation: a phenomenon that currently escapes simple explanation. Thus, Wang et al.'s3 work further contributes to our immunobiological understanding of PBC, enriching our knowledge of pathways with the potential for manipulation and the potential to improve disease prognostication. Further specific studies will hopefully expand this area through exploring links between peripheral and central Tfh populations in liver disease and dissecting the degree to which (p)Tfh are a component or product of PBC pathogenesis. Care will be needed when interpreting data from different populations because Tfh number and function vary with age.15 In addition, establishing the mechanism by which Tfh numbers are increased, perhaps with reference to key molecules, such as BCL-6, IRF-4, ICOS/L, and CD40/L, will provide valuable additional insights. With an ever-increasing repertoire of specifically targeted immunotherapeutic agents available, refined biological-based therapy of PBC, such as costimulatory molecule manipulation, continues to be attractive, particularly if treatment can be stratified to the most “at-risk/at-benefit” patients by advanced immune profiling. Gwilym J. Webb, M.A., M.R.C.P.Gideon M. Hirschfield, Ph.D., F.R.C.P. Center for Liver Research and NIHR Biomedical Research UnitUniversity of BirminghamBirmingham, UK Author names in bold designate shared co-first authorship.