- Top of page
- Materials and Methods
- Supporting Information
With the recent development of new therapeutic strategies tolerance induction in transplant patients might be achievable . Due to new improved and humanized antibodies even targeting of co-receptors such as CD4 in a nondeletional manner has experienced a renaissance (www.biotest.de). The translation of those therapeutic approaches into the clinic has been proven rather difficult [2, 3]. So far only induction of mixed chimerism by infusion of donor hematopoetic stem cells combined with recipient conditioning has resulted in immunosuppression-free graft survival in transplant patients [4-8]. Many factors interfere with tolerance induction in the clinic. Patients have a higher pool of memory particularly cross-reactive memory T cells [9, 10]. Tolerance induction protocols are therefore accompanied at least initially with IS treatment for safety reasons. CNI seem to be most effective in preventing reactivation of preformed memory T cells [3, 11]. Indeed, although clinical studies are focusing on CNI-sparing protocols, mainly because of their nephrotoxicity, some patients have to be (re)-converted to CNI because of rejection . Consequently, CNI remain the standard immunosuppression at least for high-responders [13-15].
Although CNI are clinically very effective, conflicting findings have been reported as to whether their concomitant use with tolerance induction protocols has beneficial or adverse effects on promoting long-term graft acceptance . Effect of concomitant application of CsA on allograft survival has been investigated in tolerance induction protocols based on bone marrow-mediated induction of chimerism [17-19], costimulatory blockade [16, 20, 21] but also nondeletional targeting of co-receptors such as CD4 and CD8 [22, 23]. When reviewing these studies it appears that regardless of the protocol used concomitant CsA treatment acts synergistic or has no effect when the tolerance protocol used on its own does not or to a rather low extent induce permanent acceptance [19, 22, 24, 25]. In contrast, when using other tolerance induction protocols with a high percentage of permanent graft acceptances, an additional CsA treatment can but does not always lead to tolerance abrogation [17, 20, 23, 26, 27]. In addition, the tolerance-abrogating effect of CsA is dependent on dose, time of application and the transplantation model studied [16, 28, 29]. However, especially from the recent results of clinical trials utilizing the mixed chimerism approach to achieve immunosuppression-free graft survival, we have learned that CNI do not appear to completely antagonize tolerance induction in all settings [4-8]. Furthermore, many of the operational tolerant liver and kidney transplant patients had received CNI previously [30-35].
A simultaneous CsA application can prevent the development of allo-specific regulatory T cells (Tregs) [28, 36, 37]. CsA can also inhibit activation induced apoptosis of allo-reactive effector T cells, a mechanism that plays an important role for peripheral tolerance induction .
Whether CNIs also influence other processes leading to impaired long-term graft function has not been studied. To ensure safe and successful application of CNI regarding their control of memory T cell re-activation in combination with tolerance induction protocols the following questions have to be addressed: Which mechanisms are involved in CNI-mediated tolerance abrogation? Which intervening treatments can be applied to antagonize CNI-mediated interference with tolerance induction?
In accordance to some previous published observations in experimental models we observed an abrogation of anti-CD4mAb-mediated tolerance induction by simultaneous transient CsA treatment. Simultaneous CsA application resulted in intragraft up-regulation of genes associated with B cell migration and activation such as CXCL13, CCL19 and BATF leading to B cell activation, alloantibody production and complement–mediated destruction of glomeruli. Administration of a depleting polyclonal anti-rat B cell IgG serum or splenocytes from tolerant recipients 3 weeks after transplantation could arrest B cell accumulation activation in vivo and thereby diminish complement mediated destruction.
- Top of page
- Materials and Methods
- Supporting Information
Here we describe abrogation of an anti-CD4mAb-induced tolerance by an additional transient application of CsA. The additional CsA application induced local selective expression of B cell chemokine CXCL13 resulting in B cell infiltration, B cell activation and subsequently in alloantibody production and complement-mediated destruction of the glomeruli. However, transient B cell depletion or transfer of splenocytes from tolerant recipients, at a time point when the B cell activation is already detectable, can reverse this process.
As pointed out earlier our findings on CNI-mediated tolerance abrogation are in accordance with some previously published results from other experimental transplant models utilizing costimulatory blockade or co-receptor targeting [16, 20, 23]. But it has to be pointed out again that in some reports no detrimental effect of an additional CNI treatment could be reported [22, 24, 48]. Also for the induction of microchimerism, the only so far successful approach to achieve tolerance in patients, conflicting data on synergy/compatibility or antagonism of CNIs have been published in experimental models [17, 48].
The microarray analysis of graft samples from anti-CD4mAb + CsA-treated recipients revealed an up-regulation of genes associated with B cell attraction. The selective gene expression was already detectable during the CsA treatment. CXCL13 is a B cell chemokine regulating the migration of B cells into secondary lymphoid tissues [49, 50]. CXCL13 and its receptor CXCR5 are required for normal development of secondary lymphoid organs but have also been detected in inflammatory lesions with extranodal lymphoid neogenesis [51, 52]. Steinmetz et al.  described high intragraft CXCL13 expression in B cell clusters of kidney transplants undergoing rejection. CXCL13 expression and B cell recruitment have been also detected during renal interstitial injury (interstitial nephritis and chronic IgA nephropathy) [53, 54]. However, the cellular source of increased CXCL13 expression could not been determined. Generally, follicular dendritic cells are believed to be the main CXCL13 producers in lymphoid and inflamed tissue . Interestingly, there were no follicular dendritic cells detectable in grafts from anti-CD4mAb + CsA-treated recipients (data not shown). In our experiments, CXCL13 was exclusively produced within the glomeruli. In accordance with its function, increased CXCL13 production resulted in massive intragraft B cell infiltration. Furthermore, local B cell infiltration including IgG+ cells was associated with IgG alloantibody production and complement deposition. These infiltrating B cells play an active role for the pathogenesis of CNI-mediated tolerance abrogation as transfer of graft infiltrating B cells into recipients receiving anti-CD4mAb monotherapy caused kidney graft deterioration. B cells may contribute to pathogenic alterations in two ways. Alloantibodies produced by B cells cause complement activation and deposition as presented in our manuscript. B cells may also function as local antigen presenting cells and amplify the activation of infiltrating T cells after withdrawal of CNI. Indeed we observed an increase in T cell numbers and cytokine production at later time points after transplantation (Figure 1B and data not shown). However, transfer of splenocytes from tolerant recipients on Day 20 after transplantation could prevent further production of IgG alloantibodies and diminished glomerulosclerosis.
We also showed that a CNI-based immunosuppressive maintenance therapy is associated with an up to 100-fold increase in local CXCL13 transcription. Patients on a CNI-based immunosuppressive maintenance therapy with an alloantibody-mediated rejection and C4d deposition show enhanced intragraft CXCL13 expression (Supplementary Figure 5). However, whether here similar regulatory mechanisms are operating as in our experimental transplant model needs to be further investigated. Such an investigation was clearly beyond the scope of our work.
We also determined a cellular source of CNI-induced CXCL13 expression. Primary mesangial cells isolated from glomeruli responded with an increase in CXCl13 transcription when stimulated with different CNIs (Supplementary Figure 6). Interestingly, addition of TNF-α as one major local inflammatory mediator following transplantation could further increase CXCL13 expression in mesangial cells, resulting in a perfect milieu for B cell attraction/infiltration and activation (Supplementary Figure 6). When the initial CNI concentration is weaned off freshly activated T cells can deliver help to infiltrated B cells resulting in B cell activation and antibody production.
We also tested whether a transient B cell depletion at a time point, where signs of B cell activation such as alloantibody production were already detectable, could rescue tolerance induction. Indeed, depletion of B cells could reduce intragraft B cell accumulation, inhibit differentiation of alloantibody-producing cells and ameliorate intragraft alterations.
B cells have a well-established role in allograft rejection, both in hyperacute and antibody-mediated acute rejection  as well as in promoting cellular immunity [56, 57]. Novel data suggest that B cells may also play an important role in allograft survival and the development of operational transplant tolerance [35, 58]. Whereas enhanced frequencies of naïve and transitional B cells have been associated with long-term allograft survival, memory B cells have been linked with decreased allograft survival [58, 59]. Newell et al.  described that increased level of IL-10 producing regulatory B cells are associated with positive outcome in renal transplantation patients.
Furthermore, recent results from EU- and ITN-sponsored network projects have reported a B cell signature associated with operational tolerance, suggesting that B cells may contribute to the tolerant state [35, 59]. Top-ranked significant B cell-related genes included CD79b and PNOC . CD79b is a membrane protein that forms a heterodimer with CD79a and is expressed by all B cell subsets including plasma cells [61, 62]. PNOC is a secreted protein that binds the opioid receptor-like receptor (OPRL-1). Arjomand et al.  found that human peripheral blood lymphocytes and especially CD19+ B cells express PNOC. Our own preliminary results revealed enhanced PNOC expression in naïve and transitional B cells (data not shown). Anton et al.  showed that nociceptin modulates adaptive immune responses such as antibody production. Waits et al.  found that N/OFQ could up-regulate CTLA-4 on T cells, which is known to have immunosuppressive properties.
Our results revealed higher expression of intragraft PNOC in grafts from tolerance developing anti-CD4mAb-treated recipients on Day 60 posttransplant whereas CD79b was increased in grafts from recipients receiving CsA. These data may hint to a CNI-mediated disturbance of intragraft B cell subpopulations with less PNOC expressing transitional and naïve B cells to activated memory B cells and plasma cells also expressing CD79b.
Our data represent unexpected findings of CNI, which could be relevant when designing new treatment, especially drug weaning, protocols for transplant patients.
In case of an antagonizing effect of simultaneously applied CNIs we need mechanisms, which counterbalance local induction of CXCL13 and drive the B cell compartment toward a “tolerant” phenotype supporting establishment of transplantation tolerance.
We provide evidence that early depletion of B cells or transfer of splenocytes from tolerant recipients are two treatment options that can prevent CNI-mediated B cell activation and chronic graft rejection if applied early after transplantation.