Differential Effects Of Bortezomib On Acute Graft Versus Host Disease Mediated By Dendritic Cells

Allogeneic hematopoietic stem cell transplantation (HSCT) is a potentially curative therapy for many malignant and non-malignant hematological diseases. However, graft-versus-host disease (GVHD) remains a lethal complication that limits its further applications. Acute GVHD (aGVHD) is a inflammatory response of donor T cells against host tissues. Antigen Presenting Cells (APCs), specifically host DCs, play a central role in the initiation of aGVHD. Allogeneic T cell priming in this context is significantly influenced by the state of DC activation due to'danger signals'during and after conditioning of the recipients. Immature DCs (imDCs) are specialized for antigen capture but with low APC ability, resulting in less T cell activation. Whereas, mature DCs (mDCs) upon maturation signals such as pathogen-derived products LPS or endogenous inflammatory cytokine TNF-α, migrate to the lymphoid tissue with increased APC ability and T-cell activation potential. NF-κB, a widespread transcription factor in virtually all cell types, controls the expressions of a number of genes important for immune and inflammatory responses characteristic of GVHD pathophysiology. Interestingly, signals that induce DC maturation are also strong activators of NF-κB, suggesting that NF-κB may be involved in DC maturation.Bortezomib (PS-341; Velcade), a reversible proteasome inhibitor, has been demonstrated to exert numerous biological effects including the blockade of NF-κB activation. The ability to inhibit NF-κB pathway make bortezomib a potentially attractive option for the prevention of GVHD. Evidence from the animal models as well as clinical data indicates a potential role for bortezomib in the treatment of GVHD. However, contradictory results of GVHD aggravation were observed in murine models with delayed bortezomib administration after bone marrow transplantation (BMT).Thus, to specify the conditions under which bortezomib improves GVHD, we set up aGVHD murine models of different severity fulfilled by infusion of decreasing donor splenocytes (SCs) and chose the different timing of bortezomib administration. The following are the main methods we performed and the major results we obtained in the experiments.Ⅰ. Early bortezomib administration protected mice form aGVHD in a mild aGVHD model Methods:Murine aGVHD models with different severity were set up by infusing lethally irradiated BALB/c with decreasing doses of donor C57BL/6 SCs (2×107,1×107,5×106). In each model, bortezomib was administered from dayO to day2 or from day6 to day8 post-BMT. We observed the effects of bortezomib on aGVHD by means of survival, weight changes, pathological impairment of target organs, donor T cell proliferation and levels of proinflamatory cytokines.Results:1. Bortezomib prolonged survival and improved weight loss in mice with mild aGVHD Early bortezomib administration (day0-2 post-BMT) imposed no effects on survival improvement in either an aggressive model (SC 2×107) or a modest model (SC 1×107) of aGVHD. In contrast, when the number of donor SCs was decreased to further reduce the severity of the GVHD (SC 5×106), significant (P=0.016) increases in survival were observed, with more than 70% of mice becoming long-term survivors. Bortezomib also improved weight loss of aGVHD mice from day26 post-BMT. But, even in this mild model, delaying the administration of bortezomib (day6-8 post-BMT) resulted in significantly greater mortality than did PBS treatment of GVHD control animals (Median survival:26 days for aGVHD group,7days for intervention group). Notably, animals that underwent transplantation with BM alone and then were treated with bortezomib at early or later time post-BMT had long-term survival similar to those of untreated control mice.2. Bortezomib improved pathological changes in the small intestine of aGVHD mice Histologic analysis show reduced damage to the small intestines of bortezomib-treated animals compared with PBS-treated GVHD animals. Microscopic examination revealed obvious villous blunting and fusion, crypt-cell hyperplasia and apoptosis, along with inflammatory infiltrates in the small intestine from PBS-treated GVHD control mice. While, animals administered with bortezomib had modest small intestine injury with only mild villous blunting and few inflammatory infiltrates.3. Bortezomib resulted in reduced donor T cell proliferation in aGVHD mice Mice receiving donor BM and SCs with or without bortezomib had>90% donor T cells at day 15 and>99% donor T cells at day 30 post-BMT, suggesting that donor T cell chimerism was not adversely affected. Our results showed, at day 3 post-BMT, CD4+ T cell number was (14.20±1.32)×105 for aGVHD group, (6.26±0.67)×105 for intervention group. CD8+ T cell number was (7.82±0.72)×105 for aGVHD group, (1.44±0.15)×105 for intervention group. At day 5 post-BMT, CD4+ T cell number was (21.25±1.77)×105 for aGVHD group, (3.44±0.22)×105 for intervention group. CD8+ T cell number was (60.29±5.02)×105 for aGVHD group, (13.06±0.85)xl05 for intervention group. At day 7 post-BMT, CD4+ T cell number was (47.82±2.81)×105 for aGVHD group, (15.69±1.95)×105 for intervention group. CD8+T cell number was (56.72±3.34)×105 for aGVHD group, (19.03±2.36)×105 for intervention group. Thus, bortezomib treatment resulted in significantly less donor CD4+ and CD8+ T cell proliferation at day+3,+5 and +7 post-BMT (P<0.01).4. Bortezomib induced decreases in Thl proinflamatory cytokines and increases in Th2 proinflamatory cytokines in aGVHD mice On day 7 post-BMT, TNF-αlevel was (65.83±9.63) pg/ml for aGVHD group, (27.41±4.79)pg/ml for intervention group. IFN-γlevel was (665.54±36.79) pg/ml for aGVHD group, (468.52±27.42) pg/ml for intervention group. IL-4 level was (178.01±11.66) pg/ml for aGVHD group, (225.36±14.61) pg/ml for intervention group. IL-10 level was (151.42±10.22) pg/ml for aGVHD group, (220.65±33.80) pg/ml for intervention group. Thus, serum levels of the TNF-α, IFN-γwere both significantly (P<0.01) decreased and serum levels of the IL-4, IL-10 were both significantly (P<0.05) increased in the recipients receiving SCs and bortezomib treatment compared with recipients receiving SCs and PBS.Ⅱ. Improved aGVHD by bortezomib correlates with low levels of TNF-αand LPS as well as immature state of host DCs before administrationMethods:In murine aGVHD models with different severity (SC 1×107,5×106) before bortezomib administration, TNF-αlevels in the sera and tissues were assayed by ELISA kit. Determination of serum LPS levels was performed by Tachypleus Limulus Amebocyte Lysate (LAL) assay. Flow cytometry was used to determine the number of splenic donor T cells expressing TNF-αand analyse MHCⅡexpression on host DCs. Results:1. Analysis of serum LPS level before bortezomib intervention Increased serum LPS levels were observed in recipients on day+6 post-BMT (2.53±0.24 EU/ml for SC 5×106 group,8.14±1.17 EU/ml for SC 1×107 group) compared with those on d0 post-BMT (0.68±0.03 EU/ml for SC 5×106 group, 1.13±0.04 EU/ml for SC 1×107 group) in each model (P<0.01). Besides, on d0 post-BMT, significantly higher serum LPS levels were observed in recipients with 1×107 donor SCs compared to those with 5×106 SC (P<0.05).2. Analysis of serum TNF-αlevel before bortezomib intervention Increased serum TNF-αlevels were observed in recipients on day+6 post-BMT (54.85±10.51 pg/ml for SC 5×106group,76.55±7.18pg/ml for SC 1×107group) compared with those on d0 post-BMT (10.42±2.22 pg/ml for SC 5×106 group, 12.54±3.17 pg/ml for SC 1×107group) in each model (P<0.01). However, there was no difference in serum TNF-αlevel at d0 between the two models (P>0.05)3. Analysis of localized TNF-αlevel before bortezomib intervention Significant increases in localized TNF-a in the small intestine were observed in recipients with 1×107donor SCs (603.34±59.05 pg/ml) compared with those with 5×106 SCs (399.63±26.19 pg/ml) on day+2 post-BMT(P<0.01).4. Analysis of the number of splenic donor T cells expressing TNF-αbefore bortezomib intervention Significant increases in the number of splenic donor T cells expressing TNF-αwere observed in recipients with 1×107 donor SCs (87.48±12.63)×104 compared with those with 5×106 SCs (33.05±3.41)×104 at 12 hours post-BMT (P<0.05).5. Analysis of the state of host DCs before bortezomib intervention CDllchigh/H-2b- population of DCs was gated as host splenic DCs. The mean fluorescence intensity (MFI) of MHCⅡin host splenic DCs from recipients with 1×107 donor SCs (MFI 477.5±37.43) was significantly higher than that from recipients with 5×106 SC (MFI 276.37±39.62) at 18 hours post-BMT (P<0.01). Ⅲ. Bortezomib inhibited the phenotypic and functional maturation of imDCs rather than already mDCs in vitroMethods:imDCs were prepared from mouse BM progenitors by culturing in the presence of GM-CSF and IL-4. To produce already matured DCs (mDCs), imDCs were stimulated with LPS or TNF-a. To observe the effects of bortezomib on DCs with different maturation stage, we performed flow cytometry to assay phenotypes, applied Annexin V-FITC/PI to analyze apoptosis, conducted mixed lymphocyte reaction (MLR) to assay allostimulatory capacity of DCs and impacts on regulatory T cells (Tregs), used western-blot to study the expression of IκBαin DCs.Results:1. Effects of bortezomib on phenotypic changes in DCs Bortezomib prevented imDCs maturation in response to LPS or TNF-α. Namely, imDCs pretreatment with the proteasome inhibitor prevented the up-regulation CD80, CD86, CD40 and CD83 as well as MHC-Ⅱ. The normal up-regulation of integrin CD54 and chemokine receptor CXCR4 molecules in response to these inflammatory signals was slightly reduced by the proteasome inhibitor. In contrast to pre-treated imDCs, no significant difference was found regarding the expression of surface markers on bortezomib treated versus untreated mDCs.2. Effects of bortezomib on apoptosis in DCs A dose-dependent induction of cell apoptosis in both imDCs in response to LPS and already mDCs (LPS-induced) was observed with bortezomib treatment. However, pre-treated imDCs in response to LPS were more susceptible to bortezomib-induced apoptosis than already matured DCs (LPS-induced) (OnM, imDC14.30±0.98, mDC13.33±0.78; 5nM, imDC 23.24±2.00, mDC16.31±1.55;10nM, imDC 31.83±2.10, mDC22.67±1.66; 20nM, imDC 39.87±2.21, mDC 27.33±2.49) (P <0.01). Similar results were obtained when using TNF-a as a maturation stimulus (OnM, imDC14.29±1.51, mDC 14.63±1.42; 5nM, imDC 23.51±2.10, mDC 16.39±1.69;10nM, imDC 34.43±2.57, mDC 25.23±2.30; 20nM, imDC 41.58±3.52, mDC28.74±2.57) (P＜0.01)3. Effects of bortezomib on DCs allostimulatory capacity 5nM bortezomib reduced the capacity of imDCs to induce proliferation in alloreactive CD4+ T cells (Stimulation indices were 20.12±1.45 for intervention group,35.44±1.07 for control group) (P<0.01) and CD8+ T cells (Stimulation indices were 7.35±1.67 for intervention group,18.19±0.40 for control group) (P< 0.05). Additionally, significant decreases in TNF-αin alloreactive CD4+ T cells (290.46±57.79 for intervention group,867.23±110.30 for control group) (P<0.01), but not in CD8+ T cells (145.04±34.43 for intervention group,154.85±29.95 for control group) (P>0.05), when co-cultured with LPS-primed imDCs pretreated with bortezomib. Further, significant increases were observed in Tregs in CD4+ T cells when cocultured with bortezomib-pretreated imDCs (22.00±3.75% for imDCs group, 11.89±1.89% for mDCs group) (P<0.05). However, bortezomib with the low concentration does not substantially affect mDCs capacity to prime allogeneic lymphocyte proliferation and TNF-a production.4. Effects of bortezomib on NF-κB activity in DCs In imDCs, the pretreatment with bortezomib increased the expression level of IκB-αgradually upon LPS stimulation in a time dependent fashion (The ratio of IκB-αtoβ-actin at 12h was 26.71±3.08 for intervention group,10.64±1.36 for control group. The ratio at 24h was 35.28±3.69 for intervention group,6.04±1.29 for control group.) (P<0.01), indirectly indicating an inhibition of NF-κB activity. In contrast, the expression of IκB-αremained unaffected in mDCs after bortezomib treatment (The ratio of IκB-αtoβ-actin at 12h was 11.02±1.47 for intervention group, 11.07±1.79 for control group. The ratio at 24h was 9.58±1.47 for intervention group, 10.48±2.52 for control group.) (P>0.05).In conclusion, proteasome inhibition bortezomib, as a double-edged sword in GVHD, could confer protection from GVHD, in part, due to interfering with DCs at immature stage but not at matured stage. These data provide new insights into the immunopharmacology of bortezomib and suggest a novel approach to the manipulation of DCs for therapeutic and experimental application.