Hypoxia-induced 26S proteasome dysfunction increases immunogenicity of mesenchymal stem cells

Abstract

Bone marrow-derived allogeneic (donor derived) mesenchymal stem cells (MSCs) are immunoprivileged and are considered to be prominent candidates for regenerative therapy for numerous degenerative diseases. Even though the outcome of initial allogeneic MSCs based clinical trials was encouraging, the overall enthusiasm, of late, has dimmed down. This is due to failure of long-term survival of transplanted cells in the recipient. In fact, recent analyses of allogeneic MSC-based studies demonstrated that cells after transplantation turned immunogenic and were subsequently rejected by host immune system. The current study reveals a novel mechanism of immune switch in MSCs. We demonstrate that hypoxia, a common denominator of ischemic tissues, induces an immune shift in MSCs from immunoprivileged to immunogenic state. The immunoprivilege of MSCs is preserved by downregulation or the absence of major histocompatibility complex class II (MHC-II) molecules. We found that 26S proteasome-mediated intracellular degradation of MHC-II helps maintain the absence of MHC-II expression on cell surface in normoxic MSCs and preserves their immunoprivilege. The exposure to hypoxia leads to dissociation of 19S and 20S subunits, and inactivation of 26S proteasome. This prevented the degradation of MHC-II and, as a result, the MSCs became immunogenic. Furthermore, we found that hypoxia-induced decrease in the levels of a chaperon protein HSP90α is responsible for inactivation of 26S proteasome. Maintaining HSP90α levels in hypoxic MSCs preserved the immunoprivilege of MSCs. Therefore, hypoxia-induced inactivation of 26S proteasome assembly instigates loss of immunoprivilege of allogeneic mesenchymal stem cells. Maintaining 26S proteasome activity in mesenchymal stem cells preserves their immunoprivilege.

Fig. 1. Exposure to hypoxia induces loss of immunoprivilege in MSCs.

a Rat bone marrow-derived MSCs were exposed to hypoxia for 24 h. MHC-II levels as measured by western blotting increased in hypoxic MSCs, which showed regression when inhibited by siRNA. n = 3. b Immunofluorescence images showed a significant increase in the expression of MHC-II under hypoxia compared with normoxia. n = 6. c–e To determine the immunogenicity of MSCs, normoxic and hypoxic rat MSCs (with or without siRNA) were co-cultured with allogeneic leukocytes at a ratio 1:10 for 72 h. c Leukocyte-mediated cytotoxicity in MSCs (LDH release) increased significantly in hypoxic MSCs vs. normoxic cells, which was rescued by siRNA-mediated inhibition of MHC-II. n = 10. d The effect of MSCs on Treg cell (CD4⁺CD25⁺) induction in a mixed leukocyte population was assessed by flow cytometry. The number of Treg cells decreased after co-culture with hypoxic MSCs, siRNA-mediated inhibition of MHC-II increased Treg cell number. n = 3. e The effect of MSCs on leukocyte activation and proliferation was determined using PI staining, by assessing the number of cells present in different phases of cell cycle. The % of activated and proliferating leukocytes showed a significant increase under hypoxia. The number of activated and proliferating leukocytes decreased after siRNA-mediated MHC-II inhibition in MSCs. n = 3. *p < 0.05 compared with normoxia group; #p < 0.05 compared with hypoxia group. Each experiment was repeated four to six times

Fig. 2. 26S proteasome regulates MHC-II levels and preserves immunoprivilege of MSCs.

a, b Rat MSCs were treated with 26S proteasome inhibitor (MG132, 2 µM and 5 µM for 24 h). MHC-II levels determined by western blotting (a) and immunostaining (b) showed a dose-dependent increase (n = 3). c Immunoprecipitation (IP) analysis was performed in rat MSCs with or without 26S inhibitor to determine the involvement of 26S proteasome in the degradation of MHC-II. IP data revealed a significant accumulation of ubiquitinated MHC-II protein in 26S-inhibited group. IP was performed with MHC-II antibody and blotting was performed with polyubiquitin antibody. Left panel: IP; right panel: lysate (n = 4). d, e To determine the immunogenicity of MSCs, normoxic MSCs (with or without 26S inhibitor) were co-cultured with allogeneic leukocytes at a ratio 1:10 for 72 h. d LDH levels increased significantly in 26S inhibitor-treated MSCs (n = 10). e Treg (CD4⁺CD25⁺) cell number in the mixed leukocyte population decreased significantly after co-culture with 26S inhibited MSCs n = 3. *p < 0.05 compared with normoxia group. Each experiment was repeated four to six times

Fig. 3. Exposure to hypoxia led to dissociation of 26S proteasome complex in rat MSCs.

a Model depicts 26S proteasome structure; MHC-II degradation by 26S maintains absence of MHC-II in normoxic MSCs. Hypoxia-induced dissociation of 26S proteasome (19S and 20S subunits) results in accumulation of MHC-II. b Immunoprecipitation (IP) assay was performed to monitor interaction between 19S and 20S subunits in normoxic and hypoxic MSCs. IP was performed with 20S antibody and blotted with antibodies for 19S and 20S. In normoxic MSCs, 19S and 20S subunits bind to form functional 26S proteasome. The binding of two subunits decreased in hypoxic MSCs (n = 4). c The two-dimensional (2D) blue-native polyacrylamide gel electrophoresis (BN-PAGE)/SDS-PAGE assay was performed to study protein–protein interaction between subunits of 26S proteasome. The cell lysates from normoxia- and hypoxia-exposed MSCs were subjected to 2D SDS-PAGE and immunoblotted using specific antibodies for Sug1 (one of the constituents of 19S subunit) and α3 (one of the constituents of 20S subunit). The multiprotein complex appearing in the high molecular weight range ~1200–2000 kDa (white arrows) represent 26S proteasome. The amount of 26S complex was lesser in hypoxia-exposed cells compared with normoxic group (n = 3). d Quantitative densitometric analysis of 2D immunoblots reveal that the fluorescence intensity (RFU) of 26S complex is stronger in normoxic MSCs compared with hypoxia-exposed cells (n = 3). eThe ratio of bound vs. unbound fractions of Sug1 (19S subunit) and α3 (20S subunit) involved in the formation of 26S proteasome complex were significantly higher in normoxic cells compared with hypoxia-exposed MSCs (n = 3). *p < 0.05 compared with normoxic MSC. Each experiment was repeated three to four times

Fig. 4. 26S proteasome activity and HSP90α levels were downregulated in hypoxic rat MSCs.

a To measure 26S activity, the levels of both 19S (deubiquitinating activity) and 20S (proteolysing activity) were determined. The activities were measured by using flurogenic substrates:- U-555 for 19S and SUC-LLVY-AMC for 20S. Hypoxic MSCs were found to have a marked reduction in 26S activity (n = 3). b, c NOB1, BLM10, HSP90α, and HSP90β mRNA and protein levels were determined by RT-PCR and western blotting. NOB1, BLM10, and HSP90β levels did not change in MSCs after exposure to hypoxia for 24 h. However, HSP90α mRNA and protein levels decreased in hypoxia-exposed MSCs (n = 4); *p < 0.05 compared with normoxic MSC. Each experiment was repeated four to six times

Fig. 5. HSP90α regulates 26S activity and MHC-II levels in normoxic MSCs.

a, b Rat MSCs were treated with HSP90α inhibitor (SNX-2112, 0.5 µM, 1 µM, and 2 µM for 24 h), 26S activity (19S and 20S activities) by fluorescence assay and MHC-II levels by western blotting were measured. a 26S activity decreased in HSP90α-inhibited MSCs (n = 3). b MHC-II expression increased in HSP90α-inhibited MSCs in a dose-dependent manner (n = 4). *p < 0.05 compared with normoxic MSC. Each experiment was repeated four to six times

Fig. 6. Maintaining HSP90α levels preserves immunoprivilege of MSCs under hypoxia.

a Rat MSCs were transduced with lentiviral construct to overexpress HSP90α. HSP90α and MHC-II levels were measured by western blotting in normoxic, hypoxic, and HSP90α-overexpressing hypoxic MSCs (n = 4). b 26S proteasome activity by fluorescence assay in normoxic, hypoxic, and HSP90α-overexpressing hypoxic MSCs. HSP90α overexpression rescued 26S activity in hypoxic MSCs (n = 4). c, d To determine the immunogenicity of MSCs, normoxic MSCs, hypoxic MSCs, and HSP90α-overexpressing hypoxic MSCs were co-cultured with allogeneic leukocytes at a ratio 1:10 for 72 h. c LDH levels increased significantly in hypoxic MSCs, HSP90α overexpression prevented hypoxia-induced increase in LDH levels (n = 10). d Treg (CD4⁺CD25⁺) cell number in the mixed leukocyte population decreased significantly after co-culture with hypoxic MSCs; HSP90α-overexpressing hypoxic MSCs were able to induce Treg cell number. n = 3. Each experiment was repeated four to six times. *p < 0.05 compared with normoxia group; #p < 0.05 compared with hypoxia group

Fig. 7. Loss of immunoprivilege of human MSCs after exposure to hypoxia.

a Human bone marrow-derived MSCs (hMSCs) were exposed to hypoxia for 24 h. HLA-DRα levels as measured by western blotting increased in hypoxic MSCs (n = 3). b, c To determine the immunogenicity of MSCs, normoxic and hypoxic hMSCs were co-cultured with allogeneic leukocytes at a ratio 1:10 for 72 h. b Leukocyte-mediated cytotoxicity (LDH release) increased significantly in hypoxic hMSCs vs. normoxic cells (n = 10). c The effect of hMSCs on Treg cell (CD4⁺CD25⁺) induction in a mixed leukocyte population was assessed by flow cytometry. The number of Treg cells decreased after co-culture with hypoxic hMSCs (n = 4). d 26S proteasome activity was measured by determining the activities of both 19S (deubiquitinating activity) and 20S (proteolysing activity). The exposure to hypoxia led to a significant decrease in 26S activity in hMSCs (n = 4). e hMSCs were treated with 26S proteasome inhibitor (MG132, 2 µM and 5 µM for 24 h); HLA-DRα levels determined by western blotting showed a dose-dependent increase (n = 3); *p < 0.05 compared with normoxia group. Each experiment was repeated four to six times

Fig. 8. HSP90α maintains 26S activity and preserves immunoprivilege of hMSC.

a, b hMSCs were treated with HSP90α inhibitor (SNX-2112, 0.5 µM, 1 µM, and 2 µM for 24 h), 26S activity (19S and 20S activities) by fluorescence assay, and HLA-DRα levels by western blotting were measured. a 26S activity decreased in HSP90α-inhibited MSCs (n = 3). b HLA-DRα expression increased in HSP90α-inhibited MSCs (n = 3). c, d To determine the immunogenicity of hMSCs after HSP90α inhibition, hMSCs were treated with SNX-2112 (0.5 µM, 1 µM, and 2 µM for 24 h) and then co-cultured with allogeneic leukocytes at a ratio 1:10 for 72 h. c Leukocyte-mediated cytotoxicity (LDH levels) in hMSCs increased significantly in the presence of HSP90α inhibitor (n = 10). d Treg (CD4⁺CD25⁺) cell number in the mixed leukocyte population decreased significantly after co-culture with HSP90α-inhibited hMSCs. n = 10. *p < 0.05 compared with normoxia group. Each experiment was repeated four to six times