Immune tolerance against infused FVIII in hemophilia A is mediated by PD-L1+ Tregs
Janine Becker-Gotot, … , Johannes Oldenburg, Christian Kurts
Janine Becker-Gotot, … , Johannes Oldenburg, Christian Kurts
Published September 15, 2022
Citation Information: J Clin Invest. 2022;132(22):e159925. https://doi.org/10.1172/JCI159925.
View: Text | PDF
Research Article Hematology Immunology

Immune tolerance against infused FVIII in hemophilia A is mediated by PD-L1+ Tregs

Abstract

A major complication of hemophilia A therapy is the development of alloantibodies (inhibitors) that neutralize intravenously administered coagulation factor VIII (FVIII). Immune tolerance induction therapy (ITI) by repetitive FVIII injection can eradicate inhibitors, and thereby reduce morbidity and treatment costs. However, ITI success is difficult to predict and the underlying immunological mechanisms are unknown. Here, we demonstrated that immune tolerance against FVIII under nonhemophilic conditions was maintained by programmed death (PD) ligand 1–expressing (PD-L1–expressing) regulatory T cells (Tregs) that ligated PD-1 on FVIII-specific B cells, causing them to undergo apoptosis. FVIII-deficient mice injected with FVIII lacked such Tregs and developed inhibitors. Using an ITI mouse model, we found that repetitive FVIII injection induced FVIII-specific PD-L1+ Tregs and reengaged removal of inhibitor-forming B cells. We also demonstrated the existence of FVIII-specific Tregs in humans and showed that such Tregs upregulated PD-L1 in patients with hemophilia after successful ITI. Simultaneously, FVIII-specific B cells upregulated PD-1 and became killable by Tregs. In summary, we showed that PD-1–mediated B cell tolerance against FVIII operated in healthy individuals and in patients with hemophilia A without inhibitors, and that ITI reengaged this mechanism. These findings may impact monitoring of ITI success and treatment of patients with hemophilia A.

Authors

Janine Becker-Gotot, Mirjam Meissner, Vadim Kotov, Blanca Jurado-Mestre, Andrea Maione, Andreas Pannek, Thilo Albert, Chrystel Flores, Frank A. Schildberg, Paul A. Gleeson, Birgit M. Reipert, Johannes Oldenburg, Christian Kurts

×

Figure 2

PD-L1+ Tregs are necessary and sufficient to suppress FVIII-specific B cells in vivo.

Options: View larger image (or click on image) Download as PowerPoint
PD-L1+ Tregs are necessary and sufficient to suppress FVIII-specific B c...
(A) Experimental scheme. WT (black, n = 5) and Foxp3-LuciDTR (orange, n = 5) mice were intravenously injected with 2 IU/mouse of rhFVIII at weekly intervals. Foxp3+ Tregs were depleted by injecting 15 ng/g mouse DTX intraperitoneally on day –1 and 0. (B) ELISA-based quantification of the FVIII-specific IgG antibody titer in the serum of WT and Foxp3-LuciDTR mice. (C) Number of FVIII-specific B cells in spleens of WT and Foxp3-LuciDTR mice after treatment with rhFVIII by flow cytometry. (D) Mean fluorescence intensity (MFI) of PD-1 on FVIII-specific B cells of splenocyte suspensions. (E) Early apoptotic cells are presented as percentage of annexin V+ and Hoechst– FVIII-specific B cells after in vitro restimulation with 0.25 μg rhFVIII overnight. (F) Experimental setup. Tregs (1 × 106) isolated either from WT or Pd-l1–/– mice were injected into HemA mice. Starting on the next day, HemA (red, n = 9) and HemA mice that received Tregs from WT (blue, n = 9) or Pd-l1–/– (purple, n = 9) mice were intravenously injected with 2 IU/mouse of rhFVIII at weekly intervals. (G) FVIII-specific IgG antibody titers measured by ELISA in the serum. (H) The percentage of residual active FVIII in the plasma of rhFVIII-treated mice. (I) Number of FVIII-specific B cells in spleens of HemA mice with or without Treg transfer after rhFVIII treatment. (J) Proportion of splenic PD-1+ FVIII-specific B cells and (K) PD-1 MFI on FVIII-specific B cells. (L) Early apoptotic cells presented as annexin V+ and Hoechst– FVIII-specific B cells after in vitro restimulation with 0.25 μg rhFVIII overnight. *P < 0.05; **P < 0.01 by unpaired 2-tailed Student’s t test (B–E) or 1-way ANOVA with Bonferroni’s post hoc test (G–L).