mTOR kinase inhibitors promote antibody class switching via mTORC2 inhibition

Abstract

The mammalian target of rapamycin (mTOR) is a kinase that functions in two distinct complexes, mTORC1 and mTORC2. In peripheral B cells, complete deletion of mTOR suppresses germinal center B-cell responses, including class switching and somatic hypermutation. The allosteric mTORC1 inhibitor rapamycin blocks proliferation and differentiation, but lower doses can promote protective IgM responses. To elucidate the complexity of mTOR signaling in B cells further, we used ATP-competitive mTOR kinase inhibitors (TOR-KIs), which inhibit both mTORC1 and mTORC2. Although TOR-KIs are in clinical development for cancer, their effects on mature lymphocytes are largely unknown. We show that high concentrations of TOR-KIs suppress B-cell proliferation and differentiation, yet lower concentrations that preserve proliferation increase the fraction of B cells undergoing class switching in vitro. Transient treatment of mice with the TOR-KI compound AZD8055 increased titers of class-switched high-affinity antibodies to a hapten-protein conjugate. Mechanistic investigation identified opposing roles for mTORC1 and mTORC2 in B-cell differentiation and showed that TOR-KIs enhance class switching in a manner dependent on forkhead box, subgroup O (FoxO) transcription factors. These observations emphasize the distinct actions of TOR-KIs compared with rapamycin and suggest that TOR-KIs might be useful to enhance production of class-switched antibodies following vaccination.

Keywords: B lymphocyte; class switching; differentiation; kinase; rapamycin.

Fig. 1.

High concentrations of TOR-KIs reduce mTORC1 activity and block B-cell proliferation. (A and B) Purified splenic B cells were labeled with CFSE to track cellular proliferation and cultured with media only or stimulated with LPS + IL-4 or LPS + IL-4 and inhibitors at the indicated concentrations. (A) mTORC1 activity was assessed at the indicated time points (B, Far Right) by intracellular staining (ICS) for p-S6 (S240/244 site) on proliferated cells. The graphs depict mean ± SEM of the p-S6 [median fluorescence intensity (MFI)] (n = 3). Data were normalized to the stimulated, no-drug treatment condition. (**P < 0.01, one-way ANOVA with Tukey’s multiple comparison test). Unstim, unstimulated. (B) CFSE-labeled B cells were stimulated with LPS + IL-4 without or with inhibitors at the indicated concentrations for 48 h. Data are representative of three independent experiments. Rap, rapamycin.

Fig. 2.

TOR-KIs increase in vitro B-cell CSR and decrease plasmablast differentiation. (A and B) Purified B cells were cultured with media only or stimulated with αCD40 + IL-4 in the absence or presence of the inhibitors indicated. (A) Dot plots show IgG1⁺ B cells (Top) and the cell division history of eFluor670-labeled B cells (Bottom; eFluor670 is a cell division tracking dye with a different emission spectrum than CFSE), as determined by FACS after 4 d. Other TOR-KIs in this study had similar effects on cell division at the concentrations used for differentiation experiments. (B) Percentage of live B cells that have divided at least once (based on eFluor670 division history) expressing surface IgG1 was determined by FACS after 4 d. (C) Switching to IgG1 was assessed as in B, except cells were stimulated with LPS + IL-4. (D) Representative FACS plots of plasmablast differentiation in purified B cells stimulated with LPS or LPS and indicated inhibitors for 3 d. (Bottom) B cells were labeled with eFluor670 to track division history. (E) Graph of the percentage of live B cells with a plasmablast phenotype determined by FACS after 3 d of stimulation as indicated in D. (F) Purified B cells were stimulated with LPS in the absence or presence of the inhibitors indicated. The amount of IgM in cell supernatants was quantitated by ELISA after 4 d, and data were normalized to the stimulated, no-drug treatment condition. (G) Aicda mRNA transcripts were quantitated by qPCR in B cells stimulated with αCD40 + IL-4. For A and D, data are representative of three or more independent experiments. For B, C, E, F, and G, graphs depict the mean and SEM of n ≥ 3 samples per condition. (*P < 0.05; **P < 0.01; ***P < 0.001 by one-way ANOVA with Tukey’s multiple comparison test, measured vs. the no-drug sample). Akti, Akt inhibitor VIII; AZD, AZD8055; INK, INK128; Ku, Ku-0063794; 242, PP242.

Fig. 3.

Transient treatment with AZD8055 increases class-switched antibody responses in vivo. Mice were immunized with NP-OVA and treated daily for 4 d with AZD8055 at 20 mg/kg (PID −1 to PID 2). (A and B) Total and high-affinity NP-specific IgG1 production was measured by ELISA in sera collected during both the primary (PID 21) and secondary (PID 28) immune responses. (C) Percentage of spleen cells with a T_FH phenotype (CD4⁺CXCR5^highPD-1⁺) was determined by FACS at PID 28. (D) Purified CD4⁺ T cells were cultured under skewing conditions and restimulated to determine the percentage of cells producing cytokines characteristic of T_FH cells (IL-21⁺IL-17A⁻). Cells were cultured in the continuous presence of inhibitors (Left) or for 24 h with inhibitors, followed by washout of drug from the cultures (Right). (E) Total and high-affinity NP-specific IgM was measured at PID 28. (F) Percentage of spleen cells with a GC phenotype (B220⁺, IgD^low, CD38⁻, and Fas⁺) was determined by FACS at PID 28. (*P < 0.05; **P < 0.01; ***P < 0.001 by one-way ANOVA with Tukey’s multiple comparison test, measured vs. the vehicle group).

Fig. 4.

Transient treatment with AZD8055 improves the humoral response in aged mice in vivo. Aged mice (∼1.5 y old) were transiently treated with AZD8055 as described above and immunized with NP-OVA. (A) Production of total and high-affinity NP-specific IgG1 was measured by ELISA, with vehicle-treated young immunized mice as a comparison. (B) Percentage of spleen cells with a GC phenotype was determined by FACS at PID 28. (*P < 0.05, one-way ANOVA with Tukey’s multiple comparison test, measured vs. the vehicle group.)

Fig. 5.

Rapamycin has a more profound effect on B-cell proliferation and CSR than TOR-KIs. Purified B cells were labeled with eFluor670 to measure cell division and were stimulated with αCD40 + IL-4 in the presence or absence of indicated inhibitors. Representative FACS plots are shown for the different treatment conditions (Upper), and a FACS proliferation histogram is shown for high-dose rapamycin and INK128 (Middle). (Lower) Effects of inhibitor concentration on cellular proliferation and CSR are represented on line graphs. Data are representative of two independent experiments.

Fig. 6.

Inactivation of mTORC1 vs. mTORC2 has opposing effects on CSR. (A) Purified resting B cells from either control (CD19Cre^+/−) or rictor-flox/CD19Cre (rictor^ΔB) were subjected to Western blotting to check rictor protein levels (Left). RBC-lysed total splenocytes from the indicated genotypes were activated with αCD40 + IL-4 for 24 h, and intracellular phosphoflow (p-Flow) analysis was performed to measure mTORC1 (pS6-S240/244) and mTORC2 (pAKT-S473) activity. The fold change in median fluorescence intensity was calculated and subjected to log₂ conversion to obtain p-Flow score values as previously described (9). Data are reported as mean ± SD [n = 3 (pAKT) or n = 2 (pS6)]. An unpaired Student t test was used for statistical analysis of the pAKT data. I.e., long exposure; s.e., short exposure. (B) Purified B cells were labeled with CFSE and activated with αCD40 + IL-4 for 4 d. The percentage of live B cells that had divided at least once (based on CFSE division history) expressing surface IgG1 was determined by FACS after 4 d. (C, Left) Percentage of B cells expressing surface IgG1 after 4 d of activation with αCD40 + IL-4 is plotted for several experiments. In the +TOR-KI condition, INK128 was added at a concentration of 1–5 nM. (C, Right) Percentage of B cells expressing surface CD138 after 4 d of activation with LPS is plotted. (D, Left) Western blotting was performed the same way as in A but with resting B cells from control (raptor^fl/fl) or raptor-flox/CD21Cre (raptor^ΔB). (D, Right) Intracellular p-Flow analysis to measure mTORC1 and mTORC2 activity after 24 h of αCD40 + IL-4 activation. Data are reported as mean ± SD (n = 3) for both pAKT and pS6. An unpaired Student t test was used for statistical analysis. (E) Representative FACS plots as in B. (F) Same analysis as in C. A paired Student t test was used for statistical analysis of both C and F.

Fig. 7.

Increases in B-cell CSR are mediated by FoxO transcription factors. (A) Purified B cells from polyI:C-treated FoxO triple-floxed mice lacking Mx1-Cre were treated with the indicated inhibitors, followed by stimulation with αCD40 + IL-4. The percentage of IgG1-switched cells was determined as in Fig. 2. (A and B) PolyI:C-treated FoxO triple-floxed mice (FoxO1/2/3^fl/fl) expressing Mx1-Cre were treated with polyI:C and stimulated with αCD40 + IL-4. Intracellular staining was used to distinguish FoxO-deleted cells (indicated by red − signs) from nondeleted cells (indicated by red + signs) in the same population. (B) Percentage of IgG1-switched B cells was reported as mean ± SEM (**P < 0.01; ***P < 0.001 by one-way ANOVA with Tukey’s multiple comparison test, measured on FoxO-sufficient cells vs. the no-drug sample). GDC, GDC-0941.

Fig. 8.

Roles of mTOR complexes in CSR and the effects of distinct classes of mTOR inhibitors. Activated B cells initiate CSR in response to T cell-derived signals (CD40, IL-4) and/or innate pattern recognition [Toll-like receptor (TLR) engagement]. These signals converge on the lipid kinase PI3K, leading to activation of the Ser/Thr kinase AKT. The predominant action of AKT is to suppress CSR by inactivating FoxO transcription factors that are required for this process (22, 25). Thus, inhibitors of PI3K or AKT tend to increase CSR. mTOR exists in two complexes with complex roles in CSR. mTORC1 (defined by the raptor subunit) promotes CSR through an unknown mechanism (multiple arrows) that is partially independent of the role of mTORC1 in B-cell proliferation. Consequently, deletion of raptor or treatment with the selective mTORC1 inhibitor rapamycin suppresses CSR. mTORC2 (defined by the rictor subunit) suppresses CSR by promoting AKT phosphorylation, leading to FoxO inactivation. Thus, partial deletion of rictor increases CSR. When B cells are treated with TOR-KIs at submaximal concentrations that preserve proliferation, the mTORC2 effect dominates and CSR increases. At higher concentrations, full inhibition of both mTORC1 and mTORC2 suppresses proliferation and CSR. Although AKT can promote mTORC1 activity (multiple arrows), inhibition of PI3K/AKT does not completely suppress mTORC1 signaling in activated B cells (36).