Regulation of Bcl-XL by non-canonical NF-κB in the context of CD40-induced drug resistance in CLL

Abstract

In chronic lymphocytic leukemia (CLL), the lymph node (LN) microenvironment delivers critical survival signals by inducing the expression of anti-apoptotic Bcl-2 members Bcl-XL, Bfl-1, and Mcl-1, resulting in apoptosis blockade. We determined previously that resistance against various drugs, among which is the clinically applied BH3 mimetic venetoclax, is dominated by upregulation of the anti-apoptotic regulator Bcl-XL. Direct clinical targeting of Bcl-XL by, e.g., Navitoclax is however not desirable due to induction of thrombocytopenia. Since the actual regulation of Bcl-XL in CLL in the context of the LN microenvironment is not well elucidated, we investigated various candidate LN signals to drive Bcl-XL expression. We found a dominance for NF-κB signaling upon CD40 stimulation, which results in activation of both the canonical and non-canonical NF-κB signaling pathways. We demonstrate that expression of Bcl-XL is first induced by the canonical NF-κB pathway, and subsequently boosted and continued via non-canonical NF-κB signaling through stabilization of NIK. NF-κB subunits p65 and p52 can both bind to the Bcl-XL promoter and activate transcription upon CD40 stimulation. Moreover, canonical NF-κB signaling was correlated with Bfl-1 expression, whereas Mcl-1 in contrast, was not transcriptionally regulated by NF-κB. Finally, we applied a novel compound targeting NIK to selectively inhibit the non-canonical NF-κB pathway and showed that venetoclax-resistant CLL cells were sensitized to venetoclax. In conclusion, protective signals from the CLL microenvironment can be tipped towards apoptosis sensitivity by interfering with non-canonical NF-κB signaling.

Fig. 1. CD40 stimulation strongly activates NF-κB signaling and Bcl-XL expression in CLL.

CLL cells were cultured for 24 h on fibroblasts (3T3), fibroblasts transfected with human CD40L (3T40L), stimulated with CpG, or stimulated with IL-4 and αIgM. A Protein lysates were probed for p100, p52, p-p65, Bcl-XL, Bfl-1, and actin as loading control. B Densitometric analysis of p100, p52, p-p65, Bcl-XL, and Bfl-1 are shown normalized to the actin loading control (n = 7). Bars represent the mean ± SEM, *p < 0.05, **p < 0.01 (two-way ANOVA).

Fig. 2. Bcl-XL plays an important role in venetoclax drug resistance in CLL.

A,B CLL cells were cultured for 24 h on 3T3 or 3T40L. After detachment, cells were treated with a titration of ABT-199 (0–2500 nM) for 24 h with or without inhibition of Bcl-XL (A-1331852). For inhibition of Bcl-XL, a titration of 0-30-117-468-1875-7500 nM A-1331852 was used. Viability (A) was measured using flow cytometry using DiOC6 and TO-PRO3 viability stainings (Thermo Scientific). Specific apoptosis (B) was calculated. Averaged data of 4 CLL samples are shown. Bars represent the mean ± SEM, *p < 0.05, **p < 0.01 (paired t-test (3T40L versus 3T40L + A-1331852)).

Fig. 3. CD40 stimulation induces Bcl-XL expression via canonical as well as non-canonical NF-κB signaling.

A HEK293T cells were transfected with luciferase reporter gene constructs carrying Bcl-XL promoter truncations. NF-κB binding sites were predicted using JASPAR database. Promoter activity was measured 24 h post-transfection. Bars represent the mean ± SEM of the luciferase activation normalized to the empty pGL3-basic vector (n = 3). B HEK293T cells were transfected with luciferase reporter gene constructs carrying p100 promoter truncations, or a p100 promoter, where the first NF-κB1 binding site was scrambled (depicted by an arrow). NF-κB binding sites were predicted using JASPAR database. Promoter activity was measured 24 h post-transfection. Bars represent the ± SEM of the luciferase activation normalized to the empty pGL3-basic vector (n = 7). C Bcl-XL, p100, and Mcl-1 mRNA expression by CLL cells over time after stimulation with CD40L measured by real-time PCR (n = 6). Results are shown as the mean ± SEM. D CLL cells were cultured on 3T3 or 3T40L for 6, 16, or 24 h. Protein lysates were probed for p100, p-p65, p52, Bcl-XL, and actin as loading control.

Fig. 4. Non-canonical NF-κB signaling is critical for Bcl-XL expression in CLL.

A HEK293T cells were transfected with empty pGL3-basic vector (mock) or HA-tagged NIK with or without cotransfection of p100. Protein lysates were probed for HA, p100, p-p65, p52, and actin as loading control. B HEK293T cells were transfected with Bcl-XL promoter luciferase reporter gene constructs as well as HA-tagged NIK, with or without cotransfection of p100 (n = 2). Promoter activity was measured and bars represent the mean ± SEM of the luciferase activation normalized to the empty pGL3-basic vector, **p < 0.01 (one-way ANOVA). C CLL cells were nucleofected with either a non-targeting control siRNA (siCtrl) or an siRNA targeting NIK (siNIK) and subsequently cultured on 3T40L for 24 h. After detachment, protein lysates were made and probed for p100, p-p65, p52, Bcl-XL, and actin as loading control. Blots of two representative CLL samples are shown. D Densitometric analysis of p100, p52, p-p65, and Bcl-XL are shown (n = 6). Bars represent the mean ± SEM, *p < 0.05, **p < 0.01, ****p < 0.0001 (two-way ANOVA). E–H CLL cells were nucleofected with NIK siRNA or control siRNA and cultured on 3T40L for 24 h. NIK (n = 10) (E), Bcl-XL (n = 10) (F), Bfl-1 (n = 10) (G), and IκBα (n = 4) (H) mRNA expression was measured by real-time PCR. Samples were normalized to HPRT, **p < 0.01, ****p < 0.0001 (paired sample t-test). I After detachment, cells were incubated with 0.001–10 μM ABT-199 for 24 h after which viability was measured by flow cytometry using DiOC6 and TO-PRO-3 viability dyes (n = 4). Bars represent the mean ± SEM, *p < 0.05 (paired sample t-test (3T40L + siCtrl versus 3T40L + siNIK)).

Fig. 5. Targeting non-canonical NF-κB signaling downregulates Bcl-XL and increases sensitivity to venetoclax.

A Structural formula of 4-(1-(2-aminopyrimidin-4-yl)indolin-6-yl)-2-methylbut-3-yn-2-ol (CW15337). B NIK inhibitory activity of 0.001–30 µM CW15337 was determined using the Kinase-Glo^® assay. Results are shown as mean ± SEM (n = 6). C IKKα and IKKβ inhibitory activity of up to 30 µM CW15337 was determined using the dissociation-enhanced ligand fluorescent immunoassay. To measure the kinase activity of IKKα/β, recombinant IKKα/β were incubated with a peptide substrate containing the phosphorylation motif of IκBα and ATP solution in presence/absence of CW15337. Kinase activity was quantified based on the amount of phosphorylated substrate relative to the untreated control. D CLL cells were cultured on 3T40L and simultaneously treated with a titration of CW15337 for 24 h. After detachment, viability was measured by flow cytometry using DiOC6 and TO-PRO-3 viability dyes. Bars represent the mean ± SEM (n = 9). E In addition, cells were stained intracellularly for Bcl-XL and measured by flow cytometry. Bars represent the mean ± SEM (n = 7). F MFI peaks of 1 representative sample is shown. G Protein lysates were probed for p100, p52, p-p65, Bcl-XL, and actin as loading control. Blot from 1 representative CLL sample is shown. H Densitometric analysis of p100, p52, p-p65, and Bcl-XL are shown (n = 7). Bars represent the mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 (two-way ANOVA). I After detachment, cells were incubated with a titration of 0.001–10 µM ABT-199 for 24 h after which viability was measured by flow cytometry using DiOC6 and TO-PRO-3 viability dyes. Bars represent the mean ± SEM (n = 8).