Ibrutinib therapy downregulates AID enzyme and proliferative fractions in chronic lymphocytic leukemia

Abstract

Activation-induced cytidine deaminase (AID) initiates somatic hypermutation and class switch recombination of the immunoglobulin genes. As a trade-off for its physiological function, AID also contributes to tumor development through its mutagenic activity. In chronic lymphocytic leukemia (CLL), AID is overexpressed in the proliferative fractions (PFs) of the malignant B lymphocytes, and its anomalous expression has been associated with a clinical poor outcome. Recent preclinical data suggested that ibrutinib and idelalisib, 2 clinically approved kinase inhibitors, increase AID expression and genomic instability in normal and neoplastic B cells. These results raise concerns about a potential mutagenic risk in patients receiving long-term therapy. To corroborate these findings in the clinical setting, we analyzed AID expression and PFs in a CLL cohort before and during ibrutinib treatment. We found that ibrutinib decreases the CLL PFs and, interestingly, also reduces AID expression, which correlates with dampened AKT and Janus Kinase 1 signaling. Moreover, although ibrutinib increases AID expression in a CLL cell line, it is unable to do so in primary CLL samples. Our results uncover a differential response to ibrutinib between cell lines and the CLL clone and imply that ibrutinib could differ from idelalisib in their potential to induce AID in treated patients. Possible reasons for the discrepancy between preclinical and clinical findings, and their effect on treatment safety, are discussed.

Graphical abstract

Figure 1.

Studies of proliferative CLL subsets during ibrutinib treatment of patients with CLL. Peripheral blood samples from patients with CLL before and after 1 and 4 weeks postibrutinib treatment were centrifuged in Ficoll gradient, and total MCs obtained were next analyzed by flow cytometry, using the different markers detailed here. Leukemic cells were discriminated by gating lymphocytes and by labeling CD19 plus CD5. (A) Ki-67⁺ (i), CD38 (ii), CD86 (iii), or CXCR4 (v) surface expression was evaluated in the 3 points depicted. Results are shown as percentages of CD19⁺/CD5⁺/Ki-67⁺ cells (i), CD19⁺/CD5⁺/CD38⁺ cells (ii), CD19⁺/CD5⁺/CD86⁺ cells (iii), or CD19⁺/CXCR4^low/CD5^high cells (v) by fixing the gate at the pretreated condition. (iv) Intracellular labeling of IgM plus IgG was performed where percentages of CD19⁺/IgM⁺/IgG⁺ cells before and after treatment is shown. As depicted for CD19⁺/Ki-67⁺, CD19⁺/CD38⁺; CD19⁺/CD86⁺, IgM⁺/IgG⁺, and CXCR4^lowCD5^high, the P values were P ≤ .0001, P = .001, P = .011; P = .016, and P = .005; respectively (n = 10). Each dot represents a single patient sample (*P < .05; ** P < .005; 2-tailed, Student paired t test). (B) Intracellular staining of Ki-67 was developed in CD19⁺/IgM⁺/IgG⁺ cells. Shown are representative dot plots and Ki-67 histograms in a representative patient (left) and the statistical analysis of the entire cohort evaluated (right part), where each dot represents a single patient sample. Pretreated IgM⁺/IgG⁺/Ki67⁺ was 5.6% vs 1.7% at 1 week (mean difference, 3.9; 95% confidence interval [CI], 1.5-6.3; P = .0013), whereas at 4 weeks, it was 0.3% (mean difference, 3.9; 95% CI, 2.9-7.6; P ≤ .0001; n = 10; 1-way ANOVA, multiple comparisons test). (C) Intracellular staining of Ki-67 was developed in CD19⁺CD5^high/CXCR4^low cells. Shown are representative dot plots and Ki-67 histograms in a representative patient (left part) and the statistical analysis of the entire cohort (right part), where each dot represents a single patient sample. For CXCR4^lowCD5^high, the mean proportion of cells expressing Ki-67 at pretreatment is 14.9% vs 5.8% at week 1 (mean difference, 9.1; 95% CI, 2.4-15.7; P = .042); and for week 4, 1.36%, with a mean difference of 13.5 (95% CI, 6.7-20.3; P = .017; n = 10; 1-way ANOVA, multiple comparisons test).

Figure 2.

Phosphorylation profile of CD5^high/CXCR4^low leukemic cells in patients with CLL treated with ibrutinib. Cell sorting purified CD19⁺/CD5^high/CXCR4^low cells obtained from PBMC of 3 patients with CLL before and after 4 weeks of ibrutinib therapy were used to perform a phosphoarray analysis. (A) Volcano plot showing the fold change vs the P value obtained for each evaluated phosphorylation site. Black dots refer to sites that did not change after treatment (63 phosphorylation sites). Depicted in color are the sites that were significantly dephosphorylated (green dots, 12 sites) or phosphorylated (red dots, 19 sites) during therapy (P < .05; n = 3; 2-tailed paired Student t test). The reference for each numbered dot is depicted in panel B. (B) Heat map showing the targeted proteins with their corresponding phosphorylation sites in 3 CLL patient samples. Heat map scale is shown at the bottom. (C) Statistically significant changes in phosphorylation rate (red, upper panel) or dephosphorylation rate (green, lower panel) during ibrutinib administration are shown. Each square represents the signal ratio of the antibody against the unphosphorylated form of the site before and after treatment, connected with the round dot for the signal ratio of the phosphorylated form (P < .05; n = 3; 2-tailed paired Student t test). Relevant proteins associated with inactivation of PI3K/AKT, JAK/STAT, and p53 pathway are indicated in bold at the x axes.

Figure 3.

Analysis of AID expression in B cells from ibrutinib-treated patients with CLL. PBMC samples from patients with CLL before and after 1 and 4 weeks of ibrutinib in vivo therapy were studied by flow cytometry to measure intracellular AID protein. Leukemic cells were discriminated by gating lymphocytes and by labeling CD19 plus CD5. Expression of AID as percentage of positive cells in the proliferative CD19⁺/IgM⁺/IgG⁺ subset compared with IgM⁺/IgG^- quiescent cells or from proliferative CD19⁺/CD5^high/CXCR4^low vs CD19⁺/CD5^low/CXCR4^high quiescent cells for treatment at 3 times. (A) Results for the IgM⁺/IgG⁺ subset in the whole group of patients evaluated (left graphic) and 1 representative cytometry data (right graphic). The corresponding median values were 7.2% pre- and 2.5% at week 4 posttreatment (mean difference, 4.7%; 95% CI, 2.3-7.2; P = .0007, 1-way ANOVA, multiple comparisons test). Nonsignificant differences were found between pre- and 1 week posttreatment samples (P = .810). Right panel shows representative dot plots of the gates constructed to select the subsets and from where the histograms of AID expression were obtained. PF and QF are depicted in green and red, respectively. (B) Results for the CXCR4 low/CD5 high fraction in the whole group of patients evaluated (left graphic) and 1 representative cytometry data (right graphic). Significant differences were found comparing pr-treatment mean (10%) vs mean posttreatment at week 1, 4.5%; (mean difference, 5.68%; 95% CI, 0.91-10.45; P = .278) and vs week 3, 1.13% (mean difference, 9.5%; 95% CI, 4.3%-13.8%; P = .0016 by 1-way ANOVA, multiple comparisons test). PF and QF are depicted in green and red, respectively. (C-D) MFI AID expression on the quiescent and proliferative subsets CD19⁺IgM⁺IgG⁺ compared with IgM⁺/IgG^- quiescent cells or from proliferative CD19⁺/CD5^high/CXCR4^low vs CD19⁺/CD5^low/CXCR4^high quiescent cells, at the points indicated. Intracellular staining of AID as MFI was evaluated. Shown are representative dot plots and gate criteria (higher graphics), as well as AID histograms of cytometric MFI shifts with treatment (lower left graphic) for a single patient. The statistics of the whole cohort evaluated are shown in the lower right graphics. The mean of cells expressing AID in IgM⁺/IgG⁺ PF at pretreatment was 9.7 MFI vs 6.0 MFI at week 4 (mean differences, 3.7; 95% CI, 1.4-6.0; P = .0004; n = 10). For CXCR4^lowCD5^high PF pretreatment, 7.0 MFI vs 4.5 at week 4 MFI (mean differences, 2.4; 95% CI, 1.0-3.7; P = .0004; n = 5). One-way ANOVA with Tukey's multiple comparisons test was used in all cases (***P < .0005).

Figure 4.

Analysis of AID expression in CLL cells treated in vitro with ibrutinib. (A) MEC-1 cell line was cultured in the presence of ibrutinib at 0.1 and 1 µM for 24 and 48 hours, or in DMSO as control. AID expression at the protein level was evaluated by immunoblotting and flow cytometry techniques. Shown are representative immunoblots (upper graphic) and mean fluorescence cytometry histograms (lower panels), where increasing AID levels are detected for all treated conditions. Proliferation measured by Ki-67 expression was performed at baseline and after 24 and 48 hours of ibrutinib treatment (right graphic). (B) PB CLL cells were incubated in the presence of ibrutinib at 0.3 and 1 µM. DMSO was used as basal control, and CD40L⁺IL-4 stimulation to induce AID expression was used as positive control (n = 5). AICDA mRNA and protein expression was evaluated at 24 and 48 hours, as indicated in the x-axis. Shown are the results from the whole cohort of patients studied by quantitative qPCR (left graphic) and the immunoblot obtained when studying the 2 patients colored in red and blue (right graphic). AID increase was detected only in the CD40L⁺IL-4 activated condition for both mRNA and protein determinations. (C) PBMC from 5 patients with CLL were incubated with CD40⁺IL-4 or CD40⁺IL-4⁺ibrutinib 1 μM, and then cultured for 5 days. Proliferation was evaluated by Ki-67 expression, and viability of cells was analyzed by propidium iodide as a life/death marker (upper graphics). As previously described by Slinger et al, maximum inhibition of proliferation was achieved after 24 hours. AICDA mRNA levels and AID protein were tested by qPCR and immunoblotting, respectively, at the different times and conditions (lower graphics). Significance was calculated comparing stimulated condition (CD40⁺IL-4) with stimulated condition + ibrutinib by 2-tailed, paired t-test (*P < .05; **P < .01). (D) PBMC from 3 patients with CLL were incubated in the presence of CD40L⁺IL-4 in the same conditions as mentioned here. Samples without treatment (RPMI+FBS 10%+DMSO -ctrol-), samples treated with CD40L⁺IL-4 and samples treated with CD40L⁺IL-4⁺1 μM of ibrutinib were tested. Glyceraldehyde-3-phosphate dehydrogenase (load control); AID; STAT6 and p-STAT6^-Tyr641 proteins expression were evaluated at day 5. Immunoblots depict the effect of ibrutinib on AID expression and STAT6 phosphorylation. Whereas AID and p-STAT6^-Tyr641 decrease during ibrutinib treatment, the nonphosphorylated form of STAT-6 remains unchanged. The same proteins were analyzed in the MEC-1 cell line after incubation with ibrutinib (ibru) or DMSO as control (Ctrol). As depicted, p-STAT6^-Tyr641 was undetectable in neither condition, whereas the nonphosphorylated STAT-6 form remains unchanged. As previously described, AID expression increased when MEC1 cells are cultured in presence of ibrutinib.

Figure 4.

Analysis of AID expression in CLL cells treated in vitro with ibrutinib. (A) MEC-1 cell line was cultured in the presence of ibrutinib at 0.1 and 1 µM for 24 and 48 hours, or in DMSO as control. AID expression at the protein level was evaluated by immunoblotting and flow cytometry techniques. Shown are representative immunoblots (upper graphic) and mean fluorescence cytometry histograms (lower panels), where increasing AID levels are detected for all treated conditions. Proliferation measured by Ki-67 expression was performed at baseline and after 24 and 48 hours of ibrutinib treatment (right graphic). (B) PB CLL cells were incubated in the presence of ibrutinib at 0.3 and 1 µM. DMSO was used as basal control, and CD40L⁺IL-4 stimulation to induce AID expression was used as positive control (n = 5). AICDA mRNA and protein expression was evaluated at 24 and 48 hours, as indicated in the x-axis. Shown are the results from the whole cohort of patients studied by quantitative qPCR (left graphic) and the immunoblot obtained when studying the 2 patients colored in red and blue (right graphic). AID increase was detected only in the CD40L⁺IL-4 activated condition for both mRNA and protein determinations. (C) PBMC from 5 patients with CLL were incubated with CD40⁺IL-4 or CD40⁺IL-4⁺ibrutinib 1 μM, and then cultured for 5 days. Proliferation was evaluated by Ki-67 expression, and viability of cells was analyzed by propidium iodide as a life/death marker (upper graphics). As previously described by Slinger et al, maximum inhibition of proliferation was achieved after 24 hours. AICDA mRNA levels and AID protein were tested by qPCR and immunoblotting, respectively, at the different times and conditions (lower graphics). Significance was calculated comparing stimulated condition (CD40⁺IL-4) with stimulated condition + ibrutinib by 2-tailed, paired t-test (*P < .05; **P < .01). (D) PBMC from 3 patients with CLL were incubated in the presence of CD40L⁺IL-4 in the same conditions as mentioned here. Samples without treatment (RPMI+FBS 10%+DMSO -ctrol-), samples treated with CD40L⁺IL-4 and samples treated with CD40L⁺IL-4⁺1 μM of ibrutinib were tested. Glyceraldehyde-3-phosphate dehydrogenase (load control); AID; STAT6 and p-STAT6^-Tyr641 proteins expression were evaluated at day 5. Immunoblots depict the effect of ibrutinib on AID expression and STAT6 phosphorylation. Whereas AID and p-STAT6^-Tyr641 decrease during ibrutinib treatment, the nonphosphorylated form of STAT-6 remains unchanged. The same proteins were analyzed in the MEC-1 cell line after incubation with ibrutinib (ibru) or DMSO as control (Ctrol). As depicted, p-STAT6^-Tyr641 was undetectable in neither condition, whereas the nonphosphorylated STAT-6 form remains unchanged. As previously described, AID expression increased when MEC1 cells are cultured in presence of ibrutinib.