Manuel Castro de Moura

Resistance to Bruton tyrosine kinase (BTK) inhibitors remains a major clinical challenge in B-cell lymphomas and often occurs in the absence of BTK or PLCG2 mutations. Here, we investigated non-genetic mechanisms of ibrutinib resistance in marginal zone lymphoma (MZL) and their broader therapeutic implications. Chronic ibrutinib exposure generated a resistant MZL model that showed cross-resistance to BTK inhibitors and degraders, without evidence of multidrug resistance or genetic alterations. Integrated transcriptomic, epigenetic, and proteomic analyses revealed extensive reprogramming, including activation of PI3K/AKT, MAPK, and MYC pathways, repression of apoptosis and oxidative phosphorylation, and a prominent cytokine-secretory phenotype. Interleukin-16 (IL-16) emerged as a central mediator of resistance. IL-16 was transcriptionally upregulated, actively secreted, and sufficient to induce ibrutinib resistance across multiple CD9-positive models of MZL, mantle cell lymphoma, chronic lymphocytic leukemia (CLL), and activated B-cell-like diffuse large B-cell lymphoma. Serum IL-16 levels were elevated in ibrutinib-refractory CLL patients without BTK/PLCG2 mutations. Mechanistically, IL-16 engaged CD9-enriched membrane microdomains to activate PI3Kδ, thereby sustaining AKT and ERK signaling, stabilizing MYC, inducing NF-κB-dependent programs, and upregulating antiapoptotic effectors, including BFL1 (BCL2A1). Pharmacological or genetic disruption of the IL-16/CD9/PI3K axis restored sensitivity to BTK inhibitors and R-CHOP and abrogated IL-16-induced signaling in primary CLL samples. In conclusion, an IL-16/CD9-driven, epigenetically regulated survival pathway represents one possible mechanism of resistance to BTK inhibitors and chemoimmunotherapy, supporting therapeutic targeting of this axis in refractory B-cell lymphomas.

Introducing Bruton’s tyrosine kinase (BTK) inhibitors has significantly improved outcomes for patients with B-cell malignancies and autoimmune disorders. However, resistance, either primary or acquired, remains a major clinical challenge. To better understand the underlying resistance mechanisms to BTK inhibitors, we established an ibrutinib-resistant model from a patient-derived splenic marginal zone lymphoma (MZL) cell line (VL51) through prolonged drug exposure. Resistant cells exhibited a 15-fold increase in ibrutinib’s IC50, along with distinct morphological changes, mitochondrial activation, and cross-resistance to covalent, non-covalent BTK inhibitors and BTK degraders. Integrated transcriptomic, epigenomic, and proteomic analyses identified overexpression and secretion of IL-16 as a key feature of resistance, driven by chromatin remodeling and activation of the FLI1 transcription factor. IL-16 conferred ibrutinib resistance via CD9-mediated activation of the NF-κB and AKT signaling pathways and was found to be elevated in the serum of ibrutinib-refractory CLL patients. Functional studies showed that targeting the IL-16/CD9 axis using neutralizing antibodies or CD9-binding peptides restored sensitivity to BTK inhibitors and R-CHOP chemotherapy in MZL, mantle cell lymphoma, and diffuse large B-cell lymphoma models. These findings reveal a novel, targetable resistance mechanism with potential therapeutic implications for overcoming BTK inhibitor resistance in B-cell lymphomas.

Expression levels of surface ERBB4 and pERBB4 was measured by FACS (A) and immunoblotting (B) in Karpas1718 parental and resistant. (A) Two independent experiments were performed: exp1 (left panel) and exp2 (right panel). Density plots show the median MFI values of two replicates from each experiment: negative control (dotted black), parental (grey), resistant (red). Results are presented from Welch Two Sample test performed on MFI values of the two experiments. (B) Immunobloting for total (grey) and phospho (red) ERBB4 in parental (left) and resistant (right) Karpas1718 cells. Representative image of two independent experiments. Data on the barplot represents the mean of two experiments, errors bars represent standard deviation of the mean. * for p<0.05 from a t-test.

Panel of proteins tested in the immunoblotting or immunofluorescence experiments.

Expression levels of crucial factors are correlated with resistance to idelalisib. HBEGF (A) and ERBB4 (B) expression levels are inversely correlated with idelalisib sensitivity. Conversely expression of let-7c (D) and miR-29c (F) microRNAs is associated with sensitivity to idelalisib. Expression and sensitivity data were analyzed from a previous publication of our group in a panel of 34 B cell lymphoma cell lines (4). Cell lines were split in two groups based on higher or lower values than the median expression of the corresponding gene or miRNA. Mean of idelalisib IC50 were calculated for these two groups and compared by t-test.

Parental (green) and resistant (orange) levels of DNA promoter methylation (A) and gene expression (RNA-seq, B) of DUSP4, and p-ERK upon 1µM of idelalisib (Flow cytometry).

(A) Cell viability for the combination of idelalisib and the ERBB inhibitor lapatinib in parental and resistant by MTT assay (72h). Bars correspond to the mean of two independent experiments. Error bars represent standard deviation of the mean. Table contains p-values from a moderated t-test comparing each combination to idelalisib as single agent. The benefit of the combination was assessed both as synergism according to the Chou-Talalay combination index (CI, synergistic: CI1) (B) (2) and as potency (x-axis, synergistic: potency>0.5, additive: 01, additive: 010, additive 0>ZIP>10, antagonistic: ZIP10, additive: 0>HSA>10, antagonistic HSA<0), in additional lymphoma cell lines including the DLBCL lines TMD8, and OCILY10, and the MCL models SP53 and UPN1.

Panel of antibodies used in flow cytometry experiments.

Multi-drug resistance phenotype was ruled out by the expression of MDR1 (left) and MDR2 (right) genes by real time PCR. RQ values calculated by the DDCt method. Karpas1718 (K1718) parental lines in grey and resistant in red. Data derived from two independent experiments; error bars represent standard deviation of the mean.

Expression of let-7c (A) and miR-29c (B) miRNAs was evaluated in Karpas1718 parental and resistant lines by real time PCR upon previous exposure to DMSO (control, grey), decitabine (100nM, yellow) or tazemetostat (5µM, red). RQ values calculated by the DDCt method, RNU6 was used as housekeeping miRNA. Data derived from three independent experiments.