Development and characterization of prototypes for in vitro and in vivo mouse models of ibrutinib-resistant CLL

Abstract

Although ibrutinib improves the overall survival of patients with chronic lymphocytic leukemia (CLL), some patients still develop resistance, most commonly through point mutations affecting cysteine residue 481 (C481) in Bruton's tyrosine kinase (BTKC481S and BTKC481R). To enhance our understanding of the biological impact of these mutations, we established cell lines that overexpress wild-type or mutant BTK in in vitro and in vivo models that mimic ibrutinib-sensitive and -resistant CLL. MEC-1 cell lines stably overexpressing wild-type or mutant BTK were generated. All cell lines coexpressed GFP, were CD19+ and CD23+, and overexpressed BTK. Overexpression of wild-type or mutant BTK resulted in increased signaling, as evidenced by the induction of p-BTK, p-PLCγ2, and p-extracellular signal-related kinase (ERK) levels, the latter further augmented upon IgM stimulation. In all cell lines, cell cycle profiles and levels of BTK expression were similar, but the RNA sequencing and reverse-phase protein array results revealed that the molecular transcript and protein profiles were distinct. To mimic aggressive CLL, we created xenograft mouse models by transplanting the generated cell lines into Rag2-/-γc-/- mice. Spleens, livers, bone marrow, and peripheral blood were collected. All mice developed CLL-like disease with systemic involvement (engraftment efficiency, 100%). We observed splenomegaly, accumulation of leukemic cells in the spleen and liver, and macroscopically evident necrosis. CD19+ cells accumulated in the spleen, bone marrow, and peripheral blood. The overall survival duration was slightly lower in mice expressing mutant BTK. Our cell lines and murine models mimicking ibrutinib-resistant CLL will serve as powerful tools to test reversible BTK inhibitors and novel, non-BTK-targeted therapeutics.

Graphical abstract

Figure 1.

In vitro characterization of MEC-1 cells overexpressing WT and mutant BTK. GFP-labeled MEC-1 cell lines stably overexpressing mutant BTK (BTK^C481S or BTK^C481R) and BTK^WT were generated by using standard lentiviral transduction methods. Cells were sorted to enrich the transduced GFP⁺ cell populations in each cell line. For all experiments, a population of cells >75% GFP⁺ was used. (A) Phosphorylated (Y223) and total BTK proteins were overexpressed in transduced cells. p-PLCγ2 (Tyr1217) and p-ERK (Thr202/Tyr204) levels were also increased in all BTK-overexpressing cells. Vinculin was the loading control. (B) IgM stimulation in transduced MEC-1 cells. Cells (5 ×10⁵ per milliliter) were seeded in flasks, with or without IgM (20 µg/mL) and incubated at 37°C for 15 minutes. Protein extracts were subjected to immunoblot assays to determine the levels of pBTK (Y223), BTK, pERK (Thr202/Tyr204), and ERK. Vinculin was the loading control. (C) Exponentially growing GFP⁺ cells were seeded in flasks and incubated with ibrutinib (0.01, 0.1, and 1 µM) for 3 hours. Protein extracts were subjected to immunoblot assays to determine the levels of pBTK (Y223), BTK, pERK (Thr202/Tyr204), and ERK. Vinculin was the loading control. (D) Validation of cell surface markers. Cells were stained with anti-CD19-PECy7 and anti-CD23-PE monoclonal antibodies and analyzed with a flow cytometer. Flow cytometry dot plots showing GFP⁺ cells (left) and CD19⁺CD23⁺ cells (right) after gating on the GFP⁺ cell population.

Figure 1.

In vitro characterization of MEC-1 cells overexpressing WT and mutant BTK. GFP-labeled MEC-1 cell lines stably overexpressing mutant BTK (BTK^C481S or BTK^C481R) and BTK^WT were generated by using standard lentiviral transduction methods. Cells were sorted to enrich the transduced GFP⁺ cell populations in each cell line. For all experiments, a population of cells >75% GFP⁺ was used. (A) Phosphorylated (Y223) and total BTK proteins were overexpressed in transduced cells. p-PLCγ2 (Tyr1217) and p-ERK (Thr202/Tyr204) levels were also increased in all BTK-overexpressing cells. Vinculin was the loading control. (B) IgM stimulation in transduced MEC-1 cells. Cells (5 ×10⁵ per milliliter) were seeded in flasks, with or without IgM (20 µg/mL) and incubated at 37°C for 15 minutes. Protein extracts were subjected to immunoblot assays to determine the levels of pBTK (Y223), BTK, pERK (Thr202/Tyr204), and ERK. Vinculin was the loading control. (C) Exponentially growing GFP⁺ cells were seeded in flasks and incubated with ibrutinib (0.01, 0.1, and 1 µM) for 3 hours. Protein extracts were subjected to immunoblot assays to determine the levels of pBTK (Y223), BTK, pERK (Thr202/Tyr204), and ERK. Vinculin was the loading control. (D) Validation of cell surface markers. Cells were stained with anti-CD19-PECy7 and anti-CD23-PE monoclonal antibodies and analyzed with a flow cytometer. Flow cytometry dot plots showing GFP⁺ cells (left) and CD19⁺CD23⁺ cells (right) after gating on the GFP⁺ cell population.

Figure 2.

Comparison of the transcriptomic profiles of WT and mutant-BTK–overexpressing cells. (A-C) Volcano plots comparing MEC-1 cells overexpressing BTK^WT and BTK^C481S (A), BTK^WT and BTK^C481R (B), and BTK^C481S and BTK^C481R (C). The DEGs (FDR, ≤0.05; fold change, >2 or < −2) are labeled in red. (D-F) Heat maps of the DEGs (FDR, ≤0.05; fold change, >2 or <−2) between MEC-1 cells overexpressing BTK^WT and BTK^C481S (D), BTK^WT and BTK^C481R (E), or BTK^C481S and BTK^C481R (F).

Figure 3.

Functional protein profiling of WT and mutant-BTK–overexpressing MEC-1 cells. Protein was extracted in biological triplicates from exponentially growing MEC-1 cells with either WT or mutant BTK and subjected to RPPA assays that included 426 antibodies. (A) Representative heat map of proteins with increased and decreased expression in each cell line (n = 3 per cell line). All BTK-overexpressing cell lines (BTK^WT, BTK^C481S, and BTK^C481R) were compared with the BTK^EV cell line. The fold changes of mean linear values were used to generate the heat map (fold change, ≥1.2 or ≤0.8). The color key indicates the fold change. (B) The number of increased (red) and decreased (blue) phosphorylated and total proteins in each transduced cell line compared with the cell line with the empty vector. (C) Validation of RPPA data for some proteins by immunoblot analysis. Protein extracts were subjected to immunoblot assays. To cover all of these proteins, 5 separate immunoblots were prepared, with actin used as the loading control for each. (D) The top 10 canonical pathways identified with Ingenuity Pathway Analysis and associated with the indicated BTKs.

Figure 4.

Comparison of in vivo parameters and tumor development in WT and mutant-BTK–expressing MEC-1 cells. MEC-1 cells harboring WT or mutant BTK were injected into 8-week-old Rag2^−/−γ_c^−/− female mice (n = 10 per group), and the animals were monitored for body weight and development and progression of leukemia. At the end point of the study, the spleens, livers, and femurs were collected and macroscopically evaluated. (A) Body weight of mice (n = 10 per group) in each cohort. (B) Spleen/body weight ratios of mice in each cohort (n = 8 BTK^WT, n = 9 BTK^C481S, n = 10 BTK^C481R; *P < .05). (C) Liver/body weight ratios of mice in each cohort (n = 8 BTK^WT, n = 9 BTK^C481S, n = 10 BTK^C481R). Groups were compared by fitting a linear mixed-effect model for analysis of variance. (D) Kaplan-Meier curves depict the overall survival of mice in each group (n = 10 per group; mutant vs WT, log-rank test; P = .0001).

Figure 5.

Localization of GFP⁺ and mutant-BTK–expressing MEC-1 cells in mice. Tumors were established from MEC-1 cells harboring WT or mutant BTK, as described in Figure 4. (A-B) Representative flow cytometry plots for the spleen and bone marrow samples collected from mice with BTK^WT (n = 4; left), BTK^C481S (n = 9; center), and BTK^C481R (n = 9; right). At the time of euthanization, cells were collected from the spleen (A) and bone marrow (B), stained with a monoclonal antibody against human CD19 to identify the GFP⁺ MEC-1 cells, and analyzed by flow cytometry. (C-D) The percentage of CD19⁺ MEC-1 cells in spleen (C) and bone marrow (D). One-way analysis of variance, P = .05; Student t test, *P < .05 for BTK^WT vs BTK^C481R cells; **P < .01 for BTK^WT vs BTK^C481S cells.

Figure 6.

Histopathological assessment of tissues. Histopathologic assessments of hematoxylin and eosin–stained slides were conducted by a veterinary pathologist. Representative images of tissues and lesions of lymphocytic infiltration were captured at the indicated magnifications for the main and inset images. The evaluation results, including the morphologic diagnoses of the organs and the summaries of the mean scores, are provided in supplemental Table 3. Liver and spleen: n = 8/no disease, n = 8/EV, n = 8/BTK^WT, n = 9/BTK^C481S, and n = 9/BTK^C481R. Bone: n = 8/no disease, n = 6/EV, n = 7/BTK^WT, n = 6/BTK^C481S, and n = 6/BTK^C481R.

Figure 7.

Immunohistochemical analyses of spleen, liver, and bone tissues after the establishment of disease from MEC-1 cells with WT or mutant BTK. Representative macroscopic images of spleens from each group are provided in the top row. Spleen, liver, and bone tissues were collected at the time of euthanasia. B-cell marker expression (CD19 and CD20) and proliferation marker expression (Ki-67) are shown for the spleen, liver, and bone marrow from mice with no disease and from BTK^WT, BTK^C481S, and BTK^C481R mice.