Mutant Calreticulin Requires Both Its Mutant C-terminus and the Thrombopoietin Receptor for Oncogenic Transformation

Abstract

Somatic mutations in calreticulin (CALR) are present in approximately 40% of patients with myeloproliferative neoplasms (MPN), but the mechanism by which mutant CALR is oncogenic remains unclear. Here, we demonstrate that expression of mutant CALR alone is sufficient to engender MPN in mice and recapitulates the disease phenotype of patients with CALR-mutant MPN. We further show that the thrombopoietin receptor MPL is required for mutant CALR-driven transformation through JAK-STAT pathway activation, thus rendering mutant CALR-transformed hematopoietic cells sensitive to JAK2 inhibition. Finally, we demonstrate that the oncogenicity of mutant CALR is dependent on the positive electrostatic charge of the C-terminus of the mutant protein, which is necessary for physical interaction between mutant CALR and MPL. Together, our findings elucidate a novel paradigm of cancer pathogenesis and reveal how CALR mutations induce MPN.

Significance: The mechanism by which CALR mutations induce MPN remains unknown. In this report, we show that the positive charge of the CALR mutant C-terminus is necessary to transform hematopoietic cells by enabling binding between mutant CALR and the thrombopoietin receptor MPL.

Figure 1. Mutant CALR is sufficient to engender an MPN phenotype in mice

(A) Retroviral bone marrow transplant (BMT) scheme. (B) White blood cell (WBC) count, hematocrit (HCT), platelet count (PLT) at 16 weeks post-transplantation in peripheral blood (PB) of recipient mice receiving empty vector (EV), wild-type CALR (CALR^WT) or mutant CALR (CALR^MUT)-expressing c-Kit+ bone marrow (BM) cells (mean ± SD, n=5 in each group) demonstrates significant thrombocytosis in CALR^MUT group (C) Histopathologic H&E sections of BM from representative EV, CALR^WT or CALR^MUT recipient mice demonstrates macro-megakaryocytes, megakaryocytes with hyper-lobated nuclei, megakaryocytic clustering, emperipolesis, and atypical location at sinusoids and trabecular bone in CALR^MUT recipient animals (20X magnification) (D) Megakaryocyte counts per high power field (HPF) in BM of EV, CALR^WT or CALR^MUT recipient mice demonstrates increased megakaryocyte number in CALR^MUT recipient animals. All p values were determined by unpaired two-tailed Student’s t test (*0.01 < p < 0.05; **0.001 < p < 0.01; ns, not significant).

Figure 2. MPL is required for mutant CALR-mediated cellular transformation

(A – D) Growth curves in parental Ba/F3 cells (A), Ba/F3-MPL cells (B), Ba/F3-EPOR cells (C), and Ba/F3-G-CSFR cells (D) stably expressing EV, CALR^WT, CALR^MUT, or Jak2V617F demonstrate IL-3 independent growth in CALR^MUT-expressing Ba/F3-MPL cells but not in parental Ba/F3, Ba/F3-EPOR or Ba/F3-G-CSFR cells. (E – G) Growth curve in parental UT-7 (E), UT-7-MPL (F), or UT-7-EPOR (G) cells ectopically expressing CALR variants demonstrates GM-CSF-independent growth in CALR^MUT-expressing UT-7-MPL cells only.

Figure 3. Introduction of +1 bp frameshift mutations into the endogenous Calr locus is sufficient to confer oncogenic activity to Calr

(A – D) Growth curves in parental Ba/F3-Cas9 cells (A) Ba/F3-MPL-Cas9 cells (B), Ba/F3-EPOR-Cas9 cells (C), and Ba/F3-G-CSFR-Cas9 cells (D) demonstrates IL-3 independent growth in Calr-targeted Ba/F3-MPL-Cas9 cells only (E) Sequence verification confirming on-target editing of endogenous Calr (exon 9) in Ba/F3-MPL cells. +1 bp frameshift mutations are indicated in black.

Figure 4. Mutant CALR activates the JAK-STAT signaling axis downstream of MPL

(A) Pre-ranked GSEA in CALR^MUT versus CALR^WT expressing Ba/F3-MPL cells 24 hours post-IL-3 withdrawal. Stat5 target genes are enriched in CALR^MUT compared to CALR^WT-expressing Ba/F3-MPL cells (left) and Stat3 target genes are enriched in CALR^MUT compared to CALR^WT-expressing Ba/F3-MPL cells (right) (B) Immunoblotting demonstrates differential phosphorylation of MPL and Jak2 in Ba/F3-MPL transformed by CALR^MUT (C) Immunoblotting demonstrates differential phosphorylation of Stat5 and Stat3 in Ba/F3-MPL cells transformed by CALR^MUT, but not in parental Ba/F3, Ba/F3-EPOR, or Ba/F3-G-CSFR cells stably expressing CALR^MUT.

Figure 5. Mutant CALR-transformed hematopoietic cells are sensitive to JAK2 inhibition

(A) Growth curve in CALR^MUT-transformed Ba/F3-MPL cells transduced with shRNAs targeting Jak2 (shJak2 #1 or shJak2 #2) or a non-targeting shRNA control (shLuc) demonstrates significantly decreased proliferation in cells subjected to Jak2 knockdown compared to control cells (B) Intracellular phosphoprotein flow cytometry analysis of CALR^MUT (left) or Jak2V617F (right) expressing-Ba/F3-MPL cells demonstrates abrogation of Stat3 phosphorylation in response to treatment with the JAK2 inhibitor, INCB018424. (C) Pre-ranked GSEA in CALR^MUT versus CALR^WT expressing Ba/F3-MPL cells 24 hours post-IL-3 withdrawal performed on a consensus signature of genes downregulated by JAK2 inhibitor ruxolitinib (2µM dose, 6 hours). Individual perturbational signatures (the top 100 most downregulated) for each assay (6 different cancer cell lines in total) were obtained by using the Lincscloud API (version a2, LINCS Production Phase L1000 data, Broad Institute) from the NIH LINCS Program resource. For the consensus signature, only genes from the top 100 present in at least two assays were considered for further analysis. All p values were determined by unpaired two-tailed Student’s t test (*0.01 < p < 0.05; (**0.001 < p < 0.01; ns, not significant)

Figure 6. Structure-function analysis of mutant CALR uncovers a critical oncogenic role for the positive electrostatic charge of the mutant C-terminus

(A) Schema of domain and sequence mutants generated from CALR^WT and CALR^MUT. (B) Growth curve in Ba/F3-MPL cells stably expressing CALR domain mutants demonstrates that the mutant C-terminus alone is insufficient for transformation. (C). Growth curve in Ba/F3-MPL cells stably expressing CALRΔKDEL mutant demonstrates that deleting the KDEL from CALR^WT is insufficient for transformation. (D) Growth curve in Ba/F3-MPL cells stably expressing CALR^MUT C-terminus sequence-mutants demonstrates that all transform Ba/F3-MPL cells with equal efficacy. (E) (Top) Scheme of CALR^MUT-neutral and CALR^MUT-positive charge mutants. (Bottom) Growth curve in Ba/F3-MPL cells stably expressing CALR^MUT-neutral or CALR^MUT-positive variants shows that loss of the net positive charge in the mutant C-terminus abolishes the transforming capacity of CALR^MUT.

Figure 7. Mutant CALR binds to MPL and this interaction correlates with the transforming capacity of mutant CALR

(A – C) Immunoblotting of FLAG-immunoprecipitated proteins and whole cell lysates from 293T cells co-transfected with FLAG-CALR variants and (A) MPL, (B) EPOR or (C) G-CSFR demonstrates that CALR^MUT differentially binds to MPL, but not to other type I cytokine receptors (D) Immunoblotting of GST-immunoprecipitated proteins and whole cell lysates from 293T cells co-transfected with GST-CALR domain mutants and MPL demonstrates that only full length CALR^MUT, and not the mutant C-terminus alone, binds to MPL (E) Immunoblotting of FLAG-immunoprecipitated proteins and whole cell lysates from 293T cells co-transfected with mutant FLAG-CALR^MUT charge variants and MPL demonstrates that CALR^MUT-neutral binding to MPL is markedly diminished compared to CALR^MUT-positive or CALR^MUT binding.