Oncogenic CSF3R mutations in chronic neutrophilic leukemia and atypical CML

Abstract

Background: The molecular causes of many hematologic cancers remain unclear. Among these cancers are chronic neutrophilic leukemia (CNL) and atypical (BCR-ABL1-negative) chronic myeloid leukemia (CML), both of which are diagnosed on the basis of neoplastic expansion of granulocytic cells and exclusion of genetic drivers that are known to occur in other myeloproliferative neoplasms and myeloproliferative-myelodysplastic overlap neoplasms.

Methods: To identify potential genetic drivers in these disorders, we used an integrated approach of deep sequencing coupled with the screening of primary leukemia cells obtained from patients with CNL or atypical CML against panels of tyrosine kinase-specific small interfering RNAs or small-molecule kinase inhibitors. We validated candidate oncogenes using in vitro transformation assays, and drug sensitivities were validated with the use of assays of primary-cell colonies.

Results: We identified activating mutations in the gene encoding the receptor for colony-stimulating factor 3 (CSF3R) in 16 of 27 patients (59%) with CNL or atypical CML. These mutations segregate within two distinct regions of CSF3R and lead to preferential downstream kinase signaling through SRC family-TNK2 or JAK kinases and differential sensitivity to kinase inhibitors. A patient with CNL carrying a JAK-activating CSF3R mutation had marked clinical improvement after the administration of the JAK1/2 inhibitor ruxolitinib.

Conclusions: Mutations in CSF3R are common in patients with CNL or atypical CML and represent a potentially useful criterion for diagnosing these neoplasms. (Funded by the Leukemia and Lymphoma Society and others.).

Figure 1. Sensitivity to Kinase Inhibition in Leukemia Specimens with Transforming Mutations in CSF3R

Panel A shows the location and recurrence of CSF3R mutations found in samples from 16 of 27 patients with chronic neutrophilic leukemia (CNL) or atypical chronic myeloid leukemia (CML), along with samples from patients with other types of leukemia. The mutation locations and number of observations are indicated by black circles. The Q741X mutation was found in a sample obtained from a patient with acute myeloid leukemia (AML), and one of the T618I mutations was found in a sample from a patient with early T-cell precursor T-cell acute lymphoblastic leukemia (ETP-T-ALL). Five patients with CNL or atypical CML had both membrane proximal and truncation mutations. (For details, see Table S3 in the Supplementary Appendix.) Two additional CSF3R mutations (Q739X and T618I, which are not shown) have been reported in AML specimens sequenced by the Cancer Genome Atlas. CSF3R coordinates are numbered according to the conventions of the Ensembl genome browser, a numbering system that differs from historical CSF3R numbering owing to the inclusion of the 23-amino-acid signal peptide, despite the absence of this signal peptide from the mature protein. Panel B shows the sensitivity of white cells from Patient 3, who had CNL and a CSF3R S783fs mutation (Table S3 in the Supplementary Appendix), to a panel of 66 small-molecule kinase inhibitors. The 50% inhibitory concentration (IC₅₀) of each drug is plotted as a percentage of the median IC ₅₀ for each drug from 150 samples obtained from patients with leukemia. A specimen was considered to be hypersensitive to an inhibitor if the IC₅₀ was less than 10% of the median IC₅₀ for that inhibitor for the entire cohort (as indicated by the dashed red line). This specimen was hypersensitive to dasatinib (green) and insensitive to JAK kinase inhibitors (orange). SFK denotes SRC-family kinase, and TNK2 tyrosine kinase nonreceptor 2. Panel C shows the sensitivity of white cells from Patient 3 to small interfering RNAs (siRNAs) directed against all known tyrosine kinases, as described previously.^, Silencing of TNK2 and an SRC family kinase, FGR, resulted in a substantial decrease in cell viability. All cell-viability values after silencing with each individual siRNA have been normalized to the median value of the entire panel. The bars on the graph represent the mean normalized cell viability from triplicate data points for each individual siRNA. The T bars represent standard errors. The black horizontal line indicates the mean of all values across the entire siRNA panel, and the red dashed line indicates a threshold of significance, which is calculated as the mean minus 2 SD for all values. In addition to carrying the CSF3R S783fs mutation, Patient 3 had a minority of clones with a CSF3R S783fs–T615A compound mutation, but this small percentage of cells did not have an effect on sensitivity to inhibitors in short-term assays. Panel D shows the sensitivity of white cells from a patient with ETP-T-ALL and a CSF3R T618I mutation to the same panel of 66 small-molecule kinase inhibitors that was used to test cells from Patient 3, as described in Panel B. These cells were insensitive to dasatinib (green) and sensitive to JAK kinase inhibitors (orange). Panel E shows interleukin-3–dependent Ba/F3 cells that were infected with murine retrovirus expressing wild-type CSF3R, membrane proximal mutations, or truncation mutations. Uninfected parental Ba/ F3 cells and empty-vector infected Ba/F3 cells were used as controls. Over a 10-day period, both classes of CSF3R mutations were capable of transforming Ba/F3 cells to interleukin-3–independent growth, and the membrane proximal mutations (T615A and T618I) transformed cells in this assay substantially faster than the truncation mutants (Q741X and S783fs). Panel F shows Ba/F3 cells expressing CSF3R T618I or S783fs mutations before or after interleukin-3– independent transformation (IL3- indicates transformed cells). Cell lysates were subjected to immunoblot analysis for CSF3R, TNK2, phospho-STAT3 (pSTAT3), total STAT3, phospho-JAK2 (pJAK2), total JAK2, phospho-SRC (pSRC), total SRC, and actin. Parental Ba/F3 cells or Ba/F3 cells expressing wild-type CSF3R were included as controls.

Figure 2. Use of Tyrosine Kinase Inhibitors to Treat Dysregulated Signaling Induced by CSF3R Mutations

Panel A shows the effect of dasatinib on colony formation in bone marrow cells from mice that were infected with mutant CSF3R-containing retroviruses or an empty vector; the control cells expressed endogenous wild-type CSF3R. Cells were grown in methylcellulose containing the minimal amount of granulocyte colony-stimulating factor (G-CSF) necessary to form colonies (10 ng per milliliter for the empty vector, 0.4 ng per milliliter for the S783fs mutation, and no G-CSF for the T618I mutation). Cells were plated with increasing concentrations of dasatinib (0, 1, 10, 100, and 1000 nM). The experiment was performed in triplicate with the number of colonies normalized to those in the untreated controls. Values represent the mean percent colonies; the T bars indicate standard errors. A single asterisk indicates P<0.07, and a double asterisk indicates P<0.005 for the comparison between the T618I mutation and the S783fs mutation at equivalent doses of dasatinib. Panel B shows the results of a similar colony-formation assay, in which the cells were plated with ruxolitinib (0, 10, 100, or 1000 nM). Panel C shows the results for Patient 9, who had CNL and a CSF3R T618I mutation and in whom earlier testing indicated sensitivity to ruxolitinib (Rux) in vitro (Fig. S3C and S3E in the Supplementary Appendix). This patient was treated with 500 mg of hydroxyurea (HU) daily starting on day 13. Hydroxyurea was stopped on day 21 and oral ruxolitinib (at a dose of 10 mg twice daily) was administered. On day 70, the dose of ruxolitinib was increased to 15 mg twice daily. The numbers of white cells and neutrophils (absolute neutrophil count) are shown. Panel D shows normalized platelet counts while Patient 9 was undergoing the treatment regimen shown in Panel C.

Figure 3. Model for Activation and Signaling of CSF3R Mutations

Truncation mutations in CSF3R (the receptor for G-CSF) result in increased expression levels. Downstream signaling mediators — SRC family kinases (SFKs) and TNK2 — are preferentially activated by these truncation mutations. Consequently, leukemic cells harboring the mutations are highly sensitive to dasatinib. Truncation mutations in CSF3R may also show sensitivity to JAK kinase inhibitors in the context of JAK kinase stimulation downstream of high ligand concentrations. In contrast, membrane proximal mutations in CSF3R show completely ligand-independent function. In this capacity, the dominant mode of signaling appears to operate through the JAK-STAT pathway. Thus, patients with membrane proximal mutations may be candidates for treatment with JAK kinase inhibitors, such as the JAK1/2 inhibitor ruxolitinib.