Granulocyte colony-stimulating factor receptor T595I (T618I) mutation confers ligand independence and enhanced signaling

Article metrics

Although acquired nonsense mutations occurring in the intracellular domain of the granulocyte colony-stimulating factor receptor (GCSFR, gene name CSF3R) occur in patients with severe congenital neutropenia (SCN), these and other missense mutations have rarely been reported in other disorders. Two recent reports have identified CSF3R T595I as a recurrent mutation in patients with chronic neutrophilic leukemia (CNL) and atypical (BCR-ABL1-negative) chronic myeloid leukemia (aCML). Exon sequencing of patients with CNL or aCML identified the CSF3R T595I mutation occurring exclusively in CNL. CSF3R mutations were not seen in aCML, primary myelofibrosis or chronic myelomonocytic leukemia (CMML), suggesting its strong, specific association with CNL.1 Another study revealed activating mutations of CSF3R in more than half of patients with CNL or aCML, the great majority due to T595I. This also conferred in vitro sensitivity to JAK inhibition, and administration of ruxolitinib to a patient with CNL and CSF3R T595I, which resulted in marked clinical improvement.2 Unlike those children and adolescents with SCN who are treated chronically with pharmacological doses of recombinant human GCSF, these individuals are much older and have no antecedent exposure to GCSF. Interestingly, this same mutation was observed in a patient with SCN who first acquired the classical nonsense mutation at codon 715 of the CSF3R (d715). This patient later developed acute myeloid leukemia (AML) with compound CSF3R mutations (d715+T595I) present in the leukemic blasts.3

We have screened patients with myeloid neoplasms, including sAML (n=37), myelodysplastic syndromes (MDS; n=120) and MDS/myeloproliferative neoplasms (MDS/MPNs; n=48), for the presence of CSF3R mutations using next-generation sequencing. Informed consent has been obtained according to the institutional review board. We also analyzed The Cancer Genome Atlas database of 205 AML cases. We report the detection of the recurrent CSF3R T595I mutation in a 64-year-old man who carried the diagnosis of CMML with normal cytogenetics and negative fluorescent in situ hybridization studies. The patient’s total white blood cell count was 38 290/μl; Hg 10.7 g/dl; hematocrit 32.8; mean corpuscular volume 98.5 fl; red cell distribution width 17.5%; platelet count 198 000/μl; nucleated RBC 60/μl. The peripheral blood differential was 78% neutrophils, 10% bands, 1% metamyelocytes, 6% lymphocytes and 5% monocytes with no blasts or young unidentified forms. His peripheral monocyte count varied with 10–13% monocytosis before therapeutic intervention. Hypogranular neutrophils were seen on the peripheral smear and dysplasia of the erythroid, and megakaryocytic lineages were found in the bone marrow. The bone marrow biopsy showed no fibrosis. There were no mutations in SETBP1, KRAS, NRAS or CBL. In addition to this patient, two patients were identified from The Cancer Genome Atlas with the CSF3R T595I mutation. In a separate large AML study, five patients (out of 1446 de novo AML cases) were found to harbor CSF3R T595I.4 That study also revealed two patients with CSF3R T617N (T630N), a mutation involving the transmembrane domain, which had been previously found in 2 (out of 555) patients with AML5 and in an extended family with hereditary neutrophilia.6

To determine the phenotypical and biochemical effects of the mutant GCSFR, we transfected the murine IL-3-dependent Ba/F3 cell line with pCDNA3 containing either the wild-type (wt) human CSF3R or CSF3R T595I. (Note that the codon numbering excludes the 23 signal peptide sequence; thus, GCSFR T595I is the same as GCSFR T618I.) Stable transfectants were then sorted for similar levels of ectopic GCSFR expression (data not shown). Cells expressing GCSFR T595I displayed proliferation (Figure 1a, upper panel) and greater survival (Figure 1a, lower panel) in the absence of GCSF. Cells expressing wt GCSFR died by 72 h. These results were corroborated by MTT assay, which showed a growth advantage for T595I-expressing cells at GCSF concentrations <1 ng/ml and similar growth at GCSF >1 ng/ml (Figure 1b). After 24 h of cytokine withdrawal, there was a significant drop in cell count seen only in the wt GCFSR cells with maintenance of cell density for GCSFR T595I (Figure 1c, upper panel). By 48 h, viability was maintained at 52% for T595I and only 19% for the wt GCSFR (Figure 1c, lower panel). Western blot analysis identified ligand-independent, constitutive activation of STAT3 and STAT5, and significant enhancement of Src (Figure 2a). No difference in activation status of Akt or Erk1/2 was seen (data not shown). Previously, we found increased reactive oxygen species (ROS) produced in cell lines or primary cells expressing the truncated GCSFR found in patients with SCN who developed secondary MDS/AML.7 Here, we observed that cells expressing the constitutively active GCSFR T595I produced more ROS at baseline and following GCSF stimulation (Figure 2b). The T595I receptor underwent ligand-dependent endocytosis similar to wt GCSFR (Figure 2c). Thus, in the absence of ligand, the receptor signals constitutively and without the appropriate downregulation of signaling that occurs with ligand-bound receptor.8 Proliferation for both wt GCSFR and GCSFR T595I was similarly inhibited by ruxolitinib, with an IC50 of 289 and 523 nM, respectively (Figure 2d). Even though there was greater Src family kinase activation in cells expressing GCSFR T595I, dasatinib did not affect their cell growth.

Figure 1

Gain-of-function mutation in GCSFR T595I. Ba/F3 were stably transfected with pCDNA3 containing the cDNA for either wt CSF3R or the CSF3R T595I mutant. Stable transfectants were selected by flow cytometry for comparable levels of ectopic GCSFR expression. (a) Ligand-independent proliferation and survival. Ba/F3 cells were grown in mIL-3 containing medium then washed twice and resuspended in the presence or absence of human GCSF. Cell numbers were counted by trypan blue exclusion to determine viability. (b) GCSF dose-response curves for wt GCSFR and GCSFR T595I were determined using MTT assay at 48 h. Data are representative of three independent experiments. The values represent the absorbance read at 600 nm following the manufacturer’s protocol. (c) Cells expressing GCSFR T595I displayed resistance to cytokine withdrawal-induced cell death. Cells were counted by trypan blue exclusion for number and survival.

Figure 2

Signaling properties of GCSFR T595I. (a) Constitutive signaling by GCSFR T595I. Cells were removed from murine IL-3 containing medium, washed twice, resuspended in RPMI with 1% bovine serum albumin and incubated for 8 h. The cells were then stimulated with 100 ng/ml GCSF for the indicated timepoints. Lysates were prepared and blotted with antibodies against pSTAT5 (Y694), pSTAT3 (Y705), phospho-Src (SFK, Y416) and actin for protein loading. (b) Enhanced ROS producton. Cells expressing either the wt or mutant receptor were stained with dihydrorhodamine-123, before being stimulated with indicated concentrations of GCSF and analyzed by flow cytometry. (c) Normal internalization behavior of GCSFR T595I. (d) Sensitivity to ruxolitinib. Cells were washed free of IL-3, then suspended in medium with 100 ng/ml GCSF and increasing concentrations of ruxolitinib or dasatinib. MTT assays were performed at 48 h. Parental cells died within 48 h, independently of tyrosine kinase inhibitor.

Mutations in either the intracellular, transmembrane, or extracellular domains of the GCSFR can profoundly alter its function.1, 2, 5, 6, 9 There may be some confusion in the literature because some reports include the 23 amino-acid signal peptide in their amino-acid numbering system and others do not. The extracellular domain mutation we described here is T595I, which also corresponds to T618I.1, 2 Mutations arising in the transmembrane domain have been reported as T617N, which would correspond to T640N.5, 6 Here, we show that the T595I mutation induces cell autonomous, GCSF-independent proliferation and survival with constitutive activation of proximal signaling events, particularly through the JAK–STAT pathway. Enhancement of Src and ROS may contribute to its proliferative/survival advantage when compared with the cells expressing the wt CSF3R.7 These functional and biochemical data provide additional support for the potential role of CSF3R mutations in myeloid leukemogenesis. Other factors, such as mutant SETBP1,10 may be required to permit full oncogenicity in the context of a mutated GCSFR. Recent reports have identified SETBP1 mutations in the SKI homologous region in patients with CMML, CNL and MDS/MPN.1, 10, 11, 12, 13 The functional implications of the SETBP1 mutations are not known, however, overexpression of SETBP1 may be involved in self-renewal of leukemic myeloid progenitors via increased expression of Hoxa9 and Hoxa10.14 A subset of CNL patients carried mutations in both SETBP1 and CSF3R.1 As only one individual developed MDS in a multi-generational family with chronic neutrophilia associated with GCSFR T617N,6 co-operating mutations may be required for more aggressive disease.

CMML and aCML are overlap entities of MDS/MPNs characterized by morphological dysplasia with accumulation of monocytes or neutrophils, respectively.15 Distinguishing CMML from CNL can also pose diagnostic challenges. Because our patient had monocytosis >1000/μl, but fluctuating percentage of monocytes on peripheral smear, strict WHO criteria for CMML or CNL were not fulfilled. In our cohort, we noted a p.Q749X CSF3R truncation mutation in a man with CMML, which confers enhanced proliferation with impaired differentiation (Mehta et al., Leukemia, in review). It is unlikely that the current WHO definitions of CMML, atypical CML and CNL, based largely on morphology and cell counts, are precise. Greater appreciation of how the gain-of-function mutation in the CSF3R affects the output of abnormal granulocyte/macrophage progenitors may shed light on these related disorders and redefine how we classify them.

There is a much stronger association of mutant GCSFR with 19/22 found in CNL compared with 5/28 in atypical CML.1, 2 Our own whole-exome sequencing of an additional 11 patients with aCML did not reveal any mutations in CSF3R (unpublished data and personal communication per Dr Carlo Gambacorti-Passerini, University of Milano Bicocca). Our study is an example of a case where WHO classification by mutation identification will be difficult in order to optimally classify the generally poor prognosis of the mixed MDS/MPN disorders. Further identification of molecular mechanisms responsible for altered cell behavior will help define the contribution of aberrant GCSFR signaling to myeloid leukemogenesis, and may provide candidate pathways for targeted therapy in addition to JAK inhibition.


  1. 1

    Pardanani A, Lasho TL, Laborde RR, Elliott M, Hanson CA, Knudson RA et al. CSF3R T618I is a highly prevalent and specific mutation in chronic neutrophilc leukemia. 2013; 27: 1870–1873.

  2. 2

    Maxson JE, Gotlib J, Pollyea DA, Fleischman AG, Agarwal A, Eide CA et al. Oncogenic CSF3R mutations in chronic neutrophilic leukemia and atypical CML. N Engl J Med 2013; 368: 1781–1790.

  3. 3

    Beekman R, Valkhof MG, Sanders MA, van Strien PM, Haanstra JR, Broeders L et al. Sequential gain of mutations in severe congenital neutropenia progressing to acute myeloid leukemia. Blood 2012; 119: 5071–5077.

  4. 4

    Beekman R, Valkhof M, van Strien P, Valk PJ, Touw IP . Prevalence of a new auto-activating colony stimulating factor 3 receptor mutation (CSF3R-T595I) in acute myeloid leukemia and severe congenital neutropenia. Haematologica 2013; 98: e62–e63.

  5. 5

    Forbes LV, Gale RE, Pizzey A, Pouwels K, Nathwani A, Linch DC . An activating mutation in the transmembrane domain of the granulocyte colony-stimulating factor receptor in patients with acute myeloid leukemia. Oncogene 2002; 21: 5981–5989.

  6. 6

    Plo I, Zhang Y, Le Couedic JP, Nakatake M, Boulet JM, Itaya M et al. An activating mutation in the CSF3R gene induces a hereditary chronic neutrophilia. J Exp Med 2009; 206: 1701–1707.

  7. 7

    Zhu QS, Xia L, Mills GB, Lowell CA, Touw IP, Corey SJ . G-CSF induced reactive oxygen species involves Lyn-PI3-kinase-Akt and contributes to myeloid cell growth. Blood 2006; 107: 1847–1856.

  8. 8

    Sarkar CA, Lowenhaupt K, Horan T, Boone TC, Tidor B, Lauffenburger DA . Rational cytokine design for increased lifetime and enhanced potency using pH-activated "histidine switching". Nat Biotechnol 2002; 20: 908–913.

  9. 9

    Dong F, Brynes RK, Tidow N, Welte K, Lowenberg B, Touw IP . Mutations in the gene for the granulocyte colony-stimulating-factor receptor in patients with acute myeloid leukemia preceded by severe congenital neutropenia. N Engl J Med 1995; 333: 487–493.

  10. 10

    Piazza R, Valletta S, Winkelmann N, Redaelli S, Spinelli R, Pirola A et al. Recurrent SETBP1 mutations in atypical chronic myeloid leukemia. Nat Genet 2013; 45: 18–24.

  11. 11

    Meggendorfer M, Bacher U, Alpermann T, Haferlach C, Kern W, Gambacorti-Passerini C et al. SETBP1 mutations occur in 9% of MDS/MPN and in 4% of MPN cases and are strongly associated with atypical CML, monosomy 7, isochromosome i(17)(q10), ASXL1 and CBL mutations. 2013; 27: 1852–1860.

  12. 12

    Laborde RR, Patnaik MM, Lasho TL, Finke CM, Hanson CA, Knudson RA et al. SETBP1 mutations in 415 patients with primary myelofibrosis or chronic myelomonocytic leukemia: independent prognostic impact in CMML. Leukemia, 2013; 27: 2100–2102.

  13. 13

    Damm F, Itzykson R, Kosmider O, Droin N, Renneville A, Chesnais V et al. SETBP1 mutations in 658 patients with myelodysplastic syndromes, chronic myelomonocytic leukemia and secondary acute myeloid leukemias. Leukemia 2013; 27: 1401–1403.

  14. 14

    Oakley K, Han Y, Vishwakarma BA, Chu S, Bhatia R, Gudmundsson KO et al. Setbp1 promotes the self-renewal of murine myeloid progenitors via activation of Hoxa9 and Hoxa10. Blood 2012; 119: 6099–6108.

  15. 15

    Muramatsu H, Makishima H, Maciejewski JP . Chronic myelomonocytic leukemia and atypical chronic myeloid leukemia: novel pathogenetic lesions. Semin Oncol 2012; 39: 67–73.

Download references

Author information

Correspondence to S J Corey.

Ethics declarations

Competing interests

NCP cross receives honoraria from Novartis.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Mehta, H., Glaubach, T., Long, A. et al. Granulocyte colony-stimulating factor receptor T595I (T618I) mutation confers ligand independence and enhanced signaling. Leukemia 27, 2407–2410 (2013) doi:10.1038/leu.2013.164

Download citation

Further reading