Abstract

Severe congenital neutropenia (SCN) is an inherited disorder of granulopoiesis that is associated with a markedly increased risk of developing acute myeloid leukemia (AML) or myelodysplasia (MDS). Somatic mutations of CSF3R, encoding the G-CSF receptor (G-CSFR), are strongly associated with the development of AML/MDS in SCN. These mutations invariably produce a truncated G-CSFR that, though remaining ligand-dependent, transmits a hyperproliferative signal. Transgenic mice carrying a targeted (knock-in) mutation of Csf3r (termed d715) reproducing a mutation found in a patient with SCN have an exaggerated neutrophil response to G-CSF treatment but do not develop AML/MDS. Moreover, we recently showed expression of the d715 G-CSFR confers a strong clonal advantage at the hematopoietic stem cell level that is dependent upon exogenous G-CSF. Collectively, these data suggest that CSF3R truncation mutations are an initiation or early progression factor for leukemic transformation. However, there is, as yet, scant direct evidence supporting this hypothesis. Previous studies have established that activating mutations of receptor tyrosine kinases, such as internal tandem duplications of FLT3, are able to cooperate with PML-RARα to induce AML. Since the CSF3R mutations in SCN also are “activating”, we asked whether the d715 G-CSFR could cooperate with PML-RARα to induce AML in mice. PML-RARα transgenic mice were intercrossed with d715 G-CSFR mice (all inbred > 10 generations onto a C57BL/6 background) to generate the cohorts listed in Table 1. A separate cohort for each genotype was treated chronically with pegylated G-CSF (1 mg/kg every 4–5 days for 6 months) to simulate the high level of serum G-CSF present in patients with SCN. Complete blood counts were performed at 3 months intervals and documented a similar increase in neutrophil counts in all mice treated with G-CSF. The cumulative incidence of AML and median follow-up for each cohort is shown in Table 1. None of the mice without the PML-RARα transgene, regardless of G-CSF treatment, developed AML, confirming that the d715 G-CSFR is not sufficient to induce AML. In mice carrying the PML-RARα transgene but not treated with G-CSF, a nonsignificant trend to increased AML was observed in mice expressing the d715 G-CSFR (P=0.12). However, in mice carrying the PML-RARα transgene and treated with G-CSF, the presence of the d715 G-CSFR significantly increased the penetrance (P=0.009) and reduced the latency of AML. In all cases, the leukemia was characterized by leukocytosis, splenomegaly, and a high percentage of blasts in the bone marrow and spleen that co-express Gr1 and c-Kit. These data provide the first direct evidence that the CSF3R mutations present in patients with SCN are leukemogenic and provide further support for the proposition that patients who acquire CSF3R mutations be considered for early stem cell transplantation.

The cumulative incidence of AML and median follow-up for each cohort

PML-RARCSF3RG-CSF RxNMedian follow-upCumulative AML%
No WT No 20 454 0.0% 
No d715 No 45 250 0.0% 
No WT Yes 20 488 0.0% 
No d715 Yes 47 311 0.0% 
Yes WT No 57 286 6.4% 
Yes d715 No 67 267 11.9% 
Yes WT Yes 54 322 20.4% 
Yes d715 Yes 50 311 44.0% 
PML-RARCSF3RG-CSF RxNMedian follow-upCumulative AML%
No WT No 20 454 0.0% 
No d715 No 45 250 0.0% 
No WT Yes 20 488 0.0% 
No d715 Yes 47 311 0.0% 
Yes WT No 57 286 6.4% 
Yes d715 No 67 267 11.9% 
Yes WT Yes 54 322 20.4% 
Yes d715 Yes 50 311 44.0% 

Author notes

Disclosure: No relevant conflicts of interest to declare.