Game of clones: the genomic evolution of severe congenital neutropenia

HAM-WASSERMAN MANUSCRIPT Game of clones: the genomic evolution of severe congenital neutropenia Ivo P. Touw 1 1 Department of Hematology, Erasmus MC Cancer Institute, Rotterdam, The Netherlands Severe congenital neutropenia (SCN) is a genetically heterogeneous condition of bone marrow failure usually diagnosed in early childhood and characterized by a chronic and severe shortage of neutrophils. It is now well-established that mutations in HAX1 and ELANE (and more rarely in other genes) are the genetic cause of SCN. In contrast, it has remained unclear how these mutations affect neutrophil development. Innovative models based on induced pluripotent stem cell technology are being explored to address this issue. These days, most SCN patients receive life-long treatment with granulocyte colony-stimulating factor (G-CSF, CSF3). CSF3 therapy has greatly improved the life expectancy of SCN patients, but also unveiled a high frequency of progression toward myelodysplastic syndrome (MDS) and therapy refractory acute myeloid leukemia (AML). Expansion of hematopoietic clones with acquired mutations in the gene encoding the G-CSF receptor (CSF3R) is regularly seen in SCN patients and AML usually descends from one of these CSF3R mutant clones. These findings raised the questions how CSF3R mutations affect CSF3 responses of myeloid progenitors, how they contribute to the pre-leukemic state of SCN, and which additional events are responsible for progression to leukemia. The vast (sub)clonal heterogeneity of AML and the presence of AML-associated mutations in normally aged hematopoietic clones make it often difficult to determine which mutations are responsible for the leukemic process. Leukemia predisposition syndromes such as SCN are unique disease models to identify the sequential acquisition of these mutations and to interrogate how they contribute to clonal selection and leukemic evolution. Learning Objectives Knowledge of the possible mechanisms by which mutations in ELANE or HAX1 cause SCN and the controversies that preclude a unifying hypothesis Understanding of how somatic mutations in CSF3R affect CSF3-induced signal transduction and contribute to the expansion of hematopoietic clones in SCN Knowledge of the long-term benefits and the risks of life-long CSF3 treatment Understanding the principle of clonal evolution of SCN toward MDS/AML Finnish born Rolf Kostmann, who received his medical training at the Karolinska Institute in Stockholm Sweden, was the first to describe a syndrome that he termed infantile genetic agranulocytosis. In his PhD thesis published in 1956, he described 14 cases from 9 consanguineous families from a parish in Northern Sweden with the now classical symptoms of severe congenital or chronic neutropenia (SCN). 1 Kostmann meticulously described the clinical and histologic course of the disease in 6 patients, the others had died from severe bacterial infections before the start of his investigations. The major conclusion of the work was that the syndrome was caused by a single recessively inherited gene defect. Another notable observation was that 2 of the 6 patients, although showing pathologies in the bone marrow comparable to the severe cases, only displayed a temporary agranulocytosis. Thus, Rolf Kostmann was also the first to note the considerable genotype-phenotype heterogeneity of SCN, a phenomenon that is still ill understood today. A visionary statement of Kostmann was his prediction that, rather than treatment of the infections with antibiotics, induction of (neutrophil) maturation may form the basis for a more causal therapy in the future. This became reality with the introduction of granulocyte colony-stimulating factor (G-CSF), now termed colonystimulating factor 3 (CSF3) in the treatment of SCN in the 1980s, a few years after his death in 1982. Genetic subtypes of SCN SCN is defined as a chronic state of severe neutropenia, with absolute neutrophil counts below 0.5 10 9 /L. Patients are highly susceptible to bacterial infections, with infections by staphylococcus aureus and Gram-negative bacteria being the most lifethreatening. 1-3 SCN involves multiple hereditary syndromes of which the underlying genetic defects have been identified, but there is still a significant portion of patients of which the etiology remains unknown. 3 Also, sporadic cases exist, which may not always be diagnosed shortly after birth. For these cases, SCN is used as an abbreviation to indicate severe chronic (rather than congenital) neutropenia. Congenital syndromes with multiple developmental defects, such as Shwachman Diamond syndrome, WHIM syndrome, Barth syndrome, myelokathexis, and glycogen storage disease 1b are sometimes also being referred to as SCN. 2,3 More recently, several new genetic subtypes of SCN have been discovered (Table 1). Here, I will focus mainly on the SCN subtypes that are characterized by a maturation arrest in the bone marrow at the Conflict-of-interest disclosure: The author declares no competing financial interests. Off-label drug use: None disclosed. Hematology 2015 1

Table 1. Genetic and hematologic features of SCN subtypes Hematopoietic manifestations additional to severe neutropenia Proposed disease mechanisms References SCN subtype Affected gene Inheritance 3,7,54 Leukemia predisposition Unfolded protein response, excessive apoptosis of myeloid cells Mutated neutrophil elastase ELANE Autosomal dominant (AD) or sporadic 3 Growth factor independent transcriptional GFI1 AD Lymphopenia Defective transcription, myeloid differentiation 3 repressor 1 deficiency block HAX1-deficiency HAX1 Autosomal recessive (AR) Leukemia predisposition Mitochondrial leakage, excessive apoptosis 10 Glucose-6-phosphatase deficiency G6PC3 AR Thrombocytopenia Impaired intracellular glucose homeostasis; excessive apoptosis 14 56 X-linked neutropenia (XLN) WAS X-chromosome linked Lymphopenia, leukemia predisposition Defective cytoskeleton organization, vesicle trafficking Vacuolar protein sorting-associated 45 deficiency VPS45 AR Myelofibrosis Defective lysosomal trafficking 55 Jagunal homolog 1 deficiency JAGN1 AR Neutrophil dysfunction; CSF3 hypo-responsiveness Aberrant N-glycosylation in biosynthetic pathway, reduced CSF3R signaling Bi-allelic CSF3R deficiency CSF3R AR CSF3 unresponsiveness Defective CSF3R signaling 34 promyelocyte-myelocyte stage (Figure 1). These include the SCN patients with mutations in ELANE, HAX1, and patients with X-linked neutropenia (XLN) with mutations in WAS. SCN with mutations in ELANE (ELANE-SCN) The discovery of mutations in ELANE, the gene encoding neutrophil elastase (NE), as a cause of SCN was reported in 2000 by Horwitz et al. 4 This was preceded by their publication describing ELANE mutations in families with an autosomal inheritance of cyclic neutropenia (CyN), a syndrome characterized by oscillating neutrophil counts with a periodicity of 21 days. 5 Although the genetic evidence that ELANE mutations are causative to SCN and CyN is compelling, it is far from clear how they contribute to the chronic or cyclic episodes of severe neutropenia. 6,7 Initially it was thought that the ELANE mutations in CyN were distinct from, and more benign than, those found in SCN mutations. However, later studies in extended patient series showed that, although there are trends in the spectrum of ELANE mutations toward either SCN or CyN, there is also a significant overlap. 6 The inability to identify ELANE mutations that rigorously discriminate between SCN and CyN argued in favor of the involvement of as yet unidentified modifying host factors, perhaps modifying genes. Evidence supporting this latter suggestion came from a study in a kindred of 8 from a sperm donor with an unanticipated mosaic ELANE mutation S97L in the germ line. All children shared the paternal S97L haplotype. However, whereas 7 children from 5 healthy mothers developed SCN, the child of 1 mother had CyN. 8 These findings show that, although the paternal ELANE-S97L allele preferentially caused SCN, host factors from the last mother had modified disease outcome toward CyN. To date, 100 ELANE mutations have been identified in ELANE- SCN, which are scattered over the entire 5 exome-containing gene. 6 Most of these are missense mutations in exons 2-5, but indels, frameshifts, nonsense, and splice-site mutations, mutations altering the translational start site and mutations in the 5 promoter region not affecting protein structure have also been reported. 6 It continues to be a challenge to dissect how this plethora of mutations can cause neutropenia, and no unifying hypothesis, if that indeed exists, has been offered thus far. Postulated mechanisms include: (1) altered function through mutant NE mislocalization (ie, granular versus membrane-proximal); (2) altered biochemical properties affecting substrate specificities of the mutant NE; and (3) protein misfolding in the endoplasmic reticulum (ER), triggering an unfolded protein response (UPR). 7 Given the mutation diversity, none of these potential explanations can be held responsible for causing neutropenia in all patients. Complicating matters even further, SCN patients have been reported in which ELANE expression in the myeloid progenitors (promyelocyte, myelocytes, metamyelocytes) is severely down-regulated, casting doubts on whether the mutant NE protein contributed at all to the neutropenia in these cases. 9 Despite these inconsistencies, it is likely that ELANE mutations evoke some form of cellular stress, either by the activation of the UPR or via other as yet unknown mechanism(s), as a consequence of which myeloid development stalls at the promyelocyte/myelocyte stage. SCN with mutations in HAX1 (HAX1-SCN) HAX1 mutations were first described by Klein et al in consanguineous pedigrees from the Middle East showing a recessive inheritance of SCN. 10 Quickly thereafter, it was realized that mutations in HAX1 were also the underlying cause of SCN in the Swedish families originally described by Kostmann. Ironically, the initial genetic 2 American Society of Hematology

Figure 1. Major features of severe congenital neutropenia. studies on the Kostmann pedigrees had failed to pinpoint HAX1 mainly because 1 affected family member had a de novo mutation in ELANE, but no inherited HAX1 mutations, thus confounding the outcome of the genetic linkage analysis. 11 HAX1 (HCLS1-associated protein X-1) is a ubiquitously expressed multifunctional protein. HAX1 predominantly localizes to mitochondria, but is present also in the ER and nuclear envelopes. 3 HCLS1 is a hematopoietic cell specific cortactin-like protein that modulates endosomal trafficking through its interaction with F-actin. 3 The loss of the anti-apoptotic function of HAX1 serves as a major explanation for the neutropenia in HAX1-SCN patients, in which increased apoptosis in neutrophils could be corrected by ectopic expression of HAX1. 10 By which mechanism(s) this is caused is still unclear. Mitochondrial leakage and associated pro-apoptotic cascades appear likely candidates and indeed the loss of mitochondrial membrane potential and elevated cytochrome c release, 2 key markers of mitochondrial control of apoptosis were detected in HAX1-SCN myeloid cells. 10 Intriguingly, a role of HAX1 in reducing ER stress responses involving the UPR in conditions of ischemic stress has recently been discovered. 12 This would offer a potential explanation for the overlapping phenotypes of ELANE-SCN and HAX1-SCN. As was suggested for the heterogeneous genotype-phenotype relationship between ELANE mutations and disease outcome (SCN versus CyN), host factor involvement may also play a role in HAX1-SCN. This could explain the unexpected occurrence of di-genic mutations (ELANE and HAX1) in a single SCN patient and the presence of an ELANE mutation in a patient from one of the Kostmann families, who had unaffected HAX1 genes. 11,13 These phenomena are unlikely to happen by chance, given the rarity of ELANE and HAX1 mutations in the healthy population. Similarly, host factors may have influenced the course of the disease in the two patients reported by Rolf Kostmann, who had significantly milder symptoms than others within the same pedigree. 1 X-linked neutropenia (XLN) XLN is a rare familial form of SCN caused by autosomal dominant mutations in WAS, the Wiscott Aldrich syndrome gene. Whereas deleterious mutations in WAS are responsible for the classical immune deficiency syndrome, the gain-of-function mutations in XLN cause an overactive protein, leading to elevated actin polymerization mediated by the nucleation core activity of the actin-related protein 2/3(Arp2/3) complex. 14 The actin cytoskeleton is involved in neutrophil homeostasis by controlling neutrophil differentiation, but also through the control of vesicle trafficking. This provides a potential functional connection with other neutropenia syndromes, such as Hermansky Pudlak syndrome type 2 (AP3), Griscelli syndrome type 2 (RAB27A), p14 deficiency (ROBLD3), and the more recently identified SCN subtypes with mutations in VPS45 or JAGN1, in which the affected genes are also implicated in intracellular vesicle trafficking (Table 1). 15 Because HAX1 interacts with the hematopoietic specific cortactin-like protein HCLS-1 and cortactin is involved in actin polymerization, deregulated vesicle trafficking might be involved in HAX1-SCN as well. Together, these findings merit the hypothesis that in addition to mitochondrial and endoplasmic reticulum stress, abnormalities in intracellular trafficking may contribute to defective neutrophil production in various SCN subtypes. 16 Introduction of CSF3 in the treatment of SCN CSF3 (G-CSF) is the major hematopoietic growth factor driving the production of neutrophils under homeostatic and emergency conditions, ie, when a rapid increase in neutrophil levels is needed to adequately cope with bacterial infections. 17 In 1989, Bonilla et al described the first successful outcome of CSF3 therapy in SCN. 18 This landmark study showed sustained peripheral absolute neutrophil counts (ANC) between 1.3 and 9.5 10 9 /L in all 5 patients, resolving preexisting chronic infections and reducing the number of new infectious episodes. Confirmation of these results in a randomized controlled phase III trial soon followed. 19 Since 1994, different registries have systematically monitored and regularly reported on patients with different forms of neutropenia, including SCN. 2,20 These reports generally support the long-term efficacy of CSF3 treatment, now spanning well 20 years in some patients. However, they have also raised questions about possible adverse effects, most notably the malignant transformation of SCN toward MDS or AML. Leukemic transformation of SCN and XLN In the pre-growth factor era, malignant progression of SCN was sporadically seen, but with the improved life expectancy achieved with CSF3 treatment, the numbers of SCN cases terminating in MDS or AML increased. In 2000, a first comprehensive analysis of Hematology 2015 3

the incidence of malignant transformation MDS/AML showed that among 352 SCN patients monitored for an average of 6 years on CSF3 treatment, 31 developed MDS/AML with a cumulative risk of 13% after 8 years. 21 A prospective study published in 2006 and updated in 2010 showed a 22% cumulative incidence of MDS/AML after 15 years of CSF3 treatment, whereas none of the CyN patients developed MDS or AML. Importantly, this study also showed that SCN patients who needed more than the median dosage of CSF3 (8 g/kg/d) and nonetheless did not reach median absolute neutrophil counts had a significantly increased MDS/AML incidence, ie, 34% (95% CI: 21%-47%), compared to patients who responded well to lower CSF3 dosages, in which the cumulative SCN/AML incidence was 15% (95% CI: 4%-25%). 22,23 The strongly increased leukemia risk is a feature shared between ELANE-SCN, HAX1-SCN, and XLN. For ELANE-SCN, it was initially suggested that the type of ELANE mutation might predict the severity of symptoms, including leukemia risk. 24 However, a later study did not support this idea. 6 Because current therapies, including those involving allogeneic HSC transplantation, fail to cure SCN/AML, 25,26 it is important to consider adaptations in the treatment before malignant transformation occurs. Notably, rare cases progressing to ALL or CMML have also been reported. 27 CSF3R mutations in SCN In the early 1990s, the first SCN cases with nonsense mutations in CSF3R were identified. 28 These mutations were detected both in the neutropenic and AML phase of the disease. The C-to-T transitions, converting glutamine codons into stop codons, cause truncations of the cytoplasmic domain of the CSF3R protein. Functional studies in murine factor-dependent myeloid cell lines (32D, L-GM, FDCP) revealed that these truncated CSF3R forms exert a dominant action over wild type CSF3R, leading to a hyper-proliferation and a block of neutrophil differentiation in response to CSF3. 29,30 Subsequently, it was shown that these CSF3R nonsense mutations were acquired (somatic), rather than inherited, and not the primary cause of SCN. 31 In a later study involving 148 SCN patients, 23 of which had progressed to secondary malignancies, CSF3R mutations at 17 different nucleotide positions were reported. 27 Whereas the frequency of CSF3R mutations in the non-leukemic patients was 34% (34/125), it was 78% in the SCN/AML cases (18/23), suggestive of a correlation between the acquisition of CSF3R mutations and leukemic transformation. 27 Significantly, this study also revealed a frequent existence of multiple clones with distinct CSF3R mutations, indicative of a strong selective pressure favoring the expansion of such clones. 27 In a later study, XLN patients were reported who developed leukemia with CSF3R mutations similar to those found in SCN. 32 The time between the first occurrence of detectable clones with CSF3R mutations and signs of malignant transformation is highly variable. In some patients, progression to MDS/AML occurred within a few months after the first detection of these clones. 27,31 In other patients, CSF3R mutant clones may persist for many years without progression to leukemia even at a high clonal burden, suggesting that additional cooperating events are involved in malignant transformation. 27,33 A novel somatic mutation in CSF3R (CSF3R-T595I), affecting the extracellular domain of CSF3R, was detected in the AML blasts of an SCN patient. This mutation was acquired in a clone that already harbored a CSF3R nonsense mutation and which preexisted for 15 years. 33 This is an autoactivating mutation, rendering the proliferation of hematopoietic progenitors growth factor independent. 33 Although the vast majority of CSF3R mutations in SCN are somatic and associated with hyper-proliferative responses, inherited bi-allelic CSF3R mutations causing CSF3 non-responsiveness have recently been identified in SCN patients from 2 unrelated families (Table 1). 34 Earlier, rare cases of SCN with CSF3R mutations leading to CSF3 nonresponsiveness responses with an unknown status of inheritance had been reported. 35,36 CSF3R mutations are confined to SCN and chronic neutrophilic leukemia (CNL) In a recent population study involving more than 17 000 individuals without hematologic disorders, clones were identified in the peripheral blood that contained DNMTA3, TET2, ASXL1, orjak2-v617f mutations, ie, known driver mutations of myeloid neoplasms. 37 These clones expanded with age and were associated with an increased probability of developing a hematological malignancy later in life. The prevailing somatic variants in this study were C-to-T transitions, similar to those causing the Gln to Stop mutations leading to CSF3R truncations. Importantly, none of the 17 000 individuals had mutations in CSF3R, indicating that expansion of clones with acquired CSF3R mutations is a true hallmark of SCN and not associated with a natural aging process of hematopoietic stem and progenitor cells (HSPCs). 37 In 2013, Maxson et al reported the frequent occurrence of a novel auto-activating CSF3R mutation, CSF3R-T618I. 38 With the introduction of genome-wide databases, the annotation of CSF3R mutations to amino acid (aa) positions has changed because the 23 aacontaining signal peptide is included in the new nomenclature. Hence, the CSF3R-T618I mutation is identical to the CSF3R-T595I found in a case of SCN/AML. 33 These CSF3R mutations are rare or absent in MDS, CMML, primary myelofibrosis, and de novo AML. 39,40 Strikingly, 5 CNL patients in the Maxson study also had CSF3R-truncating mutations which were present on the same allele as the CSF3R-T618I mutation, identical to what was reported in SCN/AML. 33,38 This raises the intriguing question as to why SCN/AML and CNL share identical CSF3R mutations. Because information on the timing of acquisition of CSF3R mutations in CNL is missing, a scenario of a CSF3-driven clonal expansion phase cannot be readily tested. However, it seems plausible that CNL is preceded by a CSF3 hypo-responsive pre-leukemic state, from which HSPC clones escape through the acquisition of CSF3R mutations. 41 Functional consequences of CSF3R mutations The impact of the CSF3R-truncating mutations on the CSF3 responses of myeloid progenitors is profound. It is beyond the scope of this article to comprehensively discuss how intracellular signaling from these mutant receptors is affected, but more details can be found in recent reviews. 16,42 Some of the main signaling properties of CSF3R are shown in Figure 2. An important feature of the CSF3R truncation mutants is their inability to undergo receptor endocytosis. This is because of the deletion of a critical di-leucine motif, and disturbed endosomal trafficking and lysosomal degradation, owing to defective suppressor of cytokine signaling 3-mediated ubiquitination of a critical membrane-proximal lysine residue in the CSF3R. 43 Because of their prolonged half-life and residence time at the plasma membrane, the CSF3R truncation mutants act dominantly over the wild-type CSF3R. 28 As a consequence, activation of the CSF3R truncation mutants results in an elevated proliferative response, altered kinetics of STAT5 activation, ie, from transient (minutes) to sustained (hours) activation. 44 This is particularly relevant because STAT5 is essential for the clonal 4 American Society of Hematology

Figure 2. Schematic representation of the activated CSF3R complex. Boxes 1 and 2 denote conserved regions for binding of JAK kinases. After CSF3-induced homodimerization of CSF3R chains, JAKs phosphorylate 4 receptor tyrosine residues, and STAT proteins. PI-3K/Akt, STAT5, and p21ras-erk1/2 signaling promote cell survival and proliferation, whereas STAT3, SOCS3, and SHIP promote cell cycle arrest and neutrophil differentiation in myeloid cell line models. The region affected by nonsense mutations in SCN is indicated. The resulting truncations lead to the loss of a di-leucine motif critical for receptor internalization, and of a docking site for SOCS3 (Y729) important for lysosomal routing of CSF3R. 43 expansion of hematopoietic stem and progenitor cells in mice harboring the Csf3r-d715 mutation. 45 In addition, activation of truncated CSF3R results in the increased production of intracellular reactive oxygen species (ROS) in myeloid progenitors via sustained activation of NADPH oxidase complex 2. 46 The elevated ROS levels may potentially affect myeloid progenitors by inactivation of oxidation sensitive phosphatases (PTP1B, PTEN) that negatively control growth factor signaling, by activation of stress pathways, and by inflicting DNA damage, leading to an elevated mutation rate. Despite these features, Csf3r-d715 mice do not develop leukemia, neither spontaneously nor when subjected to prolonged CSF3 administration. 47,48 The CSF3R-T618I mutant spontaneously activates JAK-mediated signaling. This can be attenuated by JAK inhibitors, such as ruxolitinib, which has therapeutic potential for CNL patients. 38 Genetic defects associated with the leukemic evolution of SCN A first systematic analysis with conventional Sanger sequencing showed that mutations that are common in de novo AML or myeloproliferative neoplasms, are rare or absent in leukemia or MDS secondary to SCN. 49 The subsequent use of next generation sequencing has greatly facilitated the identification of mutations in the leukemic evolution of SCN. A report on an ELANE-SCN patient, who developed AML 17 years after initiation of CSF3 treatment, 33 showed 12 acquired non-synonymous mutations in the AML blasts that had evolved from a clone with a CSF3R-d715 mutation. Two of these mutations, in LLGL2 and ZC3H18, already co-occurred in this clone 15 years prior to malignant transformation. This population expanded in time, whereas other 3 additional clones solely harboring distinct CSF3R mutations (d717, d725, d730) disappeared from the bone marrow. The other mutations, including the auto-activating CSF3R-T618I (595I) mutation and mutations in RUNX1 and ASXL1, were present only in the MDS/AML phase. 49 In a collaborative study involving patients from different registries, the patterns of acquisition of leukemia-associated mutations were further investigated. 50 This study revealed that 20 (64.5%) of the 31 patients had mutations in RUNX1 and that these mutations mostly occurred in clones with earlier acquired CSF3R mutations. This pattern was seen in ELANE-SCN, HAX1-SCN, as well as XLN (Figure 1). Other mutations associated with secondary malignancies, eg, in ASXL1, SUZ12, and EP300, were less frequent 50 A sequential analysis at stages prior to overt leukemia revealed that the RUNX1 mutations are late events in the leukemic transformation of SCN. The high frequency of cooperating RUNX1 and CSF3R mutations in SCN patients suggests a novel and unique molecular pathway of leukemogenesis, not seen thus far in any other type of adult or pediatric malignancy. 50 Early biomarkers of leukemia risk One of the objectives of the above-mentioned sequencing analysis was to identify genetic markers that, in addition to the somatic CSF3R mutations, might be used prospectively to detect early arising clones undergoing malignant transformation. 50 Despite their high frequency, somatic RUNX1 mutations are less useful because they appear late in the transformation process. 50 The mutations in LLGL2 and ZC3H18, reported in the SCN/AML cells of 1 patient and which could be back-tracked to a CSF3R mutant clones in the early SCN phase appeared more promising. 33 However, these mutations are far less recurrent than the CSF3R and RUNX1 mutations and therefore also of limited value as general early biomarkers of malignant transformation. Hence, the somatic CSF3R mutations currently serve as the single early clonality markers associated with leukemic transformation in 75% of SCN/AML cases. Significantly, in all informative cases in which hematopoietic clones with CSF3R mutations were present in the SCN phase, the AML blasts were derived from one of these clones. However, because CSF3R mutant clones may persist in the bone marrow of Hematology 2015 5

SCN patients for periods between months and years before MDS or AML becomes overt 27,33 ; this complicates the choice at what point in time one should consider discontinuation of CSF3 treatment and apply a preemptive therapy, eg, allogeneic stem cell transplantation (allo-sct), to prevent leukemic progression. However, as discussed above, this decision may be expedited when patients do not reach peripheral neutrophil numbers of 2 10 9 /L, while needing excess CSF3 dosages ( 10 g/kg/d) to get to these still suboptimal ANC values. 22 Outlook Although we have gained significant knowledge of the heterogeneous genetic origin of SCN, it is still uncertain how these mutations cause the neutropenia and why they predispose to leukemia. CSF3 treatment continues to be the most effective and well tolerated, albeit non-curative, therapy for SCN. Because this comes with the risk of clonal expansion of myeloid progenitors with acquired CSF3R mutations, which sooner or later may transform to MDS or leukemia, additional treatment strategies should be explored. A plausible scenario that merits exploration is to use small molecule inhibitors with selectivity for the CSF3R mutants to eradicate the mutant clones shortly after they appear. A more provocative scenario would be to correct the underlying genetic defect in hematopoietic stem cells by gene therapy. Novel genome-editing strategies using CRISPR/Cas9 or TALEN systems hold promise of facilitating this. Finally, studies aimed at a full understanding of the spectrum of consequences of the mutations in HAX1, ELANE and other mutations causative to SCN are still urgently needed, not only to explain the disease mechanisms but also to design additional therapeutic regimens. A complication is that both Hax1 deficient mice and mice with patient orthologous Elane mutations do not recapitulate the severe neutropenia seen in patients. 51 More recently, induced pluripotent stem cell (ips) lines from SCN patients have been generated. Although it is still early days for these models, these lines demonstrate the salient features of SCN, including the promyelocyte-maturation arrest. 52,53 They may thus serve as the preferred models to develop protocols for genetic correction by genome editing and to study the consequences of the combinations of mutations associated with malignant transformation. Acknowledgments I thank my colleagues at the Department of Hematology of Erasmus MC, who have contributed to the bone marrow failure program over the years and for providing a stimulatory research environment. In particular, I thank Bob Löwenberg and Ruud Delwel for their long-lasting support, and Julia Obenauer for advice on the paper. This work was supported by the Dutch Cancer Society KWFkankerbestrijding. Correspondence Ivo P. Touw, Department of Hematology, Erasmus MC Cancer Institute, PO Box 2040, 3000 CA Rotterdam, The Netherlands; Phone: 31 10 704 3837; Fax: 31 10 704 4745; e-mail: i.touw@erasmusmc.nl. References 1. Kostmann R. Infantile genetic agranulocytosis; agranulocytosis infantilis hereditaria. Acta Paediatr Suppl. 1956;45(Suppl 105):1-78. 2. Donadieu J, Beaupain B, Mahlaoui N, Bellanne-Chantelot C. Epidemiology of congenital neutropenia. Hematol Oncol Clin North Am. 2013;27(1):1-17. 3. Klein C. Genetic defects in severe congenital neutropenia: emerging insights into life and death of human neutrophil granulocytes. Annu Rev Immunol. 2011;29:399-413. 4. Dale DC, Person RE, Bolyard AA, et al. Mutations in the gene encoding neutrophil elastase in congenital and cyclic neutropenia. Blood. 2000; 96(7):2317-2322. 5. Horwitz M, Benson KF, Person RE, Aprikyan AG, Dale DC. Mutations in ELA2, encoding neutrophil elastase, define a 21-day biological clock in cyclic haematopoiesis. Nat Genet. 1999;23(4):433-436. 6. Germeshausen M, Deerberg S, Peter Y, Reimer C, Kratz CP, Ballmaier M. The spectrum of ELANE mutations and their implications in severe congenital and cyclic neutropenia. Hum Mutat. 2013;34(6):905-914. 7. Horwitz MS, Corey SJ, Grimes HL, Tidwell T. ELANE mutations in cyclic and severe congenital neutropenia: genetics and pathophysiology. Hematol Oncol Clin North Am. 2013;27(1):19-41. 8. Newburger PE, Pindyck TN, Zhu Z, et al. Cyclic neutropenia and severe congenital neutropenia in patients with a shared ELANE mutation and paternal haplotype: evidence for phenotype determination by modifying genes. Pediatr Blood Cancer. 2010;55(2):314-317. 9. Skokowa J, Fobiwe JP, Dan L, Thakur BK, Welte K. Neutrophil elastase is severely down-regulated in severe congenital neutropenia independent of ELA2 or HAX1 mutations but dependent on LEF-1. Blood. 2009;114(14):3044-3051. 10. Klein C, Grudzien M, Appaswamy G, et al. HAX1 deficiency causes autosomal recessive severe congenital neutropenia (Kostmann disease). Nat Genet. 2007;39(1):86-92. 11. Melin M, Entesarian M, Carlsson G, et al. Assignment of the gene locus for severe congenital neutropenia to chromosome 1q22 in the original Kostmann family from Northern Sweden. Biochem Biophys Res Commun. 2007;353(3):571-575. 12. Lam CK, Zhao W, Cai W, et al. Novel role of HAX-1 in ischemic injury protection involvement of heat shock protein 90. Circ Res. 2013;112(1): 79-89. 13. Germeshausen M, Zeidler C, Stuhrmann M, Lanciotti M, Ballmaier M, Welte K. Digenic mutations in severe congenital neutropenia. Haematologica. 2010;95(7):1207-1210. 14. Devriendt K, Kim AS, Mathijs G, et al. Constitutively activating mutation in WASP causes X-linked severe congenital neutropenia. Nat Genet. 2001;27(3):313-317. 15. Boztug K, Klein C. Genetics and pathophysiology of severe congenital neutropenia syndromes unrelated to neutrophil elastase. Hematol Oncol Clin North Am. 2013;27(1):43-60. 16. Touw IP, Palande K, Beekman R. Granulocyte colony-stimulating factor receptor signaling: implications for G-CSF responses and leukemic progression in severe congenital neutropenia. Hematol Oncol Clin North Am. 2013;27(1):61-73. 17. Metcalf D. Hematopoietic cytokines. Blood. 2008;111(2):485-491. 18. Bonilla MA, Gillio AP, Ruggeiro M, et al. Effects of recombinant human granulocyte colony-stimulating factor on neutropenia in patients with congenital agranulocytosis. N Engl J Med. 1989;320(24):1574-1580. 19. Dale DC, Bonilla MA, Davis MW, et al. A randomized controlled phase III trial of recombinant human granulocyte colony-stimulating factor (filgrastim) for treatment of severe chronic neutropenia. Blood. 1993; 81(10):2496-2502. 20. Dale DC, Bolyard AA, Schwinzer BG, et al. The Severe Chronic Neutropenia International Registry: 10-year follow-up report. Support Cancer Ther. 2006;3(4):220-231. 21. Freedman MH, Bonilla MA, Fier C, et al. Myelodysplasia syndrome and acute myeloid leukemia in patients with congenital neutropenia receiving G-CSF therapy. Blood. 2000;96(2):429-436. 22. Rosenberg PS, Alter BP, Bolyard AA, et al. The incidence of leukemia and mortality from sepsis in patients with severe congenital neutropenia receiving long-term G-CSF therapy. Blood. 2006;107(12):4628-4635. 23. Rosenberg PS, Zeidler C, Bolyard AA, et al. Stable long-term risk of leukaemia in patients with severe congenital neutropenia maintained on G-CSF therapy. Br J Haematol. 2010;150(2):196-199. 24. Bellanne-Chantelot C, Clauin S, Leblanc T, et al. Mutations in the ELA2 gene correlate with more severe expression of neutropenia: a 6 American Society of Hematology

study of 81 patients from the French Neutropenia Register. Blood. 2004;103(11):4119-4125. 25. Carlsson G, Winiarski J, Ljungman P, et al. Hematopoietic stem cell transplantation in severe congenital neutropenia. Pediatr Blood Cancer. 2011;56(3):444-451. 26. Choi SW, Boxer LA, Pulsipher MA, et al. Stem cell transplantation in patients with severe congenital neutropenia with evidence of leukemic transformation. Bone Marrow Transplant. 2005;35(5):473-477. 27. Germeshausen M, Ballmaier M, Welte K. Incidence of CSF3R mutations in severe congenital neutropenia and relevance for leukemogenesis: results of a long-term survey. Blood. 2007;109(1):93-99. 28. Dong F, Brynes RK, Tidow N, Welte K, Löwenberg 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(8):487-493. 29. Dong F, van Buitenen C, Pouwels K, Hoefsloot LH, Lowenberg B, Touw IP. Distinct cytoplasmic regions of the human granulocyte colony-stimulating factor receptor involved in induction of proliferation and maturation. Mol Cell Biol. 1993;13(12):7774-7781. 30. Fukunaga R, Ishizaka-Ikeda E, Nagata S. Growth and differentiation signals mediated by different regions in the cytoplasmic domain of granulocyte colony-stimulating factor receptor. Cell. 1993;74(6):1079-1087. 31. Tidow N, Pilz C, Teichmann B, et al. Clinical relevance of point mutations in the cytoplasmic domain of the granulocyte colonystimulating factor receptor gene in patients with severe congenital neutropenia. Blood. 1997;89(7):2369-2375. 32. Beel K, Vandenberghe P. G-CSF receptor (CSF3R) mutations in X-linked neutropenia evolving to acute myeloid leukemia or myelodysplasia. Haematologica. 2009;94(10):1449-1452. 33. Beekman R, Valkhof MG, Sanders MA, et al. Sequential gain of mutations in severe congenital neutropenia progressing to acute myeloid leukemia. Blood. 2012;119(22):5071-5077. 34. Triot A, Jarvinen PM, Arostegui JI, et al. Inherited biallelic CSF3R mutations in severe congenital neutropenia. Blood. 2014;123(24):3811-3817. 35. Druhan LJ, Ai J, Massullo P, Kindwall-Keller T, Ranalli MA, Avalos BR. Novel mechanism of G-CSF refractoriness in patients with severe congenital neutropenia. Blood. 2005;105(2):584-591. 36. Ward AC, van Aesch YM, Gits J, et al. Novel point mutation in the extracellular domain of the granulocyte colony-stimulating factor (G-CSF) receptor in a case of severe congenital neutropenia hyporesponsive to G-CSF treatment. J Exp Med. 1999;190(4):497-507. 37. Jaiswal S, Fontanillas P, Flannick J, et al. Age-related clonal hematopoiesis associated with adverse outcomes. N Engl J Med. 2014;371(26): 2488-2498. 38. Maxson JE, Gotlib J, Pollyea DA, et al. Oncogenic CSF3R mutations in chronic neutrophilic leukemia and atypical CML. N Engl J Med. 2013;368(19):1781-1790. 39. 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(5):e62-e63. 40. Mehta HM, Glaubach T, Long A, et al. Granulocyte colony-stimulating factor receptor T595I (T618I) mutation confers ligand independence and enhanced signaling. Leukemia. 2013;27(12):2407-2410. 41. Touw IP, Beekman R. Severe congenital neutropenia and chronic neutrophilic leukemia: an intriguing molecular connection unveiled by oncogenic mutations in CSF3R. Haematologica. 2013;98(10):1490-1492. 42. Palande K, Roovers O, Gits J, et al. Peroxiredoxin-controlled G-CSF signalling at the endoplasmic reticulum-early endosome interface. J Cell Sci. 2011;124(Pt 21):3695-3705. 43. Palande K, Meenhuis A, Jevdjovic T, Touw IP. Scratching the surface: signaling and routing dynamics of the CSF3 receptor. Front Biosci (Landmark Ed). 2013;18:91-105. 44. Hermans MH, Antonissen C, Ward AC, Mayen AE, Ploemacher RE, Touw IP. Sustained receptor activation and hyperproliferation in response to granulocyte colony-stimulating factor (G-CSF) in mice with a severe congenital neutropenia/acute myeloid leukemia-derived mutation in the G- CSF receptor gene. J Exp Med. 1999;189(4):683-692. 45. Liu F, Kunter G, Krem MM, et al. Csf3r mutations in mice confer a strong clonal HSC advantage via activation of Stat5. J Clin Invest. 2008;118(3):946-955. 46. 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(5):1847-1856. 47. Hermans MH, Ward AC, Antonissen C, Karis A, Lowenberg B, Touw IP. Perturbed granulopoiesis in mice with a targeted mutation in the granulocyte colony-stimulating factor receptor gene associated with severe chronic neutropenia. Blood. 1998;92(1):32-39. 48. McLemore ML, Poursine-Laurent J, Link DC. Increased granulocyte colony-stimulating factor responsiveness but normal resting granulopoiesis in mice carrying a targeted granulocyte colony-stimulating factor receptor mutation derived from a patient with severe congenital neutropenia. J Clin Invest. 1998;102(3):483-492. 49. Link DC, Kunter G, Kasai Y, et al. Distinct patterns of mutations occurring in de novo AML versus AML arising in the setting of severe congenital neutropenia. Blood. 2007;110(5):1648-1655. 50. Skokowa J, Steinemann D, Katsman-Kuipers JE, et al. Cooperativity of RUNX1 and CSF3R mutations in severe congenital neutropenia: a unique pathway in myeloid leukemogenesis. Blood. 2014;123(14):2229-2237. 51. Schaffer AA, Klein C. Animal models of human granulocyte diseases. Hematol Oncol Clin North Am. Feb 2013;27(1):129-148. 52. Hiramoto T, Ebihara Y, Mizoguchi Y, et al. Wnt3a stimulates maturation of impaired neutrophils developed from severe congenital neutropenia patient-derived pluripotent stem cells. Proc Natl Acad Sci U S A. 2013;110(8):3023-3028. 53. Morishima T, Watanabe K, Niwa A, et al. Genetic correction of HAX1 in induced pluripotent stem cells from a patient with severe congenital neutropenia improves defective granulopoiesis. Haematologica. 2014; 99(1):19-27. 54. Nanua S, Murakami M, Xia J, et al. Activation of the unfolded protein response is associated with impaired granulopoiesis in transgenic mice expressing mutant Elane. Blood. 2011;117(13):3539-3547. 55. Vilboux T, Lev A, Malicdan MC, et al. A congenital neutrophil defect syndrome associated with mutations in VPS45. N Engl J Med. 2013;369(1):54-65. 56. Boztug K, Jarvinen PM, Salzer E, et al. JAGN1 deficiency causes aberrant myeloid cell homeostasis and congenital neutropenia. Nat Genet. 2014;46(9):1021-1027. Hematology 2015 7