Author Affiliations
- Paula Aranaz1,
- Cristina Hurtado1,
- Ignacio Erquiaga1,
- Itziar Miguéliz1,
- Cristina Ormazábal1,
- Ion Cristobal2,
- Marina García-Delgado1,
- Francisco Javier Novo1 and
- José Luis Vizmanos1⇓
- 1Department of Genetics, School of Sciences, University of Navarra, Pamplona
- 2CIMA, Center for Applied Medical Research, University of Navarra, Pamplona, Spain
- Correspondence: José Luis Vizmanos, Ph. D., Department of Genetics, School of Sciences, Research Building. University of Navarra, C/ Irunlarrea 1, E-31008 Pamplona, Spain. E-mail: jlvizmanos{at}unav.es
Abstract
Background Despite the discovery of the p.V617F in JAK2, the molecular pathogenesis of some chronic myeloproliferative neoplasms remains unclear. Although very rare, different studies have identified CBL (Cas-Br-Murine ecotropic retroviral transforming sequence) mutations in V617FJAK2-negative patients, mainly located in the RING finger domain. In order to determine the frequency of CBL mutations in these diseases, we studied different regions of all CBL family genes (CBL, CBLB and CBLC) in a selected group of patients with myeloproliferative neoplasms. We also included V617FJAK2-positive patients to check whether mutations in CBL and JAK2 are mutually exclusive events.
Design and Methods Using denaturing high performance liquid chromatography, we screened for mutations in CBL, CBLB and CBLC in a group of 172 V617FJAK2-negative and 232 V617FJAK2-positive patients with myeloproliferative neoplasms not selected for loss of heterozygosity. The effect on cell proliferation of the mutations detected was analyzed on a 32D(FLT3) cell model.
Results An initial screening of all coding exons of CBL, CBLB and CBLC in 44 V617FJAK2-negative samples revealed two new CBL mutations (p.C416W in the RING finger domain and p.A678V in the proline-rich domain). Analyses performed on 128 additional V617FJAK2-negative and 232 V617FJAK2-positive samples detected three CBL changes (p.T402HfsX29, p.P417R and p.S675C in two cases) in four V617FJAK2-positive patients. None of these mutations was found in 200 control samples. Cell proliferation assays showed that all of the mutations promoted hypersensitivity to interleukin-3 in 32D(FLT3) cells.
Conclusions Although mutations described to date have been found in the RING finger domain and in the linker region of CBL, we found a similar frequency of mutations in the proline-rich domain. In addition, we found CBL mutations in both V617FJAK2-positive (4/232; 1.7%) and negative (2/172; 1.2%) patients and all of them promoted hypersensitivity to interleukin-3.
Introduction
BCR-ABL1-negative chronic myeloproliferative neoplasms (MPN) are a heterogeneous group of clonal hematologic malignancies characterized by abnormal proliferation and survival of one or more myeloid lineage cells. In some cases these diseases evolve to acute myeloid leukemia (AML). These hematologic neoplasms include both classic MPN [essential thrombocythemia (ET), polycythemia vera (PV) and primary myelofibrosis (PMF)] and atypical MPN (such as chronic eosinophilic leukemia, chronic neutrophilic leukemia, hypereosinophilic syndrome, mast cell disease and myeloid neoplasms with eosinophilia, among others).1
In the late 1990s some genetic aberrations were described as molecular disease-causing events in these neoplasms, most of them via fusion genes resulting from reciprocal chromosomal translocations. Such fusions activate tyrosine kinases, playing a role similar to ABL1 in chronic myeloid leukemia.2,3 However these fusions are very rare and most of them have been reported in one or two cases worldwide.4
This situation changed in 2005 with the description of the p.V617F mutation (valine to phenylalanine in amino acid 617) in JAK2, found not only in classic MPN but also in a small number of atypical MPN and other myeloid neoplasms.5 Furthermore, it was found that most of the V617FJAK2-negative cases of PV had other transforming mutations in exon 12 of JAK2. Other gain-of-function mutations have also been described in genes coding for JAK-STAT receptors, such as MPL or EPOR in familial and sporadic cases of MPN.6–10 However, to date it is not known whether these mutations cause the full phenotype or whether they cooperate with other still uncharacterized mutations. Thus, there is still a significant proportion of patients in whom the molecular disease-causing event remains to be discovered.
Recently, the application of single nucleotide polymorphism and comparative genomic hybridization array technologies has led to the identification of new mutations in loss of heterozygosity regions affecting genes such as TET2,11 ASXL1,12 IKZF1,13 RUNX1,14 IDH1 and IDH2,15 EZH2,16 NF1,17 and CBL.18–23
CBL (11q23) codes for a protein of the Cbl family of E3-ubiquitin ligases (CBL, CBLB and CBLC) that acts as a negative regulator of some cell signaling pathways, by promoting the ubiquitination of several signaling molecules including some tyrosine kinases. CBL proteins share a common structure, with a highly conserved tyrosine kinase-binding domain in the amino-terminal region that determines substrate specificity. The catalytic E3-ubiquitin ligase activity resides in the RING finger domain, which is separated from the tyroskine kinase binding domain by a linker region. CBL and CBLB have two other domains that are not well conserved in CBLC: a proline-rich region involved in the recognition of SH3-proteins, and the carboxy-terminal UBA domain that interacts with ubiquitin molecules allowing dimer formation.24 CBL and CBLB play an important role in cell signaling in the majority of tissues, while CBLC activity seems to be restricted to epithelial cells.25–27
Over the last few years several groups have identified CBL mutations in different hematologic neoplasms, although most commonly in myelodysplastic syndromes (MDS)/MPN such as chronic myelomonocytic leukemia (CMML) and juvenile myelomonocytic leukemia.18–23,28–39 These changes cause the loss of E3-ubiquitin ligase activity, resulting in deregulation of downstream targets and an increase in cell proliferation rates. To our knowledge, CBL mutations seem to be mutually exclusive with other mutations frequently found in these diseases such as Ras mutations, FLT3-ITD or V617FJAK2. In this study we searched for mutations in CBL, CBLB and CBLC in a group of 172 V617FJAK2-negative and 232 V617FJAK2-positive MPN patients not selected for loss of heterozygosity, using a denaturing high performance liquid chromatography (dHPLC) method. Although most of the mutations described previously have been found in the RING finger domain and in the linker region of CBL, we found novel mutations also in the proline-rich domain, both in V617FJAK2-positive and -negative patients.
Design and Methods
Samples
Blood samples were collected from 404 different Caucasian MPN patients without the BCR-ABL1 fusion from several hospitals from the north of Spain. Informed consent was obtained from individual patients and the study was approved by the internal Ethics Committee. The first series of patients included 44 with V617FJAK2-negative MPN (4 diagnosed as PV, 15 as ET, 4 as PMF and 21 as atypical MPN). Later, a second series of 128 V617FJAK2-negative MPN patients (16 PV, 81 ET and 31 PMF) and 232 V617FJAK2-positive MPN patients (69 PV, 149 ET and 14 PMF) were included. The presence/absence of V617FJAK2 mutation was determined in all patients by amplification refractory mutation system polymerase chain reaction (ARMS-PCR).40 In addition, all 404 samples were negative for the presence of MPL p.W515 mutations by dHPLC. Human leukemia cell lines HEL, M07e, UKE-1 and SET-2 were also included in the study (Table 1).
Initial mutational screening by dHPLC included 20 healthy (no disease) samples used as controls in order to check the frequency of sequence changes observed in our population. For those fragments in which we found sequence variants in patients, we also included 180 additional control samples in order to rule out that the changes detected were population polymorphisms.
Cell lines
Cell proliferation assays were performed on 32Dcl3 (32D) murine myeloid cells (DSMZ N. ACC411) incubated at 37ºC in 5% CO2 and maintained in 90% RPMI 1640 medium with 10% fetal bovine serum supplemented with 10 ng/mL murine interleukin-3 (Recombinant Mouse IL3, Cat #PMC0035, Gibco®, Invitrogen Ltd., Paisley, UK).
Frequency of CBL mutations found in our series.
Plasmids
Plasmids with tagged human open reading frames in pCMV6-AC-GFP vectors were purchased from Origene Technologies (Cat #RG214069 for CBL, RG206047 for CBLB and RG205130 for CBLC). The tagged human cDNA clone for FLT3 was also purchased from Origene Technologies as pCMV6-Entry vector (Cat #RC211459) and subcloned into pCMV6-AC-RFP vector (Cat #PS100034). These pCMV6-AC vectors carried the NeoR gene. The pCMV–HA ubiquitin vector was a gift from Dr. Francis Grand from Wessex Regional Genetics Laboratory (Salisbury, UK).
Denaturing high performance liquid chromatography analysis
Genomic DNA was obtained from all the samples and amplified with GenomiPhi v2.0 (GE Healthcare, Piscataway, NJ, USA) in order to obtain enough material for mutational screening. All mutations were confirmed using the original unamplified sample and no discrepancies were observed with whole-genome amplified DNA.
We designed primers with Primer341 to amplify all coding exons of the three screened genes (CBL, CBLB and CBLC) in flanking introns. For each fragment we also designed a mutant primer introducing a nucleotide change in the forward or reverse primer, depending on the corresponding melting profile, to create a control mutated fragment to validate each dHPLC assay. Melting profiles for PCR fragments, solvent gradients and temperature conditions were calculated by Navigator™ Software v1.6.2 (Transgenomic Ltd., Omaha, NE, USA) and validated experimentally. All the analyses were performed on a WAVE® 4500HT System (Transgenomic Ltd., Omaha, NE, USA) with a DNASep® HT cartridge. Online Supplementary Table S1 contains a list of primers, the sizes of the amplified fragments and dHPLC conditions.
PCR reactions were performed with AmpliTaq™ Gold (Applied Biosystems, Foster City, CA, USA) using standard protocols. After cycling, samples were subjected to several cycles of heating and cooling in order to create heteroduplex molecules to improve mutation detection by dHPLC. For each fragment, we sequenced two samples of each different elution profile. Results were analyzed with Mutation Surveyor v3.10 (SoftGenetics LLC, State College, PA, USA) and compared to genomic reference sequences (ENSG00000110395 for CBL, ENSG00000114423 for CBLB and ENSG00000142273 for CBLC).
All coding exons of CBL, CBLB and CBLC were initially analyzed in a group of 44 V617FJAK2-negative patients (4 PV, 4 PMF, 15 ET and 21 atypical MPN). In light of the results we analyzed the RING finger domain coding exons (exons 8 and 9 from CBL, exons 9 and 10 from CBLB and exons 7 and 8 from CBLC) in 128 V617FJAK2-negative (16 PV, 31 PMF, 81 ET) and in 232 V617FJAK2-positive MPN (69 PV, 14 PMF, 149 ET), as well as in human leukemia cell lines M07e, HEL, SET-2 and UKE-1. We also included CBL exon 12 in this extended analysis because we observed a p.A678V change in one sample from the initial series.
CBL exon 8 deletions
Some of the mutations described for CBL are large deletions involving exon 8 (RING finger domain)18,21–23,29,30,32,37,42 and the design of our mutation screening assay was not able to detect some of them. We, therefore, designed a new PCR assay with primers located in exon 7 and intron 9 (E7Fw: 5’-TCCTGATGGAC-GAAATCAGA-3’; E9-Rv: 5’-CTCACAATGGATTTTGCCAGT-3’) which would amplify a normal fragment of 989 bp. With this assay, any large deletion of exon 8 would be detected as a product of smaller size.
Site-directed mutagenesis
All missense mutations detected for each gene were functionally tested. Mutants p.R420Q (used as the control mutant), p.C416W, p.P417R, p.T402HfsX29, p.S675C and p.A678V for CBL; p.R462W for CBLB and p.Q419PfsX81, p.P435S and p.E392K for CBLC were obtained using the QuikChange II Site-Directed Mutagenesis Kit (Agilent Technologies, La Jolla, CA, USA) from the original plasmids.
Transfection
Transfections were performed with Amaxa Nucleofector® Device II technology (Lonza Cologne GmbH, Basel, Switzerland) according to standard protocols. Cells of the 32D cell line in exponential growth were first transfected with FLT3 vector and maintained in medium until a second CBL/Ubi transfection. From the first transfection with FLT3 vector, cells were grown with Geneticin (G-418 sulfate, Cat #11811 Gibco®, Invitrogen Ltd., Paisley, UK) to select those clones that had incorporated the vector.
Cell proliferation assays
Proliferation analysis was performed with the CellTiter 96® AQueous One Solution Cell Proliferation Assay (MTS, Ref #G3580, Promega Corp, Madison, WI, USA) according to standard protocols, comparing cells transfected with wild-type CBL (CBL, CBLB or CBLC in each case) with cells transfected with mutant CBL during 3 or 4 days, in triplicate. In each case we carried out four different experiments, also including 32D(FLT3) cells transfected with pCMV6-AC-GFP and mock-transfected 32D(FLT3) cells as controls. In all cases cells were supplemented with 10 ng/mL recombinant human FLT3-ligand (Cat #GF038, Millipore, Temecula, CA, USA) and with 10 ng/mL murine interleukin-3 (Recombinant Mouse IL3, Cat #PMC0035, Gibco®, Invitrogen Ltd., Paisley, UK). As the positive mutant control we used p.R420Q, a previously described CBL mutant with an effect on cell proliferation.28
Statistical analysis
Results from MTS proliferation assays were compared using the Student’s t-test implemented in UNStat (a free tool available at http://www.unav.es/departamento/genetica/unstat).
Results
Mutational screening
In the initial screening of all coding exons of CBL, CBLB and CBLC in 44 patients with V617FJAK2-negative MPN we detected two missense changes not previously described in CBL (2/44; 4.5%). In this first series we also found three missense changes in CBLC described as single nucleotide polymorphisms (rs35457630, rs3208856, rs116023028) in the RING finger domain and proline-rich region. These changes were detected in samples from both patients and controls. No missense changes were detected in CBLB.
CBL changes (p.C416W or g.72251T>G and p.A678V or g.81664C>T) were found in patients diagnosed with atypical MPN, although the disease in the patient with p.A678V later evolved to CMML because of the development of dysplastic features. Whereas p.C416W affected the RING finger domain, like other mutations previously reported, p.A678V was located in exon 12, which codes for the proline-rich domain of CBL (see Figure 1). For this reason, we decided to include this exon in the analysis of CBL in an additional group of patients.
When we analyzed exon 12 of CBL and the RING finger domains of CBL, CBLB and CBLC in the additional series of samples (128 V617FJAK2-negative and 232 V617FJAK2-positive patients), we found three CBL changes in four V617FJAK2-positive patients (4/232, 1.7%). The first one (detected in a patient with ET) was a not previously reported g.71955_71955A deletion in exon 8. This is a frameshift change that truncates the RING finger domain with loss of the proline-rich and UBA carboxy-terminal domains (p.T402HfsX29) (Figure 1). The second change was a substitution g.72253C>G (p.P417R) in exon 9 in a patient diagnosed with PMF, also affecting the RING finger domain and previously identified in a patient with juvenile myelomonocytic leukemia.37 Finally, the third CBL mutation was a not previously reported g.81655C>G substitution (p.S675C) in exon 12 (proline-rich region). Remarkably, this change was detected in two different V617FJAK2-positive patients, one with ET and the other with PV. None of the 200 control samples analyzed showed any of these changes. CBL exon 8 deletions were not observed in any case.
(A) CBL mutations described to date in myeloid malignancies. Mutations found in our analysis are colored in red. In contrast to other studies in which mutations were found in the linker region (LR) and the RING finger domain, we also found two novel changes in CBL exon 12 (proline-rich domain). (B) Sequencing results observed for each mutation. p.T402HfsX29 is a frameshift mutation (g.71955_71955delA) from the amino acid Thr402 that results in a truncated protein lacking carboxy-terminal domains. This change was identified in patient with ET. p.C416W was due to a g.72251T>G substitution identified in a patient with an atypical MPN. We also found a previously reported change, p.P417R (g.72253C>G) in a patient with PMF. Changes found in exon 12 were p.S675C (g.81655C>G) in two cases (one with PV and another with ET) and p.A678V (g.81664C>T) in an atypical MPN that later evolved to CMML. While p.C416W and p.A678V changes were detected in patients negative for the p.V617FJAK2 mutation, p.T402HfsX29, p.P417R and p.S675C changes were found in patients with the p.V617FJAK2 mutation.
We also detected a not previously reported substitution (g.149486C>T, p.R462W) in the RING finger domain of CBLB in a sample from a V617FJAK2-positive patient with PV (1/232; 0.4%). In CBLC, we detected one frameshift change (g.15702_15703insC, p.Q419PfsX81), in a patient with V617FJAK2-negative PV, which has been described as a polymorphism (rs66944506).
(A to F) All missense mutations detected in CBL were functionally tested. Left, results of cell proliferation assays corresponding to four different transfections for each mutation. Two transfection controls (cells mock-transfected and cells transfected with GFP control vector) were included. Right, proliferation rates (final absorbance - initial absorbance) obtained for each mutant compared to wild-type (Student’s t-test) pooling data from four separate experiments. (A) Left, results of proliferation assays corresponding to the CBL p.R420Q mutant control. Seventy-two hours after the start of the assay, cells transfected with the mutant vector showed a significantly higher absorbance value than those transfected with the wild-type vector (P=0.034). Right, the proliferation rate of cells with p.R420Q was significantly higher than cells with wild-type (P=0.040). (B) Left, results of assays for CBL p.C416W mutation (P=0.003). Right, proliferation rate compared that of wild-type cells (P=0.023). (C) Left, results of assays for CBL p.P417R mutation (P=0.003). Right, proliferation rate compared to wild-type (P=0.003). (D) Left, results of assays for CBL p.T402HfsX29 mutation (P=0.001). Right, proliferation rate compared to wild-type (P<0.001). (E) Left, results of assays for CBL p.S675C mutation (P<0.001). Right, proliferation rate compared to wild-type (P<0.001). (F) Left, results of assays for CBL p.A678V mutation (P<0.001). Right, proliferation rate compared to wild-type (P=0.001).
None of the cell lines included in these analyses (HEL, SET-2, UKE-1 and M07e) showed any CBL, CBLB or CBLC mutation.
CBL mutations promote hypersensitivity to interleukin-3 in 32D(FLT3) cells
A significantly higher number of cells was observed in 32D(FLT3) cells transfected with CBL mutants than with wild-type CBL (P<0.05, Figure 2), grouping data from four independent cell proliferation assays. In addition, cells transfected with mutant vectors showed significantly higher proliferation rates in all cases (P<0.05, Figure 2) and with stronger effects than those observed for the p.R420Q control mutation.
By contrast, assays for p.R462W in CBLB and for p.Q419PfsX81, p.E392K and p.P435S in CBLC (Online Supplementary Figure S1) showed no significant differences (P>0.05) in proliferation rates.
Discussion
In the last few years the detection of regions with acquired loss of heterozygosity in some patients, mainly caused by acquired uniparental disomy, has allowed the identification of candidate genes that may be mutated in myeloid neoplasms. One of these genes is CBL, which codes for an E3-ubiquitin ligase protein. Cbl family proteins (CBL, CBLB and CBLC) play an important role as regulators of several signaling pathways promoting the ubiquitination and degradation of some RTK and CTK,44 many of which are involved in these diseases.4,45,46
The first CBL mutation identified was p.R420Q, affecting the RING finger domain in a patient with AML.28 Subsequently, other mutations have been reported with variable frequencies in myeloid neoplasms, affecting not only the RING finger domain but also the linker region (Figure 1). These events have been observed in 1–33% cases of secondary AML, 1–7% of MPN and 2–33% of MDS/MPN and AML,18–23,28–39 but their frequencies could be as high as 85–90% in patients with loss of heterozygosity in 11q.18–20,22,23,33 Some reports have also described that 7% of patients with non-small cell lung cancer have CBL mutations, so this gene can be mutated in other types of tumor.47 CBLB mutations and CBLC missense polymorphisms affecting the RING finger domain have also been described in myeloid neoplasms but at a lower frequency and with unknown effects.22,23,29
Evolutionary conservation of the CBL regions in which mutations have been detected. Sequence comparison between humans (Homo sapiens) and other mammals (Mus musculus, Rattus norvegicus, Pan troglodytes, Bos taurus, Macaca mulatta and Equus caballus) in the RING finger domain (A) and in the proline-rich domain (B). Both regions are highly conserved, suggesting an important functional role.
In this study we searched for mutations in CBL, CBLB and CBLC in a cohort of 404 V617FJAK2-negative and -positive MPN patients not selected for the presence of loss of heterozygosity in 11q. Our results show that CBL is mutated in V617FJAK2-negative MPN at a frequency similar to that previously reported (p.C416W and p.A678V; 1.2%, 2/172, Table 1).49 Both patients with mutations were initially diagnosed as having atypical MPN (2/21; 9.5%), although in one of them the disease evolved to CMML due to the development of dysplastic features. CMML is the disease with the highest frequency of CBL mutations reported to date.18–20,22,34 None of these mutations had been previously described and, notably, p.A678V was located in the proline-rich domain. In V617JAK2-positive MPN we found two mutations affecting the RING finger domain (p.T402HfsX29 in a patient with ET and p.P417R in a patient with PMF) and a recurrent change in the proline-rich domain (p.S675C in a patient with ET and in another one with PV) of CBL. Although TET2, ASXL1 and JAK2 mutations have been found concurrently,50 CBL mutations and V617FJAK2 seemed to be mutually exclusive events.13,21,26,51 However, we have found a similar frequency of CBL mutations in both V617FJAK2-positive and V617FJAK2-negative patients (Table 1), suggesting that the prevalence of CBL mutations could increase if V617FJAK2-positive patients were also included in CBL mutational studies. Unfortunately, we cannot know whether both mutations are in the same or in different clones or whether they are monoallelic or biallelic because of the type of sample available. None of the V617FJAK2-positive cell lines analyzed (HEL, SET-2 and UKE-1) showed CBL mutations that might help us to elucidate how both events could cooperate to drive the disease.43
In order to determine the effect of all these mutations on cell proliferation, in vitro functional assays were performed. All CBL mutations induced a hyperproliferative response to interleukin-3 in the 32D(FLT3) model, similar to that induced by the well-characterized p.R420Q mutation.28,31 This effect was not observed for the mutations detected in CBLB and CBLC. CBL was initially described as a putative tumor suppressor gene because of its negative regulatory function as an E3 ubiquitin ligase. Most of the mutations reported are located in conserved residues of the linker region and RING finger domain and could impair this regulatory function.44,46,52 RING finger domain mutations p.T402HfsX29, p.C416W and p.P417R described in this work also affect conserved residues of the protein (Figure 3) with a similar effect on the loss of activity of CBL.
Notably, we found two additional, novel mutations (one of them recurrent) affecting conserved residues in the pro-line-rich region (p.A678V and p.S675C, Figure 3) which also promote cell proliferation. In fact, we found similar frequencies of mutations in RING finger and proline-rich domains. The proline-rich region is essential for the interaction of CBL with the adaptor proteins (such as Grb2 and FRs2α) needed to maintain a stable attachment between CBL and its substrate,52 with proteins involved in the endocytosis of target receptors (such as SH3KBP1)46 and with several signaling proteins (such as the Src family).46
As in previous studies, we observed that mock-transfected cells showed greater growth than cells transfected with wild-type CBL, but less than cells transfected with mutant CBL. This fact is concordant with the proposal by some authors of a dominant negative effect of CBL mutant forms on endogenous wild-type CBL, making it unable to perform its negative regulatory function and promoting intracellular signaling and higher cell proliferation rates.18,26,27 However, in vivo studies have shown that the presence of gain-of-function mutants with a dominant effect over endogenous CBL is not enough to develop a myeloproliferative disease.18,26,27 A possible explanation for this phenomenon could be the activity of wild-type CBL as a positive regulator of cell growth contributing to the activation of pathways such as PI3K, Ras/MAPK and Src.52,53 Under normal conditions, the negative regulatory activity of CBL could mask its activity as a positive regulator, but the lack of E3-ubiquitin ligase activity could reveal its signaling enhancing activity.27 This could be the reason for the non-transforming effect of the p.R462W CBLB mutant in 32D(FLT3) cells. CBLB does not show the positive regulatory effects of CBL52 and perhaps the loss of its E3-ubiquitin ligase activity is not enough to promote cell proliferation. In fact, although CBLB activity seems to be similar to that of CBL in hematologic cells, very few cases of myeloid neoplasms with CBLB mutations have been reported.25,29
Finally, the results obtained in CBLC suggest that missense single nucleotide polymorphisms do not increase cell proliferation in our model. Wild-type CBLC induced higher proliferation rates in 32D(FLT3) cells than wild-type CBL in all assays (Online Supplementary Figure S1) possibly due to the absence of an inhibitory role of CBLC in hematologic cells.25,54,55
In conclusion, we have identified mutations in the proline-rich region of CBL in patients with MPN and also in V617FJAK2-positive patients. Although the entire CBL coding sequence has been investigated in some studies (by sequencing, not by dHPLC),18,19,22,32 most research in recent years has focused only on exons coding for the linker region and RING finger domain20,21,23,28,30,31,33–36,38 and in patients without other frequent genetic aberrations, such as mutations in JAK2.26 Proline-rich domain mutations (p.S675C and p.A678V) confer hypersensitivity to cytokines in the 32D(FLT3) model in a similar way to RING finger domain mutations (p.T402HfsX29, p.C416W and p.P427R), suggesting that they should also be considered in analyses of CBL. Although these events seem to be rare in MPN, our data highlight the importance of reevaluating the prevalence of CBL mutations in other regions of the gene in myeloid neoplasms. This could be of special interest in MDS/MPN because of the high incidence of CBL mutations in these diseases. In addition, further functional analyses of these genetic events could help us to understand the cellular functions of CBL and the role of the different protein domains. It is well known that CBL activity is mediated by the activation of different RTK, so the use of tyrosine kinase inhibitors (such as anti-FLT3) or other signal transduction inhibitors could also be effective in the treatment of patients with CBL mutations.27,32,56
Footnotes
-
Funding: this work was funded with the help of the Spanish Ministry of Science and Innovation (SAF 2007-62473), the PIUNA Program of the University of Navarra, the Caja Navarra Foundation through the Program “You choose, you decide” (Project 10.830) and ISCIII-RTICC (RD06/0020/0078).
-
PA received a predoctoral grant from the Government of Navarra.
-
The online version of this article has a Supplementary Appendix.
-
Authorship and Disclosures
The information provided by the authors about contributions from persons listed as authors and in acknowledgments is available with the full text of this paper at www.haematologica.org.
-
Financial and other disclosures provided by the authors using the ICMJE (www.icmje.org) Uniform Format for Disclosure of Competing Interests are also available at www.haematologica.org.
- Received July 26, 2011.
- Revision received January 23, 2012.
- Accepted January 26, 2012.
- Copyright© Ferrata Storti Foundation
References
Vol 97 Issue 8
