Original Articles

NOTCH1 and/or FBXW7 mutations predict for initial good prednisone response but not for improved outcome in pediatric T-cell acute lymphoblastic leukemia patients treated on DCOG or COALL protocols

Article metrics

Abstract

Aberrant activation of the NOTCH1 pathway by inactivating and activating mutations in NOTCH1 or FBXW7 is a frequent phenomenon in T-cell acute lymphoblastic leukemia (T-ALL). We retrospectively investigated the relevance of NOTCH1/FBXW7 mutations for pediatric T-ALL patients enrolled on Dutch Childhood Oncology Group (DCOG) ALL7/8 or ALL9 or the German Co-Operative Study Group for Childhood Acute Lymphoblastic Leukemia study (COALL-97) protocols. NOTCH1-activating mutations were identified in 63% of patients. NOTCH1 mutations affected the heterodimerization, the juxtamembrane and/or the PEST domains, but not the RBP-J-κ-associated module, the ankyrin repeats or the transactivation domain. Reverse-phase protein microarray data confirmed that NOTCH1 and FBXW7 mutations resulted in increased intracellular NOTCH1 levels in primary T-ALL biopsies. Based on microarray expression analysis, NOTCH1/FBXW7 mutations were associated with activation of NOTCH1 direct target genes including HES1, DTX1, NOTCH3, PTCRA but not cMYC. NOTCH1/FBXW7 mutations were associated with TLX3 rearrangements, but were less frequently identified in TAL1- or LMO2-rearranged cases. NOTCH1-activating mutations were less frequently associated with mature T-cell developmental stage. Mutations were associated with a good initial in vivo prednisone response, but were not associated with a superior outcome in the DCOG and COALL cohorts. Comparing our data with other studies, we conclude that the prognostic significance for NOTCH1/FBXW7 mutations is not consistent and may depend on the treatment protocol given.

Introduction

T-cell acute lymphoblastic leukemia (T-ALL) accounts for approximately 10–15% of all leukemias in children. Despite improved therapy, still 30% of these cases relapse and ultimately die.1, 2

Various chromosomal aberrations are known in T-ALL and some have been associated with prognosis.3, 4, 5 NOTCH1 may be important for T-ALL pathogenesis and was initially identified as part of rare t(7;9) translocations.6, 7 A role for NOTCH1 is now more clear as nearly 60% of T-ALL cases have NOTCH1 mutations affecting the heterodimerization (HD), the juxtamembrane domain (JM) or the proline, glutamic acid, serine, threonine-rich (PEST) domains.8, 9 HD or JM mutations result in ligand-independent proteolytical cleavages (reviewed in Grabher et al.10), resulting in the release of intracellular NOTCH1 (ICN). ICN is a transcription factor that regulates differentiation and proliferation through the activation of various target genes including cMYC, HES1 and PTCRA.10, 11, 12

As an alternative NOTCH1 activation mechanism, inactivating mutations in the F-Box WD40 domain containing protein 7 gene (FBXW7) were identified in 8–30% of T-ALL patients.13, 14, 15, 16 FBXW7 is part of the E3 ubiquitin ligase complex that controls the turnover of various proteins including ICN. FBXW7 interacts with phosphodegron domains located in the PEST domain of ICN. Therefore, inactivating mutations in FBXW7 or loss of the phosphodegron domains through truncating NOTCH1 PEST mutations both result in the stabilization of ICN in the nucleus. Mutations in FBXW7 and NOTCH1 PEST mutations are mutually exclusive,13, 14, 16 indicating that they seem to exert an equivalent oncogenic effect.

Mutations in NOTCH1 or FBXW7 may have prognostic relevance in T-ALL. Breit et al.17 reported that NOTCH1 mutant pediatric patients in the German ALL-BFM 2000 study show a good in vivo prednisone response and have an improved event-free survival (EFS). In contrast, Zhu et al.18 published an unfavorable outcome for NOTCH1-mutated adult T-ALL patients, but not for pediatric patients. We could not confirm a favorable prognostic effect for NOTCH1-mutated pediatric T-ALL patients treated on Dutch Childhood Oncology Group (DCOG) protocols,19 and this was confirmed by children treated on POG protocols for which no relation was identified between the presence of NOTCH1 mutations and relapse.20 These initial studies investigated the relevance for NOTCH1 HD and PEST mutations,17, 18, 19 but did not include NOTCH1 JM mutations or FBXW7 mutations. We now extended our initial study by examining the prognostic effect of NOTCH1 and FBXW7 mutations in 141 pediatric T-ALL patients treated on DCOG or German Co-Operative Study Group for Childhood Acute Lymphoblastic Leukemia study (COALL-97) protocols. The functional consequences of NOTCH1/FBXW7 mutations in relation to ICN levels and activation of target genes in primary leukemia samples were investigated.

Patients/materials and methods

Patient samples

This study comprised 146 primary pediatric T-ALL patients, of which 72 were treated on DCOG protocols ALL-7/8,21, 22 (n=30) or ALL-9 (n=42).23 This cohort had a median follow-up of 67 months, and included 51 male and 21 female patients. As the overall disease-free survival for patients treated on these DCOG protocols are comparable, these patients will be analyzed as one cohort as carried out before.19, 24 Of these, 70 patients were part of our previous study.19 For ALL7/8 patients, in vivo prednisone response was monitored at day 8 following 7 days of BFM-like prednisone monotherapy and one intrathecal dose of methotrexate. A clearance to less than 1000 blasts per μl blood at day 8 was considered as an initial prednisone good response (PGR). In total, 74 patients were enrolled in the German COALL-97 protocol19 with a median follow-up of 52 months. This cohort included 49 male and 25 female patients. The patients’ parents or legal guardians provided informed consent to use leftover diagnostic biopsies for research in accordance with the institutional review board and the Declaration of Helsinki Principles. Isolation of leukemia cells from blood or bone marrow samples has been described before,25 and all samples contained >90% of leukemic blasts. Clinical and immunophenotypic data were supplied by both study centers. Classification into T-cell development stages was based on EGIL criteria:26 pro-/pre- (CD7+, CD2+ and/or CD5+ and/or CD8+), cortical (CD1+) or mature T-cell stage (sCD3+/CD1).

Genomic DNA and RNA extraction

Isolation of genomic DNA and RNA from 5 × 106 leukemic cells using the TRIzol reagent (Invitrogen, Breda, the Netherlands) and copy DNA synthesis were carried out as described before.24, 25

Mutational detection

NOTCH1 exons 25–34 were screened for mutations that include all relevant domains (Supplementary Table S1). For FBXW7, the F-box and WD40 domains (exon 5, exons 7–11) were amplified, covering all FBXW7 mutations as reported so far. PCR reactions were carried out as described before.19 Primers are shown in Table 1. PCR products were sequenced using the BigDye Terminator version 3.1 Cycle Sequencing Kit (Applied Biosystems, Foster City, CA, USA) on a 3130 DNA Analyzer (Applied Biosystems).

Table 1 Distribution of wild-type and NOTCH1 mutations within categorized subgroups of T-ALL patients

Identification of recurrent rearrangements by FISH, RQ-PCR or array-CGH

SIL-TAL, CALM-AF10 or rearrangements of LMO2, TLX1, TLX3, TAL1, CALM-AF10, SET-NUP214, HOXA or MLL were determined with fluorescence in-situ hybridization (FISH) as previously described.24, 27, 28 NOTCH1 translocations were detected using bacterial artificial chromosomes clones RP11-769N4, RP11-1008C19, RP11-83N9 and RP11-662J2 covering both sides adjacent to the NOTCH1 locus. Bacterial artificial chromosomes were obtained from BAC/PAC Resource Center (Children's Hospital, Oakland, CA, USA). Expression levels of TLX1, TLX3, TAL1, LMO2 or HOXA or CALM-AF10 and SET-NUP214 fusion products were measured relative to the expression of glyceraldehyde-3-phosphate dehydrogenase as described before.27, 28 Array-CGH analysis was performed as previously described,27 on the human genome CGH Microarray 105 or 400K dual arrays (Agilent Technologies, Santa Clara, CA, USA), which consists of 105 000 or 400 000 60-mer oligonucleotide probes that span both coding and noncoding sequences with an average spatial resolution of 15 or 5 kb. Microarray images were analyzed using feature extraction software (version 8.1; Agilent Technologies) and the data were subsequently imported into array-CGH analytics software version 3.1.28 (Agilent Technologies).

Gene expression array analysis

RNA integrity testing, copy DNA and copy-copy RNA (ccRNA) syntheses, hybridization and washing to Human Genome U133 plus2.0 microarrays (Affymetrix, Santa-Clara, CA, USA), extraction of probeset intensities from CEL-files and normalization with RMA or VSN methods were performed as described before.28 Differentially expressed genes between NOTCH1 mutant versus wild-type T-ALL patients were determined by Wilcoxon statistics and corrected for multiple testing error29 using the Bioconductor package ‘Multtest’ in R. Heatmaps based on the TOP50 most significant differentially expressed genes were performed in Dchip software, Harvard University, Boston, MA, USA.30 Microarray data are available at http://www.ncbi.nlm.nih.gov/geo/.

Reverse-phase protein microarray analysis and western blot

Reverse-phase protein microarray construction and analysis was performed essentially as previously described.31, 32 To isolate proteins from 10 × 106 leukemic cells, lysis was performed in 20 μl tissue protein extraction reagent (Pierce Biotechnology, Rockford, IL, USA) with 300 nM NaCl, 1 mM orthovanadate and protease inhibitors. Cells were incubated at 4 °C for 20 min and subsequently centrifuged at 10 000 r.p.m. for 5 min in an Eppendorf centrifuge. Supernatants were stored at −80 °C before printing on the microarrays. Lysates were diluted to 1.0 mg/ml protein concentration and mixed 1:1 with 2 × SDS Tris-glycine buffer (Invitrogen) containing 5% 2-mercaptoethanol (Sigma, Zwijndrecht, the Netherlands) (FC=0.5 mg/ml). Lysates were spotted at a concentration of 0.5 μg/μl (neat spot) and 0.125 μg/μl in duplicate with 350 μm pins on glass-backed nitrocellulose coated array slides (FAST slides; Whatman, Kent, UK) using an Aushon Biosystems 2470 (Aushon Biosystems, Billerica, MA, USA). Printed slides were stored at −20 °C or directly used. The first of each 25 slides printed were subjected to Sypro Ruby Protein Blot staining (Invitrogen) to determine total protein amount. These slides were visualized on a NovaRay CCD fluorescent scanner (Alpha Innotech, San Leandro, CA, USA). The remaining slides were used for staining with a specific antibody. Before this, slides were incubated with 1 × Reblot (Chemicon, Temecula, CA, USA) for 15 min and subsequently washed with phosphate-buffered saline twice. This was continued with a blocking procedure for 5 h using 1gr I-Block (Applied Biosystems) diluted in 500 ml phosphate-buffered saline with 0.5% Tween 20. Slides were stained with an automated slide stainer (Dako, Glostrup, Denmark) according to the manufacturer's instructions using the Autostainer catalyzed signal amplification kit (Dako). In each staining run, a negative control slide was stained with the secondary antibody only for background subtraction. Briefly, endogenous biotin was blocked for 10 min with the biotin blocking kit (Dako), followed by application of protein block for 5 min; primary antibodies were diluted in antibody diluent and incubated on slides for 30 min and biotinylated secondary antibodies were incubated for 15 min. Signal amplification involved incubation with a streptavidin/biotin/peroxidase complex provided in the catalyzed signal amplification kit for 15 min, and amplification reagent (biotinyl-tyramide/hydrogen peroxide, streptavidin/peroxidase) for 15 min each. A signal is generated using streptavidin-conjugated IRDye680 (LI-COR Biosciences, Lincoln, NE, USA). Slides were allowed to air dry following development. Stained slides were scanned individually on the NovaRay scanner (Alpha Innotech) and files were saved in TIF format in Photoshop 7.0. All slides were subsequently analyzed with the MicroVigene version 2.8.1.0 program (VigeneTech, Carlisle, MA, USA). To screen for ICN protein levels, we have used and optimized the conditions for the ICN Val1744 antiserum (catalog no. 2421; Cell Signaling Technology, Beverly MA, USA). Slides were scanned in a NovaRay scanner (Alpha Innotech) and analyzed with the MicroVigene version 2.8.1.0 program (VigeneTech). For western blot validation,28 protein loading was validated by staining for Actin (Sigma, catalog no. 2547)

Statistics

Statistics were performed using SPSS 15.0 software (SPSS Inc., Chicago, IL, USA). The Pearson's χ2-test or the Fisher's exact test was used to test differences in the distribution of nominal data as indicated. Statistical significance for continuous distributed data was tested using the Mann—Whitney's U-test. Differences between patient populations in EFS and relapse-free survival (RFS) were tested by using the log-rank test. For RFS, an event is defined as relapse or nonresponse toward induction therapy at day 56 (COALL) or at start of consolidation therapy (DCOG). An event for EFS is defined as relapse, nonresponse toward induction therapy, death in remission because of toxicity or development of a secondary malignancy. Data were considered significant when P0.05 (two sided).

Results

NOTCH1 and/or FBXW7 mutations in pediatric T-ALL patients

Bone marrow or blood DNA samples for 146 primary T-ALL patients were analyzed for NOTCH1 (exons 25–34) and/or FBXW7 mutations (exons 5, 7–11) and 141 samples were successfully amplified and sequenced. The locations of mutations in specific NOTCH1 or FBXW7 domains are shown in Figure 1a.

Figure 1
figure1

NOTCH1 and FBXW7 mutations in pediatric T-ALL patients. (a) Schematic representation of identified mutations in the heterodimerization (HD), juxtamembrane (JM) and PEST domains in NOTCH1 and in the WD40-repeats of FBXW7. Missense mutations are indicated by an open triangle, a silent mutation is indicated by a filled gray triangle and nonsense mutations are indicated by a filled black triangle. (b) The distribution of NOTCH1 and FBXW7 mutation types in the DCOG and COALL cohorts.

Heterozygous mutations in NOTCH1 were detected in 79 out of 141 cases (56%), whereas 23 T-ALL patients (16%) harbored a point mutation in FBXW7. In total, 89 patients (63%) contained NOTCH1 and/or FBXW7 mutations. In total, 35 patients (39%) had a missense mutation or an in-frame insertion/deletion in the HD-domain of NOTCH1, whereas 9 (10%) and 13 (15%) patients harbored a combination of HD and PEST or FBXW7 mutations, respectively. Seventeen patients (19%) had a single NOTCH1 PEST mutation and ten (11%) had a single FBXW7 mutation (Figure 1b). We confirmed that NOTCH1 PEST domain mutations and FBXW7 mutations were nearly mutual exclusive,14, 16 but one patient carried a FBXW7 and a NOTCH1 PEST mutation. Five patients had a mutation in the JM domain of NOTCH1 (5.6%) of which one also had a NOTCH1 HD mutation. It is not known whether these JM and HD mutations occurred in cis or affected different alleles.

In total, 66 NOTCH1 mutations were found and 10 HD and 9 PEST mutations were not reported before to the best of our knowledge (Supplementary Figure S1). Ten FBXW7 point mutations were found, five of which have not been observed before in T-ALL (Supplementary Figure S2). These are H379L in exon 7, R465P in exon 8 and K622STOP, G687V and E693K in exon 11. The E693K mutation was previously identified in a gastric carcinoma patient.33

NOTCH1 and/or FBXW7 mutations activate ICN and downstream target genes in primary T-ALL samples

As published for T-ALL cell lines,8, 9, 11, 12, 14, 16 we demonstrated by using reverse-phase protein microarrays that NOTCH1 and/or FBXW7 mutations result in enhanced levels of ICN in primary T-ALL cells. The specificity of the NOTCH1 antibody was validated on the T-ALL cell line HPB-ALL, and ICN detection was lost on treatment with a γ-secretase inhibitor (Figure 2a). NOTCH1 and/or FBXW7-mutated patients showed about twofold higher ICN levels compared with wild-type patients (Figure 2b, P=0.0015). Strikingly, four wild-type patients also showed high ICN levels despite the absence of NOTCH1 and/or FBXW7 mutations (Figure 2b). Subsequent FISH and array-CGH analyses ruled out potential NOTCH1 translocations or other chromosomal NOTCH1 rearrangements in these four patients (data not shown).

Figure 2
figure2

NOTCH1/FBXW7 mutations activate the NOTCH1 pathway in primary T-ALL patient biopsies. (a) Western blot analysis of lysates from the HPBALL T-ALL cell line, which is NOTCH1-mutated. Treatment for 96 h with γ-secretase inhibitors including compound E (100 nM) or DAPT (5 μM), results in loss of activated intracellular NOTCH1 expression (ICN). Actin was used as loading control. (b) NOTCH1 ICN levels in wild-type and NOTCH1 and/or FBXW7-mutated T-ALL patients and T-ALL cell lines analyzed with reverse-phase protein microarray. NOTCH1/FBXW7 wild-type patient samples with high ICN levels are marked by an asterisk (c) Heatmap showing the TOP50 most differentially expressed genes between NOTCH1 and/or FBXW7 mutant patients versus wild-type patients. NOTCH1 direct target genes are indicated. Annotations indicated are genetic rearrangements, Gender and NOTCH1/FBXW7 mutation status. Genetic rearrangements indicated are: T, TAL1 or SIL-TAL1; L, LMO1 or LMO2 (includes del(11)(p12p13)); A, HOXA-activated (includes cases with SET-NUP214; CALM-AF10 or Inv(7)(p15q34)); 1, TLX1; 2, TLX2; 3, TLX3; O, Other; U, Aberration unknown. Gender is indicated F, female or M, male. NOTCH1/FBXW7 mutation status is indicated 0, wild-type and 1, NOTCH1- and/or FBXW7-mutated; NOTCH1/FBXW7 wild-type patients with high ICN levels are marked by an asterisk; NOTCH1/FBXW7 wild-type patients having a NOTCH1 signature that cluster with NOTCH1-activated patients based upon hierarchical clustering based on the TOP50 probeset are indicated with a filled triangle.

We investigated whether NOTCH1/FBXW7 mutations would result in the activation of specific genes. Expression array data28, 34 were available for 111 T-ALL patients with a known NOTCH1/FBXW7 mutation status. The TOP50 most significant and differentially expressed genes (probesets) between NOTCH1/FBXW7 mutant and wild-type patients comprised previous published and validated NOTCH1 direct target genes including HES1, HES4, DTX1, PTCRA, NOTCH3, PTPRC, CR2, LZTFL1, TASP1, SHQ1 and RHOU (Figure 2c).10, 11 Although cMYC is a NOTCH1 target gene in T-ALL cell lines, this gene did not appear in our TOP50 nor TOP200 gene lists (not shown). Eight wild-type patients also seemed to express genes from this NOTCH1 signature (Figure2c; data not shown). For six out of these eight patients for which ICN levels were available, two patients were among the four wild-type cases having the highest ICN protein levels. Similarly to these two cases having a NOTCH1 signature and high ICN levels, none of the remaining six patients with a NOTCH1 signature carried NOTCH1 translocations or alternative chromosomal abnormalities based on FISH and array-CGH results (data not shown).

NOTCH1/FBXW7 mutations in relation to clinical, immunophenotypic and cytogenetic parameters

We did not observe a relationship between NOTCH1/FBXW7 mutations with gender, age or white blood cell counts (Table 1). For 23 patients, the in vivo prednisone response was known. NOTCH1-activated patients were correlated with a good in vivo prednisone response as 14 out of 16 patients with an initial PGR contained NOTCH1 mutations, in contrast to 2 out of 7 cases with a poor response (P=0.01). This observation was stronger by including FBXW7 data where 15 out of 16 cases with a PGR had a NOTCH1/FBXW7 mutation in contrast to only 2 out of 7 prednisone poor response cases (P=0.003, Table 1). Classification into T-cell development stages on EGIL criteria26 revealed that NOTCH1/FBXW7 mutations were less frequently identified in mature T-ALL cases (P=0.05, Table 1). In relation to molecular cytogenetic data, NOTCH1/FBXW7 mutations were identified in all cytogenetic T-ALL subgroups (Table 1). Considering TAL1- or LMO2-rearranged cases as a single TAL/LMO entity based on their identical expression profiles,28, 34 and including an additional 19 TALLMO-like patients with a TAL/LMO signature that lack TAL1 or LMO2 rearrangements,28 NOTCH1 mutations were less frequent. Only 25 out of 60 TALLMO patients (42%) had a NOTCH1 mutation (P=0.002, Table 1). This remained significant when including FBXW7 mutations as only 30 out of 60 cases (50%) had a NOTCH1/FBXW7 mutation (P=0.004). NOTCH1 mutations were more prevalent in TLX3-rearranged cases, in which 21 out of 27 cases (86%) had a NOTCH1 mutation (P=0.02). This remained significant when taking FBXW7 mutations into account (P=0.01).

Prognostic relevance of NOTCH1 and/or FBXW7 mutations

We then investigated the relevance of NOTCH1 and/or FBXW7 mutations in relation to treatment outcome. For the DCOG cohort, mutations in NOTCH1 and/or FBXW7 tended toward poor treatment outcome. The 5-year EFS rates for patients with NOTCH1 mutations only compared with wild-type patients were 57±8% versus 76±8% (P=0.08) for the DCOG cohort but 63±8 versus 64±10% for the COALL cohort (P=0.99, Figures 3a and b). Inclusion of FBXW7 mutations resulted in 5-year EFS rates of 58±7 versus 74±9% (P=0.16) for the DCOG cohort and 63±8 versus 68±10% for the COALL cohort (P=0.90; data not shown).

Figure 3
figure3

NOTCH1-activating mutations have no prognostic implication in pediatric T-ALL. Event-free survival (EFS) (a, b) and Relapse-free survival (RFS) (c, d) for the pediatric T-ALL patients treated on DCOG protocols (a, c) or the COALL protocol (b, d). Patients carrying NOTCH1 and/or FBXW7 mutations and wild-type patients have been indicated.

Events in both cohorts are summarized in Table 2. NOTCH1 mutations tended toward a lower RFS in the DCOG (P=0.068) and COALL cohorts (P=0.094) with 5-year RFS of 83±7 versus 62±8% for the DCOG cohort and 89±6 versus 67±8% for the COALL cohort for wild-type and NOTCH1-mutated patients, respectively (Figures 3c and d). These trends became less evident when including FBXW7 mutation data, with an RFS of 82±8 versus 62±8% in the DCOG cohort (P=0.101) and an RFS of 86±7 versus 70±8% in the COALL cohort (P=0.23) for wild-type or NOTCH1-activated patients (data not shown).

Table 2 Events in DCOG and COALL cohorts

We also investigated the effect for specific NOTCH1 and/or FBXW7 mutations on the activation of downstream target genes and outcome. As reported by the group of Pear and co-workers, specific NOTCH1 mutations or combinations of NOTCH1/FBXW7 mutations may have strong NOTCH1-activating effects, whereas others may only have modest activating effects.35 For this, we distinguished weak NOTCH1-activating mutations, that is, NOTCH1 HD or PEST mutations or FBXW7 mutations, and strong NOTCH1-activating mutations, that is, NOTCH1 JM mutations or combinations of NOTCH1 HD mutations with PEST mutations or FBXW7 mutations. Although ICN protein levels were significantly higher for NOTCH1- and/or FBXW7 -mutated cases versus wild-type cases, there was relation to the types of NOTCH1-activating mutations investigated (Supplementary Figure S3). To investigate differential activation of downstream target genes between patients with weak or strong NOTCH1-activating mutations, we first calculated the most significantly and differently expressed genes (probesets) between patients with strong NOTCH1-activating mutations versus wild-type patients, which again revealed mostly bonafide NOTCH1 target genes. However, these genes were expressed at intermediate levels for patients having weak NOTCH1-activating mutations (Supplementary Figure S4), indicating that these types of mutations indeed differ in their potential to activate downstream target genes in primary leukemic samples. Distinction between these types of mutations may also have prognostic significance as patients from the DCOG cohort with strong NOTCH1-activating mutations had a significant poor outcome relative to wild-type patients (P=0.012) as well as to patients carrying weak NOTCH1-activating mutations (P=0.048) (Supplementary Figure S5a). However, this observation could not be substantiated for COALL-97 T-ALL patients (Supplementary Figure S5b). We also investigated whether ICN protein levels itself had prognostic significance. As 55 out of 66 patients for whom ICN protein levels were available were treated on the COALL cohort, we divided these patients into quartiles and determined their RFS and EFS rates. However, no relationship between ICN protein levels and RFS or EFS was present (P=0.98 and 0.97, respectively).

Discussion

Activation of NOTCH1 as a consequence of activating NOTCH1 mutations or inactivating FBXW7 mutations is a frequent phenomenon in T-ALL.8 We screened for NOTCH1 and FBXW7 mutations in 141 pediatric T-ALL patient samples and identified NOTCH1 mutations in 56% and FBXW7 mutations in 16% of the patients. In total, 63% of the patients had an aberrantly activated NOTCH1 pathway due to mutations. In line with previous studies,14, 16 we observed that NOTCH1 PEST domain mutations and FBXW7 mutations occurred in a mutually exclusive manner with the exception of one patient. This patient had a nonsense mutation in FBXW7 in contrast to missense mutations that are normally observed in FBXW7-mutated patients. This implies that mutant FBXW7 but not truncated FBXW7 proteins exert a dominant-negative effect in the E3-ubiquitin ligase complex. Interestingly, Park et al.15 also discovered a nonsense mutation due to a 5 bp insertion in FBXW7 in combination with a NOTCH1 PEST mutation in a non-Hodgkin's lymphoma patient.

The frequency of NOTCH1-activating mutations is in line with other studies also comprising adult T-ALL patient series.36, 37 In adult studies, NOTCH1 and FBXW7 mutations were identified in 60–62% and 18–24% of the T-ALL patients, respectively. This indicates that the oncogenic role for NOTCH1/FBXW7 during T-cell oncogenesis remains conserved over age. We did not find evidence for mutations outside the NOTCH1 HD, JM and PEST domains in any of the 141 pediatric T-ALL patients indicating that reported mutations in the LNR region, the RAM, ANK and TAD domains are very rare.18, 36, 38

We found that NOTCH1 and FBXW7 mutations resulted in increased levels of cleaved NOTCH1 (ICN) in primary leukemia cells and was associated with the activation of NOTCH1 target genes,10, 11, 12 including HES1, HES4, DTX1, PTCRA, NOTCH3, PTPRC, CR2, LZTFL1, TASP1 and RHOU. This confirms that the mutations manifest functionally at the protein level in patient samples. We identified 10 patients that lacked NOTCH1 and/or FBXW7 mutations that either expressed high levels of ICN or that expressed NOTCH1 target genes. As we did not find chromosomal translocations or other types of rearrangements involving the NOTCH1 locus, this implies that additional mutation mechanisms in NOTCH1 or directly downstream regulatory genes must exist that so far been left unnoticed in T-ALL. Although cMYC was identified as a prominent NOTCH1 target in T-ALL cell lines,11, 12 it was not identified as target gene in primary samples. However, two cases expressed ectopic cMYC levels due to a t(8;14)(q24;q11) translocation, which were both wild-type for NOTCH and FBXW7, supporting a role for MYC as NOTCH1 target. Further research will be required to establish whether cMYC is generally upregulated by means of other oncogenic mechanisms in addition to activated NOTCH1 in primary samples and therefore left undetected, or that the expression of cMYC is rapidly lost on isolation of primary leukemic cells.

NOTCH1/FBXW7 mutations were identified at a lower frequency in T-ALL cases with a mature immunophenotype. This may explain the low incidence of NOTCH1/FBXW7 mutations in the TAL/LMO subgroup because TAL1 rearrangements, which are the most recurrent abnormality in this subgroup, are associated with a mature T-cell development arrest.24, 39 This is an interesting finding and suggests that the oncogenic role of NOTCH1 is less prominent in T-ALL cases arrested at a relative mature T-cell developmental stage. Interestingly, NOTCH1/FBXW7 mutations were identified at a higher frequency in TLX3-rearranged T-ALL. The oncogenic activation of NOTCH1 thus far has been regarded as one of the earliest acquired abnormalities in a preleukemic progenitor cell that therefore becomes committed to the T-ALL.10, 40 In this perspective, our data indicate that the importance of deregulated NOTCH1 as initiating event during T-cell oncogenesis depends on additional collaborating events like TLX3 or TAL1 rearrangements. It also suggests that the oncogenic program that is followed by T-ALL cases that eventually arrest at the mature development stage may be less dependent on NOTCH1. Whether NOTCH1-activating mutations represent truly initiating leukemic events or not needs to be established, as evidence is emerging that NOTCH1 activation in some T-ALL cases may have occurred as a secondary event which may be acquired or lost at relapse.41

In the study of Breit et al.,17 NOTCH1 mutations were associated with an initial PGR and a significantly lower minimal residual disease content at day 78. Our study supports this association with initial prednisone response for NOTCH1/FBXW7 mutant patients. This association is also validated for patients of the EORTC-CLG study. In that study, NOTCH1-activating mutations were also associated with reduced minimal residual disease during therapy.42 The association for NOTCH1-activating mutations with PGR seems to be in contrast with the finding that γ-secretase inhibitors can sensitize for glucocorticoids in glucocorticoid-resistant cells.43 It may be that the NOTCH pathway has opposing effects in the glucocorticoid response in responsive against resistance patients, but it now seems clear that activation of NOTCH1 by mutations does not drive glucocorticoid resistance. Further research will be required to clarify this seeming contradiction.

NOTCH1 mutations are not associated with a superior outcome for patients treated on the BFM-like DCOG protocols or the COALL-97 protocol. The survival rate of NOTCH1-activated patients was actually less than for wild-type patients. Separating patients carrying strong NOTCH1-activating mutations from those with weak NOTCH1-activating mutations35 or patients that were wildtype showed a significant poor outcome for patients having strong NOTCH1-activating mutations in the DCOG cohort. This could not be reproduced for T-ALL patients treated on the German COALL-97 protocol. In the accompanying article of Clappier et al.,42 NOTCH1-activating mutations did not predict improved outcome for patients treated on the BFM-derived EORTC-CLG protocols either. These observations are in contrast to the findings by the BFM study group.17 In the accompanying article of Kox et al.,44 this finding is now validated in an extended series comprising 301 pediatric T-ALL patients treated on the ALL-BFM 2000 protocol. A favorable prognostic effect of NOTCH1 and/or FBXW7 mutations was also identified in a recent study by Park et al.,15 although the overall incidence of identified NOTCH1 mutations was only 31%. No favorable outcome of NOTCH1 and/or FBXW7-mutated cases has been observed neither for adult T-ALL patients treated on GMALL 05/93 and 06/99 multicenter protocols,45 nor for patients treated on the MRC UKALLXII/ECOG E299337 or LALA-9436 protocols. A significant association with improved outcome for NOTCH1-activating mutations has only been observed for adult T-ALL patients treated on the GRAALL-2003 multicenter protocol.36 These results indicate that the prognostic effect of NOTCH1/FBXW7 mutations may strongly depend on the treatment protocol given.

Compared with the ALL-BFM-2000 protocol, the DCOG ALL-7/8 protocol in general showed an inferior outcome.22 Although both protocols are highly related, part of the patients treated on the DCOG ALL-7/8 cohort received less chemotherapy and none of them received prophylactic cranial irradiation, except for patients with initial central nervous system involvement. NOTCH1-activating mutations may provoke central nervous system relapse because of the activation of the CCR7 chemokine.46 This study therefore predicts that NOTCH1-activating mutations would result in increased risk for central nervous system relapse through the CCL19-CCR7 axis in the absence of cranial irradiation. However, the numbers of central nervous system relapses in our cohorts were too low to substantiate this notion. In addition, neither the CCR7 gene nor its ligand CCL19 was identified as significantly differentially expressed genes that were activated in NOTCH1/FBXW7-mutated T-ALL patients based on our microarray expression data set (data not shown). As our patient biopsies were all obtained from peripheral blood or bone marrow samples, we cannot exclude that these genes are only upregulated in malignant blasts in the context of a neuronal environment. Cranial radiation may contribute to the differences in prognostic value for NOTCH1-activating mutations between the DCOG and ALL-BFM-2000 cohorts, but this does not apply for the COALL-97 cohort that includes cranial irradiation. Therefore, other differences among treatment protocols seem important.

In conclusion, NOTCH1/FBXW7 mutations that activate the NOTCH1 pathway are identified in >60% pediatric T-ALL patients and result in elevated ICN levels and activation of NOTCH1 target genes. Mutations were more often found in association with TLX3-rearranged T-ALL, but were less frequently identified in TAL/LMO T-ALL patients and T-ALL patients with a mature T-cell phenotype. NOTCH1/FBXW7 mutations predict for an initial PGR, which does not translate into a superior outcome of T-ALL on DCOG ALL-7/8, ALL-9 or COALL-97 protocols.

References

  1. 1

    Pieters R, Carroll WL . Biology and treatment of acute lymphoblastic leukemia. Pediatr Clin North Am 2008; 55: 1–20, ix.

  2. 2

    Pui CH, Evans WE . Treatment of acute lymphoblastic leukemia. N Engl J Med 2006; 354: 166–178.

  3. 3

    Cave H, Suciu S, Preudhomme C, Poppe B, Robert A, Uyttebroeck A et al. Clinical significance of HOX11L2 expression linked to t(5;14)(q35;q32), of HOX11 expression, and of SIL-TAL fusion in childhood T-cell malignancies: results of EORTC studies 58881 and 58951. Blood 2004; 103: 442–450.

  4. 4

    Meijerink JP, den Boer ML, Pieters R . New genetic abnormalities and treatment response in acute lymphoblastic leukemia. Semin Hematol 2009; 46: 16–23.

  5. 5

    Van Vlierberghe P, Pieters R, Beverloo HB, Meijerink JP . Molecular-genetic insights in paediatric T-cell acute lymphoblastic leukaemia. Br J Haematol 2008; 143: 153–168.

  6. 6

    Ellisen LW, Bird J, West DC, Soreng AL, Reynolds TC, Smith SD et al. TAN-1, the human homolog of the Drosophila notch gene, is broken by chromosomal translocations in T lymphoblastic neoplasms. Cell 1991; 66: 649–661.

  7. 7

    Suzuki S, Nagel S, Schneider B, Chen S, Kaufmann M, Uozumi K et al. A second NOTCH1 chromosome rearrangement: t(9;14)(q34.3;q11.2) in T-cell neoplasia. Leukemia 2009; 23: 1003–1006.

  8. 8

    Weng AP, Ferrando AA, Lee W, Morris JPt, Silverman LB, Sanchez-Irizarry C et al. Activating mutations of NOTCH1 in human T cell acute lymphoblastic leukemia. Science 2004; 306: 269–271.

  9. 9

    Sulis ML, Williams O, Palomero T, Tosello V, Pallikuppam S, Real PJ et al. NOTCH1 extracellular juxtamembrane expansion mutations in T-ALL. Blood 2008; 112: 733–740.

  10. 10

    Grabher C, von Boehmer H, Look AT . Notch 1 activation in the molecular pathogenesis of T-cell acute lymphoblastic leukaemia. Nat Rev Cancer 2006; 6: 347–359.

  11. 11

    Palomero T, Lim WK, Odom DT, Sulis ML, Real PJ, Margolin A et al. NOTCH1 directly regulates c-MYC and activates a feed-forward-loop transcriptional network promoting leukemic cell growth. Proc Natl Acad Sci USA 2006; 103: 18261–18266.

  12. 12

    Weng AP, Millholland JM, Yashiro-Ohtani Y, Arcangeli ML, Lau A, Wai C et al. c-Myc is an important direct target of Notch1 in T-cell acute lymphoblastic leukemia/lymphoma. Genes Dev 2006; 20: 2096–2109.

  13. 13

    Malyukova A, Dohda T, von der Lehr N, Akhoondi S, Corcoran M, Heyman M et al. The tumor suppressor gene hCDC4 is frequently mutated in human T-cell acute lymphoblastic leukemia with functional consequences for Notch signaling. Cancer Res 2007; 67: 5611–5616.

  14. 14

    O’Neil J, Grim J, Strack P, Rao S, Tibbitts D, Winter C et al. FBW7 mutations in leukemic cells mediate NOTCH pathway activation and resistance to gamma-secretase inhibitors. J Exp Med 2007; 204: 1813–1824.

  15. 15

    Park MJ, Taki T, Oda M, Watanabe T, Yumura-Yagi K, Kobayashi R et al. FBXW7 and NOTCH1 mutations in childhood T cell acute lymphoblastic leukaemia and T cell non-Hodgkin lymphoma. Br J Haematol 2009; 145: 198–206.

  16. 16

    Thompson BJ, Buonamici S, Sulis ML, Palomero T, Vilimas T, Basso G et al. The SCFFBW7 ubiquitin ligase complex as a tumor suppressor in T cell leukemia. J Exp Med 2007; 204: 1825–1835.

  17. 17

    Breit S, Stanulla M, Flohr T, Schrappe M, Ludwig WD, Tolle G et al. Activating NOTCH1 mutations predict favorable early treatment response and long-term outcome in childhood precursor T-cell lymphoblastic leukemia. Blood 2006; 108: 1151–1157.

  18. 18

    Zhu YM, Zhao WL, Fu JF, Shi JY, Pan Q, Hu J et al. NOTCH1 mutations in T-cell acute lymphoblastic leukemia: prognostic significance and implication in multifactorial leukemogenesis. Clin Cancer Res 2006; 12: 3043–3049.

  19. 19

    van Grotel M, Meijerink JP, van Wering ER, Langerak AW, Beverloo HB, Buijs-Gladdines JG et al. Prognostic significance of molecular-cytogenetic abnormalities in pediatric T-ALL is not explained by immunophenotypic differences. Leukemia 2008; 22: 124–131.

  20. 20

    Larson Gedman A, Chen Q, Kugel Desmoulin S, Ge Y, LaFiura K, Haska CL et al. The impact of NOTCH1, FBW7 and PTEN mutations on prognosis and downstream signaling in pediatric T-cell acute lymphoblastic leukemia: a report from the Children′s Oncology Group. Leukemia 2009; 23: 1417–1425.

  21. 21

    Kamps WA, Bokkerink JP, Hahlen K, Hermans J, Riehm H, Gadner H et al. Intensive treatment of children with acute lymphoblastic leukemia according to ALL-BFM-86 without cranial radiotherapy: results of Dutch Childhood Leukemia Study Group Protocol ALL-7 (1988–1991). Blood 1999; 94: 1226–1236.

  22. 22

    Kamps WA, Bokkerink JP, Hakvoort-Cammel FG, Veerman AJ, Weening RS, van Wering ER et al. BFM-oriented treatment for children with acute lymphoblastic leukemia without cranial irradiation and treatment reduction for standard risk patients: results of DCLSG protocol ALL-8 (1991–1996). Leukemia 2002; 16: 1099–1111.

  23. 23

    Veerman AJ, Kamps WA, van den Berg H, van den Berg E, Bokkerink JP, Bruin MC et al. Dexamethasone-based therapy for childhood acute lymphoblastic leukaemia: results of the prospective Dutch Childhood Oncology Group (DCOG) protocol ALL-9 (1997–2004). Lancet Oncol 2009; 10: 957–966.

  24. 24

    van Grotel M, Meijerink JP, Beverloo HB, Langerak AW, Buys-Gladdines JG, Schneider P et al. The outcome of molecular-cytogenetic subgroups in pediatric T-cell acute lymphoblastic leukemia: a retrospective study of patients treated according to DCOG or COALL protocols. Haematologica 2006; 91: 1212–1221.

  25. 25

    Stam RW, den Boer ML, Meijerink JP, Ebus ME, Peters GJ, Noordhuis P et al. Differential mRNA expression of Ara-C-metabolizing enzymes explains Ara-C sensitivity in MLL gene-rearranged infant acute lymphoblastic leukemia. Blood 2003; 101: 1270–1276.

  26. 26

    Bene MC, Castoldi G, Knapp W, Ludwig WD, Matutes E, Orfao A et al. Proposals for the immunological classification of acute leukemias. European Group for the Immunological Characterization of Leukemias (EGIL). Leukemia 1995; 9: 1783–1786.

  27. 27

    Van Vlierberghe P, van Grotel M, Beverloo HB, Lee C, Helgason T, Buijs-Gladdines J et al. The cryptic chromosomal deletion del(11)(p12p13) as a new activation mechanism of LMO2 in pediatric T-cell acute lymphoblastic leukemia. Blood 2006; 108: 3520–3529.

  28. 28

    Van Vlierberghe P, van Grotel M, Tchinda J, Lee C, Beverloo HB, van der Spek PJ et al. The recurrent SET-NUP214 fusion as a new HOXA activation mechanism in pediatric T-cell acute lymphoblastic leukemia. Blood 2008; 111: 4668–4680.

  29. 29

    Hochberg Y, Benjamini Y . More powerful procedures for multiple significance testing. Stat Med 1990; 9: 811–818.

  30. 30

    Li C, Wong WH . Model-based analysis of oligonucleotide arrays: expression index computation and outlier detection. Proc Natl Acad Sci USA 2001; 98: 31–36.

  31. 31

    Paweletz CP, Charboneau L, Bichsel VE, Simone NL, Chen T, Gillespie JW et al. Reverse-phase protein microarrays which capture disease progression show activation of pro-survival pathways at the cancer invasion front. Oncogene 2001; 20: 1981–1989.

  32. 32

    Petricoin 3rd EF, Espina V, Araujo RP, Midura B, Yeung C, Wan X et al. Phosphoprotein pathway mapping: Akt/mammalian target of rapamycin activation is negatively associated with childhood rhabdomyosarcoma survival. Cancer Res 2007; 67: 3431–3440.

  33. 33

    Lee JW, Soung YH, Kim HJ, Park WS, Nam SW, Kim SH et al. Mutational analysis of the hCDC4 gene in gastric carcinomas. Eur J Cancer 2006; 42: 2369–2373.

  34. 34

    Homminga I, langerak AW, De Laat W, de Rooij JJ, Stubbs A, Buijs-Gladdines J et al. NKX2-1 and MEF2C oncogenes delineate novel subtypes in T-cell acute lymphoblastic leukemia. (in preparation).

  35. 35

    Chiang MY, Xu L, Shestova O, Histen G, L′Heureux S, Romany C et al. Leukemia-associated NOTCH1 alleles are weak tumor initiators but accelerate K-ras-initiated leukemia. J Clin Invest 2008; 118: 3181–3194.

  36. 36

    Asnafi V, Buzyn A, Le Noir S, Baleydier F, Simon A, Beldjord K et al. NOTCH1/FBXW7 mutation identifies a large subgroup with favorable outcome in adult T-cell acute lymphoblastic leukemia (T-ALL): a Group for Research on Adult Acute Lymphoblastic Leukemia (GRAALL) study. Blood 2009; 113: 3918–3924.

  37. 37

    Mansour MR, Sulis ML, Duke V, Foroni L, Jenkinson S, Koo K et al. Prognostic implications of NOTCH1 and FBXW7 mutations in adults with T-cell acute lymphoblastic leukemia treated on the MRC UKALLXII/ECOG E2993 protocol. J Clin Oncol 2009; 27: 4352–4356.

  38. 38

    Gordon WR, Roy M, Vardar-Ulu D, Garfinkel M, Mansour MR, Aster JC et al. Structure of the Notch1-negative regulatory region: implications for normal activation and pathogenic signaling in T-ALL. Blood 2009; 113: 4381–4390.

  39. 39

    Asnafi V, Radford-Weiss I, Dastugue N, Bayle C, Leboeuf D, Charrin C et al. CALM-AF10 is a common fusion transcript in T-ALL and is specific to the TCRgammadelta lineage. Blood 2003; 102: 1000–1006.

  40. 40

    Eguchi-Ishimae M, Eguchi M, Kempski H, Greaves M . NOTCH1 mutation can be an early, prenatal genetic event in T-ALL. Blood 2008; 111: 376–378.

  41. 41

    Mansour MR, Duke V, Foroni L, Patel B, Allen CG, Ancliff PJ et al. Notch-1 mutations are secondary events in some patients with T-cell acute lymphoblastic leukemia. Clin Cancer Res 2007; 13: 6964–6969.

  42. 42

    Clappier E, Collette S, Grardel N, Girard S, Suarez L, Brunie G et al. Prognostic significance of NOTCH1 and FBXW7 mutations in childhood T-cell acute leukemia: results from the EORTC Children Leukemia Group. Leukemia 2010, (this issue).

  43. 43

    Real PJ, Tosello V, Palomero T, Castillo M, Hernando E, de Stanchina E et al. Gamma-secretase inhibitors reverse glucocorticoid resistance in T cell acute lymphoblastic leukemia. Nat Med 2009; 15: 50–58.

  44. 44

    Kox C, Zimmermann M, Stanulla M, Leible S, Schrappe M, Ludwig WD et al. The favorable effect of activating NOTCH1 receptor mutations on long-term outcome in T-ALL can be seperated from NOTCH pathway activation by FBXW7 loss of function. Leukemia 2010, (this issue).

  45. 45

    Baldus CD, Thibaut J, Goekbuget N, Stroux A, Schlee C, Mossner M et al. Prognostic implications of NOTCH1 and FBXW7 mutations in adult acute T-lymphoblastic leukemia. Haematologica 2009; 94: 1383–1390.

  46. 46

    Buonamici S, Trimarchi T, Ruocco MG, Reavie L, Cathelin S, Mar BG et al. CCR7 signalling as an essential regulator of CNS infiltration in T-cell leukaemia. Nature 2009; 459: 1000–1004.

Download references

Acknowledgements

LZ and WKS were financed by the Stichting Kinderen Kankervrij (KiKa; Grant no. KiKa 2007-012). IH was financed by the Dutch Cancer Society Dutch Cancer Society (KWF-EMCR 2006-3500) CK was financed by KiKa (Grant no. KiKa 2008-029). We thank the German Jose Carreras Leukemia Foundation (Grant no. SP 04/03 to MH).

Author information

Correspondence to J P P Meijerink.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Author contributions

LZ designed experiments, performed research and wrote the article; IH performed research and wrote the article; VC performed RPMA analysis; MLW performed research; JB-G performed NOTCH1 and FBXW7 mutation analysis; CK performed western blot analysis; WS prepared samples for RPMA analysis; ES, AJPV, WK and MH provided patient samples and clinical and immunophenotypic data; EP supervised study and wrote the article; RP designed and supervised study and wrote the article; JPPM was principal investigator, designed and supervised the study and wrote the article.

Supplementary Information accompanies the paper on the Leukemia website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Zuurbier, L., Homminga, I., Calvert, V. et al. NOTCH1 and/or FBXW7 mutations predict for initial good prednisone response but not for improved outcome in pediatric T-cell acute lymphoblastic leukemia patients treated on DCOG or COALL protocols. Leukemia 24, 2014–2022 (2010) doi:10.1038/leu.2010.204

Download citation

Keywords

  • pediatric T-ALL
  • NOTCH1
  • FBXW7
  • TLX3
  • prednisone response
  • outcome

Further reading