
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Human Cancer Biology |
Authors' Affiliations: 1 Department of Haematology, University College London; 2 Department of Haematology, Royal Free Hospital; and 3 Great Ormond Street Hospital, London, United Kingdom
Requests for reprints: Marc R. Mansour, Department of Haematology, University College London, 98 Chenies Mews, London WC1E 6HX, United Kingdom. E-mail: m.mansour{at}ucl.ac.uk.
| Abstract |
|---|
|
|
|---|
Experimental Design: We used denaturing high-performance liquid chromatography, sequencing, and fragment analysis to analyze Notch-1 mutational status and mutant level in 62 patients at presentation as well as 16 matched presentation-relapse samples.
Results: We detected Notch-1 mutations in 47 patients (76%). Seven of these were low-level mutations (quantified at
10%), despite high blast counts, suggesting that they were acquired as a secondary event in a subclone. Of 16 matched presentation-relapse samples studied, 7 were wild-type at both presentation and relapse. Five of nine mutant-positive patients at presentation relapsed with the same mutation(s) at the same high level. Four patients had evidence of a change in mutant at relapse. One lost a PEST mutation and became wild-type. Two others lost mutations at relapse but acquired different mutations, despite unchanged T-cell receptor rearrangements, suggesting that the latter event predated the acquisition of the Notch-1 mutation. One relapsed with a secondary T-cell leukemia and different Notch mutation.
Conclusions: These results suggest that Notch-1 mutations can sometimes be acquired as secondary events in leukemogenesis and must be used cautiously as solitary minimal residual disease markers.
-secretase at the transmembrane domain. This liberates intracellular Notch-1 (ICN) to translocate to the nucleus and induce downstream signaling events (5–7). The first evidence implicating aberrant Notch signaling in leukemogenesis came from identification of the rare t(7;9) translocation in three pediatric patients with T-cell acute lymphoblastic leukemia (T-ALL), which juxtaposed the T-cell receptor (TCR) β promoter with ICN and led to constitutive Notch-1 activation (8). Subsequently, activating Notch-1 mutations have been found as a frequent occurrence in both pediatric and adult T-ALL, being detected in >50% of patients at disease presentation (9–11). Mutations tend to cluster in two hotspots: the HD, where they are associated with ligand-independent cleavage of ICN (12), and the PEST domain, where they lead to a COOH-terminally truncated ICN. The latter results in loss of the negative regulatory domain, escape from FBXW7-mediated degradation and prolongation of the half-life of ICN (9, 13, 14). Approximately 20% of patients have mutations in both domains, usually occurring in cis, which results in synergistic Notch activation (9). Increased ICN levels have recently been shown to activate important oncogenes, including c-Myc and mammalian target of rapamycin (15–17).
The role of Notch as a primary event in leukemogenesis has been recapitulated in murine models, where animals transplanted with cells transduced with either ICN or the Notch ligand Delta-like 4 developed a lethal T-cell leukemia (18–20). Interestingly, Notch-1 mutations have also been found in the majority of mice with TAL-1–induced T-ALL as well as in mice deficient in various combinations of TP53, HAXA2, and Rag2 genes, suggesting that Notch-1 mutations can occur as secondary events and play a collaborative role in T-ALL pathogenesis (21, 22).
In human T-ALL while Notch-1 mutations occur in all cytogenetic and molecular subgroups, it is unclear whether they are primary initiating or secondary cooperating events. To address this issue, we have assessed the relative level of Notch-1 mutated alleles in a cohort of human patients at presentation with T-ALL and examined stability of these mutations at disease relapse.
| Materials and Methods |
|---|
|
|
|---|
Notch-1 mutational analysis. Notch-1 mutation screening and identification were carried out using heteroduplex analysis and sequencing as previously reported (10). The HD-N, HD-C, transactivation (TAD), and PEST domains were amplified by 35 cycles of the PCR using the primer sequences shown in Table 1 . The proofreading enzyme Optimase (Transgenomic) was used according to the manufacturer's specifications for the TAD and HD-C domain with the inclusion of 1 mol/L betaine due to the high GC content of these fragments. However, products of adequate quality could not be obtained using Optimase for the HD-N and PEST domains, and these fragments were amplified using BioTAQ DNA polymerase (Bioline). PCR products were analyzed by denaturing high-performance liquid chromatography (WAVE DNA Fragment Analysis System, Transgenomic). Abnormal chromatograms were confirmed by repeat analysis and samples were sequenced using the Dye Terminator Cycle Sequencing Quick Start kit and analyzed on the CEQ8000 Genetic Analysis System (Beckman Coulter).
|
TCR rearrangements. TCR rearrangements were determined at the TCR
locus using heteroduplex analysis and fluorescently labeled primers as previously reported (23–25). For those patients where no suitable marker was detected at the
locus, the
locus was analyzed.
Same patient identity. Same patient identity was confirmed in the paired presentation-relapse samples using size analysis at four highly informative polymorphic short tandem repeat loci (VWA, FES, D11S-554, and F13I; ref. 26). PCRs were carried out using BioTAQ DNA polymerase, a fluorescently labeled forward primer, and 30 cycles of amplification with annealing temperatures of 54°C for F131, FES, and VWA and 61°C for D11S-554. Products were sized by fragment analysis.
| Results |
|---|
|
|
|---|
60%). However, for 10 mutations, the mutant level was <30% and in six of these it was
10% (Fig. 2), indicating that
20% of cells carried the mutation (Table 2
). Two of these low-level mutations were in the same patient (#50; Table 2). The remaining 27 mutations were nucleotide substitutions, which were not amenable to quantification. Of these, only one case (#40) had an apparent low-level mutation by visual estimation of the WAVE chromatogram, and all others were high level. The low-level mutants were not due to the presence of normal cells, as in each patient morphologic examination revealed at least 75% leukemic blasts. Furthermore, three of these patients had a high-level mutation in another domain. This suggested that the low-level mutations were acquired in a subclone as a later event. None of the patients with low-level mutations had evidence of ploidy or loss of genomic material of chromosome 9q by G-banding cytogenetics.
|
|
|
|
|
1 clone was identified at both presentation and relapse. Case 15 had a low-level HD-C mutation at presentation (delV1677; 9% mutant) that was undetectable at relapse. However, the patient relapsed with a new high-level HD-N mutation (H1592Q;F1593T; Fig. 3C). To determine whether the HD-N mutation was present at diagnosis but below the level of detection of the WAVE analysis, mutant-specific quantitative PCR was done on the presentation sample. The HD-N mutation could not be detected at a sensitivity of 1 x 10–4. Furthermore, the same TCR V
1 clone was seen at both presentation and relapse. Another patient (#14) had an 11-bp insertion in the PEST domain at presentation, which was not detected at first or second relapse (22 and 29 months, respectively). However, a 39-bp insertion in the HD-C domain was present at a low level (6%) at diagnosis and progressively increased at each relapse (21% and 33%, respectively; Fig. 3D-F). Of note, at each time point, the patient had the same TCR V
11 clone; no new clone was detectable. In these three cases, the results indicate that the Notch-1 mutations occurred after the TCR rearrangement.
Case 16 presented at 5 months of age and relapsed 20 months later. Two high-level mutations (H1545P and S2514F), which were detected at presentation, were lost at relapse, but a new high-level (34% mutant) 2-bp deletion in the PEST domain was acquired. TCR analysis at the
locus showed loss of the original clone and acquisition of a new unrelated clone (V
4 to V
2 clone, respectively), with different CDR3 sequence. Therefore, in this patient, the results probably reflect development of a secondary leukemia, similar to cases described previously (25).
| Discussion |
|---|
|
|
|---|
In murine models of T-ALL, Notch activation can both directly induce leukemia and collaborate with other initiating genetic events to perpetuate leukemic growth (18–22). Mice transgenic for SCL/LMO1 usually acquire a Notch-1 mutation by the time of overt transformation to T-ALL (22). Oligoclonality predates acquisition of the mutation, but it is not until after the latter event that these thymocytes can form transplantable tumors. Whether Notch-1 mutations are an initiating or secondary collaborating event in human T-ALL is unclear. Our results suggest that both events may occur.
Of the 59 mutations detected in the total cohort of 62 patients, 49 were present at a high level, which was consistent with the presence of a heterozygous mutation in the majority of cells. Furthermore, in five of nine mutant-positive patients in our paired presentation-relapse cohort, the same mutation was present at relapse. In these patients, it is likely that the mutation had occurred early in disease pathogenesis. Two of these patients had mutations in both the HD and PEST domains at presentation and relapse, suggesting they are in the same cell. In vitro studies have shown that double mutations of this nature cause synergistic Notch activation when in cis (9). It is also noteworthy that two patients (#14 and #9) had evidence of increasing mutant level and one (#15) acquired a new high-level mutation at relapse, suggesting that Notch activation may offer a selection advantage to these cells. The two mutations identified in patient 16 deserve particular mention. The H1545P mutation in exon 26 lies upstream of the HD-N in the third Lin12/Notch repeat (LNR-C); this domain has been shown to play an important role in preventing Notch activation in invertebrates and in vitro (29, 30). This mutation was undetectable in the remission sample and thus a true disease-associated mutation rather than constitutional polymorphism and is the first mutation to be described in this domain in T-ALL. Additionally, the S2514F mutation in the PEST domain of this patient involves a highly conserved serine residue, mutation of which prevents cyclin-dependent kinase 8–mediated Notch-1 hyperphosphorylation and leads to stabilization of ICN (31). Whereas most PEST mutations introduce frameshifts or stop codons upstream of S2514, point mutations of this residue have not been reported and provide a potentially novel mechanism of Notch activation.
Results in some of our patients suggested that the Notch-1 mutations occurred not as a primary event but later in the pathogenesis of the disease. Seven were present at a particularly low level. This was not due to large numbers of normal hematopoietic cells, as review of the morphology showed that all cases had blast counts of
74%. A low mutant level could occur if the majority of leukemic cells either lost genomic material, including a mutated Notch-1 allele, or gained copies of the WT allele. However, genomic loss of 9q34 has not been described by either genome-wide analysis or fluorescent in situ hybridization in two substantial cohorts of T-ALL patients (32–34). Furthermore, the finding that T-ALL tumors frequently acquire a second synergistically activating Notch-1 mutation suggests that leukemic cells favor increased, rather than decreased, Notch signal strength, making genomic loss of mutated Notch alleles unlikely. Although homozygosity of the Notch-1 mutated allele has been found in some T-ALL lines as a result of loss of heterozygosity (22), the fact that no patient in our cohort had a mutant level of >50% supports the finding of others that this is not a common phenomenon in human T-ALL (32–34). Notch-1 gene duplication has been described in T-ALL, occurring in a small population of cells as a subclonal phenomenon (35), supporting the hypothesis that Notch activation can be a late secondary event. However, the low mutant levels observed in our cases could only be explained by duplication of the WT gene in the majority of cells with the mutant-carrying clone in the minority. Our results therefore suggest that the low-level mutants described here are due to the gain of a mutation in a subclone, which, in the presence of high blast counts, must be a secondary event. These results were supported by the demonstration that three mutations were lost at relapse, whereas the TCR rearrangements remained unchanged. This suggests the Notch-1 mutation was acquired after the TCR rearrangement in these three patients, a situation analogous to the secondarily acquired Notch-1 mutations in murine models of T-ALL.
The acquisition of Notch-1 mutations at different stages of disease may influence the role of Notch in disease pathogenesis. Early development of a mutation may lead to expansion of the T-cell population before acquisition of other pathogenic events, whereas with later development its effect on activation of other oncogenic pathways may be more important (15–17). Irrespective of the underlying mechanisms, however, results of the current study have implications for Notch as a target both for minimal residual disease analysis and for therapy. The high incidence of Notch-1 mutations in both pediatric and adult patients makes it an attractive target for minimal residual disease analysis, although its application would also be challenging due to the complexity of some of the mutations. Nevertheless, the lack of mutation stability at relapse and the evident acquisition after the TCR rearrangement observed in a proportion of our patients means that caution should be exercised in using them as solitary markers for minimal residual disease analysis. In addition, there may be implications for Notch pathway inhibitors that are currently in clinical trials, which, even if effective against constitutively activated mutant-positive cells, may select out remaining WT leukemic cells.
| Footnotes |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received 6/15/07; revised 8/ 7/07; accepted 8/15/07.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
W. R. Gordon, M. Roy, D. Vardar-Ulu, M. Garfinkel, M. R. Mansour, J. C. Aster, and S. C. Blacklow Structure of the Notch1-negative regulatory region: implications for normal activation and pathogenic signaling in T-ALL Blood, April 30, 2009; 113(18): 4381 - 4390. [Abstract] [Full Text] [PDF] |
||||
![]() |
A.-C. Tien, A. Rajan, and H. J. Bellen A Notch updated J. Cell Biol., March 9, 2009; 184(5): 621 - 629. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. S. Levy, J. A. Kahana, and R. Kumar AKT inhibitor, GSK690693, induces growth inhibition and apoptosis in acute lymphoblastic leukemia cell lines Blood, February 19, 2009; 113(8): 1723 - 1729. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |