Activating Notch1 mutations in mouse models of T-ALL

TAL1 is a basic helix-loop-helix (bHLH) transcription factor that is normally expressed in hematopoietic cells, endothelial cells, and cells of the central nervous system. Through chromosomal translocation, interstitial deletion, or biallelic activation, TAL1 is misexpressed in thymocytes of 60% and 45% of pediatric and adult patients with T-cell acute lymphocytic leukemia (T-ALL), respectively.^1,2  Improved results in the treatment of pediatric T-ALL have been achieved in recent years because of intensified chemotherapy regimens, leading to a 5-year event-free survival rate approaching 80%^3,4 ; however, patients whose lymphoblasts overexpress the TAL1 oncogene have less favorable prognoses than do patients with activation of other oncogenes.^5,6 

Double-strand breaks occur in mammalian cells as a result of exposure to DNA-damaging agents such as ionizing radiation (IR) or during V(D)J recombination in lymphocytes. Many T-ALL tumors harbor chromosomal translocations or rearrangements that activate oncogenes or create oncogenic fusion genes.⁷  These translocations and rearrangements likely occur as a result of errors in the repair of double-strand breaks. The histone H2A variant, H2AX, plays a role in the cellular response to IR-induced double-strand breaks.^8,9  H2AX deficiency alone causes only a modest predisposition to cancer; however, mice deficient for both H2AX and p53 rapidly develop T- and B-cell lymphomas and solid tumors, demonstrating that H2AX acts as a tumor suppressor in mice.^10,11  The fact that human H2AX (H2AFX) maps to 11q23, a region that is frequently altered in human cancer, suggests that the human gene may also function as a tumor suppressor.¹⁰ 

The NOTCH genes encode single-pass transmembrane receptors that regulate apoptosis, proliferation, and cell fate determination in multicellular organisms. Binding of NOTCH ligands initiates a series of proteolytic cleavages in NOTCH1. The last of these cleavages, which is catalyzed by γ-secretase, results in the release of the intracellular domain of NOTCH1 (ICN), permitting it to translocate to the nucleus and form part of a multiprotein complex that regulates gene transcription.¹²  Recent work from our laboratories has revealed that activating mutations in NOTCH1 occur in more than 50% of human T-ALL.¹³  Previous studies have demonstrated that the Notch1 gene is a frequent site of retroviral insertional mutagenesis in mouse models of T-ALL^14-17  (see also http://TCGD.ncifcrf.gov). To determine whether Notch mutations are acquired in mouse models of T-ALL, we sequenced the heterodimerization domain and the PEST domain of all 4 Notch genes in tumors from our previously established models of T-ALL.

FVB/N Lck-Tal1 transgenic mice and Tal1/⁺HEB^+/- mice have been previously described.^18,19 Ink4a/Arf^+/- mice were obtained from the MMHCC mouse repository²⁰  and mated to Tal1 transgenic mice to obtain Tal1/⁺Ink4a/Arf^+/- mice.²⁴  129Sv/ev Tp53^-/-, H2AX^-/-, H2AX^-/-Tp53^-/-, and H2AX⁺/^-Tp53^-/- mice were described previously.¹⁰  RAG2-deficient mice²¹  were mated to the above mice to obtain H2AX^-/-Tp53⁺/^-RAG^-/- and H2AX⁺/^-Tp53^-/-RAG^-/- mice.

Exons 26 and 27 of Notch1 were amplified using the following primers: exon 26 forward, 5′-ACGGGAGGACCTAACCAAAC-3′; exon 26 reverse, 5′-CAGCTTGGTCTCCAACACCT-3′; exon 27 forward, 5′-CGCTGAGTGCTAAACACTGG-3′; and exon 27 reverse, 5′-GTTTTGCCTGCATGTACGTC-3′. Exon 34 was amplified in 2 fragments using the following primers: forward 1, 5′-GCTCCCTCATGTACCTCCTG-3′; reverse 1, 5′-TAGTGGCCCCATCATGCTAT-3′; forward 2, 5′-ATAGCATGATGGGGCCACTA-3′; reverse 2, 5′-CTTCACCCTGACCAGGAAAA-3′. The products were direct sequenced at Agencourt Bioscience Corporation (Beverly, MA), and the results were analyzed using Mutation Surveyor (State College, PA).

Murine Tal1 tumor cell lines either were treated with 1 μM DAPT (N-[N-(3,5-difluorophenacetyl-L-alanyl)]-(S)-phenylglycine t-butyl ester) (catalog number 565770; Calbiochem, San Diego, CA) or were mock treated with DMSO for 6 days. The cells were fixed with 70% ethanol, stained with propidium iodide, and analyzed by flow cytometry.

Tal1 tumor cell lines were untreated or treated with 1 μM DAPT. Cell lysates were fractionated through sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred to PVDF. Intracellular Notch1 was detected by probing the blot with the V1744 antibody (Cell Signaling Technology, Beverly, MA). The blot was then stripped and reprobed with an anti-β-actin antibody (Sigma, St Louis, MO) to ensure equal loading.

RNA was extracted from Tal1 tumor cells that were untreated or treated with DAPT using TRIzol (Invitrogen, Carlsbad, CA). cDNA was prepared with Superscript II reverse transcriptase (Invitrogen). Reverse transcription-polymerase chain reaction (RT-PCR) was then performed using primers specific for Deltex, Hes1, and GAPDH.²² 

We sequenced the heterodimerization domain and the PEST domain of all 4 Notch genes in tumors from Tal1/⁺, Tal1/⁺HEB⁺/-, and Tal1/⁺Ink4a/Arf⁺/- mice. Of the 27 tumors we analyzed, we found activating mutations in the Notch1 gene in 20 (74%) samples (Table 1). One tumor had a point mutation in the heterodimerization domain leading to a leucine-to-proline change at residue 1668 (1679 in human NOTCH1). Point mutations causing identical leucine-to-proline substitutions have been observed in 4 patients with T-ALL.¹³  Most mutations that we detected in the mouse tumors were in the PEST domain. Here, as in the human cell lines and samples, we found insertions, deletions, and point mutations, resulting in premature stop codons and loss of the notch1 PEST domain. In contrast, we found no mutations in Notch2, Notch3, or Notch4.

In addition to the Notch1 mutations found in tal1 transgenic mice, we also found Notch1 mutations in 9 of 29 (31%) of T-cell tumors that developed in H2AX^-/-, Tp53^-/-, H2AX^-/-Tp53^-/-, H2AX⁺/-Tp53^-/-, Tp53^-/-RAG^-/-, H2AX^-/-Tp53^-/-RAG^-/-, and H2AX^-/-p53⁺/-RAG^-/- mice. One tumor (455) had an alanine-to-proline missense mutation in the heterodimerization domain of notch1. This same mutation at the homologous residue in human NOTCH1 (1702) was also observed in one primary sample from a patient with T-ALL.¹³  However, as in the Tal1 transgenic mice, most of the mutations were in the PEST domain. These data indicate that Notch1 mutations are not specific to leukemias arising in Tal1 transgenic mice but that they arise in diverse T-ALL-prone backgrounds. Of note, Notch1 mutations are significantly more common in Tal1 transgenic mice than in mice that are heterozygous or deficient for Tp53 (P = .001), H2AX (P = .006), or RAG (P = .001) using a 2-tailed Fisher exact test. Because of the complex genotypes of the mice analyzed in this study, further experiments will be necessary to determine the individual contributions of Tp53, H2AX, or RAG deficiency to susceptibility to Notch1 mutations.

The mutations we have found affecting full-length notch proteins in murine T-ALL are predicted to activate notch pathway signaling in way that is dependent on cellular γ-secretase activity. Therefore, to determine whether the tumor cells depend on notch signaling, we treated tumor cell lines derived from these mice with the γ-secretase inhibitor DAPT (Figure 1, Table S1; see the Supplemental Table link at the top of the online article, at the Blood website). After treatment of the cell lines with the inhibitor for 6 days, we found that most cell lines exhibited G₀/G₁ arrest, an increase in apoptosis, or both, as indicated by cells with 45% to 70% sub-G₀/G₁ DNA content (similar results were also seen after 3 days). In addition, we demonstrate that in sensitive and resistant cell lines, DAPT treatment inhibits the production of activated notch (Figure 1E) and the transcription of the notch target genes Deltex and Hes1 (Figure 1F). Some Tal1 tumor cell lines were resistant to the γ-secretase inhibitor treatment. Two of 4 γ-secretase-resistant cell lines did not have mutations in Notch1. Resistant cell lines with mutations in Notch1 have likely incurred additional mutations, rendering them independent of notch pathway signaling for growth and survival because DAPT treatment does decrease notch1 signaling in these cells. In fact, our previous studies have demonstrated that tumor 5146 displays constitutive NFκB activation; therefore, it may be dependent on NFκB signaling rather than on notch signaling for its growth and survival.²³  One cell line was sensitive to the γ-secretase inhibitor but did not have a mutation in Notch1. In this case, we hypothesize that there might have been activating mutations in one or more other components of the notch signaling pathway. However, we cannot rule out the possibility that this cell line has a mutation in another substrate of γ-secretase. This work provides further evidence that Notch1 activation plays a key role in the pathogenesis of T-ALL in humans and in murine models and provides model systems incorporating clinically relevant oncogenes and tumor suppressors for testing therapeutics that target the NOTCH signaling pathway.