Notch3/Jagged1 circuitry reinforces notch signaling and sustains T-ALL

Abstract

Deregulated Notch signaling has been extensively linked to T-cell acute lymphoblastic leukemia (T-ALL). Here, we show a direct relationship between Notch3 receptor and Jagged1 ligand in human cell lines and in a mouse model of T-ALL. We provide evidence that Notch-specific ligand Jagged1 is a new Notch3 signaling target gene. This essential event justifies an aberrant Notch3/Jagged1 cis-expression inside the same cell. Moreover, we demonstrate in Notch3-IC-overexpressing T lymphoma cells that Jagged1 undergoes a raft-associated constitutive processing. The proteolytic cleavage allows the Jagged1 intracellular domain to empower Notch signaling activity and to increase the transcriptional activation of Jagged1 itself (autocrine effect). On the other hand, the release of the soluble Jagged1 extracellular domain has a positive impact on activating Notch signaling in adjacent cells (paracrine effect), finally giving rise to a Notch3/Jagged1 auto-sustaining loop that supports the survival, proliferation, and invasion of lymphoma cells and contributes to the development and progression of Notch-dependent T-ALL. These observations are also supported by a study conducted on a cohort of patients in which Jagged1 expression is associated to adverse prognosis.

Figure 1

Cis-expression and interaction of Jagged1 and Notch3 in N3-232T lymphoma cells. (A). Jagged1, Notch3, and Notch1 expression was determined in WCEs by immunoblot analysis using specific antibodies raised against the C-terminal region of the analyzed proteins and able to detect their full-length and/or the processed forms. To reveal the activated form of Notch1, an anti-Notch1 Val1744 antibody was used. Anti–β-actin was used as a loading control (B) Jagged1 (green) and Notch3 (red) cis-expression (merge) was analyzed by double immunofluorescence staining. Magnification: × 40. (C) sJag1-ECD sheds in N3-232T CCM revealed by immunoblot analysis. (D–F) sJag1-ECD triggers Notch signaling on neighboring cells. (D) SCB29 cells were cultured with N3-232T CCM for the indicated time, in the presence of vehicle alone (DMSO) or GSI-I. The expression of cleaved Notch3-IC (N3-IC) and activated Notch1 (N1-Val¹⁷⁴⁴) was analyzed by Western blot. (E and F) qRT-PCR analysis of pTα and Jag1 target genes in the presence of DMSO or GSI-I in SCB29 cells grown in N3-232T CCM. The data are represented as fold of activation with respect to starting time. The graph bars display the means ± SD of three independent experiments, made in triplicate. ***P < .001. (G) Expression levels of pTα mRNA evaluated by qRT-PCR in SCB29 cultured in N3-232T CCM in the presence of control IgG or JAG1 neutralizing antibody (anti-Jag1). The level of pTα mRNA from SCB29 cells cultured in CCM plus IgG was set as 100%. Mean values ± SD are shown and they were obtained from three independent experiments, each in triplicate. *P < .05. (H) N3-232T WCEs were subjected to immunoprecipitation with anti-Jagged1 antibody and revealed in Western blot (WB) with anti-Notch3. Heavy chain IgGs (IgG H) are used as a loading control. (I) Raft (R) and non-raft (NR) fractions from N3-232T cells were analyzed in immunoblot assays to detect the localization of Jag1-FL, Jag1-TMICD, Jag1-ICD, pro-ADAM17, and mature ADAM17. (J) R and NR fractions derived from MβCD-treated and untreated (DMSO) N3-232T cells were used for immunoblot assay with anti-Jagged1 antibody. Anti-616 p56^lck and anti–α-tubulin were used as fraction markers; anti–β-actin was used as a loading control. All data are representative of at least three independent experiments, each in triplicate.

Figure 2

Jag1-ICD interacts with and empowers the Notch3-driven transcriptional complex. (A) Extracts from non-nuclear (NN) and nuclear (N) fractions were probed with anti-Jag1 antibody. Anti–lamin B and anti–α-tubulin were used as fraction controls. (B) WCEs from HEK-293T cells co-transfected with different combinations of Notch3-IC-HA, RBP-Jκ, MAML1-Flag, and Jag1-TMICD-Myc expression vectors were immunoprecipitated with anti–c-Myc and immunoblotted with antibodies as labeled. (C and D) Luciferase assay on preT 2017 cells co-transfected with pTα-luc and several combinations of plasmids encoding for MAML1, RBP-Jκ, Jag1-TMICD, and Notch3-IC or Notch1-IC. All luciferase data are graphed as fold of activation with respect to pcDNA3 negative control. The bars represent the mean of three independent experiments ± SD, all made in triplicate; ***P < .001.

Figure 3

Jagged1 is a Notch3-signaling target gene. (A) Schematic representation of murine Jagged1 putative promoter sequence. The genomic DNA region of Jagged1 promoter ranging between − 1351 and − 237 bp upstream the start site (pJ1pro^{− 1351/− 237}) was cloned into pGL3 Basic luciferase vectors as described in the Materials and Methods section. (B) Activation of the Jagged1 promoter constructs by Notch3 signaling. HEK-293T cells were co-transfected with a luciferase reporter construct containing pJ1pro^{− 1351/− 237} and Notch3 transcriptional activator complex (Notch3-IC + RBP-Jκ + MAML1). (C) ChIP assay in N3-232T cells using anti–RBP-Jκ or anti-Notch3 antibodies or IgG as control. Immunoprecipitated DNA was analyzed by PCR using a primer set that amplifies the CSL/RBP-Jκ consensus binding site spanning between − 997 and − 991 bp on pJ1pro^{− 1351/− 237} promoter. (D) preT cells were transfected with Notch3-IC expression vector and Jag1 mRNA was analysed by RT-PCR. Anti–β-actin was used as a loading control. (E) preT 2017 cells were transfected with pJ1pro^{− 1351/− 237}-luc and Notch3-IC, MAML1, and RBP-Jκ, with or without Jag1-TMICD expression vectors. All data presented were collected from three independent experiments, each made in triplicate. The luciferase graph bars show the means ± SD and analyzed as fold of activation with respect to pcDNA3 empty control. ***P < .001; *P < .05.

Figure 4

Relationship between Notch3 and Jag1 in human T-ALL cell lines. (A) WCEs from KE37, Molt3, Dnd4.1, Cem, Jurkat, P12-Ichikawa, SIL-ALL, and Kopkt cell lines were analyzed by immunoblot analysis with anti-Notch3, anti-Notch1-Val¹⁷⁴⁴, and anti-Jagged1 antibodies, as labeled. Anti–β-actin was used as a loading control. (B) mRNA derived from the same human T-ALL cell lines was analyzed by qRT-PCR using a primer set specific for Jagged1. The data are analyzed as fold of activation with respect to KE37 negative control. The graph bars display the means ± SD of three similar experiments, each in triplicate. (C) Immunoblot WCE from Molt3, P12-Ichikawa, SIL-ALL, and Kopkt cells after silencing of Notch3 (Notch3 siRNA) compared to the control (CTR siRNA). Anti–β-actin was used as a loading control. (D) qRT-PCR was used to determine the effects of Notch3 knockdown on Jagged1 target gene expression in Notch3-silenced cells after 48 hours. The graph bars display the means ± SD of at least three independent experiments, analyzed as fold of activation with respect to scramble control. **P < .01.

Figure 5

Inhibition of Jag1 processing impairs mouse and human lymphoma cell proliferation, apoptosis, and invasiveness. (A) WCEs of N3-232T cells treated with TAPI-2 or vehicle alone (EtOH) were immunoblotted with anti-Jag1 antibody. Anti–β-actin was used as a loading control. (B and C) The TAPI-2 effect on growth was tested by the MTT assay, whereas apoptosis was measured by the analysis of Annexin V staining (Annexin V +, gray bars; Annexin V −, dark bars). (D) Molt3, Kopkt, and Jurkat cell lines were treated with TAPI-2 or EtOH and the effect on cell growth was tested by the MTT assay. (E) The Jurkat cell line treated with TAPI-2 or EtOH was used in invasion Matrigel assay. All data presented were collected from three independent experiments. The graph bars display the means ± SD, analyzed as fold of activation with respect to EtOH negative control. ***P < .001; **P < .01; *P < .05. (F) The graph represents gene expression profiling (GEP) data for the expression of probe set 216268_s_at (U133 2.0 plus array; Affymetrix) representing the Jagged1 gene in a cohort of 76 pediatric T-ALL patients. Patients were stratified in three final risk groups based on Minimal Residual Disease (MRD) levels, response to the first week of steroids, and resistance to induction therapy. Each dot corresponds to one patient and the expression value of Jagged1 is given in log₂ scale after normalizing GEP data with justRMA algorithm normalization. JustRMA is an algorithm fulfilling two steps, namely, background adjustment of all the probe sets present on the GeneChip and quantile normalization to make the values of all the GeneChips comparable. The X-axis represents patient stratification according to final risk groups, and the Y-axis represents Jagged1 gene expression level.

Figure 6

Schematic representation of the reciprocal auto-sustaining loop between Jagged1 ligand and Notch3 receptor in T-ALL. (A) The Jagged-1–triggered transactivation induces cleavage of Notch, by sequential activation of ADAM10/TACE and γ-secretase, to generate Notch-IC. In a T ALL cell context, a Notch3-IC–driven transcriptional complex activates the Jagged1 target gene, by constitutively binding the canonical RBP-Jκ binding site spanning between − 997 and − 991 bp in the Jagged1 promoter. (B) The constitutive expression of Notch3 intracellular domain in thymocytes and lymphoma cells from N3-ICtg mice determines a Notch3/Jagged1 cis-expression within the same cell and a constitutive raft-dependent Jagged1 processing, whose result is the empowering of Notch3 signaling. The Jagged1 intracellular domain is able to translocate into the nucleus where it is recruited specifically by the Notch3 transcriptional complex (Notch3-IC + MAML1 + RBP-Jκ), acting as a co-activator able to sustain the activation of specific target genes, such as pTα and Jagged1 itself (autocrine effect). On the other hand, the shedding of soluble extracellular domain of Jagged1 (sJag1-ECD) has a positive impact on triggering Notch signaling pathway in neighboring cells (paracrine effect).