NOTCH1 Is Aberrantly Activated in Chronic Lymphocytic Leukemia Hematopoietic Stem Cells

To investigate chronic lymphocytic leukemia (CLL)-initiating cells, we assessed NOTCH1 mutation/expression in hematopoietic stem cells (HSCs). In NOTCH1-mutated CLL, we detected subclonal mutations in 57% CD34+/CD38− HSCs. NOTCH1 mutation was present in 66% CD34+/CD38+ progenitor cells displaying an increased mutational burden compared to HSCs. Flow cytometric analysis revealed significantly higher NOTCH1 activation in CD34+/CD38− and CD34+/CD38+ cells from CLL patients, regardless NOTCH1 mutation compared to healthy donors. Activated NOTCH1 resulted in overexpression of the NOTCH1 target c-MYC. We conclude that activated NOTCH1 is an early event in CLL that may contribute to aberrant HSCs in this disease.

To investigate chronic lymphocytic leukemia (CLL)-initiating cells, we assessed NOTCH1 mutation/expression in hematopoietic stem cells (HSCs). In NOTCH1-mutated CLL, we detected subclonal mutations in 57% CD34+/CD38− HSCs. NOTCH1 mutation was present in 66% CD34+/CD38+ progenitor cells displaying an increased mutational burden compared to HSCs. Flow cytometric analysis revealed significantly higher NOTCH1 activation in CD34+/CD38− and CD34+/CD38+ cells from CLL patients, regardless NOTCH1 mutation compared to healthy donors. Activated NOTCH1 resulted in overexpression of the NOTCH1 target c-MYC. We conclude that activated NOTCH1 is an early event in CLL that may contribute to aberrant HSCs in this disease.
Keywords: notch signaling, chronic lymphocytic leukemia, hematopoietic stem cells, nOTch1 mutation, cD34+ cells inTrODUcTiOn Chronic lymphocytic leukemia (CLL) is a mature B cell malignancy characterized by accumulation of clonal B cells in blood, bone marrow (BM), and lymphoid tissues. The search for CLL-initiating cells has never been successful. While several cell types have been suggested as giving rise to CLL, yet, there is no consensus as to its normal cellular counterpart (1). CLL cells have monoclonal immunoglobulin gene rearrangements, suggesting that lymphoid malignant stem cells originate after cells have committed to the lymphoid lineage. More recently, it has been reported that hematopoietic stem cells (HSCs) from CLL patients display the propensity to generate clonal B cells, suggesting the involvement of HSCs in lymphoid leukemogenesis (2).
In 2009, we first identified NOTCH1 mutations in CLL (3) and provided data on the adverse prognostic outcome associated with mutated NOTCH1 (4). More recently, independent studies confirmed the presence and the prognostic relevance of NOTCH1 mutations in CLL patients (5). All mutations resulted in NOTCH1 impaired degradation that led to NOTCH1 deregulated signaling, indicating that mutations could contribute to increase NOTCH activation in CLL (6,7). Recently, analyses of peripheral blood CD34+CD19− cells and BM hematopoietic progenitors revealed NOTCH1 Expression in CLL HSCs Frontiers in Oncology | www.frontiersin.org April 2018 | Volume 8 | Article 105 NOTCH1 mutation in some CLL samples (8,9). Conversely, others failed to demonstrate the presence of NOTCH1 mutations in circulating CD34+ cells from NOTCH1 mutated patients (10). In lymph node, CLL cells show NOTCH1 activation independent of mutation (11) and recent evidence have shown non-mutational NOTCH1 signaling with anti-apoptotic effects in peripheral blood CLL cells (12). However, the role of NOTCH1 signaling in the HSC compartment of CLL is still unknown. Flow cytometric analysis and cell-sorting were performed using the following antibodies: PE anti-CD34 and PC5 anti-CD38 (Beckman Coulter), PerCP-Cy5.5 anti-CD34, FITC anti-CD19, PE-Cy7 anti-CD38, APC-H7 anti-CD10, V450 anti-CD3, and V500 anti-CD45 (BD Biosciences). NOTCH1 ICN on sorted populations was performed using PE anti-NOTCH1 (mN1A) (eBioscience). Cells were analyzed using a FACSCanto II and sorted using a FACS Aria III cell sorter (BD Biosciences).

Direct sanger sequencing
NOTCH1 gene mutational analysis was performed by directional sequencing of PCR fragments from genomic DNA. Primers and PCR conditions were as previously described (13).

Western Blot analysis
Whole-cell lysates extracted from BM CD34+ cells (5 × 10 5 ) of CLL patients and HDs were analyzed by western blot using an anti-NOTCH1 antibody (clone bTAN20) able to detect the 300-kDa inactive precursor (FL), the 120-kDa transmembrane/ cytoplasmic/cytoplasmic (TM) subunit, and the active 100-kDa intracellular domain. Whole-cell lysates (3 µg) isolated from peripheral blood CD5+CD19+ CLL cells of NOTCH1 mutated patients were used as positive control.

statistical analysis
Statistical analyses were performed with GraphPad (GrapdhPad Software Inc., La Jolla, CA, USA). In the text, data are presented as mean ± SD and statistical differences between mean values were evaluated using the Student's t-test and Mann-Whitney test.

resUlTs hscs From cll Patients showed nOTch1 Mutation
The mean proportion of BM CD34+ cells before enrichment was 0.75 ± 0.44%. After the FACS sorting procedures, the mean purity of CD34+/CD38− cells was 94.58 ± 3.52% and CD34+/ CD38+ cells were 98.12 ± 1.34% ( Figure 1A). We initially analyzed the NOTCH1 mutational hotspot by Sanger sequencing. The CD34+/CD38− fraction did not contain NOTCH1-mutated cells. Then, a high sensitivity AS-PCR assay for the NOTCH1 mutation (13) indicated the presence of small HSCs mutated clones in 57% CLL samples. Densitometric analysis revealed a mean 6.4 ± 4.7% NOTCH1 mutant allelic burden. Moreover, we used a ddPCR assay to validate NOTCH1 mutational data in HSCs by a more quantitative method. We confirmed NOTCH1 mutations in two HSCs samples for which DNA was available that display an allelic ratio of 2.6 and 8.5%, respectively ( Figure 1B). Altogether, these data confirm that NOTCH1 mutation is an early event in CLL hematopoiesis in a fraction of patients. Additionally, we measured the NOTCH1 mutational burden along specific stages of HSC differentiation in NOTCH1-mutated CLL patients. The mean percentage of the mutant allele progressively increased from 6.4 ± 4.7% in CD34+CD38− to 14.9 ± 11.3% in CD34+CD38+CD10+CD19+ cells, 22.7 ± 6.5% in CD34−CD38+CD10+CD19+ cells and 40.5 ± 4.3% in neoplastic CD5+CD19+ cells ( Figure 1C). The analysis of the rearrangement status of the IgH gene revealed in both NOTCH1 mutated and NOTCH1 unmutated CD34+CD38− HSCs the presence of a germline configuration in the half of the samples while the other 50% showed a clonal VDJ. hscs From cll Patients have nOTch1 aberrantly activated also in Unmutated nOTch1 Patients Thus, we analyzed the NOTCH1 signaling status in HSCs and progenitor cells of NOTCH1-mutated and unmutated CLL samples. Physiologically, the active intracellular domain (ICN) of NOTCH1 accumulates in cells with activated NOTCH1 signaling as a result of a cleavage of the transmembrane (TM) subunit made by the y-secretase complex (14). Here, we used flow cytometry to quantitate the percentage of active NOTCH1-ICN in CD34+/ CD38− HSCs and CD34+CD38+ progenitors from the BM of CLL patients and HDs, used as control. As shown in Figure 1Di, (12). Thus, using quantitative reverse transcription-PCR, we found significantly higher mRNA expression levels of c-MYC in CD34+ cells from NOTCH1 mutated and WT CLL samples compared to HD (3.5 ± 0.7 and 2.6 ± 0.08 vs 1.3 ± 0.1) (Figure 1Diii). In addition, we showed higher Hes1 expression in CD34 + cells from CLL patients compared to HD (4.2 ± 1.1 vs 1.1 ± 0.2) (Figure 1Diii), in line with upregulated NOTCH1 pathway.
Next, we analyzed whether the higher levels of NOTCH1 activation in HSCs and progenitors CLL samples were accompanied with increased NOTCH1 expression. We analyzed by western blot the expression levels of NOTCH1-TM subunit in CD34+ cells from BM aspirates of four NOTCH1-mutated CLL patients and three HDs (6). The median purity of immunoselected CD34+ cells was 97.3% (range 74-99.5%; Figure 2A). Results revealed that CD34+ samples from CLL patients always expressed the NOTCH1-TM protein. Conversely, in CD34+ cells from HDs, NOTCH1-TM was either absent or expressed at lower levels than those observed in CLL samples (Figure 2Bi,ii). These data demonstrated that high levels of NOTCH1 signaling activation correlated with NOTCH1-TM overexpression in the CD34+ hematopoietic compartment of CLL.

DiscUssiOn
The Notch pathway is genetically altered in a large number of hematopoietic and solid tumors (15). We recently reported that activating mutations of NOTCH1 are recurrently associated with CLL and predict poor outcome (3,4). The identification of a somatic NOTCH1 mutation could help providing additional information on the cellular origin of CLL. Sanger sequencing analysis (13) failed to detect mutations in CD34+/CD38− HSCs fractions of CLL patients. However, the sensitivity of this method does not allow the identification of a mutation whose allelic representation is <10%. Thus, we used more sensitive PCR based methods in order to verify results obtained with direct sequencing.
The analysis of CD34+/CD38+ progenitors detected the NOTCH1 mutation in 57% of the samples. Damm et al. (8) described a similar frequency of NOTCH1 mutation in multipotent progenitors of CLL samples performing NGS analysis on rare CD34+CD19− peripheral blood cells. Recently, the same analysis was conducted in immunoselected CD34+ BM cells that resulted NOTCH1-mutated in 8 out of 13 CLL samples (9). Here, we used two high sensitive PCR assays specific for NOTCH1 mutation and employed FACS sorted CD34+CD38+ cells from BM aspirates. Additionally, we showed here that the NOTCH1 mutational burden increased along specific stages of HSC differentiation in NOTCH1-mutated CLL patients. This suggests that the NOTCH1 lesion is selected and expands during HSC differentiation toward a B neoplastic cell, thus strengthening the hypothesis that the genetic alteration is an initial event associated with the stepwise malignant transformation of CLL.
In normal human BM, CD34+CD38− populations do not rearrange the IgH gene (16), thus raising issues on the molecular purity of the sorted hematopoietic cell fractions presenting a VDJ rearrangement. Nevertheless, the high purity of HSC doublesorted samples together with the sensitivity of the IgH method (5%) limited the risk of detecting small fractions of contaminating neoplastic B cells. The alternative option is that IgH rearrangement is the expression of a B neoplastic transactivation at the HSC level.
Recently, common nonmutational NOTCH1 activation has been described in mature CLL cells (12) raising the question of whether the same condition is present in HSCs to determine their aberrant behavior. The Notch1 signaling has been extensively analyzed in the contest of embryonic hematopoiesis. Ex vivo approaches suggest that Notch signaling can expand HSCs, raising the question of whether this is a physiologic Notch function. Gerhardt et al. (17) correlated NOTCH1 with hematopoiesis in animal models and identified cell-autonomous functions for Notch1 signaling in fetal HSCs homeostasis. The present study indicated that the pool of CD34+ cells, including HSC and progenitor compartments, tend to have NOTCH1 aberrantly expressed and activated in CLL patients compared to HDs. NOTCH1 deregulation and overepression of c-Myc are independent of NOTCH1 mutational status. These data clearly show that expansion of the leukemic stem cell clone does not necessarily require a mutation to upregulate the NOTCH1 signaling, suggesting the presence of extrinsic factors from the BM HSC niche that are capable of stimulating and promoting CLL-initiating cell clone expansion. In CLL patients, BM mesenchimal cells express different ligands, which might play a role in NOTCH1 activation (18). However, additional studies are warranted to compare the levels and type of these ligands in the BM of healthy people vs CLL patients. Alternatively, CLL-HSCs may have cell-intrinsic mechanisms activating NOTCH1, which involve alterations of NOTCH1 pathway regulators (19,20) or aberrant regulation of NOTCH1 receptor recycling (21). This selective pressure might contribute to the onset of specific NOTCH1 mutations in a DNA context that is prone to spontaneous microdeletion (5) (Figure 2C). Our discovery of NOTCH1 deregulated signal and mutations in CLL-HSC have significant therapeutic implications in this disease. A variety of approaches was used to inhibit NOTCH1 for cancer therapy, including presenilin γ-secretasi inhibtors, trafficking modulators (22), and blocking antibodies. In the next future, it will be interesting to evaluate the effects of these anti-NOTCH1 drugs in the development of CD34+CD38− and CD34+CD38+ CLL populations.
In conclusion, our data confirmed the presence of NOTCH1 mutations in HSCs of CLL patients and showed for the first time a common nonmutational NOTCH1 activation occurring early in CLL hematopoiesis and represent a rationale for the use of therapies targeting the NOTCH1 signaling in CLL aimed to inhibit the survival of CLL-initiating cells.

eThics sTaTeMenT
This study was carried out in accordance with the recommendations of University/Hospital of Perugia guidelines, Ethics Committee of Perugia, with written informed consent from all subjects. All subjects gave written informed consent in accordance with the Declaration of Helsinki. The protocol was approved by the Ethics Committee of Perugia.