Abstract
In chronic lymphocytic leukemia (CLL), mutations in the NOTCH1 receptor occur in 4-10% of newly diagnosed patients and 15–20% of multiply relapsed patients. Using next-generation sequencing, our group previously reported NOTCH1 p.P2514fs mutations in 15 CLL patients (9.4%) in an initial cohort of 160 CLL patients in which NOTCH1 mutations were associated with IGHV unmutated (UM) CLL (p=0.0001). Further analysis using a three-group comparison (NOTCH1 mut, IGHV UM vs. NOTCH1 wild-type [wt] IGHV UM vs. NOTCH1 wt IGHV mut) showed that NOTCH1 mutations associated with both trisomy 12 (p=0.049) and 17p deletion (p=0.0008) and poor overall survival (HR 2.99, p=0.008).
Given that targeting activating mutations has proven an effective therapeutic strategy in many cancers, we explored the therapeutic potential of a Notch1 inhibitor, PF-03084014, in CLL. Previous studies in T-cell acute lymphoblastic leukemia cells harboring NOTCH1 mutations have shown that gamma secretase inhibitors can induce apoptosis by blocking Notch1 receptor activation. When we tested the gamma secretase inhibitor (GSI) PF-03084014 in 18 CLL samples with NOTCH1 mutations, it consistently induced apoptosis of all CLLs after 48 hours in culture across all cytogenetic groups tested (13.3-47.2% death with 5 μM GSI, p<0.0001)(Figure A). The induction of apoptosis was similar (GSI vs. ibrutinib, p=ns) to that of ibrutinib (n=11,11.9-74.4% death with 5 μM ibrutinib, p<0.0001). In contrast, GSI treatment only induced apoptosis in some (n=10), but not all NOTCH1 wt CLLs (n=6) (p=0.0137). We next tested the effect of GSI PF-03084014 in the context of a stromal environment. Co culture of CLL cells with CD40L-expressing fibroblasts partially mimics the lymph node and bone marrow microenvironments, which are known sites of drug-resistance and proliferation of CLL in vivo. We tested whether the interaction with stromal cells protects CLL cells from Notch1 inhibitor-induced apoptosis, as seen with other drugs including ibrutinib. We found that co culture with CD40L-expressing 3T3s decreased GSI-induced apoptosis in NOTCH1 mutant CLLs (p=0.0006) and in the majority of the NOTCH1 wt CLLs that responded to the GSI (Figure A).
Since NOTCH1 mutations have been reported to be an independent marker of aggressive disease in CLL, we tested whether CLL cells with NOTCH1 mutations were more proliferative in vitro compared to NOTCH1 wt CLL cells. We showed that CLL cells upregulate Ki67 expression in co culture with 3T3-CD40L cells and in a cohort of 10 NOTCH1 mutants and 11 NOTCH1 wt, we found the NOTCH1 mutants to be more proliferative than the NOTCH1 wt (median of 7.6% vs. 2.3%, p=0.015). To then address whether blocking of the Notch1 pathway decreases proliferation, we treated CLL cells in co culture with 3T3-CD40L cells with 5 μM GSI for 7 days. GSI treatment decreased the percentage of Ki67+ CLL cells in all but one NOTCH1 mutant (median decrease 28.3%, p=0.044) as wells as in the majority of NOTCH1 wt samples (median decrease 38.7%, p=0.037).
Having established that inhibition of Notch1 can reduce proliferation and induce apoptosis in CLL cells in vitro, we were interested in determining the downstream genes that may be the effectors of this activity. We therefore compared the gene expression profiles (GEP) of NOTCH1 mut vs. NOTCH1 wt CLLs, and found upregulation of genes involved in the Notch1 pathway, in apoptosis and in chemokine signaling in the NOTCH1 mutants. Furthermore, comparing GEP of high Ki67 vs. low Ki67 expressing CLL cells revealed higher expression of a range of both upstream and downstream Notch1 pathway genes in high Ki67 expressing CLL cells.
In conclusion, we show that PF-03084014 induces apoptosis and decreases proliferation in both NOTCH1 mutant and wt CLL cells. We find NOTCH1 mutant CLL cells to be more proliferative than NOTCH1 wt and show upregulation of Notch1 pathway genes in NOTCH1 mutants compared to wt CLL cells and in high Ki67 expressing compared to low Ki67 expressing CLL cells. Taken together, these results emphasize the important role of Notch1 signaling in CLL in general, perhaps particularly in proliferative compartments like lymph nodes, and demonstrate that Notch1 pathway inhibitors are worthy of therapeutic investigation in CLL.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.