Slit-Robo Pathway Is Clinically Relevant and May Represent a Potential Target in Acute Promyelocytic Leukemia

Background: The Slit-Robo pathway has been shown to participate in the pathogenesis of several malignant diseases in addition to its physiologic role during the development of the central nervous system (CNS). However, the relevance of the Slit-Robo pathway in acute promyelocytic leukemia (APL) is presently unknown. We investigated the status of the Slit-Robo pathway in APL following the recent demonstration by Amodeo et al (CellRep. 2017) that the PML protein induces neural stem/progenitor cells migration by inhibiting the SLIT2 gene, which was associated with an increased presence of H3K27me3 in the SLIT2 promotor region. Moreover, sensitivity towards the PML-targeting drug arsenic trioxide in primary glioblastoma cells was shown to be regulated by the PML/SLIT axis. Aims: Here, we quantified SLIT2 transcript levels in bone marrow (BM) samples from APL patients and healthy subjects and determined whether they are associated with clinical and laboratory features at diagnosis and treatment outcomes. In addition, we evaluated the effect of increased SLIT2 protein on leukemic cell survival, proliferation and clonogenecity. Methods: Cell cycle distribution, proliferation index (Ki-67/proliferation curve), colony formation and rate of apoptotic cells (annexin V/PI) were evaluated in both APL cell lines NB4 (ATRA-sensitive), NB4R2 (ATRA- resistant) and four primary APL samples following the treatment with the human recombinant SLIT2 protein (50 ng/ml) for 24 to 72 hours. In addition, 88 patients (age, 18-73y) with newly diagnosed APL enrolled in the International Consortium on Acute Leukemia (ICAL) study - ICAPL2006 were included. For comparison, BM mononuclear cells from five healthy volunteers (age, 18-60y) were also included. SLIT2 transcripts was determined by qPCR and expressed as comparative Ct method. Patients were dichotomized into "low" and "high" SLIT2 expression groups using a cut off value of 1.53 based on survival receiver operating characteristic (ROC) curve and the C index analyses. The following parameters were used to evaluate treatment outcome: complete hematological remission (CHR); 5-year Disease-Free Survival (DFS) and 5-year Overall Survival (OS) rates. Results: At the end of the sixth day, in vitro treatment with SLIT2 significantly reduced the proliferation rate (P=.01) and clonogenicity (P=.01) in primary APL samples, while significantly increasing the apoptotic rate after 72 hours treatment with SLIT2 (P=.03). In accordance, SLIT2 treatment decreased cell proliferation and clonogenicity (all P<.01) and increased the drug-induced apoptosis in a time-dependent manner in NB4 and NB4R2 cells (P<.01). Cell cycle analysis revealed that the cell cycle was arrested in the S-phase (P=.001) with a reduction of the G₂/M phase (P=.001) in NB4 after 72 hours of treatment but no such impact was observed on NB4R2. In the clinical setting, patients with APL had a higher-than-normal expression of SLIT2 (6.7-fold higher than BMMC; P=.01). Baseline characteristics were similar between patients with low and high SLIT2 levels, except for a higher white blood cell count in patients with low SLIT2 expression (P=.024). Overall, 80/94 (85%) patients achieved complete hematological remission (CR). SLIT2 transcript levels had no impact on CR rate (P=.099). The estimated 5y DFS and the CIR rates were 87% (95% CI: 80-92%) and 12% (95% CI: 7-17%), respectively. Patients with low SLIT2 expression presented a lower 5y DFS rate (79%, 95% CI: 48-92%) than those with high SLIT2 expression (96%, 95% CI: 64-97%) (P=.026). In addition, patients with low SLIT2 expression exhibited an enhanced CIR rate compared to high SLIT2 expression (21% vs 11%) (P=.03). With a median follow up of 32 months (1-101 months), the estimated 5y OS rate was 79% (95% CI: 72-84%). Patients with lower SLIT2 expression had a lower 5y OS rate (68%, 95% CI: 45-83%) compared to those with higher SLIT2 expression (84%, 95% CI: 73-91%) (P=.008). This result was consistent with the multivariable proportional hazards analysis (HR: 1.02; 95% CI: 1.01-1.03; P=.001). Conclusion: Our results suggest that SLIT2 transcript levels may predict outcomes in APL patients treated with ATRA and anthracycline-based chemotherapy. Our data show that SLIT2 treatment represses APL cell growth, colony formation and induces apoptosis in vitro and we are currently assessing the role of the Slit-Robo pathway in the hCG-PML/RARA transgenic mouse model.