Identification of DLK1 As a Novel Therapeutic Target in Down Syndrome Myeloid Leukemia

The current standard-of-care for myeloid leukemia associated with Down syndrome (ML-DS) relies on cytotoxic chemotherapy. Although primary ML-DS has favorable prognosis, the 10-20% of patients who develop relapsed leukemia or have refractory disease show a 3-year event-free survival less than 21% (Raghuram et al., 2023). Novel treatment options are needed to improve outcome in these patients with relapsed or refractory disease. The genomic landscape of ML-DS has been characterized by multiple studies (Labuhn et al., 2019; Nikolaev et al., 2013; Sato et al., 2024; Yoshida et al., 2013). In addition to trisomy 21, ML-DS samples possess mutations in the major hematopoietic transcription factor GATA1 and co-operating mutations in three major classes of proteins – cohesin complex components, epigenetic modulators and signaling molecules. However, there are no studies on differential gene expression analysis in ML-DS samples compared to normal bone marrow cells.

We have generated ML-DS patient-derived xenograft (PDX) models using primary bone marrow samples from patients with ML-DS (Barwe et al., 2019). We conducted transcriptome analysis using these PDX models (n=4) supplemented with an ML-DS cell line (CMK) to identify differentially expressed genes in ML-DS compared to CD34+ cells isolated from normal bone marrow specimens (n=4). Filtering the differentially expressed genes to shortlist cell surface resident proteins, DLK1 was revealed as one of the top ten targets overexpressed in ML-DS. Analysis of a larger dataset also showed that DLK1 is overexpressed in ML-DS. In this dataset, the mean DLK1 TPM scaled counts in ML-DS bone marrow samples (257.20±73.81, n=77) were significantly greater than the normal bone marrow specimens (2.23±0.24, n=68, p<0.0001). DLK1 expression on the cell surface of ML-DS cell and PDX lines was confirmed by flow cytometry.

To determine the role of DLK1 in ML-DS, we generated CMK cells with CRISPR/Cas9 mediated DLK1 knockout. The percentage of EdU positive cells determined by flow cytometry was significantly reduced by 67% and 65% in two distinct knockout clones (P<0.01), indicating that DLK1 indcues ML-DS cell proliferation. When injected in NSG-SGM3 mice, the CMK cells with DLK1 knockout showed minimal engraftment with less than 0.1% human cells in bone marrow, while the mice injected with wild-type CMK cells showed 63% human cell population 5 weeks post cell injection (P<0.001, n=3 each). Consistent with low bone marrow load, the median survival of mice engrafted with DLK1 knockout cells was significantly longer than those with CMK wild-type cells (P<0.05, n=5 each). Thus, our results demonstrate that DLK1 knockout in ML-DS cells suppressed cell proliferation and delayed in vivo engraftment in the bone marrow.

We used DLK1-targeting antibody drug conjugate (ADCT-701, referred to as DLK1-ADC) with a DNA intercalating agent pyrrolbenzodiazepine. Isotype control antibody (Iso-ADC) with the same payload was used as control. While DLK1-ADC failed to induce cell death in DLK1 knockout cells, the IC50 for CMK wild-type cells was 0.027 nM. CMK cells with shRNA mediated DLK1 knockdown showed increased IC50 (0.172 nM). Iso-ADC did not have a significant effect on cell viability in any of these lines within the tested concentration range. These results highlight the target-dependent specificity and efficacy of DLK1-ADC in inducing ML-DS cell death. DLK1-ADC also suppressed the viability and colony forming ability of CPCT-0010, a PDX model generated from a patient with refractory ML-DS. In this PDX model in vivo, DLK-1ADC showed a dose-dependent improvement in median survival by 17, 22.5 and 24 days respectively when treated with 0.25, 0.5 or 1 mg/Kg DLK1-ADC i.v. in comparison with Iso-ADC at the same dose (P<0.05). DLK1-ADC dosed twice at 1 mg/Kg also prolonged survival in three distinct PDX models generated using primary cells from patients with refractory ML-DS. Finally, DLK1-ADC, but not Iso-ADC cured 2 of 3 mice injected with a refractory ML-DS PDX model while the third mouse showed an 84-day prolongation of survival compared to that of mice treated with Iso-ADC. Thus, taken together, we have identified DLK1 as a novel therapeutic target for refractory ML-DS and show that targeting DLK1 may be a novel treatment option for patients with refractory ML-DS.