DNA methyltransferase inhibition overcomes diphthamide pathway deficiencies underlying CD123-targeted treatment resistance
Katsuhiro Togami, … , Cory M. Johannessen, Andrew A. Lane
Katsuhiro Togami, … , Cory M. Johannessen, Andrew A. Lane
Published November 1, 2019; First published August 22, 2019
Citation Information: J Clin Invest. 2019;129(11):5005-5019. https://doi.org/10.1172/JCI128571.
View: Text | PDF
Research Article Hematology Oncology

DNA methyltransferase inhibition overcomes diphthamide pathway deficiencies underlying CD123-targeted treatment resistance

Abstract

The interleukin-3 receptor α subunit, CD123, is expressed in many hematologic malignancies including acute myeloid leukemia (AML) and blastic plasmacytoid dendritic cell neoplasm (BPDCN). Tagraxofusp (SL-401) is a CD123-targeted therapy consisting of interleukin-3 fused to a truncated diphtheria toxin payload. Factors influencing response to tagraxofusp other than CD123 expression are largely unknown. We interrogated tagraxofusp resistance in patients and experimental models and found that it was not associated with CD123 loss. Rather, resistant AML and BPDCN cells frequently acquired deficiencies in the diphthamide synthesis pathway, impairing tagraxofusp’s ability to ADP-ribosylate cellular targets. Expression of DPH1, encoding a diphthamide pathway enzyme, was reduced by DNA CpG methylation in resistant cells. Treatment with the DNA methyltransferase inhibitor azacitidine restored DPH1 expression and tagraxofusp sensitivity. We also developed a drug-dependent ADP-ribosylation assay in primary cells that correlated with tagraxofusp activity and may represent an additional novel biomarker. As predicted by these results and our observation that resistance also increased mitochondrial apoptotic priming, we found that the combination of tagraxofusp and azacitidine was effective in patient-derived xenografts treated in vivo. These data have important implications for clinical use of tagraxofusp and led to a phase 1 study combining tagraxofusp and azacitidine in myeloid malignancies.

Authors

Katsuhiro Togami, Timothy Pastika, Jason Stephansky, Mahmoud Ghandi, Amanda L. Christie, Kristen L. Jones, Carl A. Johnson, Ross W. Lindsay, Christopher L. Brooks, Anthony Letai, Jeffrey W. Craig, Olga Pozdnyakova, David M. Weinstock, Joan Montero, Jon C. Aster, Cory M. Johannessen, Andrew A. Lane

×

Figure 6

Impaired ADP-ribosylation activity in resistant cells and effectiveness of tagraxofusp plus azacitidine in vivo.

Options: View larger image (or click on image) Download as PowerPoint
Impaired ADP-ribosylation activity in resistant cells and effectiveness ...
(A) Log2(fold change) (log2FC) in RNA expression for diphthamide genes in individual PDXs (left) and patients (right) after exposure to tagraxofusp in vivo compared to vehicle (PDXs) or baseline (patients). Dotted line represents no change (log2FC = 0) and gray area is less than 2-fold change in expression (log2FC –1 to 1). *Genes below 0.1 fragments per kilobase of transcript per million mapped reads (FPKM) after tagraxofusp, in which case an FPKM of 0.1 was used to calculate fold change. Genes that had FPKM less than 0.1 before treatment are not shown. (B) Flow cytometry–based in vitro enzymatic labeling assay for tagraxofusp-induced ADP-ribosylation activity in parental and tagraxofusp-resistant CAL1 cells, showing decreased labeling with biotin-NAD+ in the setting of tagraxofusp resistance. Staining detected with streptavidin-BV510 fluorescence. (C) Flow cytometry–based enzymatic labeling assay with increasing doses of tagraxofusp measuring ADP-ribosylation activity as in panel B for BPDCN PDXs harvested from spleen or bone marrow at the time of disease progression after in vivo treatment with vehicle (n = 8) or tagraxofusp (n = 10). Curves compared by regression modeling of least-squares fit. (D) Flow cytometry–based ADP-ribosylation activity over increasing concentrations of tagraxofusp added in vitro measured in single CD45+CD123+ cells collected from the bone marrow of 5 patients before and after in vivo treatment with tagraxofusp (curves compared as in panel D). (E) Disease burden measured by peripheral blood human CD45+CD123+ flow cytometry in animals engrafted with BPDCN PDX cells after 2 cycles of treatment (at days 0 and 30) with vehicle, azacitidine, tagraxofusp, or the combination of tagraxofusp and azacitidine (Tag + Aza). (F) Kaplan-Meier overall survival curves from the time of treatment start for animals after treatments described in panel E (n = 10 mice per arm for panels E and F). Curves compared pairwise by log-rank test.