Targeting ornithine decarboxylase reverses the LIN28/Let-7 axis and inhibits glycolytic metabolism in neuroblastoma

Abstract

LIN28 has emerged as an oncogenic driver in a number of cancers, including neuroblastoma (NB). Overexpression of LIN28 correlates with poor outcome in NB, therefore drugs that impact the LIN28/Let-7 pathway could be beneficial in treating NB patients. The LIN28/Let-7 pathway affects many cellular processes including the regulation of cancer stem cells and glycolytic metabolism. Polyamines, regulated by ornithine decarboxylase (ODC) modulate eIF-5A which is a direct regulator of the LIN28/Let-7 axis. We propose that therapy inhibiting ODC will restore balance to the LIN28/Let-7 axis, suppress glycolytic metabolism, and decrease MYCN protein expression in NB. Difluoromethylornithine (DFMO) is an inhibitor of ODC in clinical trials for children with NB. In vitro experiments using NB cell lines, BE(2)-C, SMS-KCNR, and CHLA90 show that DFMO treatment reduced LIN28B and MYCN protein levels and increased Let-7 miRNA and decreased neurosphere formation. Glycolytic metabolic activity decreased with DFMO treatment in vivo. Additionally, sensitivity to DFMO treatment correlated with LIN28B overexpression (BE(2)-C>SMS-KCNR>CHLA90). This is the first study to demonstrate that DFMO treatment restores balance to the LIN28/Let-7 axis and inhibits glycolytic metabolism and neurosphere formation in NB and that PET scans may be a meaningful imaging tool to evaluate the therapeutic effects of DFMO treatment.

Figure 1. A) Patient PET scans

Patient treated on Phase I study of DFMO tablets with 1000mg/m2 twice daily shows reduced tumor metabolic activity after one cycle of DFMO treatment, indicated by decreased ¹⁸F-FDG uptake in tumors. After starting DFMO treatment, the patient's tumor was still present, although showed negative PET avidity. Decreased PET activity is an indication of reduced glycolytic metabolism. B) Proposed pathway of DFMO effects on LIN28/Let-7 axis and glycolytic metabolism in NB. DFMO reversibly inhibits ODC, leading to reduced hypusination of eIF5A and a reversal of the LIN28/Let-7 axis. Changes in LIN28/Let-7 axis likely contribute to reduced glycolytic metabolic activity seen in NB cells treated with DFMO.

Figure 2. Sensitivity to DFMO treatment in three NB cell lines

DFMO IC₅₀ values were calculated after 72 hours of treatment; results are the average of 4 independent experiments. BE(2)-C cells were most sensitive to DFMO treatment with an IC₅₀ of 3.0 mM, SMS-KCNR cells exhibited moderate sensitivity with an IC₅₀ of 10.6 mM, and CHLA90 cells were the least sensitive to DFMO treatment with an IC₅₀ of 25.8 mM.

Figure 3. DFMO treatment reverses the LIN28B/Let-7 axis in NB

A) Western blot analysis of LIN28B and MYCN protein levels in cells treated with DFMO. Cells were treated with 5 mM or 10 mM DFMO for 48-96 hours. LIN28B and MYCN protein expression decreased with DFMO treatment in BE(2)-C and SMS-KCNR cells at all three timepoints. LIN28B protein expression decreased in CHLA90 cells with DFMO treatment, but MYCN protein expression did not change. B) qRT-PCR analysis of Let-7 miRNA expression in NB cells after 6 hours of treatment with DFMO. qRT-PCR data is presented as fold change compared to untreated. BE(2)-C and SMS-KCNR Let-7 miRNA expression was significantly increased in cells treated with DFMO. Let-7 miRNA expression significantly decreased in CHLA90 cells treated with DFMO (*p < 0.05, **p < 0.01, ***p < 0.001 relative to untreated. #p < 0.05 comparing doses).

Figure 4. Neurosphere Formation Assay

A) Both BE(2)-C and SMS KCNR cell lines show inhibition of neurosphere formation in the presence of 5 mM DFMO. B) Treatment with DFMO shows a dose dependent decrease in neurosphere formation in BE(2)-C cells (n=3 separate experiments).

Figure 5. ATP analysis for NB cells treated with 10 mM DFMO for 72 or 96 hours

All data are presented as percent of untreated control. A) Total ATP per sample treated with 10 mM DFMO. B) Total cell number per sample treated with 10 mM DFMO. C) Ratio of total ATP over total cell number (ATP/cell). BE(2)-C cells had the greatest decrease in ATP/cell with DFMO treatment with a 40.8% decrease in ATP/cell at 72 hours and a 60.5% decrease at 96 hours when compared to untreated. SMS-KCNR cells also had decreased ATP/cell after treatment with DFMO at 72 (15.5% decrease) and 96 hours (26.9% decrease). There were no significant change in ATP/cell levels in CHLA90 cells after DFMO treatment at either timepoint. *p < 0.05, ***p < 0.001 relative to untreated. #p < 0.05 comparing timepoints.

Figure 6. DFMO treatment in vivo

Mice began treatment with 2% DFMO in drinking water after tumors reached approximately 0.2 cm³. PET/CT scans were done at baseline, 19 days, and 32 days after treatment was initiated. A) Average NB tumor volume was quantified using CT scans. There was no significant difference in tumor volume between vehicle and DFMO treatment groups. B) Average tumor SUVmax was quantified using PET scans. Average SUVmax was lower in mice treated with DFMO at 19 days and 32 days of treatment when compared to vehicle. C) PET images of one vehicle treated mouse and one DFMO treated mouse at all three timepoints. The DFMO treated mouse showed less ¹⁸F-FDG uptake in the tumor at both 19 days and 32 day days when compared to the vehicle (*p < 0.05).