UCSF

Introduction

The pentose phosphate pathway (PPP) is essential for NADPH generation and redox homeostasis in cancer, including glioblastomas. However, the precise contribution to redox and tumor proliferation of the second PPP enzyme 6-phosphogluconolactonase (PGLS), which converts 6-phospho-δ-gluconolactone to 6-phosphogluconate (6PG), remains unclear. Furthermore, non-invasive methods of assessing PGLS activity are lacking. The goal of this study was to examine the role of PGLS in glioblastomas and assess the utility of probing PGLS activity using hyperpolarized δ-[1-¹³C]gluconolactone for non-invasive imaging.

Methods

To interrogate the function of PGLS in redox, PGLS expression was silenced in U87, U251 and GS2 glioblastoma cells by RNA interference and levels of NADPH and reduced glutathione (GSH) measured. Clonogenicity assays were used to assess the effect of PGLS silencing on glioblastoma proliferation. Hyperpolarized δ-[1-¹³C]gluconolactone metabolism to 6PG was assessed in live cells treated with the chemotherapeutic agent temozolomide (TMZ) or with vehicle control. ¹³C 2D echo-planar spectroscopic imaging (EPSI) studies of hyperpolarized δ-[1-¹³C]gluconolactone metabolism were performed on rats bearing orthotopic glioblastoma tumors or tumor-free controls on a 3T spectrometer. Longitudinal 2D EPSI studies of hyperpolarized δ-[1-¹³C]gluconolactone metabolism and T2-weighted magnetic resonance imaging (MRI) were performed in rats bearing orthotopic U251 tumors following treatment with TMZ to examine the ability of hyperpolarized δ-[1-¹³C]gluconolactone to report on treatment response.

Results

PGLS knockdown downregulated NADPH and GSH, elevated oxidative stress and inhibited clonogenicity in all models. Conversely, PGLS expression and activity and steady-state NADPH and GSH were higher in tumor tissues from rats bearing orthotopic glioblastoma xenografts relative to contralateral brain and tumor-free brain. Importantly, [1-¹³C]6PG production from hyperpolarized δ-[1-¹³C]gluconolactone was observed in live glioblastoma cells and was significantly reduced by treatment with TMZ. Furthermore, hyperpolarized δ-[1-¹³C]gluconolactone metabolism to [1-¹³C]6PG could differentiate tumor from contralateral normal brain in vivo. Notably, TMZ significantly reduced 6PG production from hyperpolarized δ-[1-¹³C]gluconolactone at an early timepoint prior to volumetric alterations as assessed by anatomical imaging.

Conclusions

Collectively, we have, for the first time, identified a role for PGLS activity in glioblastoma proliferation and validated the utility of probing PGLS activity using hyperpolarized δ-[1-¹³C]gluconolactone for non-invasive in vivo imaging of glioblastomas and their response to therapy.