Category: Preclinical Development
Purpose: Hypoxia is a hallmark of tumor progression and drives an aggressive program that promotes acidity, immunosuppression, and resistance to standard therapies. Previously, anti-angiogenic therapies were unsuccessful in attempts to minimize breast tumor progression as this induced hypoxia, propagating resistance to radiotherapy and chemotherapy. A novel therapeutic aiming to rescue tumor oxygen and pH rather than starving of oxygen may relieve hypoxia-induced resistance mechanisms and enhance overall therapeutic response when administered as a pre-treatment.
Granulocyte-Macrophage Colony-Stimulating Factor (GM-CSF) is known primarily for its immune-stimulant properties and capability of inducing an anti-tumor response. However, we were the first to report an additional role for GM-CSF in stimulating dose-dependent macrophage expression of a soluble isoform of VEGFR-1 (sVEGFR-1). In our previous attempt to align with the anti-angiogenesis field, we discovered that intratumoral delivery of 100 ng GM-CSF abrogated angiogenesis via sVEGFR-1 and led to near anoxic tumor microenvironment (TME). Because anti-angiogenic therapies have proven unsuccessful and the culprit appears to be hypoxia and not angiogenesis, we hypothesize that low-dose (5 ng) GM-CSF may induce sufficient sVEGFR-1 to sequester excessive VEGF and re-establish vessel integrity to rescue blood perfusion, oxygen delivery, and pH, as well as alter the TME profile from immunosuppressive to immunosupportive.
The goal of our current ongoing study is for “proof-of-concept” that GM-CSF can regulate tumor oxygen by repairing vessel permeability and increasing perfusion to drive a more immunosupportive environment. In parallel, we have synthesized PLGA/PEG-PLGA nanoparticles loaded with rmGM-CSF with the goal of systemic but targeted tumor-specific nanoparticle delivery of low-dose GM-CSF by PLGA/PEG-PLGA to alleviate hypoxia.
Methods: PEG-PLGA [Poly(ethylene-glycol)-poly(lactic-co-glycolic acid))] and PLGA-COOR were synthesized using ring-opening polymerization in which DL-lactide and glycolide were mixed at a molar ratio of 3:1 with either monomethoxy PEG or benzyl alcohol, respectively, as an initiator. Lyophilized rmGM-CSF was precipitated at 4 mg/ml in 0.544 M NaCl on ice in glycofurol. PLGA-COOR, PLGA-PEG, and PLGA-COOH (Resomer ®) were dissolved in 12% w/v dimethyl isosorbide and mixed in equal volumes under stirring at 1300 RPM. Glycofurol, either alone or containing GM-CSF protein precipitate, was then added to the mixture. Phase separation was initiated under continual stirring by addition of aqueous phase buffer, pH=5.9. Nanoparticle suspensions were diluted and analyzed by Nanoparticle Tracking Analysis (NTA, NanoSight) for size and Electrophoretic Light Scattering (ELS, Zetasizer) for zeta potential. Nanoparticles were centrifuged at 10,000xg for 30 minutes and supernatant was collected and quantified by ELISA as the unentrapped protein fraction. Theoretical entrapment efficiency = [(protein added-free unentrapped protein) / protein added] x 100. To determine morphology, a drop of diluted sample was placed on a silicon wafer overnight, placed in a Denton Benchtop Desk V Sputter for a thin coat of gold-palladium, and imaged with a Hitachi S-4700 Scanning Electron Microscope running at 5 KV. For in vivo intratumoral (IT) studies, wild type C57Bl/6J female PyMT breast tumor-bearing mice were treated with saline or rmGM-CSF by IT injection 3x per week for 2.5 weeks, followed by Electron Paramagnetic Resonance (EPR) spectroscopy to determine IT oxygen (pO2) and extracellular acidosis (pHe) using an injectable multi-functional soluble pTAM trityl compound, yielding spectra in Graphic 1 panel A.
Results: 5 ng intratumoral GM-CSF rescued tumor oxygen by ~38% and raised pH ~0.4 units in orthotopic PyMT tumor-bearing mice as detected by EPR, n=2 mice per group (saline: mmHg=59, pH=6.80. GM-CSF: mmHg=82, pH=7.30). Further, we have demonstrated that 5 ng/ml GM-CSF decreases IL-10 expressing tumor macrophages, ex vivo. We synthesized PLGA-PEG and PLGA-COOR diblock copolymers which were structurally confirmed by 1H-NMR Spectroscopy, with sizes of 30.2 kD and 3.8 kD, respectively. Both empty and GM-CSF-loaded PLGA/PEG-PLGA nanoparticles were then formed and characterized by size and zeta potential as in Graphic 2, yielding an average size appropriate to target the tumor by the Enhanced Permeation and Retention Effect (blue region). Average entrapment efficiency value was 87.9% +/- 4.2 SD. Morphology of nanoparticles was determined by SEM to be spherical and uniform.
Conclusion: We have provided rationale for nanoparticle-mediated delivery of GM-CSF to the tumor by demonstrating the potential of GM-CSF as an anti-cancer pretreatment modality that may address the problem of hypoxia-induced treatment resistance. Our “proof-of-concept” studies demonstrate that intratumoral administration of low dose (5 ng) GM-CSF can 1) reverse tumor hypoxia and acidity, and 2) reduce immunosuppressive IL-10 expression in ex vivo tumor macrophages in our PyMT breast cancer models. Therefore, we have formulated PLGA/PEG-PLGA nanoparticles loaded with recombinant murine GM-CSF and characterized these particles by size, zeta potential, and morphology. Overall, our GM-CSF-loaded PLGA/PEG-PLGA nanoparticles exhibit potential for future systemic and tumor-specific delivery of GM-CSF to alleviate hypoxia.
Andrey Bobko– Research Assistant Professor, West Virginia University, Morgantown, West Virginia
Valery Khramstov– Morgantown, West Virginia
Benoit Driesschaert– Morgantown, West Virginia
Tim Eubank– Associate Professor, School of Medicine, West Virginia University, Morgantown, West Virginia