P24 - Lipoprotein hitchhiking peptide amphiphiles for MRI and fluorescent imaging of brain tumors

Introduction: : Treatment of brain cancer, including glioblastoma multiforme (GBM), requires surgical resection followed by chemotherapy and radiotherapy. It has been shown that >98% resection of malignant tissues yields a better prognosis, but the infiltrative nature of GBM combined with the limitations of current optical imaging agents makes this difficult. Intraoperative imaging, such as MRI and fluorescence imaging, assists surgeons in visualizing tumor margins, but current imaging agents are limited by rapid clearance, low tumor-specific accumulation, and difficulty crossing the blood brain barrier (BBB). There is need for improved MRI and near-IR (NIR) fluorescence imaging agents for detection and removal of GBM tumors that allows for maximum resection with minimal loss of function. Our group has developed a peptide amphiphile (PA), SA-E, that has shown selective targeting capabilities towards various solid tumors via association with endogenous circulating lipoproteins. Lipoprotein binding allows SA-E to cross the blood brain barrier and be trafficked to overexpressed lipid rafts on cancer cells. SA-E was conjugated to NIR fluorophore indocyanine green (ICG) to identify fluorescent tumor margins in multiple rat glioblastoma models with tumor-to-healthy brain tissue ratios reaching 16 and a retention time of over 72 hours. We modified SA-E to add gadolinium chelator DOTA for MRI imaging, allowing for tumor-specific gadolinium enhancement with low background signal and high retention time. Herein, we show novel PA conjugates representing a promising advancement in the development of a multimodal optical imaging agent for improved resection of GBM tissues through MRI and fluorescence guided surgery (FGS).

Materials and

Methods: : The structure of SA-E was modified to allow for conjugation of MRI and NIR fluorescence contrast agents, gadolinium (Gd) and ICG. The lipoprotein binding profile of SA-E-ICG and SA-E-Gd-ICG was examined by incubating the probes in 20% mouse plasma and running FPLC. BBB crossing of the SA-E was investigated using a bEnd.3 brain epithelial cell transwell model to monitor fluorophore activity over time. Finally, rats were intracerebrally implanted with different GBM cell lines. Once the tumors reached a sufficient size, our peptides were injected via tail vein. Animals injected with SA-E-Gd-ICG conjugation were evaluated for MRI enhancement at 24-hours and 48-hours post-injection. For all models, brain, kidney, liver, and spleen were harvested for fluorescent imaging and probe quantification 48-hours post-injection. Tumor was bisected coronally, and one half of the brain was immediately frozen in O.C.T. embedding medium. The other half of the brain and harvested tissues were immediately imaged with IVIS. Tissues were frozen and homogenized to quantify fluorescent signal. Brain slices were taken via cryostat (16µm) then fixed in 4% formaldehyde for 6 minutes and stained with Hoechst (1:4000) for 6 minutes. 20X and 60X confocal images were obtained and then processed with ImageJ.

Results, Conclusions, and Discussions:: FPLC analysis confirmed that both SA-E-ICG and SA-E-Gd-ICG associated primarily with HDL when incubated with plasma, thus allowing for stable blood circulation. BBB transcytosis measurements showed that over four hours, approximately 15% of SA-E crossed the bEnd.3 membrane. In vivo experiments showed significant tumor cell uptake of SA-E up to 48-hours post-injection using syngeneic RG2-mCherry and patient derived GBM6 tumor models with relatively low fluorescent signal in kidney, liver, and spleen. Additionally, RG2-mCherry tumors injected with SA-E-Gd-ICG showed gadolinium enhancement with MRI imaging starting at 24-hours post-injection, with a peak at 48-hours post-injection. At 48-hours post-injection, IVIS and FGS setups showed tumor to healthy tissue ratios ranging from 5-16 in different tumor models. Quantification via homogenization showed that ICG signal was up to 67 times higher in RG2 tumors compared to healthy brain tissue. Confocal images showed cellular internalization of SA-E-ICG and SA-E-Gd-ICG in all tumor models. Overall, SA-E binds to endogenous HDL in circulation and crosses the blood brain barrier before being internalized by cancer cells. This system allows for high accumulation of MRI and NIR imaging agents in multiple GBM tumor models over extended periods of time with low accumulation in healthy tissues, demonstrates that SA-E has promising potential as an intraoperative multimodal imaging agents for the surgical resection of GBM.

Acknowledgements and/or References (Optional)::