Mapping Oxygen Tension in Engineered 3D Tumor Microenvironments with Ruthenium-complex Nanomicelle Optical Sensors

Introduction:: Oxygen tension and hypoxia play an essential role in oncogenesis and tumor progression. It has been shown that tumor hypoxia correlates with poor prognosis of cancer patients, including breast, lung, and head and neck cancers, as it may enhance the resistance to chemotherapy and radiotherapy [1, 2]. Previous methods to measure oxygen tension (pO₂) within patient tumors and 3D tumor models have utilized micro-oxygen electrodes. Although these have been the gold standard to measure pO₂ within tissues, they lack high spatial resolution and consume oxygen in the process. An emerging approach utilizes oxygen-sensing optical probes in conjunction with two-photon phosphorescence lifetime microscopy (2P-PLIM) to permit pO₂ reporting with high spatial resolution and high sensitivity in biologically-representative tissue culture models, with negligible oxygen consumption in the process. Of the oxygen-sensing optical probes, [Ru(dpp)₃]²⁺ (Ru) is an excellent candidate as it is two-photon excitable and has a phosphorescence lifetime in the microsecond (µs) regime. However, Ru is hydrophobic which limits its potential in vitro and in vivo. The objective of our study is to encapsulate Ru within a polymer micelle to create hydrophilic Ru-complex micelles and incorporate these micelle sensors within 3D matrices containing multicellular spheroids. We aim to characterize these micelle complexes and implement them to report pO₂ within cancerous and non-cancerous spheroid models.

Materials and Methods:: Ru-complex nanomicelles were prepared in a manner similar to that described by Khan et al. [3] by complexing [Ru(dpp)₃]Cl² with poloxamer-407 via ultrasonication (Fig. 1A). Encapsulation efficiency of the Ru as well as equivalent Ru concentration was determined via fluorescence readouts with known standards on a plate reader (Tecan). Micelle size was characterized via nanoparticle tracking analysis (NTA) and transmission electron microscopy (TEM). Localization of micelles within and surrounding breast epithelial (MCF10A) spheroids was performed via two-photon microscopy to image collagen by second harmonic generation (SHG) as well as Ru fluorescence imaging. Viability was assessed via Hoechst and ethidium homodimer staining of MCF10A spheroids at various concentrations of Ru-complex micelles in the media. To generate lifetime images from the Ru-complex nanomicelles, 2P-PLIM was performed via confocal laser scanning microscope (Zeiss, LSM880) with a Becker&Hickl time-correlated single photon counting module to perform simultaneous two-photon fluorescence lifetime imaging (2P-FLIM) and 2P-PLIM. Excitation of the Ru-complex nanomicelles occurred at 740 nm and emission was filtered via 640/20 nm bandpass filter to a BiG detector. A pO₂ calibration curve for the Ru-complex micelles was created by measuring the phosphorescence lifetime of the sensors in sealed solutions of known oxygen concentrations. To report real-time pO₂ of MCF10A spheroids, 15 µM equivalent Ru was added to the media of MCF10A spheroids embedded in collagen hydrogels and 2P-PLIM was performed at designated time intervals to produce phosphorescence lifetime images. Utilizing the oxygen calibration curve, the lifetime images were converted to pO₂ images.

Results and Discussions:: The Ru-complex micelles created were soluble in water with a mean encapsulation efficiency of 93 ± 1.5 %. NTA revealed that the mean diameter of the Ru-complex micelles was 124 nm. TEM showed successful encapsulation of the Ru to form nanomicelles (Fig. 1B). To visualize the localization of the Ru-complex micelles, fluorescent imaging of Ru was performed in 1 mm thick collagen gels embedded with MCF10A spheroids in the center of the gel. Ru-complex nanomicelles added to media showed no disruption in collagen structure and showed localized fluorescence signal within and surrounding spheroids for over 1 week of culture. MCF10A spheroid viability analyses demonstrated that effective Ru concentrations of ~7.5 µM were similar to a control absent of micelles with mean viability of about 85 %. Thus, this micelle concentration was used for all additional spheroid studies. Viability decreased to around 47 % with effective Ru concentration of ~150 µM. Calibration of the oxygen sensor showed a two-phase lag decay fit curve of the data (Fig. 1C) and a linear Stern-Volmer plot. Longitudinal analyses of pO₂ within embedded spheroids showed that at days 1, 3, and 5 post-seeding, the core remained hypoxic with ~1% O₂ present, as is expected due to the diffusion-limited supply of oxygen. pO₂ image slices of the spheroids showed a cyclical gradient of pO₂ ranging from low oxygen (~1% O₂) to slightly higher oxygen (~4% O₂) along the radial direction of the spheroid (Fig. 1D-G).

Conclusions: Hydrophobic Ru was successfully encapsulated with amphiphilic poloxamer-407 to create hydrophilic complex nano-sized micelles that can be visualized within MCF10A spheroids while maintaining high viability. These sensors demonstrated the ability to report oxygen tension in hydrogel-based 3D spheroid cultures utilizing 2P-PLIM. This method to spatially resolve oxygen concentrations in real time will allow for future development and expansion of the platform to determine pO₂ gradients and sites of hypoxia within additional 3D models, including breast cancer spheroids and organoids.

References (Optional): :
1. Höckel, M. and P. Vaupel, Tumor hypoxia: definitions and current clinical, biologic, and molecular aspects. J Natl Cancer Inst, 2001. 93(4): p. 266-76.
2. Hompland, T., C.S. Fjeldbo, and H. Lyng, Tumor Hypoxia as a Barrier in Cancer Therapy: Why Levels Matter. Cancers, 2021. 13(3): p. 499.
3. Khan, A.A., S.K. Fullerton-Shirey, and S.S. Howard, Easily prepared ruthenium-complex nanomicelle probes for two-photon quantitative imaging of oxygen in aqueous media. RSC Advances, 2015. 5(1): p. 291-300.