SUPPLEMENTARY INFORMATION DOI: 10.1038/NNANO.2015.17 Breaking the depth dependency of phototherapy with Cerenkov radiation and low-radianceresponsive nanophotosensitizers Nalinikanth Kotagiri, Gail P. Sudlow, Walter J. Akers, Samuel Achilefu* NATURE NANOTECHNOLOGY www.nature.com/naturenanotechnology 1
Supplementary Discussion Interaction of diverse materials with CR In addition to TiO 2, studies have shown that some materials which are capable of absorbing energy in the 400-700 nm wavelength range can harvest CR for imaging applications 1-4. For example, quantum dots (QDs) are known to generate singlet oxygen radicals upon excitation with light, albeit with a low quantum yield 5. In conjunction with CR, QDs have appreciable fluorescence quantum yield for in vitro and in vivo imaging applications. Thus, it is entirely plausible to use QDs as nanophotosentizers (NPS) for CRIT. Enhancement of TiO 2 CRIT with Tc Although TiO 2 can serve as NPS under both aerobic and hypoxic conditions, intrinsic antioxidants in cells such as superoxide dismutase (SOD) are capable of inhibiting the oxidative damage caused by radicals such as of superoxide species. This could reduce the contribution of superoxide to therapy 6,7. Thus complementing TiO 2 cytotoxicity with other sources of free radicals could further enhance the effectiveness of CRIT. To accomplish this goal, we incorporated titanocene (Tc), a photoinitiator, into the NPS. Physical characterization of TiO 2 -Tf A strong signature of carbon in the Energy-dispersive X-ray spectroscopy (EDX) spectrum of TiO 2 -Tf compared to TiO 2 alone (Supplementary Fig. 1a) confirmed the presence of Tf on the surface of TiO 2 (Supplementary Fig. 1b). Electron diffraction analysis confirmed that the crystal lattice structure of anatase TiO 2 remained unchanged as a result of processing with Tf (Supplementary Fig. 1c,d). Phase transformation between anatase and rutile forms of TiO 2 2
typically occur at temperatures exceeding 700 C 8. Due to the relatively mild nature of the processing used to generate TiO 2 -Tf adducts, the photocatalytic properties of anatase TiO 2 employed in this study was thus maintained. The ring measurements are as follows: TiO 2-5.7 1/nm/2 = 0.351 nm = 3.51 A 8.4 1/nm/2 = 0.238 nm = 2.38 A 10.51/nm/2 = 0.189 nm = 1.89 A 11.81/nm/2 = 0.169 nm = 1.69 A TiO 2 -Tf - 5.7 1/nm/2 = 0.351 nm = 3.51 A 8.4 1/nm/2 = 0.238 nm = 2.38 A 10.5 1/nm/2 = 0.190 nm = 1.90 A 11.8 1/nm/2 = 0.169 nm = 1.69 A Supplementary Fig. 1. Composition and phase characterization of TiO 2 -Tf. a, EDX spectra of unprocessed TiO 2 with the peaks labelled as Ti for titanium, O for oxygen and C for carbon. 3
b, EDX spectra of TiO 2 -Tf with a pronounced C peak suggesting presence of the protein Tf on the surface of TiO 2. c, Electron diffraction of unprocessed TiO 2 with ring measurements matching the crystal pattern of anatase form of TiO 2, from diffraction file: 21-1272. d, Electron diffraction pattern of TiO 2 -Tf with ring measurements and crystal structure identical to that of TiO 2. Serum stability of TiO 2 -Tf To determine the serum stability of the TiO 2 -Tf interaction, we incubated the nanoparticles in foetal bovine serum for 24 h. We found that the amount of bound AlexaTf to TiO 2 surface did not significantly change with time (Supplementary Fig. 2a). Further analysis of data showed that serum components such as albumin, using Alexa 680 labelled BSA, did not form appreciable protein corona on the TiO 2 -Tf surface relative to pristine TiO 2 (Supplementary Fig. 2b), confirming that Tf does not readily exchange with serum proteins. Supplementary Fig. 2. Serum stability of TiO 2 -Tf NPS. a, Comparison of fluorescence intensity between TiO 2 -AlexaTf NPS incubated in foetal bovine serum for 24 h and untreated samples (ns: not significant). b, Comparison of unlabelled TiO 2 -Tf and unprocessed TiO 2 incubated with Alexa 680 labelled albumin. Values are means ± s.e.m. (experiments for each group were run in triplicates). **P < 0.01. 4
Optical properties of TiO 2 and Tc Coincidentally, the excitation energy for TiO 2 (Supplementary Fig. 3a) and Tc (Supplementary Fig. 3b) is in the UV spectrum, where CR quantum efficiency (Supplementary Fig. 3c) is the highest. Supplementary Fig. 3. Spectral characterization of TiO 2 and Tc. a, Absorption spectrum of TiO 2 in water. b, Absorption spectrum of Tc in water/dmso (95/5%). c, Emission spectrum of CR from 64 Cu. d, Fluorescence spectrum of TiO 2, excited at 275 nm. CPS, counts per second. e, In vitro luminescence studies carried out with 0.1 mci/100 l of 64 Cu, 0.1 mci/100 l 64 Cu admixed with 1 mg/ml TiO 2, and 0.1 mci/100 l of 99m Tc admixed with 1mg/ml TiO 2, in each 5
well. Images were captured in the GFP channel (515-575 nm) in an IVIS Lumina XR multimodal imaging system (Caliper Life Sciences Inc.) using Living Image software. In vivo blocking study using TiO 2 -Tf labelled with Alexa 680 Supplementary Fig. 4. In vivo blocking of TiO 2 -Tf uptake by HT1080 tumours. a, Organ biodistribution of TiO 2 -AlexaTf. b, Organ biodistribution of TiO 2 -AlexaTf after administration of holo-tf to block Tf receptors. c, Comparison of biodistribution of TiO 2 -AlexaTf with and without blocking. Intrinsic toxicity of TiO 2, Tc, FDG, and 64 Cu 6
Our cytotoxicity analysis revealed that TiO 2 NPS and Tc did not induce apoptosis at 4 g/ml and 10 g/ml, respectively (Supplementary Fig. S5a). Similarly, viability in 64 Cu ( 0.5 mci (18.5 MBq)/0.1 ml) and FDG (1 mci (37 MBq)/0.1 ml) treated cells was >95% relative to untreated control (Supplementary Fig. S5b). Supplementary Fig. 5. Assessment of the intrinsic and stimulated cytotoxicity of the photoactive products using MTS cytotoxicity assay. a, Viability of cells treated with Tc-Tf, TiO 2 -Tf and TiO 2 -PEG. The control group was considered to be 100% viable. b, Viability of cells treated with 64 Cu and FDG. HT1080 cells were used and the values are means ± s.e.m. (experiments for each group were run in triplicates). Free radicals generated by TiO 2 Hydroxyl radicals, are highly reactive short lived species, non-diffusible across cell membranes, causing a highly pronounced local therapeutic effect (~ 20 nm) that typically culminates into necrosis 9,10. In contrast, superoxide radicals are more stable species that can travel across cell membranes with a long diffusion distance of ~320 nm and result in disruptions of lipid membranes such as plasma and mitochondrial membranes leading to apoptosis 31. 7
Supplementary Fig. 6. Confocal imaging of TiO 2 free radicals. Images of merged bright-field and fluorescence images comparing the degree of hydroxyl radical generation caused by 0.1, 0.25, 0.5, 1 mci/100 l of 64 Cu on tumour cells with 2.5 g/ml TiO 2 -PEG after 4 h. HPF dye was used to stain the cells. The green fluorescence from cells depicts increased hydroxyl radical generation. Highest fluorescence intensity was recorded from cells incubated with 0.25 mci/100 l of 64 Cu. Scale bar, 20 m. In vitro assessment of mitochondrial membrane potential as a result of CRIT To further understand the role of free radicals and the mechanism of cell death, we evaluated whether CRIT induces alterations in mitochondrial membrane potential. There was a significant decrease in mitochondrial membrane potential in cells treated with both TiO 2 and Tc coincubated with FDG, indicating damaged and leaky membranes. Typically, damage to mitochondrial membranes initiates the intrinsic signalling pathway for apoptosis, characterized by loss of membrane potential and a cascade of events involving caspases leading to nuclear fragmentation and cell death. 8
Supplementary Fig. 7. Loss of mitochondrial membrane potential due to CRIT. Mitochondrial membrane potential changes detected by Mitotracker Green dye as a result of CRIT. Values are means ± s.e.m. *P < 0.05, **P < 0.01. Assessment of body weight and remission after achieving CRIT induced regression of tumours 9
Supplementary Fig. 8. Change in murine weights in untreated and treated groups of mice. TiO 2 - PEG and chelated 64 Cu were administered intratumourally and monitored over 4 months. Assessment of off-target toxicity in liver and kidneys 10
Supplementary Fig. 9. Histologic images of tissues. H&E stained liver and kidney sections before and after treatment did not show significant lesions in these organs, indicating the absence of systemic toxicity from CRIT. References 1. Dothager, R.S., Goiffon, R.J., Jackson, E., Harpstrite, S. & Piwnica-Worms, D. Cerenkov radiation energy transfer (CRET) imaging: a novel method for optical imaging of PET isotopes in biological systems. PLoS ONE 5, 0013300 (2010). 2. Liu, H., et al. Radiation-luminescence-excited quantum dots for in vivo multiplexed optical imaging. Small 6, 1087-1091 (2010). 3. Carpenter, C.M., et al. Radioluminescent nanophosphors enable multiplexed smallanimal imaging. Opt. Express 20, 11598-11604 (2012). 4. Ran, C., Zhang, Z., Hooker, J. & Moore, A. In vivo photoactivation without "light": use of Cherenkov radiation to overcome the penetration limit of light. Mol. Imaging Biol. 14, 156-162 (2012). 5. Bakalova, R., Ohba, H., Zhelev, Z., Ishikawa, M. & Baba, Y. Quantum dots as photosensitizers? Nat. Biotech. 22, 1360-1361 (2004). 6. Kono, Y. & Fridovich, I. Superoxide radical inhibits catalase. J. Biol. Chem. 257, 5751-5754 (1982). 7. Morel, D.W., Hessler, J.R. & Chisolm, G.M. Low density lipoprotein cytotoxicity induced by free radical peroxidation of lipid. J. Lipid Res. 24, 1070-1076 (1983). 8. Li, W., Ni, C., Lin, H., Huang, C.P. & Shah, S.I. Size dependence of thermal stability of TiO2 nanoparticles. J. Appl.Phys. 96, 6663-6668 (2004). 9. Apel, K. & Hirt, H. Reactive oxygen species: metabolism, oxidative stress, and signal transduction. Annu. Rev. Plant Biol. 55, 373-399 (2004). 10. Rozhkova, E.A., et al. A high-performance nanobio photocatalyst for targeted brain cancer therapy. Nano Lett. 9, 3337-3342 (2009). 11