Animal heat activated cancer therapy by a traditional catalyst TiO2-Pd/graphene composites

Cancer therapy is one of the most important challenges in clinical medicine. So far different methods have been developed for cancer therapy, such as radiation therapy, surgery, chemotherapy and photodynamic therapy. Here we propose a new concept for cancer therapy, i.e., killing the cancer cells simply via reactive oxygen species (ROS) generated by TiO2-Pd/graphene composites. Activated by animal heat of 37 °C, the electrons in the valence band can be excited to the conduction band of TiO2 via the energy levels of Pd species and graphene, generating ROS without light irradiation or electric excitation. The tumors in BALB/c mice are successfully regressed at animal heat without any other external conditions, such as radiation, UV, visible and IR irradiation. Our results suggest that the design of animal heat activated cancer therapy is a feasible concept for practical applications of cancer treatments.


Results and discussion
Raman spectra, XPS and HRTEM are applied to investigate the structure of TiO 2 -Pd/graphene . As shown in Fig. 1A, the Raman peaks of anatase at 144 cm −1 (Eg), 194 cm −1 (Eg), 396 cm −1 (B1g), 516 cm −1 (A1g and B1g), and 638 cm −1 (Eg) are observed in TiO 2 and TiO 2 -Pd samples. For TiO 2 /graphene and TiO 2 -Pd/Graphene, besides the Raman peaks of anatase, the peaks of reduced graphene at 1315 and 1585 cm -1 are detected 12,13  www.nature.com/scientificreports/ results are obtained from the XRD patterns (Fig. S1). No other XRD peaks, such as PdO can be observed. The cell volume and lattice parameters of TiO 2 -Pd x% and TiO 2 -Pd/Graphene x% derived from XRD data remain almost unchanged ( Fig. S2 and Table S1) compared with TiO 2 . This suggests that Pd ions are not doped into TiO 2 lattice in substitutional or interstitial mode, as the ionic radius of Pd 2+ (86 pm) is much larger than that of Ti 4+ ions (68 pm). Therefore, it can be deduced that the Pd ions might exist on the surface of TiO 2. Moreover, according to the XPS results, the Cl 2p 3/2 peaks (198.5 eV) for TiO 2 -Pd/Graphene (Fig. S3A) locates between that of TiCl 4 (198.2 eV) and PdCl 2 (198.9 eV), ascribed to the surface O-Pd-Cl structure 14 . Two pairs of doublet Pd 3d peaks can be observed for TiO 2 -Pd/Graphene (Fig. 1C). One Pd3d 5/2 peak at 337.7 eV is attributed to -O-Pd-Cl structure (i.e., one Pd 2+ ion is linked with one Clion and one unsaturated oxygen ion) on the surface. The other peak at 336.2 eV for Pd 3d 5/2 is ascribed to -O-Pd-O-species on the surface (i.e., one Pd 2+ ion is linked with two unsaturated oxygen ions, which has been confirmed by our previous work 14 ). The molar ratio of Pd/Ti for TiO 2 -Pd and TiO 2 -Pd/graphene is 4.19%/100% and 4.14%/100%, respectively. Moreover, the C1s peak at 284.3 eV (Fig. S3B) is ascribed to the graphene. HRTEM image of TiO 2 -Pd/Graphene (Fig. 1B) confirms TiO 2 -Pd nanoparticles are attached on the surface of graphene. In addition, the peak of 1216 cm -1 in the FR-IR spectra of TiO 2 -Pd/Graphene is ascribed to vibration of C-O bond (Fig. S5). The XPS Ti 2p spectra ( Fig. S24 and S25) also confirms the graphene and TiO 2 -Pd are connected via the Ti-O-C bonds. The surviving fraction of the A549 cells was measured by standard 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay to evaluate the ability of killing cancer cells for all samples at 37 °C, as shown in Fig. 2A. The surviving fraction of the A549 cells is about 96% for TiO 2 , 85% for graphene, 71% for TiO 2 /graphene, 53% for TiO 2 -Pd and 16% for TiO 2 -Pd/graphene, respectively, when the concentration is 25 μg/ ml after 16 h of incubation at 37 °C. Moreover, it is found that the surviving fraction of cancer cells (16%) for TiO 2 -Pd/graphene is about 1/5 times as that for pure TiO 2 . With the increase of sample concentration, the performance of killing cancer cells falls in order of TiO 2 < graphene < TiO 2 /graphene < TiO 2 -Pd < TiO 2 -Pd/graphene. Furthermore, we have also checked the surviving fraction of cancer cells treated with TiO 2 -Pd x% and TiO 2 -Pd/ graphene x% and the surviving fraction is the lowest for TiO 2 -Pd1.5% and TiO 2 -Pd/graphene40% (Fig. S6). Furthermore, the increased Pd content (> 1.5%) and graphene content (> 40%) would inhibit the performance of killing cancer cells. Thus, the introduced Pd species and graphene into TiO 2 system is the important key factor for the improved ability of killing cancer cells. The optical microscopy images of A549 cancer cells in the presence of samples (100 μg/mL) are shown in Fig. S9. Most of the A549 cells survive well in the blank experiment (without the addition of any samples) (Fig. S9a). The TiO 2 can hardly influence the surviving of the cancer cells (Fig. S9b). It is found that the relative amount of healthy cancer cells decreased for TiO 2 /graphene (Fig. S9c) and TiO 2 -Pd (Fig. S9d). As we expected, seldom healthy cancers cell can be observed in Fig. S9e for TiO 2 -Pd/ graphene samples, suggesting its high efficiency on killing cancer cells. These results are in good agreement with MTT results in Fig. 2A.
The effect of temperature on the surviving fraction of cancer cells treated with various concentration of TiO 2 -Pd/graphene for 4 h is evaluated in Fig. 2B. The surviving fraction of cancer cells for TiO 2 -Pd/graphene decreased with the increased system temperature. When the concentration of TiO 2 -Pd/Graphene is about 25 μg/ mL, the surviving fraction of A549 cells are approximately 80%, 70%, 60%, 55%, and 50% at 4, 12, 25, 37 and 39 °C, respectively. Similar results are observed in the synovial cells (Fig. S7). This indicates the system www.nature.com/scientificreports/ temperature has great influence on the activity of killing cancer cells for TiO 2 -Pd/graphene, and TiO 2 -Pd/graphene can effectively inhibit the growth of cells at animal heat condition (37 °C). Fig. S8 shows the relationship between the surviving fraction of the different (A549 and smooth muscle cells) cells and concentration of TiO 2 -Pd/graphene. It is found that only about 10% of cancer cells survived and 40% of smooth muscle cells survived at 37 °C for 16 h with a concentration of 25 μg/mL. This suggests a possible selectivity on cancer therapy for TiO 2 -Pd/graphene. It is expected that further investigation should be carried out by combining other techniques, such as targeted therapy, tumor injection, magnetic control to avoid the side effects on normal cells.
For in vivo animal heat activated therapy study, we employed subcutaneous 4T1 xenograft model in BALB/c mice to study the efficacy of animal heat activated cancer therapy (37 °C) after treatment by PBS (phosphate buffer saline), TiO 2 , TiO 2 -Pd and TiO 2 -Pd/graphene. Twelve tumor-bearing mice were randomly and evenly divided into four groups, which were intratumorally injected with PBS, TiO 2 TiO 2 -Pd and TiO 2 -Pd/graphene, respectively. Figure 2C shows the thorough regression of tumor volume is observed only in the group with intratumoral injection of TiO 2 -Pd/graphene, while the volume of tumor increases for the group with PBS and TiO 2 and remains unchanged for the group with TiO 2 -Pd. For the group with intratumoral injections of TiO 2 -Pd/graphene, the average relative tumor volume decreased to approximately 50% and for the group with intratumoral injections of TiO 2 -Pd, the average relative tumor volume remains almost unchanged. Figure 2D displays the representative photographs of tumors on mice that the intratumorally injected with PBS control and TiO 2 -Pd/graphene. The growth of tumor is inhibited when injected with TiO2-Pd/graphene. These experiment results indicate that TiO 2 -Pd/graphene is an effective functional material to inhibit cancer growth under animal heat (37 °C) without external irradiation or heating.
To get the physical insight of the band structure and density of states for TiO 2 -Pd/graphene, the theoretical calculation results of pure TiO 2 , TiO 2 -Pd and graphene are shown in Fig Fig. 3E.
The Raman spectra, cyclic voltammetry curves and fluorescence spectra of the DCF are applied to investigate the relationship between the generation of ROS and temperature. The Raman spectra of TiO 2 -Pd and TiO 2 -Pd/ graphene (Fig. S19) indicates the peak intensity was enhanced with the increase of temperature, suggesting an enhanced vibrating energy of lattice, benefiting the electrons' transfer to surface of sample.
The cyclic voltammetry curves of pure TiO 2 , graphene, TiO 2 /graphene, TiO 2 -Pd and TiO 2 -Pd/graphene at 37 °C (in Fig. 3A) indicate the ability to give and receive electrons increases when Pd species and/or graphene are introduced into TiO 2 system. For the cyclic voltammetry curves of the TiO 2 -Pd/graphene at different temperatures, the redox peaks of TiO 2 -Pd/graphene are improved with the increase of temperature (Fig. 3B). This suggests that TiO 2 -Pd/Graphene may exhibit enhanced ability to give and receive electrons under thermal excitation.
The amounts of ROS ( . OH, O 2 − , et al.) generated by TiO 2 , graphene, TiO 2 /graphene, TiO 2 -Pd and TiO 2 -Pd/ graphene at 37 °C were evaluated by the photoluminescence (PL) intensity of the DCF (Fig. 3C). No PL peak is detected in the blank experiment. For TiO 2 and graphene, a weak PL peak is found, suggesting a small amount of ROS formed. Moreover, for TiO 2 /graphene and TiO 2 -Pd, the PL intensity further increases, suggesting more ROS can be generated than TiO 2 . The PL intensity for TiO 2 -Pd/graphene is the strongest among all the samples, indicating TiO 2 -Pd/graphene is most effective to generate ROS among all the samples at 37 °C. Figure 3D shows generation amount of ROS for TiO 2 -Pd/graphene at different temperatures. It is noted that the PL intensity significantly increases with the increase of temperature, indicating the increased temperature is benefit for the generation of reactive oxidation species. This also confirms that the electrons and holes in TiO 2 can be excited by heat and transfer to surface to form ROS. www.nature.com/scientificreports/ Based on the aforementioned discussion, the mechanism of generating ROS (such as ·OH, O 2 − ) and kill the cancer cells for all samples can be explained using the schematic band structure shown in Fig. 3E. The electrons in the conduction band (− 0.45 eV, vs NHE) excited by animal heat (37 °C), whose potential is more negative than that of O 2 /O 2 − (− 0.33 eV, vs NHE), can transfer to the surface and directly captured by the adsorbed O 2 molecules on the surface to form O 2 − active species as shown in Eq. (1) [16][17][18] . The holes in the valence band (+ 2.65 eV, vs NHE), whose potential is more positive than H 2 O /H + (0.82 V, vs NHE), are captured by surface absorbed H 2 O molecules to form O 2 and H + , which can further react with thermal electrons to produce the hydroxyl free radical OH·as shown in Eqs. (2), (3) and (4)  In summary, we have demonstrated a new concept for developing high efficient animal heat activated cancer treatment for TiO 2 -Pd/graphene. The electrons and holes can be excited through the energy levels of Pd species and graphene at animal heat, generating ROS which can further kill the cancer cells. This may afford a feasible and efficient approach for cancer therapy application, without any other external conditions such as radiation, UV, visible and IR irradiation that may cause serious body damage.

Methods catalyst preparation. Synthesis of Pd-TiO 2 .
All chemicals used were of analytical grade and the water was deionized water (> 18.2 MΩ cm). At room temperature, certain volume of PdCl 2 (0.1036 mol/L) solution were mixed with 40 mL of ethanol. Then 1 mL of HCl solution (12 mol/L) and 12 mL of Ti(OC 4 H 9 ) 4 was added dropwise into the mixture under vigorous stirring. The mixture was stirred until the formation of TiO 2 gel, followed by being aged for 24 h. The obtained gels were dried at 373 K for 12 h and annealed at 723 K in a muffle for 2.5 h. The resultant samples were denoted as Pd-TiO 2 . Pure TiO 2 was prepared using the same procedure, while without the addition of PdCl 2 solution. Unless stated otherwise, the nominal molar ratio of Pd 2+ to Ti 4+ is fixed at 1.5% in the precursor and. For comparison, other molar ratios were also used for Pd 2+ to Ti 4+ (such as 0.5%, 1.0%, 2.0%, 2.5% and 3.0%).

Synthesis of graphene oxide (GO)
. GO was prepared from crystalline flake graphite powder according to the modified method reported by Hummers and Offeman 19 . In brief, 10 g of graphite powder and 5 g of NaNO 3 were added into 230 mL of cooled (273 K) concentrated H 2 SO 4 . Then, 30 g of KMnO 4 was added gradually with continuous stirring and cooling, and the temperature of the mixture was maintained below 293 K. After the ice bath was removed, the mixture was stirred at 308 K for 30 min. 460 mL of distilled water was added slowly to cause an increase in temperature to 371 K, and the mixture was maintained at that temperature for 15 min. The reaction was terminated by addition of 1.4 L of distilled water followed by 25 mL of 30% H 2 O 2 aqueous solution. The solid product was separated by centrifugation and washed repeatedly with 5% HCl solution (2L) and deionized water until sulfate anion could not be detected with BaCl 2 . The resultant solid was dried in vacuum at 323 K to obtain GO.
Preparation of TiO 2 -Pd/grapheme. GO was first dissolved in deionized water by ultrasonic treatment for 20 min. Then, Pd-TiO 2 was added into the GO colloidal solution and the mixture was ultrasonic for another 1 h to obtain a homogeneous suspension. The resultant composite was collected by drying at 333 K and then www.nature.com/scientificreports/ triturated to powder in an agate mortar. Finally, the powder was calcined at 300 °C for 2 h under Ar atmosphere. The resulting products wereTiO 2 -Pd/graphene.
Characterizations. The XRD patterns were acquired using a Rigaku D/max 2500 X-ray diffraction spectrometer (Cu Kα, λ = 1.54056 Å). The average crystallite size was calculated according to the Scherrer formula (D = k λ/B cosθ). Raman spectra were taken on a Renishaw inVia Raman microscope by using the 785 nm line of a Renishaw HPNIR 785 semiconductor laser. The Fourier Transform Infra-Red (FT-IR) spectra were recorded for KBr disks containing the powder sample with an FT-IR spectrometer (MAGNA-560). The BET surface areas of the samples were determined by nitrogen adsorption-desorption isotherm measurements at 77 K (Micromeritics Automatic Surface Area Analyzer Gemini 2360, Shimadzu). XPS measurements were carried out by using a Thermo ESCALAB 250 spectrometer with an Al Kα monochromator source and all the binding energies were calibrated to the adventitious C1s peak at 284.8 eV. Diffuse reflectance UV-visible (UV-Vis) absorption spectra were recorded on a UV-Vis spectrometer (U-4100, Hitachi). ROS in the presence of samples was qualitatively detected by the H2DCF-DA assay. Photoluminescence (PL) spectra were acquired by using the 325 nm line of a nano-second Nd: YAG laser (NL303G) as excitation source. The experimental setup consists of a spectrometer (Spex 1702), a photomultiplier tube (PMT, Hamamatsu R943), a lock-in amplifier, and a computer for data processing. The cyclic coltammetry curves were measured using an electrochemical workstation (CorrTest, Wuhan, Inc.) in a conventional three-electrode cell at different temperature. The samples/ITO is used as working electrode, Pt is used as counter electrode and the saturated calomel electrode (SCE) was used as reference electrode. 0.5 mmol/L K 4 [Fe(CN) 6 ] + + 0.05 mmol/L K 3 [Fe(CN) 6 ] + + 0.1 mol/L KCl aqueous solution was used as electrolyte.
The generation ability of reactive oxygen species (ROS) for thermal catalysts can be estimated by measuring the fluorescent intensity of 2′,7′-dichlorofluorescein (DCF). The 2′,7′-dichlorodihydrofluorescein (DCFH, nonfluorescent) can rapidly react with ROS in the thermal catalysis system to form 2′,7′-dichlorofluorescein (DCF, fluorescent). By measuring the fluorescent intensity of DCF, the generation ability of ROS can be determined for thermal-catalyst. Experimental process is as follow: 5 mg of catalysts were added into 5 ml centrifuge tube, then 1 ml working solution (975 μl of diluents and 25 μl staining fluid containing DCFH) was added. The mixture was shock and heated at different temperature for 30 min (room temperature, 65 °C). Then, the mixture was centrifuged and the supernatant was taken to detect the fluorescent intensity. The exciting wavelength for exaction is 491nm 20 .
Apoptosis assay. Human Lung Carcinoma cells (A549), smooth muscle cells and synovial cells were cultured in RPMI 1640 medium in 96-well plates, containing 10% fatal calf serum (FCS) in a humidified incubator with an atmosphere of 5% CO 2 in air at 37 °C. The cell density was 2 × 10 4 cells per well. After being seeded for 24 h, the media were replaced by culture media containing a series of TiO 2 , Pd-TiO 2 and TiO2-Pd/Graphene nanoparticles with increasing concentrations in RPMI 1640 medium and then the plates were placed into the humidified incubator. After another 16 h for the interaction between the cancer-cells and the nanocomposite particles, cell viabilities were measured by the standard MTT assay, a colorimetric assay based on the ability of viable cells to reduce 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide. The survival rate and the error bar are shown in Fig. 2B (the case of incubating with nanocrystals only).
In vivo thermal therapy. BALB/c mice bearing A549 lung carcinoma tumors were intratumorally injected with TiO 2 , Pd-TiO 2 or graphene /Pd-TiO 2 (80 ul of 4 mg/ml solution for each mouse), respectively. The images were taken by an high definition camera. The tumor sizes were measured by a caliper every other day and calculated as volume = (tumor length) × (tumor width) 2 /2. Relative tumor volumes were calculated as V/V0 (V0 is the tumor volume when the treatment was initiated). The animal protocol in this study conformed to the Guide for the Care and Use of Laboratory Animals (the Guide, NRC 2011), and it was also approved by the Institutional Animal Care and Use Committee at Nankai University (Approval ID 201009080081).