Fig 1.
Labelling of TiO2-NTs with fluorophore Alexa via the two steps-reaction of the TiO2-NTs surface modification.
Firstly, 3-(2-aminoethylamino)propyltrimethoxysilane (AEAPMS) is attached to free–OH groups of the nanotubes’ surface (fTiO2-NT). Secondly, the Alexa 488 SDP ester is covalently linked to the free amino groups of silane molecules (A-TiO2-NT).
Fig 2.
Characterization of the TiO2-NTs and their photo-catalytic activity.
(A) TEM image of the TiO2-NTs. Inset in Fig 2A shows interlayer spacing of about 0.35 nm that agree well with [100] diffraction peak of anatase TiO2. (B) X-ray powder diffraction of TiO2-NTs, anatase peaks are marked with A. (C) OH• radical production of TiO2-NTs (closed black circles) as a function of wavelength compared to the radical production by P25 from Degussa (open circles). At 365 nm production of radicals decreases to half of their maximal activity at lower wavelengths.
Fig 3.
Toxicity and phototoxicity of TiO2-NTs.
MCF-7 cells were grown during the first day, exposed to 1000 μg/mL of nanotubes during the second day and incubated in dark or under UV irradiation (1.6 W/m2, wavelength 356 nm, every 3 h for 15 min in a period of 24 h) during the third day. Control experiment were performed without the appropriate stressor (nanotubes during the second day or UV irradiation during the third day). On the fourth day: (A) optical density of samples was measured by MTS assay, (B) cells were trypsinized and counted manually using Trypan Blue, (C) the same number of MCF-7 cell were seeded in pools to grow for additional 14 days and colonies were counted. Survival factor (SF) was calculated as percentage of plating efficiency (PE) of the treated cells over PE of the untreated cells. Since normality test (Shapiro-Wilk) failed, we used Mann-Whitney U statistics instead. ** denotes the significant difference in the median values at P = <0.001.
Fig 4.
Cell viability versus concentration of TiO2-NTs in absence or presence of UV radiation.
Absorbance of MTS was measured as described in Materials and methods section. (A) Each non-irradiated data set (red symbols) is compared to a sample irradiated with UV light (blue symbols) at given concentration of TiO2-NTs. All Measurements are presented in a boxplot representing the distribution of the measured cell viability and described in Materials and methods. Differences in the cell viability between irradiated and non-irradiated cells are shown with the black arrows at each concentration of TiO2-NTs. The differences in the median values between the two groups at 0 and 1000 μg/mL and the two groups at 1 and 50 μg/mL of nanotubes are considered to be significant at P = <0.001, and P = <0.05, respectively. (B) Differences of median values of MTS absorbance between irradiated and non-irradiated cells, shown with the black arrows within the frame A, are plotted with the open bars. Solid black line represents the prediction of the difference in cell viability, based on the effect of the UV irradiation and phototoxic effect due to irradiated nanotubes not covered by serum proteins. Both contributions are shown separately in the frame C. (C) Solid dark blue line shows the transmittance of UV light versus the concentration of the nanotubes obtained experimentally (S2 Information. Optical properties of TiO2-NTs dispersion.), while dashed red line represents the concentration of the free nanotubes calculated as described in the S3 Information. Model of albumin binding to TiO2-NTs and best fit parameter values. (D) The relative amount of TiO2-NTs covered by the serum proteins (red line) and the relative amount of serum proteins bound to TiO2-NTs (yellow line) are shown as predicted by the model described in the S3 Information. Model of albumin binding to TiO2-NTs and best fit parameter values.
Fig 5.
Stability of the TiO2-NTs in cell medium at different concentrations of FBS.
(A) Rate of sedimentation measured through the absorbance at 400 nm on a UV-VIS spectrometer up to one day. (B) Relative proportion of the remaining TiO2-NTs in dispersion after centrifugation at 15000 rpm (r = 6.2 cm) for 20 min. (C) Average hydrodynamic diameter of the remaining small aggregates and coated single nanotubes in dispersion of the TiO2-NTs after centrifugation.
Fig 6.
Fluorescence quenching due to binding of the bovine serum albumin to fluorescently labelled A-TiO2-NTs.
Intensity of fluorescence emission of Alexa labeled nanotubes excited at 250 nm was measured on spectrofluorimeter. Fluorescently labeled nanotubes (A-TiO2-NTs) were dispersed in water (300 μg/mL) and sonicated on ultrasonic bath, then 100 μL of dispersion was titrated with FBS in 2 μL steps to final volume 150 μL. The analog experiment with 0.6 μM Alexa has been done for reference to prove that the free Alexa has a distinguishable titration curve from the bound Alexa.
Fig 7.
Measurements of hydroxyl radical formations by UV irradiation of the TiO2-NTs.
Spin trap DMPO was used to detect production of the hydroxyl radicals generated by UV irradiated TiO2-NTs. TiO2-NTs were mixed with DMPO spin traps and cell medium with 10% FBS or without FBS. The sample was irradiated for 5 min with UV light (wavelength of 356 nm) or left in dark (control), followed by EPR measurements immediately after the addition of the cell medium. In parallel experiments the dispersion of the TiO2-NTs in the cell medium without or with the serum proteins (FBS) was put in the dark for one hour, than the spin trap DMPO was added and samples were irradiated with the UV light. (A) Representative EPR spectrum of a trapped hydroxyl radical in the presence of the FBS or (B) absence of the FBS. (C) The experimental EPR spectrum was simulated with hyperfine splitting constants: AN = 1.49 G and AH = 1.49 mT typical for OH radical. Simulation was done with EPRSim Wizard [29]. EPR spectrum intensity peak normalized to the experiment with highest intensity peak, of nanomaterial in the cell medium without the FBS and with the FBS.