Inactivation and spike protein denaturation of novel coronavirus variants by CuxO/TiO2 nano-photocatalysts | Scientific Reports – Nature.com
Preparation and characterization of CuxO/TiO2
The colour of the rutile-based CuxO/TiO2 suspension was a greenish gray (Fig. 1a), and its difference spectrum after the deposition was characterized by an absorption band at 400 − 500 nm and a broad peak at ~ 800 nm (Fig. 1b). The latter broad peak suggests that the suspension contains excess Cu2+ ions. Since the former absorption band appears to be due to a semiconductor, we examined Tauc plots and obtained the band-gap values of ~ 3.0 and ~ 2.8 eV on the assumption of the direct and indirect transitions, respectively. Because Cu2O and CuO have been reported to have direct allowed transition band-gap of 2.1 − 2.6 eV and indirect allowed transition band-gap of 1.2 − 1.6 eV, respectively6,7,8,9,10, we conclude that the optical behaviour observed in the short wavelength range, from which absorption of TiO2 has been excluded, is attributed mainly to Cu2O. The wider band-gap in comparison with bulk Cu2O could be due to the quantum-size effect, as discussed later. Since the CuxO contains Cu2O, it is expected to exhibit an antivirus effect.
Colour and spectral changes of the photocatalysts. (a, c) Colour changes of the (a) rutile- and (c) anatase-based photocatalyst suspensions after leaving in the dark and under irradiation with simulated solar light. (b) Spectra of the as-prepared photocatalyst suspensions. (d–g) Photographs of the (d, f) rutile- and (e, g) anatase-based photocatalyst coating (d, e) with or (f, g) without white pigments. (h) Spectral changes of the rutile- and anatase-based coatings in the dark and under illumination (fluorescent light, > 380 nm, 500 lx).
The anatase-based suspension showed a brownish gray colour (Fig. 1c) and an absorption band at 400 − 600 nm (Fig. 1b). The corresponding Tauc plots show that the band-gap is ~ 2.9 eV for direct transition and ~ 1.7 or ~ 2.3 eV for indirect transition. The optical behaviour could therefore be ascribed to both Cu2O and CuO.
We also subjected the rutile-based suspension to STEM and EDS analyses after thorough evaporation of water from the suspension (Fig. 2). The STEM image (Fig. 2a) shows that the primary size of the nanoparticles is smaller than 10 nm. As a result of elemental mapping based on STEM-EDS analysis (Fig. 2d–f), we found that the major component was TiO2 nanoparticles of ~ 4 − 8 nm in size, and that clusters of Cu compounds (1 − 2 nm or less) were deposited on TiO2. High resolution HAADF-STEM analysis proved that the TiO2 nanoparticles were in rutile phase and that the Cu compound clusters contained both Cu2O and CuO (Fig. 2b, c). The quantum-sized Cu2O11,12 justifies its widened band-gap of ~ 3.0 eV mentioned above due to a quantum-size effect. Small particles generally give large specific surface area, high optical transmittance and good suspendability. Actually both rutile- and anatase-based suspensions showed no sedimentation for at least 1 year.
Photocatalyst nanoparticles. (a–d) HAADF-STEM images of the photocatalysts. (e, f) STEM-EDS elemental mapping images for (e) Ti and (f) Cu.
When the suspensions were left in the dark, their colour was gradually changed to green in 24 h (Fig. 1a, c), suggesting that Cu2O was oxidized to Cu2+ by dissolved oxygen (Fig. 3, Process A). However, when we irradiated the oxidized suspensions with light from a solar simulator for 24 h, their colour was changed again to greenish gray (Fig. 1a) and brownish gray (Fig. 1c) for rutile- and anatase-based photocatalysts, respectively, indicating that Cu2+ was reduced back to Cu2O. This reduction can be explained in terms of photo-induced interfacial charge transfer from the TiO2 VB to Cu2+ at the TiO2 surface (Fig. 3, Process B)2 Resultant holes in the TiO2 VB should be consumed by oxidation of water to oxygen. Electrons in the TiO2 VB could also be excited to the conduction band (CB) under the simulated solar light, which contains weak UV light, and the excited electrons could also contribute to the reduction of Cu2+ (Fig. 3, Process C). In the case where the mol ratio of Cu to Ti was 1/100 or lower, a small absorption peak was observed at ~ 580 nm after irradiation of the anatase-based photocatalyst. This could be due to localized surface plasmon resonance (LSPR) of over-reduced, metallic Cu nanoparticles. Plasmonic metal nanoparticles in contact with TiO2 inject electrons to the TiO2 conduction band, and metal is oxidized to metal ions, in the case of Ag or less noble metals13,14 including Cu15 (Fig. 3, Process D).
Photoinduced chemical processes involved in the present photo-renewable system. (A) Aerobic oxidation of Cu(I) to Cu(II). (B) Photo-induced interfacial charge transfer from the TiO2 VB to Cu(II). (C) Photo-excitation of electrons in the TiO2 VB to CB. (D) Plasmonic excitation of over-reduced, metallic Cu nanoparticles, which inject electrons to the TiO2 CB. Process B is the major photo-process and Processes C and D are minor processes.
Photocatalytic coatings
Either (rutile or anatase) of the CuxO/TiO2 suspensions was added to a latex paint containing inorganic white pigments and organic binders, followed by 5-min stirring and 1-min degassing. Each photocatalyst slurry thus obtained was applied onto a glass plate (0.01 mL cm-2). Solvent in the coating was evaporated at 25 °C for 7 days, and white films were obtained (Fig. 1d, e). Colourless translucent films without the white pigment (Fig. 1f, g) were also prepared for spectroscopic measurements. The small particle size of CuxO/TiO2 is responsible for the good transmittance. Their average thickness was 50 µm for both photocatalyst films.
After preparation of the coatings, those were left in the dark. As a result, their absorption in the visible wavelength range was decreased gradually (Fig. 1h). In marked contrast, the absorption was gradually increased under irradiation with fluorescent light (< 380 nm light was cut off). The peak at 440 nm reflect photo-induced interfacial charge transfer from the TiO2 VB to Cu2+16,17. The absorption decrease in the dark and the increase under illumination can be explained in terms of aerobic oxidation (Fig. 3A) and photocatalytic re-reduction (Fig. 3B), respectively, of CuxO clusters. Since Cu+ in CuxO exhibits antivirus effects, the photocatalysts are expected to retain their antivirus activities, if any, under fluorescent light, even in the paint coatings.
Antivirus effects of the coatings
The photocatalyst paint coatings were subjected to inactivation tests against novel coronavirus (SARS-CoV-2, original variant) according to the procedures given in International Organization for Standardization ISO 21,702 with some modifications. Coronaviruses in 5% FBS DMEM medium (25 µL) were applied onto a glass plate (20 × 20 mm) coated with the rutile- or anatase-based photocatalyst coating. The plate was covered with a polypropylene film of the same size and was incubated under fluorescent lamp illumination (1000 lx) for 3 h, followed by evaluation of viral infectivity V (in pfu mL-1) by a plaque assay. Figure 4a shows the logV values together with those for control experiments in which a bare glass plate or a glass plate coated with the paint without photocatalyst was used instead of the photocatalytic plate. The rutile-based coating strongly inactivated the novel coronavirus and the obtained viral infectivity was lower than the detection limit of 5 pfu. Its antivirus activity [= (logV)ave – (logV0)ave, where V0 is viral infectivity for a bare glass plate and subscript ave stands for averaged values) is 3.8 or higher. The coronavirus was also inactivated by the anatase-based coating, whereas its antivirus activity was lower (2.0). This difference should be due to the lower interfacial charge transfer absorption band at 440 nm for the anatase-based photocatalyst in comparison with the rutile-based one (Fig. 1h). The larger band-gap of anatase TiO2 (3.2 eV) than rutile (3.0 eV), which lowers the contribution of the Cu2+ reduction pathway via the TiO2 CB (Fig. 3C), may also be responsible for the activity difference.
Antivirus effects of the photocatalyst coatings on novel coronaviruses. Viral infectivity (V) values for (a) the original and (b) alpha variants of novel coronavirus after incubation under fluorescent light (1000 lx for 3 h) are shown. Raw data are summarized in Table S1 in Supplementary Information.
Next we examined an antivirus effect of the rutile-based coating on alpha variant (also known as lineage B.1.1.7 or VOC-202012/01) of the novel coronavirus, which has mutations including N501Y (Fig. 4b). Its antivirus activity was 3.0; the photocatalyst coating is also effective against alpha variant. Possible mechanisms of inactivation by the photocatalysts are discussed at the end of this section.
We also subjected the photocatalyst coatings to antivirus assays for bacteriophage Qβ by Kitasato Research Center for Environmental Science and assays for bacteriophage M13 according to the procedures given in Japanese Industrial Standard JIS R1756 and those in Ref.18, respectively. Figure 5a shows the results for bacteriophage Qβ after fluorescent lamp irradiation (500 lx) for 4 h. The antivirus activities of the rutile- and anatase-based coatings were ≥ 5.0 and 3.4, respectively. The activity values were lowered to 1.6 (rutile) and 0.1 (anatase) for the control experiments without illumination (Fig. 5b). Those results show that the photo-renewing effect is very important to keep the high antivirus activities. On the other hand, the rutile-based photocatalyst keeps the significant antivirus activity even in the dark, reflecting that Cu2O remaining in the coating causes the antivirus effects because Cu2O has been known to inactivate bacteriophage Qβ19,20. In addition, both of the rutile- and anatase-based photocatalysts exhibited high inactivation effects on bacteriophage M13 (Fig. 5c); the antivirus activities were 3.6 and 5.0, respectively.
Antivirus effects of the photocatalyst coatings on bacteriophages. Viral infectivity (V) values for bacteriophage Qβ (a) after incubation under fluorescent light (500 lx for 4 h) or (b) in the dark and (c) those for bacteriophage M13 after incubation under fluorescent light (500 lx for 24 h) are shown. Raw data are summarized in Table S1 in Supplementary Information.
Antivirus effects on influenza A virus and feline calicivirus, which is often used as a surrogate for norovirus because of the similarity in terms of a capsid-enveloped structure, were also investigated in the dark by Kitasato Research Center for Environmental Science, according to the protocol of ISO 21702. The coatings were subjected to assay 18 days after the synthesis of the photocatalysts. We observed significant inactivation effects of the rutile- and anatase-based photocatalyst coatings after 24-h incubation, and the antivirus activities were > 3 for influenza A virus, and > 4 for feline calicivirus. The remaining Cu2O may also be effective for those viruses.
In previous studies, it has been shown that Cu2O inactivates influenza virus through denaturation of hemagglutinin20, which is a protein at the virus surface and binds to glycans with terminal sialic acid on host cells. We infer that Cu2O might also attack surface proteins of the novel coronaviruses, the bacteriophages Qβ and M13 and feline calicivirus. In the case of the novel coronavirus, it is known that there are spike proteins at the virus surface. The spike proteins bind ACE2, which is a receptor protein at the host cell surface, and lead to infection21. We therefore examined possible effects of the present photocatalyst on the binding ability of the spike proteins to human ACE2, in the following section.
Effects on spike proteins of novel coronavirus
We investigated if the photocatalysts containing Cu2O developed in this study denature the surface spike protein of novel coronavirus and thereby suppress its infectivity. To assess this, we prepared RBD of spike protein derived from the original, alpha and delta variants of novel coronavirus, and evaluated the denaturation effects of the photocatalysts on the protein. Since the binding of RBD to human ACE2 is an essential step in the infection21, the 5 µM recombinant RBD protein (70 µL) was mixed with the photocatalyst suspensions (70 µL) and left for 2 h at 4 °C and the binding activity toward ACE2 was examined by ELISA according to literature22. As shown in Fig. 6, the binding activity of RBD to ACE2 was significantly diminished by the incubation with the photocatalysts, indicating that the photocatalysts denatured the RBD protein. These results strongly support our conclusion that the antiviral effect of the photocatalyst on coronavirus relies, at least in part, on the protein denaturing activity. Importantly, RBD proteins derived from alpha and delta variants were also inactivated as was the original variant, indicating that the photocatalysis is effective to denature the spike protein regardless of mutation. It was suggested that the denaturation is due to disorder of electrostatic interaction in protein20. Further study to reveal the denaturing process of the proteins will provide a strategy to develop photocatalysts with even higher antiviral activity.
Denaturation effect of the rutile- and anatase-based photocatalysts on RBD domain of spike protein from (a) the original, (b) alpha and (c) delta variants of novel coronavirus. The binding activity of RBD to human ACE2 protein at various concentrations was assessed by ELISA (n = 3). Raw data are summarized in Table S2 in Supplementary Information.
