TiO2 supported single Ag atoms nanozyme for elimination of SARS-CoV2

Graphical Abstract ga1


Introduction
In December 2019, a new infectious respiratory disease emerged, causing fever, severe respiratory illness, and pneumonia [1,2].
Various studies showed that it is caused by a new severe acute respiratory syndrome β-coronavirus (SARS-CoV2), and hence the disease was named coronavirus disease 2019 (COVID-19) [3]. COVID-19 is the third large-scale pandemic caused by coronavirus in the last two decades after SARS and Middle East Respiratory Syndrome (MERS) [4,5]. The number of laboratory-confirmed COVID-19 cases is already more than 90 times higher than the total number of confirmed cases of SARS and MERS [6]. Thus, SARS-CoV2 is posing a serious threat to global public health. There is an urgent need for specific anticoronavirus therapeutics and prophylactics for treatment and prevention.
Nanomaterials possess considerable potential for antimicrobial application [7][8][9][10]. Recently, Bonam [12,13]. Compared to traditional materials, nanomaterials are superior because of their small size, large specific surface area, high adsorption and catalytic capacity [14]. Many nanomaterials have been reported to have a binding preference for amino acids or peptides. For instance, peptides have been reported to interact with the surface of gold nanoparticles (NPs) through cysteine residues or attach to the surface by a combination of several non-cysteine amino acids (i.e., tyrosine, serine, lysine, aspartic acid) via multidentate binding [15]. Sarikaya et al. used a set of experimentally determined quartz binding peptides with different affinities to quartz (10 strong, 14 moderate, and 15 weak peptide binders to quartz) [16]. Of note, Yacaman et al. designed 1-10 nm Ag NPs, which specifically interact with sulfur-bearing residues on the gp120 glycoprotein of human immunodeficiency virus 1 (HIV-1) [17], indicating that nanomaterials have advantages in antimicrobial application. Importantly, nanomaterials harboring intrinsic enzymelike activity have been identified [18][19][20][21]. These characteristics of nanomaterials prompted us to design a novel material that specifically absorbs to SARS-CoV2 and subsequently eliminate the virus. Among various inorganic materials, Ag and TiO 2 NPs are considered effective antimicrobial agents. Nano-Ag is one of the promising nanomaterial microbicides due to its effectiveness (even at small doses) and minimal toxicity and side effects [22]. Nano-Ag shows high adsorption to microbes and subsequently produces Ag + and reactive oxygen species (ROS), which are responsible for microbial shell breakdown [23][24][25]. Nano-TiO 2 is another biocompatible nanomaterial that can be excited by near-ultraviolet rays and possesses photochemical sterilization and self-cleaning capabilities [26,27]. The photochemical sterilization effects of nano-TiO 2 have been intensively investigated on a wide spectrum of organisms, including virus, bacteria and tumor cells [28][29][30]. Although significant progress has been made in nanomaterial research, improving the biocatalytic sites of nanomaterials poses a major challenge due to the inhomogeneity of traditional nanomaterials.
Single-atom nanozymes (SANs), which are defined as mimic enzyme containing only isolated single metal atoms on various substrates with atomically dispersed metal centers, maximize the atomic utilization efficiency and density of active sites [31]. In addition, the homogeneous structure of SANs facilitates the accurate discernment and characterization of active sites, thus providing deep insight into the construction of desirable SANs for high-performance catalysis at the atomic level in the biological milieu [32,33]. Consequently, the intrinsic physiochemical and catalytic features of SANs make them intriguing candidates for the design of highly efficient biocatalysts and for pioneering new paradigms in promoting beneficial catalytic reactions for desirable biomedical applications. In addition, such atomically dispersed metal catalysts have also found applications in the chemical engineering industry due to their superior catalytic performance compared with corresponding metal nanocatalysts, especially in value-added chemical production, energy transformation and environmental protection [34]. Among these applications, quite a few metal single-atom catalysts have demonstrated intriguing potential for activating O 2 molecules, which offers the possibility to catalyze the formation of singlet O 2 , thus being promising for biomedicines. Our group is devoted to exploring the strategies for synthesizing single-atom catalysts with precise and uniform gram-scale structures, which is key for future applications [35,36]. Many studies have shown that SANs exhibit very strong antibacterial and antiviral effects [37][38][39], which led us to consider SANs as a nanomaterial against SARS-CoV2.
Here, we successfully designed Ag-TiO 2 SAN, and demonstrated that the SAN exhibits excellent anti-SARS-CoV2 activity. This novel SAN exhibits enhanced in vitro absorption with a rate of up to 99.65% for SARS-CoV2 pseudovirus, which highly expresses spike protein of SARS-CoV2 on HIV-1 surface. Theoretical calculations and experimental evidences indicate that Ag atoms of SAN strongly bind to cysteine and asparagine, which are the most abundant amino acids on the surface of spike 1 protein. In vivo, the SAN efficiently promotes virus phagocytosis by macrophages. The intrinsic peroxidase-like activity of Ag-TiO 2 SAN exerts promising antiviral effects in lysosomes. With viral elimination models, we found that Ag-TiO 2 SAN significantly eliminates pseudovirus, which could protect host cells from infection (Scheme 1). In addition, gram-scale SANs can be produced via an energy-and cost-efficient approach. This study provides new insight for employing anti-SARS-CoV2 nanomaterials.

Characterization of Ag-TiO 2 SAN
One gram of Ag-TiO 2 SAN was prepared via a wet chemistry method in one batch, with the loading of Ag determined as 1.0 wt% via inductively coupled plasma optical emission spectrometry (ICP-OES). Anatase TiO 2 with a size of 5-10 nm was selected as the support, due to the uniform crystalline structure and surface arrangement, which is beneficial for precisely conducting sitespecific studies and mechanistic investigations. TiO 2 has a small size advantage over larger ones, resulting in easier passive targeting and larger surface area, which could load more metals or drug molecules. The geometric structure of Ag-TiO 2 SAN was characterized (Fig. 1a). Scanning electron microscopy (SEM) and dynamic light scattering (DLS) experiments were first employed for checking whether the decoration of Ag led to aggregation of TiO 2 ; the results showed that the SAN is uniformly distributed and the size of SAN is 10-15 nm (Fig. 1b-c). As shown in Fig. 1d-e, Ag was atomically dispersed on the TiO 2 support in Ag-TiO 2 SAN, with Ag, Ti, and O elements well distributed across the space. In comparison, when observing Ag NPs, particles with a size of 3-5 nm were detected ( Fig. 1f-g). Further, the Ag-TiO 2 interaction was investigated by Raman and solid ultraviolet−visible (UV-vis) spectroscopy (Fig. 1h). The peak around 450 nm was ascribed to Ag species whose size is larger than 3 nm, further confirming the existence of Ag NPs. Fig. 1i manifests that the characteristic peak of TiO 2 was maintained before and after Ag doping, but the Eg value increased, which is more prominent in SAN. Furthermore, the UV-vis spectra indicate that the bandgaps of TiO 2 decreased obviously in both Ag/TiO 2 (Ag-TiO 2 SAN) and Ag NPs/TiO 2 , with Eg values of 2.99 eV and 2.84 eV, respectively. These observations serve as solid and unambiguous evidence for the strong metalsupport interaction in these two materials, and different electronic structures offer supported Ag materials good potential for outperforming TiO 2 in catalytic reactions.

Ag-TiO 2 SAN exhibits high adsorption ability for SARS-CoV2 pseudovirus
In order to evaluate anti-virus effects of Ag-TiO 2 SAN, we firstly introduced the pseudovirus, which expresses spike protein of SARS-CoV2 on HIV-1 surface and carries luciferase reporter protein internally. We overexpressed human angiotensin converting enzyme 2 (hACE2) in 293T cells (hACE2 OE -293T). The overexpression of hACE2 was confirmed by the mRNA and protein levels (Figure S1a-c). In our in vitro viral infection assay, the ability of pseudovirus to infect hACE2 OE -293T was obviously higher than wild type 293T cells (hACE2 WT -293T, Figure S2).
To determine the proper concentrations of Ag-TiO 2 SAN, nano-TiO 2 and nano-Ag for the experiments, we evaluated the viability of cells treated by these nanomaterials in the concentration range from 0 to 300 μg mL −1 for 72 h. The results showed that cell viability was more than 95% when treated with 1 wt% Ag loaded SAN, nano-TiO 2 , or nano-Ag at a concentration of 50 μg mL −1 ( Figure S3a). Cell integrity was not affected at this concentration ( Figure S3b). In addition, we evaluated the dissolution of Ag during the storage of 1 wt% Ag loaded SAN. ICP-MS result showed that the silver content in Ag NPs solution (10 mg mL −1 ) was 11.4 ± 1.25 ng mL −1 , while the silver level (1.0 ± 0.46 ng mL −1 ) in Ag-TiO 2 SAN solution (10 mg mL −1 ) was much lower due to Ag being only 1 wt% of the SAN ( Figure S4), indicating the limited toxic of Ag-TiO 2 SAN.
Next, we analyzed the adsorption ability of 0.2 wt% and 1 wt% Ag atom loaded SANs for SARS-CoV2 pseudovirus. The result showed that 1 wt% Ag loaded SAN exhibits higher adsorption of pseudovirus compared with 0.2 wt% Ag loaded SAN (Fig. 2a). Thus, the high adsorption of 1 wt% Ag loaded SAN is due to surficial Ag atoms ratio. We also analyzed the adsorption ability of 1 wt% Ag loaded SAN, nano-TiO 2 and nano-Ag for pseudovirus. Surprisingly, 1 wt% Ag loaded SAN exhibits 99.65% adsorption ability for pseudovirus compared with 82.71% for nano-TiO 2 and 90.62% for nano-Ag (Fig. 2b-c). Consistently, 50 μg mL −1 Ag-TiO 2 SAN (1 wt% Ag) also exhibited promising adsorption ability for spike 1 protein of SARS-CoV2 ( Figure S5). Thus, 1 wt% Ag loaded SAN was used for further experiments. Finally, we confirmed the direct interaction between the SAN and pseudovirus by SEM. Interestingly, we observed that the Ag-TiO 2 SAN (smaller particles) was absorbed to the pseudovirus (bigger particles) surface (Fig. 2d-e). In the skin test assay, Ag-TiO 2 SAN and nano-TiO 2 have no influence on mouse skin. However, nano-Ag severely damages the cuticle of mouse skin ( Figure S6), suggesting that Ag-TiO 2 SAN has a broad application in viral adsorption, such as face mask coating and disinfectant.

Ag-TiO 2 SAN interacts with spike protein of SARS-CoV2
Spike protein is the outermost structure on SARS-CoV2, and is responsible for binding to host cells. Furthermore, 1 wt% Ag loaded SAN exhibited better adsorption of pseudovirus compared with 0.2 wt% Ag loaded SAN. Thus, we speculate that this adsorption is partially dependent on the interaction of Ag atoms and spike protein.
To test this hypothesis, we incubated RBD of spike 1 protein (S1 RBD, 51 KD) with Ag-TiO 2 SAN, nano-TiO 2 , nano-Ag and a mass ratio as 1:1 mixture of nano-TiO 2 /Ag, and collected the supernatant for cell immunofluorescence assay. As shown in Fig. 3a-b, Ag-TiO 2 SAN exhibited higher adsorption of S1 RBD than nano-TiO 2 , nano-Ag and the nano-TiO 2 /Ag mixture. To confirm this result, we measured the S1 RBD levels in precipitates and supernatants of nanomaterial incubated by western blot. The results show that Ag-TiO 2 SAN binds to more S1 RBD than nano-TiO 2 and nano-Ag (Fig. 3e). The opposite results were detected in the supernatant (Fig. 3f). In addition, we performed immuno-precipitation assay to evaluate the ratio of the SANs that are bound to S1 RBD. The results show that 0.85 ± 0.12 μg SANs bound to 1 μg S1 RBD ( Figure S7).
Spike protein of SARS-CoV2, unlike that of other β-B coronaviruses, harbors a special S1/S2 furin-recognizable site, indicating that its spike protein might possess some unique infectious properties. Lu et al. reported that a typical syncytium structure naturally formed by SARS-CoV2 infected cells, which is mediated by spike protein [40]. To investigate whether Ag-TiO 2 SAN interferes with spike-mediated cell-cell fusion, we transfected 293 T cells with spike protein-GFP reporter as the effector cells ( Figure S8a-c) and used hACE2 OE -293T as the target cells. We first sorted GFP positive cells and incubated with Ag-TiO 2 SAN, nano-TiO 2 , nano-Ag, and nano-TiO 2 /Ag mixture (mass ratio as 1:1). After effector cells and target cells were cocultured at 37 °C for 2 h, the fused cells were at least twice large as normal cells. In the Ag-TiO 2 SAN treated group, spike-mediated cell-cell fusion was significantly inhibited compared with the PBS treated group, even though nano-TiO 2 , nano-Ag, and nano-TiO 2 /Ag treatment also inhibit cell-cell fusion to different degrees (Fig. 3c-d), indicating that the SAN serves as a blocker for virus entering into host cells. To directly demonstrate the interaction between Ag-TiO 2 SAN and S1 RBD of SARS-CoV2, we cocultured S1 RBD (2 μg) and Ag-TiO 2 SAN (50 μg mL −1 ) at room temperature for 1 h. Energy disperse spectroscopy (EDS) analysis showed that the spike-Ag-TiO 2 SAN complex was formed (Fig. 3g), suggesting a direct interaction between Ag-TiO 2 SAN and spike protein of SARS-CoV2 takes place.
Human angiotensin converting enzyme 2 (hACE2) is considered to be one major receptor of spike protein mediating SARS-CoV2 infection to host cells [41]. However, whether the adsorption by Ag-TiO 2 SAN of SARS-CoV2 could interfere with the interaction between S1 RBD and hACE2 remains unclear. To answer this question, we performed pulldown assay in vitro. Interestingly, the ability of S1 RBD binds to hACE2 is significantly inhibited by Ag-TiO 2 SAN (Fig. 3h). Furthermore, enzyme linked immunosorbent assay (ELISA) were performed to further evaluate this inhibition. As shown in Figure S9a, the interaction of Ag-TiO 2 SAN and S1 RBD is negligibly affected by serum proteins. Importantly, compared with nano-TiO 2 and nano-Ag, the SAN exhibited the best blocking effect for binding of S1 RBD and hACE2 ( Figure S9b). Together, these data demonstrate that Ag-TiO 2 SAN possesses high affinity for S1 RBD of SARS-CoV2, and this adsorption protects host cells from SARS-CoV2 infection.

Peroxidase-like activity and kinetic analysis of Ag-TiO 2 SAN
Nanomaterials can catalyze the production of high levels of reactive oxygen species (ROS), depending on intrinsic nanozyme activity, which is another approach to achieve antimicrobial effects [38,42]. Therefore, we next tested whether Ag-TiO 2 SAN possesses peroxidase-like activity, catalyzing the conversion of hydrogen peroxide (H 2 O 2 ) into hydroxyl radicals. As shown in Fig. 4a and Figure  S10a-c, Ag-TiO 2 SAN effectively catalyzed the oxidation of 3,3′,5,5′tetramethylbenzidine (TMB) in the presence of H 2 O 2 in a concentration, pH and temperature-dependent manner. The optimal pH and temperature were approximately 5.0 and 35 °C, respectively, which are similar to those of natural horseradish peroxidase (4.5 and 35 °C). In addition, the optimal pH and temperature of Ag-TiO 2 SAN are similar to those of nano-Ag, suggesting the peroxidase-like activity of Ag-TiO 2 SAN is mainly dependent on Ag atoms. Of note, 1 wt % Ag-doped Ag-TiO 2 SAN exhibited higher peroxidase-like activity than traditional nano-Ag, suggesting the efficient utilization of SAN. To analyze the catalytic activity of Ag-TiO 2 SAN, we calculated the Michaelis-Menten constant (K M ) (Fig. 4b-c). K M , which represents the affinity of substrates to the enzyme. The K M value was 39.9 mM for H 2 O 2 and 0.032 mM for TMB (Table S1).
To the best of our knowledge, hydroxyl radicals can cause damage to viral components, inactivating the viral proteins and nucleic acids [43][44][45]. Therefore, it is important to determine the generation of hydroxyl radicals by Ag-TiO 2 SAN. We added H 2 O 2 to the Ag-TiO 2 SAN system and measured the amounts of hydroxyl radicals by electron spin resonance (ESR) (Fig. 4d). Based on these results, we determined the ability of Ag-TiO 2 SAN to generate hydroxyl radicals. Terephthalic acid (TA) was used to indirectly detect the concentration of hydroxyl radicals. As shown in Fig. 4e, the generation of hydroxyl radicals is increased with time, even though the pH was close to neutral. These results suggest that Ag-TiO 2 SAN could steadily induce the generation of hydroxyl radicals by catalyzing the decomposition of H 2 O 2 .

Pharmacokinetics and biocompatibility of intravenously injected Ag-TiO 2 SAN
The pharmacokinetics of the intravenously injected sequential Ag-TiO 2 SAN were assessed to reveal its in vivo behavior. Initially, the time-dependent biodistribution of Ag-TiO 2 SAN in the major organs was studied. The results show that Ag-TiO 2 SAN is mainly distributed in liver, spleen, and lung tissue because of the capture by the reticuloendothelial and vascular system (Fig. 5a). In the blood-circulation experiment, a circulating half-life in the bloodstream of 3.76 h was obtained (Fig. 5b).
Toxic side effects for normal organs have been the main problem in the application of nanomaterials. Therefore, we first tested the biocompatibility of Ag-TiO 2 SAN. The results show no significant difference in mouse body weight after persistent treatment with PBS (control) and Ag-TiO 2 SAN (2-10 mg kg −1 ) for 24 days (Fig. 5c). Furthermore, the bleeding and clotting times were not affected by 10 mg kg −1 Ag-TiO 2 SAN (Fig. 5d-e). To investigate whether Ag-TiO 2 SAN inflicts damage to organs, histological examination (HE) staining of various major organs (heart, liver, spleen, lung, and kidney) from mice injected with PBS and Ag-TiO 2 SAN was conducted to investigate the biological toxicity (Fig. 5f). No obvious pathological abnormality or inflammation was observed from both groups. Overall, these results firmly show the in vivo biocompatibility of Ag-TiO 2 SAN, potentiating its further application as antiviral agent.

Ag-TiO 2 SAN eliminates SARS-CoV2 pseudovirus in vivo
Abundant proteins in body fluids/blood such as albumin may affect the in vivo adsorption ability of Ag-TiO 2 SAN for virus particles. To exclude these effects, the SAN was treated with human serum/S1 RBD and PBS/S1 RBD mixture, respectively. We found that the SAN remained exhibiting high adsorption of S1 RBD in the presence of serum (Fig. 6a). Moreover, the interaction between Ag-TiO 2 SAN with S1 RBD, or pseudovirus was negligibly affected by serum proteins, even in the case of directly mixing with serum ( Figure S11a-b). Furthermore, the peroxidase-like activity was negligibly affected by serum proteins ( Figure S12). With antiviral assay in vitro, we found that the residual virus in supernatant of virus/SAN/H 2 O 2 significantly decreased compared with those of only virus and virus/H 2 O 2 groups. Moreover, antiviral activity of Ag-TiO 2 SAN was not affected by serum ( Figure S13a). Similarly, we cocultured S1 RBD, SAN and H 2 O 2 , with or without serum for 1 h at room temperature. The results showed that S1 RBD was significantly degraded regardless of the presence of serum or not ( Figure S13b). These results suggest that the anti-viral effect of Ag-TiO 2 SAN is negligibly affected by serum proteins.
To better understand the binding characteristics of Ag-TiO 2 SAN and S1 RBD, X-ray photoelectron spectroscopy (XPS) was conducted to compare the binding energies of N 1s and S 2p electrons. For N 1s, the peak locations of S1 RBD ( Figure S14a) were unchanged after incubation with nano-TiO 2 ( Figure S14b) and Ag-TiO 2 SAN ( Figure  S14c). Of note, XPS results of S 2p help to characterize the interaction between Ag-TiO 2 SAN and S1 RBD. Compared with the XPS of S1 RBD or S1 RBD-TiO 2 complex ( Figure S14d-e), the electronic state of S changed only with the S1 RBD-SAN complex ( Figure S14f). The intrinsically strong coordinating interaction between Ag and S assisted the active targeting of Ag-TiO 2 SAN toward the SARS-CoV2 pseudovirus. The interaction between Ag and S is consistent with other reports [46,47]. To further confirm the specific combination between Ag-TiO 2 SAN and S1 RBD, we compared the distribution of superficial amino acids on S1 RBD and serum albumin (67 KD). As shown in Fig. 6b, the percentages of asparagine, glycine, and cysteine are higher on the S1 RBD surface than on the albumin. Subsequently, we investigated the binding conformations of Ag-TiO 2 SAN and the side chains of asparagine, glycine, and cysteine, then calculated their binding energies. In this study, the experimental pH was 7.45 (1% PBS), which is higher than the isoelectric points of asparagine (5.41), glycine (5.97), and cysteine (5.02). Thus the −COOH groups are present in their anion forms and the amino acids are charged with one negative charge. The Ag@Ti 6 O 9 (OH) 6 cluster (Fig. 6c) was used to simulate the NPs with Ag atoms doped on the anatase TiO 2 , which was also used in former reports [48]. The binding energies between the side chains of asparagine (-CH 2 C(O)NH 2 ), glycine (-H), and cysteine (-SH) with Ag@Ti 6 O 9 (OH) 6 are −11.62, −1.43 and −26.60 kcal mol −1 , respectively, indicating that Ag atoms have the strongest interaction with cysteine because of the strong Ag-S bond. Furthermore, asparagine has medium binding energy because the N-atom of the amide has a lone pair of electrons, which forms a coordination bond with the Ag atoms. Glycine, with the -H side chain, has weak interactions with Ag atoms (Fig. 6d). These data provide solid evidence for the dominant affinity between Ag-TiO 2 SAN and S1 RBD.
SARS-CoV2 infects lung tissues through the respiratory tract, then enters into bloodstream to infect hACE2 + tissues [49]. Thus, we used two models, lung administration and tail injection of the pseudovirus, to evaluate the therapy efficacy of Ag-TiO 2 SAN at different stages of SARS-CoV2 infection. hACE2 transgenic mice were used to evaluate the anti-viral activity of Ag-TiO 2 SAN in vivo. The gene expression of hACE2 was confirmed with qPCR analysis ( Figure S15a). With Cy5.5 labeled S1 RBD and SARS-CoV2 pseudovirus, we detected high levels of S1 RBD and SARS-CoV2 pseudovirus in the lung tissues of hACE2 transgenic mice from 0 to 24 h. In contrast, the infection of S1 RBD and pseudovirus significantly decreased when lung tissues were pretreated with Ag-TiO 2 SAN. Importantly, the S1 RBD and pseudovirus are almost completely eliminated after 24 h (Fig. 7a−d). In addition, the results of tail injection of the pseudovirus showed that Ag-TiO 2 SAN is effectively anti-SARS-CoV2 in vivo ( Figure S15b-c), indicating that the SAN is effective antiviral activity both in the early and late stage of SARS-CoV2 infection.
To understand the mechanisms underlying the antiviral effects of Ag-TiO 2 SAN, we tracked the biological location of SAN in vivo. As known, macrophages are strategically located throughout the body tissues and liquids, where they non-specifically ingest and process foreign materials [50]. During pathogen infection, macrophages kill or degrade microorganisms by producing ROS and reducing the intracellular pH value [51,52]. Furthermore, many studies have shown that the induction of ROS production by various nanomaterials depends on the acidic environment of lysosomes in vivo [53,54]. In order to investigate whether Ag-TiO 2 SAN enhances SARS-CoV2 pseudovirus uptake by macrophages, a viral phagocytosis assay was performed. The mouse macrophage cell line RAW264.7 and the human macrophage cell line THP-1 were induced to M1 phase macrophages. As shown in Figure S16a, the adherent ability of M1 macrophages was increased. In addition, the markers of M1 macrophages were also upregulated compared with M0 macrophages (Figure S16b-c). In the phagocytosis assays, the concentration of SARS-CoV2 pseudovirus in the culture supernatant of mouse and human macrophages was significantly decreased in Ag-TiO 2 SAN treated group than PBS control (Fig. 7e-f), indicating that Ag-TiO 2 SAN promotes SARS-CoV2 pseudovirus uptake of macrophages and subsequent degradation. Finally, in order to understand the mechanisms underlying the anti-SARS-CoV2 effects of Ag-TiO 2 SAN in vivo, we tracked the intracellular localization of Ag-TiO 2 SAN by transmission electron microscope (TEM). Fig. 7g-h shows that Ag-TiO 2 SAN colocalized with lysosomes of mouse and human macrophages. More importantly, flow cytometry analysis showed that Ag-TiO 2 SAN significantly increased ROS generation in macrophages ( Fig. 7i-j), and this result was confirmed by immunofluorescence assay ( Figure S17). Thus, Ag-TiO 2 SAN initially promotes virus uptake by macrophages. Under acid condition of lysosomes of macrophages, the SAN subsequently induces high level of ROS to degrade virus.
According to our results and other studies [51,55], macrophages possess a higher tolerance to ROS than other cells. In this antiviral system, Ag-TiO 2 SAN efficiently eliminated SARS-CoV2 pseudovirus in lysosomes of macrophages. However, this effect did not affect the viability of macrophages. The pseudovirus was degraded by the SAN depending on its intrinsic peroxidase-like activity in lysosomes of macrophages. However, the concentration of Ag-TiO 2 SAN converts free radicals is not enough to kill host cells. Cellular viability is also not affected by other nanomaterials with peroxidase-like activity at appropriate concentrations [56,57].

Conclusion
In this study, we designed a TiO 2 supported single atom nanozyme containing 1 wt% atomically dispersed silver atoms (Ag-TiO 2 SAN), which could serve as a highly efficient anti-SARS-CoV2 nanomaterial. First, Ag-TiO 2 SAN displays a high adsorption ability for S1 protein of SARS-CoV2. The pull-down assay suggested that Ag-TiO 2 SAN significantly inhibits the interaction between S1 RBD and its receptor hACE2. By theoretical calculation, we found the interaction between Ag atoms and asparagine and cysteine of S1 RBD. This combination of Ag atoms and amino acids explains the mechanism of high adsorption of SARS-CoV2 pseudovirus by the SAN and inhibition of S1 RBD binding to hACE2. Meanwhile, this nanomaterial exhibits limited toxicity due to its TiO 2 carrier and low Ag atom load, suggesting Ag-TiO 2 SAN can be applied in vivo as a nanozyme-based biosafety material. Second, we found that Ag-TiO 2 SAN possesses peroxidase-like activity, producing high level of ROS under acid condition. Third, Ag-TiO 2 SAN enhanced phagocytosis of SARS-CoV2 pseudovirus in macrophages and colocalized with lysosomes, including the production of high levels of ROS due to its peroxidase-like activity. As a result, SARS-CoV2 pseudovirus is broken down. Finally, with viral elimination models, Ag-TiO 2 SAN