Fmoc-DOPA(acetonide)-OH, Fmoc-Phe-OH, and Fmoc-His(Trt)-OH were purchased from GL Biochem (Shanghai, China). Triisopropylsilane (TIPS) was purchased from TCI Europe N.V. (Zwijndrecht, Belgium). 1-[Bis(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-b]pyridinium 3-oxide hexafluorophosphate (HATU) was purchased from Matrix Innovation, Inc. (Saint-Hubert, QC, Canada). Dimethylformamide (DMF), dichloromethane (DCM), piperidine, diethyl ether, N,N-Diisopropylethylamine (DIPEA), acetonitrile, and tri-fluoroacetic acid (TFA) were purchased from Bio-Lab (Jerusalem, Israel). Sodium phosphate monobasic monohydrate was purchased from Mallinckrodt Chemicals (Chesterfield, United Kingdom). Difco Nutrient Agar and Difco LB broth, Lennox were purchased from BD Biosciences (San Jose, CA, United States). HCL fuming (37%), tryptic soy broth (TSB), lecithin, tween 80, 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) diammonium salt, peroxidase from horseradish (HRP), and hydrogen peroxide solution (30 wt. %) were purchased from Merck (Darmstadt, Germany). Lysogeny broth (LB) was purchased from Becton Dickinson (BD) (Franklin Lakes, NJ, United States). Agarose was purchased from Lifegene (Mevo Horon, Israel). Sodium dodecyl sulfate (SDS) was purchased from J.T. Baker Inc. (Phillipsburg, NJ, United States). Pierce BCA protein assay kit was purchased from Thermo Fisher Scientific (Waltham, MA, United States). NaCl was purchased from Fisher Scientific (Hampton, NH, United States). D₂O was purchased from Cambridge Isotope Laboratories, Inc. (Tewksbury, MA, United States). NaBH₄, CuCl₂, and CuCl were purchased from Acros Organics (Fair Lawn, NJ, United States). 2-Chlorotrityl chloride resin (1.0–1.6 mmol/g, 100—200 mesh) was purchased from Chem-Impex International, Inc. (Wood Dale, IL, United States). Escherichia coli (Migula) Castellani and Chalmers strain B (ATCC 11303) and FDA strain Seattle 1946 (ATCC 25922) and E. coli bacteriophage T4 (ATCC 113030-B4) were obtained from American Type Culture Collection (Manassas, VA, United States).

Ti substrates: Si wafers with a diameter of 2 inches were coated with 50 nm Ti (as measured by a quartz crystal microbalance) using electron beam evaporation (TFDS-141E, VST) at a rate of 2 Å/sec. Si substrates: Si wafers with a diameter of 10 cm were used. Mica surfaces: mica discs with a 9.9 mm diameter (Ted Pella Inc., Redding, CA, United States) were used. Glass substrates: glass microscope slides (76 mm × 26 mm × 1 mm, Paul Marienfeld GmbH & CO. KG, Lauda-Königshofen, Germany) were diced into 1 × 1 cm² pieces before use (7100 2 in. Pro-Vectus, ADT).

The peptides were synthesized manually by Fmoc solid-phase peptide synthesis (SPPS). 2-Chlorotrityl chloride resin was used (0.25 mmol scale). The amino acids were used with a 5-fold excess, and they were activated using a DIPEA/HATU mixture (4 equiv. and 3.9 equiv., respectively) for 4 min. The coupling procedure was carried for 1 h and a Kaiser test was performed after each coupling. The Fmoc protecting group was removed by stirring with a solution of 20% piperidine in DMF for 20 min. The washing procedure included washing twice with DMF, methanol, DCM, and DMF again. Before the cleavage of the peptide, the resin was washed three times with DMF and DCM. The cleavage reaction was performed using a mixture of TFA/TIPS/TDW (10 ml, 95:2.5:2.5) for 3 h. The cleaved solution was evaporated via nitrogen bubbling, precipitated with diethyl ether, and centrifuged. The crude peptide product was dissolved in an ACN/TDW mixture (5 ml, 1:1) and lyophilized.

Mica disks were cleaved before each use. 1 × 1 cm² Ti or Si surfaces were sonicated for 15 min in ethanol, washed with TDW, and dried with nitrogen. 1 × 1 cm² SiO₂ surfaces were cleaned with Oxygen/Plasma (Atto, Diener Electronic) for 10 min, immersed in SDS (2%) for 30 min, washed with TDW and dried under N₂. Immediately before coating with the peptide, the surfaces were treated by O₂ plasma for 1 min (Atto, Diener Electronic, Ebhausen, Germany). The surfaces were immersed in the peptide solution (1.5 mg/ml, 1.1 mM, in a filtered Tris buffer at pH = 8.5, 10 mM, with an ionic strength of 154 mM with NaCl) for 4 h, at room temperature. The coated surfaces were washed three times with 1 ml of the same Tris buffer solution and dried with nitrogen. After that, the surfaces were immersed in the CuCl₂ solution (0.8 mg/ml, 4.6 mM, in the same Tris buffer solution) for 2 h, at room temperature. The washing step was repeated. Lastly, the surfaces were immersed in an ice-cold NaBH₄ solution (0.3 mg/ml, 9.1 mM in the same Tris buffer) for 1 h, at room temperature, and the washing step was repeated.

The data from the plaque and antibacterial assays represent three independent experiments. In each plaque assay experiment, three uncoated and three coated surfaces were used. Each plaque assay was performed in duplicates and each antibacterial assay was performed in triplicates. The results for the control plaque assay, as well as the assays meant to display the effect of NaCl, were produced from one experiment. All the graphs were plotted using Origin software (OriginLab Corporation). The plaque assay and antibacterial test results of the PCN-coated surfaces were compared using Student’s two-sample t-test assuming unequal variances. The plaque assay results for the control surfaces were compared using one-way ANOVA test with Tukey-Kramer post hoc analysis. Statistically significant results were marked with one asterisk for p < 0.01 (*) and two asterisks for p < 0.001 (**).

We designed the trifunctional peptide NH₂-DOPA-(Phe)₂-(His)₆-OH. This design incorporates three elements: 1) the amino acid DOPA, which provides the peptide with adsorption capabilities to most substrates (Lee et al., 2007); 2) a diphenylalanine segment, which facilitates the peptide’s self-assembly process through π stacking, can aid the adhesion of the DOPA, and stabilize the coacervation of the peptide (Reches and Gazit, 2003); and 3) a hexahistidine part, which can generate coordinate bonds with various metal ions, e.g., Ni²⁺ and Cu²⁺ (Figure 1A) (Liu et al., 2003; Blankespoor et al., 2005). We suggest that the employment of a single peptide sequence with the three functions mentioned can be beneficial in designing an antiviral and antibacterial coating.

The surface modification was performed using a three-step procedure in which the surfaces were immersed in a peptide solution (P), a Cu solution (PC), and an NaBH₄ solution (PCN) to reduce the copper (Figure 1B) (Banerjee et al., 2003). These steps were meant to allow the formation of a peptide layer on the surface, form bonds between the histidine residues of the peptide and Cu ions, and finally, reduce these ions to cover the peptide layer with Cu crystals. Additionally, since DOPA can also bind metal ions, separating the adhesion of the peptide on the surface via DOPA from the addition of Cu is important to avoid competition between DOPA and hexahistidine. The three steps were performed in a Tris buffer solution with an ionic strength of 154 mM with NaCl. After every immersion, the surface was rinsed three times with 1 ml of the buffer solution, to ensure the removal of weakly attached residues and to preliminarily demonstrate the stability of the coating.

The oxidation of DOPA is considered to assist its adhesion by allowing multiple interactions with the surface. To promote this process, the procedure was performed at alkaline conditions using Tris buffer (Lee et al., 2006; Xi et al., 2009; Das and Reches, 2016; Mirshafian et al., 2016; Yuran et al., 2018). Additionally, these conditions were favored because at lower pH values the imidazole groups on the histidine residues are protonated and can no longer bind metal ions (Moyer et al., 2014; Zheng et al., 2016).

The interactions of the salt ions with water could abate the interactions between water and the polar amino acids, i.e., histidine and DOPA. This could encourage the coacervation of the peptide via π stacking interactions between the phenylalanine residues, thus forming a peptide layer more easily in a solution with higher ionic strength (Wei et al., 2014). The effect of the salt on the coacervation of the peptide can be seen in the solutions of the peptide, Cu(II), and their mixture, with and without NaCl (Supplementary Figure S2). While no turbidity can be seen in the solutions without NaCl, the coacervation is clearly visible in the peptide-Cu mixture with NaCl. Additionally, NaCl is used in His-tag purification as it can reduce weak ionic interactions and prevent nonspecific binding, which could be formed in this case between the peptide and the surface or other components in the solution.

The adhesion of the peptide and its ability to bind Cu ions were investigated by QCM-D analysis. The peptide flowed for ∼4 h and caused a substantial change in frequency, which indicated the adsorption of the peptide to the SiO₂ sensor (Figure 1C). After washing the sensor with the buffer solution, only a small change in frequency was observed, which points to the stability of the peptide coating. The changes in dissipation were ∼1 × 10^–6, which suggests the formation of a rigid film (Supplementary Figure S3) and supports the use of the Sauerbrey equation to resolve the adsorbed mass on the sensor (Bordes et al., 2010). The loaded mass/area for the peptide was ∼96 ng/cm². Afterward, a Cu(II) solution was introduced and caused a large change in frequency, which decreased after washing with the buffer. However, a smaller change remained, corresponding to a loaded mass/area value of ∼9 ng/cm². This value is expected given that the Cu ions bound to the peptide have a lower molar mass and it corresponds to a ratio of ∼2 Cu ions per each peptide molecule, as is expected from previous studies (Watly et al., 2014).

To demonstrate the importance of DOPA for the peptide’s adhesion, we synthesized the peptide NH₂-(Phe)₂-(His)₆-OH (without DOPA) and investigated its adhesion to the surface using QCM-D. This peptide caused a small frequency change, which was further reduced by washing with the buffer. The loaded mass/area for this peptide was ∼12 ng/cm², an eighth of the value for the DOPA-containing peptide, which indicates the significant effect of DOPA on the adhesion.

Surfaces coated with the peptide (P), peptide and Cu (PC), and those coated with the peptide and Cu and treated with NaBH₄ (PCN) were analyzed using SEM. The SEM images show a network-like nanostructure of the peptide (Figure 2A). Similar structures were previously received from a hexahistidine peptide’s assembly with Zn ions (Huang et al., 2019). The structure does not differ greatly with the addition of Cu ions but appears rougher after the reduction treatment with NaBH₄. Additionally, structures that appear to be Cu nanoparticles (CuNPs) embedded in the peptide network-like nanostructures were visible on some areas of the PCN-coated surface. EDS analysis showed the presence of the peptide on all three surfaces and of Cu bound to the peptide on the PC-and PCN-coated surfaces (Figure 2B).

To view the differences between a clean surface to areas that are not covered with the peptide nanostructures on a PCN-coated surface, freshly cleaved and PCN-coated mica surfaces were imaged using AFM. The topography of the bare mica was unobservable due to the low signal-to-noise ratio, while that of the PCN-coated surface was distinct (Supplementary Figure S4). The difference can be seen also from the different roughness parameters (Rq) measured for the bare mica surface (0.3 nm) and for the PCN-coated surface (1.3 nm). This indicates that there is a presence of a thin peptide layer on areas that appeared in the SEM analysis to be uncovered.

XPS was used to further examine P-, PC-, and PCN-coated surfaces. The results supported the presence of the peptide and of the Cu on the surfaces (Figure 3A). The reduction step with NaBH₄ did not appear to remove the peptide layer. Additionally, the resulting Cu 2p spectra indicated that the reduction step was successful as it increased the area below the Cu(0) and Cu(I) related peaks from 58 to 92% at the expense of the area below the Cu(II) related peaks (Figure 3B). This also suggested that there is some level of Cu(II) reduction done by the peptide itself, before introducing the reducing agent. In addition, the XPS measurements allow calculating the average thickness of the PCN coating, which was established at 6.5 nm (Razvag et al., 2013).

To obtain additional information on the interactions present in the coating, we performed FT-IR analyses. We measured CaF₂ plates coated with the peptide, a mixture of the peptide and CuCl₂, and a mixture of the peptide, CuCl₂, and NaBH₄. The broad peak at 1,600 cm⁻¹ and the narrower one at 1,606 cm⁻¹ for the PC-and PCN-coated surfaces respectively, could be attributed to the degree of crystallinity of the structures on the plate (Huang et al., 2019). This could be explained due to the formation of Cu crystals on the PC-and PCN-coated plates and is supported by the XPS results, which show a majority of reduced Cu atoms also on the PC-coated surface (Figure 3B). The peak at 1,513 cm⁻¹ in the spectrum of the P-coated plate could be indicative of imidazole ring stretching (Figure 3C). The shift and narrowing of this peak at the spectra of the PC-and PCN-coated plates could point to interactions between the imidazole ring of the histidine residues and the Cu. Additionally, the peak at 1,126 cm⁻¹ in the spectrum of the PCN-coated plate could be associated with the stretching vibration of the CN group in the imidazole (Nivedhini Iswarya et al., 2017). These data support the existence of the peptide’s interactions with Cu ions or with metallic Cu through the histidine residues.

The coating procedure resulted in an optically transparent coating, as expected from the coating’s thickness (Figure 4A). The difference between a clean surface and the coated surface, while not visible to the naked eye, can be revealed in contact angle measurements. A clean glass surface presented a contact angle of 7° and the coated surface showed an increase of the contact angle to 62° (Figures 4B,C). Transmittance spectra of a clean glass surface and a PCN-coated surface were measured and confirmed the transparency of the coating in the visible light range (Figure 4C). A transparent coating could be advantageous for covering touch screens, phone cases and screen protectors, office partitions and dividers, face shields, windows, etc.

To assess the antiviral activity of the PCN-coated surface we performed plaque assays using T4 bacteriophages (ATCC 113030-B4) and E. coli (ATCC 11303) as their host. The number of bacteriophage plaques after incubation on a coated surface, when compared to those that were cast on uncoated surfaces, can provide insight to the coating’s antiviral and virucidal capabilities. As presented in Figure 5A, the coating reduced the number of phages by more than 5 orders of magnitude (100%) when compared to uncoated SiO₂ surfaces. The phages that are incubated on the surfaces are washed and then placed on agar plates. This means that if the surface was simply preventing the adhesion of the phages, they would still be detected as plaques. Thus, these assays reveal that the PCN-coated surfaces are virucidal as they kill the phages.

To identify the necessary factors in the coating for the antiviral activity, P-and PC-coated SiO₂ surfaces, as well as surfaces coated with the peptide and treated with NaBH₄ (PN) were also examined with bacteriophage plaque assays. From these assays, presented in Figure 5B, it appears that the peptide coating by itself could promote the survivability of the phages on the surface. This may be explained due to the emergence of niches in the peptide’s network structure, in which the phages can shelter from the exposed SiO₂ surface. On the other hand, PC-and PN-coated surfaces did not display an explicit reduction of the bacteriophage numbers. The results for these surfaces were not consistent, resulting in relatively large standard deviation values, which strengthens their inferiority to the PCN-coated surface.

The importance of the salt to the formation of a coating with strong antiviral activity can be inferred from the plaque assays of PCN-coated surfaces with and without NaCl in the buffer solutions used in the surface modification steps. Without NaCl, the number of phages was reduced by 39% and with NaCl the number was reduced by 80% (Supplementary Figure S7). These assays were performed on Ti surfaces, with a 3 h incubation time of the surfaces with the phages instead of 24 h.

The PCN-coated surface was tested for bactericidal activity on E. coli (ATCC 25922). The number of bacterial colonies on the coated surface was reduced by more than 6 orders of magnitude (100%) when compared to a clean SiO₂ surface (Figure 5C). As with the bacteriophage plaque assays, these antibacterial assays also reveal the bactericidal activity of the PCN-coated surfaces. These results, when combined with the virucidal activity, reveal the effectiveness of the coating’s antiviral and antibacterial capabilities.

Different Cu materials display a pleiotropic effect as they involve several antiviral and antimicrobial pathways, negating the development of antibiotic resistance. The exact antiviral and antimicrobial mechanism of Cu remains uncertain, but several processes have been proposed (Balasubramaniam et al., 2021). Cu ions were shown to bind bacterial cells, causing membrane depolarization that leads to leaks and ruptures. Another process associated with Cu ions, which is also apparent in CuNPs, involves the production of ROS that can damage the genetic material and the lipid membranes. Specifically, Cu(II) ions have a lesser effect on viruses while the more potent Cu(I) ions were shown to possess significant antiviral abilities (Sunada et al., 2012; Kim et al., 2015). The antiviral and antimicrobial abilities of metallic Cu via contact killing were proposed by Grass et al. to involve four processes: 1) Cu dissolution from the surface causes cell damage, 2) the cellular membrane loses its integrity and its contents, 3) the Cu ions induce ROS formation, and 4) the genetic material undergoes degradation (Grass et al., 2011).

To clarify the origin of the coating’s antiviral and antibacterial activity, we performed an assay that will reveal the release of Cu(I) ions released from the surface. The assay was carried out with BCA that forms a complex with Cu(I) ions, which retains a prominent linear absorption at 562 nm. A calibration curve was made using CuCl solutions with known concentrations and the amount of Cu(I) ions released from the PCN-coated surface after different time periods was determined (Figure 6A; Supplementary Figure S8A). The ratio between the Cu⁺ ions released from the new surface after 24 h, to the number of phages or bacteria placed on the surfaces in the plaque assays and antibacterial assays is ∼2 × 10¹².

In addition, we performed an assay with ABTS, to reveal the production and release of H₂O₂ from the surface. ABTS^2- can be oxidized by the enzyme horseradish peroxidase (HRP) in the presence of hydrogen peroxide to ABTS^•−, which absorbs at 734 nm (Nir et al., 2019). A calibration curve was made using H₂O₂ solutions with known concentrations and the amount of H₂O₂ released from the PCN-coated surfaces after different time periods was determined (Figure 6B; Supplementary Figure S8B). While the amount of H₂O₂ released may not seem high enough to kill the phages or the bacteria in a solution, one should note that the H₂O₂ is released from the surface onto the bacteria or phages that contact the surface. Furthermore, the ratios between the H₂O₂ molecules released from the new surface after 24 h, to the number of phages or bacteria placed on the surfaces in the plaque assays and antibacterial assays are ∼7 × 10¹⁰ and ∼5 × 10¹⁰, respectively.

Absolute values for the release of Cu(I) ions and H₂O₂ should not be concluded from these assays, as the variation between different PCN-coated surfaces is possible. However, the trend of the release of Cu(I) and H₂O₂ is clear. In addition, the possibility of renewing the activity of an old PCN-coated surface is indicated by these assays. A freshly prepared surface released ∼60 nmol Cu(I) and ∼1.7 nmol H₂O₂ after 24 h. After several months, these values decreased to ∼37 and ∼0.3 nmol, respectively. Yet, after repeating the coating steps with the copper and NaBH₄, the old surfaces restored their Cu(I) and H₂O₂ release to ∼75 and 1.4 nmol, respectively. The possibility of reactivating old PCN-coated surfaces without repeating the peptide coating simplifies the real-life employment of this coating and its effectiveness.

Since both Cu(I) and H₂O₂ have well-studied antibacterial and antiviral properties, these results could provide insight to the mechanism of those properties in the PCN coating (Figure 6C) (Sunada et al., 2012; Kim et al., 2015). The inclusion of the potent antiviral and antibacterial abilities of Cu(I) ions and H₂O₂ with the possible dissolution of the ROS-generating, bacteria-binding Cu(II) ions, and the properties of metallic Cu and CuNPs, could result in a strong antiviral and antibacterial coating that kills both viruses and bacteria. Additionally, the release of Cu(I) and H₂O₂ from the old PCN-coated surface, as well as the possibility of renewing these surfaces, strengthen the potential of this coating.