Membrane Mask with Enzyme-Coating for Capture and Deactivation of SARS-CoV-2 Spike Glycoprotein

The current pandemic shows how susceptible global communities are to infectious respiratory viruses and highlights shortcomings of existing facemasks. Our research looked at increasing one’s protection from infection with a membrane-based facemask and enclosed space filter to deactivate SARS-CoV-2.

Like Comment
Read the paper

The current coronavirus pandemic, caused by SARS-CoV-2 (severe acute respiratory syndrome coronavirus 2), has created a severe societal health issue, resulting in significant economic problems across the globe. These novel coronavirus particles (~100-120 nm) are covered in club-shaped “S-protein” spikes, which give it its crown-like, or coronal, appearance. These protein spikes allow the virus to readily enter host cells once in the body via binding to the angiotensin-converting enzyme 2 (hACE2 in humans), resulting in a highly infectious and readily transmissible disease.

The main prevention method from viral spread is personal protective equipment (PPE), specifically respiratory face masks, such as N95 or surgical masks, but many studies have found shortcomings in these commercial options, especially upon capturing coronavirus-sized particles. This innovative project involved the development of layered membrane-based materials that can overcome such mask limitations and increase individual protection from viral infection via two main ways: (1) The precise, asymmetric porous structure of the membrane, as opposed to the non-woven fibers of an N95 mask, can more precisely capture coronavirus-sized particles, and (2) a non-toxic enzyme coating (as opposed to toxic silver nanoparticles) can denature the spike proteins of the virus, rendering it unable to infect its host. This same technique can also be valid for enclosed space air filtration to minimize viral spread.

A microporous commercial poly(vinylidene fluoride) (PVDF) membrane, referred to as PVDF400 (by Solecta Membrane Inc.) was utilized as the basis for the developed material. The difference in structure compared to an N95 mask is apparent, and is highlighted in Figure 1.

Figure 1. SEM of PVDF400 commercial membrane. (a) full thickness cross-section of PVDF400 and (b) PVDF-only layer cross-section. Surface of (c) PVDF layer and (d) polyester support layer. SEM images of N95 layers (e) 1, (f) 2, (g) 3, and (h) 4. In mask orientation, layer 1 is exposed to the open environment and layer 4 is exposed to the inside of the mask. Layer 3 is referred to as the “separating” layer, as it had the highest flow resistance.

Initially, the membrane system performance was verified with polystyrene latex (PSL) particles that were easily aerosolized to mimic the size range of coronavirus particles. After filtration with PSL particles (~100 nm on average), the permeate was analyzed to determine size range of captured particles for each material (Figure 2). The developed PVDF membranes were found to filter more selectively smaller particles (<100-150 nm) than a commercial N95 and surgical mask, validating that the precise pore structure of the membrane plays a large role in coronavirus-sized particle capture. Another important factor of the membrane, though, is the pore tortuosity. Studies from the Centers for Disease Control and Prevention (CDC) have shown that membranes with non-tortuous pores have a much lower filtration efficiency than that of tortuous ones, indicating that the longer path the penetrating particle must take increases the chance of deposition and capture. Overall, Occupational Safety and Health Administration (OSHA) has a protection factor (PF) requirement of at least 10 for N95 masks, while our developed membranes were found to have a PF of 540 ± 380.

Figure 2. PSL aerosol filtration through membrane material. (a) PSL particle air-filtration and permeability drop of unfunctionalized PVDF400 in normal (blue diamond and dashed line) and reverse orientation (orange square and dashed line). This result is from a single sample with <0.5% deviation in triplicate measurements. Feed air concentration was ~37,000 0.3-0.5 μm aerosol particles/L with RH of 21%. Particle count measured using Met One Instruments' GT-526S particle counter. Flow rate measurements normalized at STP. (b) SEM image of Subtilisin-PMAA-PVDF membrane surface after 100-nm PSL aerosol filtration for 52 minutes.

After validating the increased aerosol capture, the membrane system was functionalized with an enzyme for S-protein deactivation. A significant hurdle of this membrane system, though, was maintaining enzymatic activity on minimally hydrated surfaces. Enzymes require some degree of hydration to interact and cleave a substrate, as well as maintain active conformation. Our team tackled this problem from multiple directions. First, subtilisin Carlsberg was selected as the immobilized enzyme, as studies have shown its ability to fold and return to an active conformation upon rehydration. Second, the membrane system was functionalized with poly(methyacrylic acid) (PMAA), as the polymeric hydrogel’s ability to retain moisture/hydration has been shown. The resulting functionalized membrane showed higher enzymatic activity with the presence of PMAA even after 21 days of dry storage in ambient lab conditions, indicating ease of use and storage.

Identifying spike glycoprotein denaturation on a minimally hydrated surface, though, was another challenge of this research. Several methods were initially tried, yet the hydrophilicity of the membrane made it difficult to recover any minute solution on the membrane surface, as well as recover a concentration high enough for differential scanning calorimetry in solution-phase analysis. Through collaboration with other departments at the university, two methods were determined. First, Sypro Orange, a hydrophobic-binding fluorescent dye, was utilized on the remaining hydration of the membrane system upon S-protein exposure (0.02-0.2% hydration by mass of system). After 21 days of dry storage, the functionalized membrane was able to denature S-protein on a minimally hydrated surface with only 30 seconds of exposure time, identified by a statistically significant increase in fluorescent dye signal (from exposed hydrophobic domains of protein) (Figure 3).

Figure 3. Protein denaturation via enzyme interaction identified using Sypro Orange. Average fluorescent emission (RFU) of SGP before and after 30 second exposure to Subtilisin-functionalized (batch) PMAA-PVDF membrane in the presence of hydrophobic-binding fluorescent dye, Sypro Orange. This membrane was functionalized using convective mode and was left in ambient conditions for 21 days after enzyme functionalization before experiment. Analyzed using Synergy H1 Hydrid Reader. 0.3 mg/mL of S-Protein and 50x Sypro Orange was utilized. Minimum hydration of membrane with 1.35 μl of solution per cm2 of membrane surface during denaturation process (about 0.02 % water content). Reactions were at pH of 7.8 and 23°C. Error bars represent the standard deviation of 3 different measurements taken on triplicate samples.

Second, this experiment was repeated but using PSL particles coated with coronavirus S-Protein to investigate the degree of interaction between the functionalized membrane and the particle-bound proteins. The minute volume extracted after 30 seconds was analyzed using gel electrophoresis (Figure 4), which showed a complete disappearance of the full-length 180 kDa S-Protein band. This indicates that, above the limit of detection, the enzymatic coating of the membrane interacts with all the S-protein on coronavirus-sized particles.

Figure 4. Characterization of particle-bound protein denaturation via enzyme interaction of functionalized membrane. Analysis of 100 nm SGP-functionalized PSL after 30 second reaction with different membrane surfaces via Sodium dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE). Approximately 2.6 μg of protein was loaded in each well (except the SGP-only lane with ~5.2 μg) with reactions carried out at total protein concentrations of ~87.5 μg/mL at pH of 7.8 and 23°C. These experiments were performed in duplicate.

Overall, this project allowed for a significant improvement in the efficacy and safety of the filtration mechanisms and (subsequent deactivation parameters) for PPE and for enclosed space filtration. New, even thinner, membrane materials, which incorporate integration of easily-adaptable viral-deactivating agents, can be easily scaled up, as the functionalization methods are simple and do not require special equipment. The developed membranes were able to deactivate the S-Protein of the SARS-CoV-2 virus in realistic situations, indicating a promising technological advance towards the next generation of respiratory face masks and enclosed space virus spread mitigation.

Rollie G. Mills

PhD Candidate, University of Kentucky