Nanoparticles with adaptable self-assembled shells sense and inhibit influenza virus at ultra-low concentrations

doi:10.21203/rs.3.rs-2836718/v1

Download PDF

Article

Nanoparticles with adaptable self-assembled shells sense and inhibit influenza virus at ultra-low concentrations

https://doi.org/10.21203/rs.3.rs-2836718/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Multivalent inhibitors that mimic the polysaccharide array on cells represent a new paradigm in the development of antiviral agents and antibiotics. Covalent ligand anchoring limits the affinity and in turn potency of these inhibitors with dissociation constants (Kd) commonly found in the micromolar or upper nanomolar range. Addressing this deficiency we here report on easily accessible gold core-shell nanoparticles (rSAM-NPs) featuring adaptable reversible self-assembled monolayer (rSAM) -based shells. The rSAMs are anchored by non-covalent amidinium-carboxylate interaction on gold nanoparticles at slightly alkaline pH resulting in laterally mobile pH-responsive assemblies that are functional at physiological pH. Introducing sialic acid ligands in the shell, we show that the rSAM-NPs strongly interact with the influenza virus surface protein hemagglutinin (limit of detection LoD < 2 nM) and deactivated bird flue virus (LoD < 1 HAU) in allantoic liquid. Finally, we show that the rSAM-NPs effectively inhibit the interaction of the virus with red blood cells at concentrations in the low picomolar range. This represents a more than 1000-fold increased potency with respect to monosaccharide-based multivalent inhibitors.

Physical sciences/Chemistry/Supramolecular chemistry

Biological sciences/Microbiology/Infectious-disease diagnostics

Physical sciences/Nanoscience and technology/Nanomedicine

Biological sciences/Biological techniques/Sensors and probes/Surface plasmon resonance

Biological sciences/Microbiology/Virology/Influenza virus

Multivalent interactions govern numerous biorecognition events at nanoscopic or microscopic length scales.^1-5 Like a biological Velcro, these contacts rely on multiple individually weak but rapidly exchanging interactions between surface-confined receptors or ligands on the two interacting partners. One of the most ubiquitous multivalent interactions involves carbohydrates and lectins, where a single lectin´s intrinsically weak affinity for the carbohydrate (K_d ≈ 10^-3 – 10^-6 M) translates into an exceptionally strong and selective interaction when multiplied.^6-10 These interactions govern cell and virus recognition in numerous key biological processes such as cell-cell communication, fertilization, biofilm formation, cell differentiation, cancer metastasis, viral infection and bacterial adhesion.

As in low molecular inhibitor design, polyvalent inhibitors have been designed to block or modulate the aforementioned interactions.^3,11-13 For instance, ligand decorated polymers,^14-16 nanoparticles^17-20 or dendrimers ²¹ have been reported to display a high affinity for cell surface receptors. These systems provide a globular homogenous ligand array that mimics the carbohydrate representation on the cell surface and allows control over the ligand’s chemical composition, ligand density, spatial arrangements, and accessibility. Careful tuning of carbohydrate separation distance to match the lectin site separation and adjustments of the carbohydrate presentation and orientation have led to particularly powerful binders and inhibitors.^13,18-23 Such scaffolds can interrupt pathogen-host cell interactions and represent an attractive target for the proper design of new anti-bacterial or anti-viral drug candidates.^{7,11-13,15,18}

Despite the aforementioned advances, the majority of multivalent inhibitor designs exhibit only modest affinity enhancements, dominated by affinities with K_d:s in the micromolar range. An important reason for this limitation is their failure to mimic the dynamics and adaptable nature of the biological interactions. To compensate for this deficiency, a challenging Ångstrom precise tuning of ligand separation distance and presentation (minimizing conformational and rotational loss of entropy upon binding) is required to boost inhibitor affinity albeit at the expense of demanding synthetic procedures.¹³ To overcome this problem the use of supported lipid bilayers (SLBs) or liposomes exhibiting mobile ligands that can diffuse laterally to optimize receptor binding is an attractive approach for generating strong multivalent interactions.^5,10,24 This mimics the lateral diffusion of receptors and ligands occuring in cell membranes, a feature cleverly used by nature to promote so-called “superselective” multivalent interactions using only a few receptor copies.⁹ Supported lipid bilayers (SLBs) or liposomes are the established model systems used to mimic the fluidic properties of cell membranes. Although significant progress has been made toward simplifying the sample preparation and substrate compatibility, SLBs lack air stability and a successful deposition depends on high-quality vesicle preparation typically restricted to a narrow range of lipid compositions.^25-28

We have introduced reversible self assembled monolayers (rSAMs) as an abiotic approach to mimic the fluidity of biological cell membranes and show in this work how it can be used to construct highly potent multivalent virus inhibitors (Fig. 1). rSAMs offer robust and adaptable receptors featuring strongly enhanced affinities toward proteins, ²⁹ viruses ³⁰ and human cells³¹ compared to receptors based on stiffer scaffolds. The formation of the rSAMs is driven by the spontaneous assembly of benzamidine-terminated amphiphiles on alkanoic acid functionalized surfaces (e.g. SAMs of thiols on gold) in a neutral or alkaline aqueous solution.^32-34 This leads to monolayers featuring thicknesses correlating with the molecular length of the amphiphile accompanied by a lipid bilayer-like lateral diffusivity of the layer components.²⁹ In contrast to lipid-based models, the benzamidine monolayer is pH-responsive, resistant to plasma protein exchange and stable to rinsing and air exposure. Based on this reversible sensor platform we have recently constructed glycan-based sensors for influenza virus³⁰ and lectins²⁹ featuring picomolar and femtomolar detection limits respectively. We ascribed the high sensitivity to laterally mobile sialic acid ligands promoting superselective multivalent interactions. This leads to affinities exceeding 10000 times the affinity of covalent sensor constructs.³⁰

Here we describe the development of rSAM coated gold nanoparticles rSAM-NPs (Fig. 1). The study comprise a detailed optimization of their colloidal stability and a demonstration of their ability to bind low levels of H5N1 lectins and virus-like particles. Finally, we demonstrate how these scaffolds can interrupt pathogen-host cell interactions at ultralow concentrations outperforming previously reported glyconanomaterial-based inhibitors. Based on our findings we believe this dynamic ligand construct to represent a significant step towards high affinity multivalent receptors and pathogen inhibitors.

System design and characterization

The design of multivalent ligand architectures for virus inhibition requires attention to the structural features of the receptor in terms of ligand receptor binding affinity and specificity, receptor valency, site separation, virus size and virus mechanical properties.^3,9,10 Influenza A type viruses (IAV) are pleomorphic enveloped RNA virus with diameters of 100-150nm containing 290-340 hemagglutinin (HA) trimers and 24-50 neuraminidase (NA) tetramers on their surface. The SA binding site separation has been estimated to ca 4 nm within the HA trimer and on average 12 nm between individual HA trimers. Infection is initiated by the virus attaching to the host cell through formation of multiple (estimated to ca 10-20 per particle) weak (Kd ≈ 10^-3 M) interactions between HA and abundant cell surface sialic acids (SA), the latter modulated by the SA cleaving enzyme NA. Use of fluidic models of the host cell membrane has demonstrated a strong dependence of virus adhesion on SA surface coverage and tether length with an onset of virus binding occurring at an SA coverage of 1-5 pmol/cm² beyond which superselective virus binding is seen.^8,9 This coverage corresponds to an approximate SA-lipid/lipid ratio of 1-2 %.

The key parameters investigated to optimize avidity for this virus have therefore been the density of ligands, linker length and rigidity, nature of scaffold and its rigidity/fluidity as well as the contact area between virus and inhibitor.^3,7,13,15 Knowledge based on other host cell receptor models and inhibitor designs made us reason as follows for the rational design of rSAM based inhibitors.

Nature of the scaffold. Gold nanoparticles (AuNPs) have been extensively investigated as core for immobilizing appropriate ligands in a controlled manner exploiting the ease of synthesis and precisely controlled surface chemistry achievable in SAMs of thiols on gold.^35,36 Moreover AuNPs are biocompatible scaffolds of tunable size and optical properties with a surface plasmon band position and width strongly influenced by the gold core size and nature of ligands on its surface. The latter also influence their colloidal stability in addition to their concentration, pH and ionic strength. AuNPs have also been used as scaffolds for multivalent pathogen inhibition.^13,17,18 In one notable example, multivalent inhibition of hemagglutination in the nanomolar range was demonstrated using hyperbranched SA terminated dendron-coated AuNPs.¹⁷ As a logic extension of our previously demonstrated 2D rSAM based IAV receptors using planar gold ³⁰ we wanted to explore whether this chemistry could be transferred to 3D in the form of rSAM modified gold nanoparticles. This was inspired by our previous findings that bisbenzamidines interact strongly and rapidly precipitate negatively charged carboxylated AuNPs.³⁷ The particle size is expected to strongly influence the inhibitor performance. Since the number of ligand-receptor pairs increase with the contact area, a tighter binding has been found for size matched inhibitors.¹⁶ We therefore decided to investigate 50nm and 100nm gold as carriers for the rSAM shell which we anticipated would lead to sizes in the same range as the IAV particles (100-150nm). Also, we included differently shaped particles (rods, cubes) to account for the pleomorphic nature of the IAV particles. To optimize colloidal stability of the NPs in aqueous media we decided to perform ligand exchange of the citric acid stabilized particles with two different thiols, an aromatic (MBA) and a mercaptoundecane-tetraethylenglycol-based (MUA-TEG) carboxylic acid thiol (Fig. 1). The latter combines a long hydrophobic aliphatic chain promoting the formation of an ordered SAM, and a short oligo(ethylene glycol) spacer that increases the NP colloidal stability in aqueous media and reduces non-specific binding. The carboxylic headgroups allow the interaction with the amidinium ion promoting the formation of the reversible self-assembled monolayers (rSAMs).

Ligand density and presentation. The minimum ligand density required for superselective virus binding depends on whether the ligands are attached to static or fluidic scaffolds and the length of the linker connecting the SA with the scaffold.^8,9 As discussed above the density in fluidic scaffolds typically show superselective binding at low ligand densities i.e. at 1-2% ligand modified lipids or even lower using SLB scaffolds whereas static scaffolds require higher densities. With regards to the linker length, long linkers allow the SA ligands to better access the lectin binding site further reducing the ligand density onset for superselective binding. We previously showed that mixed rSAMs of the SA functionalized amidines (Fig. 1) and filler amidines display strongly enhanced affinity for HA and VLPs at SA coverages of 15 mol % (χ_E4-SA=0.15) and linkers containing 4 ethylene glycol repeat units such as in E4-SA (Fig. 1).^29,38 In contrast, rSAMs of E4 lacking E4-SA or with slightly increased E4-SA densities (χ_E4-SA=0.20) led to a complete suppression of binding, the latter ascribed to a paralell abrupt drop in ligand lateral diffusivity. As a starting point we therefore decided to compare these rSAM compositions and whether similar effects would be observed for the 3D format.

rSAM-modification results in stable colloids of ligand decorated nanoparticles

To first confirm formation, structure and properties of the films on planar substrates we used in situ ellipsometry (ISE) and infrared reflection absorption spectroscopy (IRAS). Fig. 2a shows the average film thickness during adsorption of a mixture of E2 and E4-SA (χ_E4-SA=0.15) on a SAM of MUA-TEG and Fig. S1 the corresponding IRAS spectra after rinse with two different buffers. The rapid adsorption kinetics, the limiting film thickness and the rinse stability (Fig. 2a) are indicative of well ordered films of densely packed amphiphiles oriented perpendicularly to the surface.²⁹ Fig. S1 and Table S1 support this picture revealing the significant vibrations with orthogonal transition dipole vectors of the anchor SAM (MUA-TEG) and the two component rSAMs as we reported previously.

The functionalization of the nanoparticles was then carried out in two steps, the success of each step being verified by DLS, FTIR, MALDI-TOF-MS and UV-Vis spectroscopy (Fig. 1c-g, Fig. S2-S4, Table S2, S3). First, the citrate stabilized nanoparticles were functionalized with MUA-TEG via ligand exchange. This led to a 4 nm increase in the hydrodynamic diameter and a pronounced redshift (Δλ=5 nm) of the surface plasmon band (Table 1; Table S2) manifested in a color change from red to bourdaux. The subsequent attachment of the amidines induced an additional red shift in the plasmon peak position of about 2 nm resulting in a change in color of the colloidal solution from bordaux to purple (Fig. 1 c, d).

Table 1. Plasmon peak position and hydrodynamic size of the AuNPs in HEPES buffer (10 mM, pH 8) before and after functionalization with MUA-TEG (SAM-NPs) and an rSAM (χ_E4-SA=0.15) (rSAM-NP). ^a

Ligand	λ (nm)	D_h(nm)	ζ-potential (mV)
Au-NP	533 ± 2	67± 2	-24 ± 4
SAM-NP	538 ± 0.3	71 ± 1	-40 ± 2
rSAM-NP	540 ± 0.4	73 ± 1	-25 ± 1

a. Error bars represent standard deviations based on three independent measurements.

This was accompanied by a 2 nm increase in the NPs hydrodynamic diameter. These particles formed stable colloids with no signs of agglomeration. Interestingly, the particle surface charge density reflected in the Z-potential (Table 1) decreased for these particles from -40 for the SAM-NPs to -25 after rSAM modification, the latter corresponding to the original Z-potential of the citrated capped NPs. This indicates that the negatively charged carboxylate anchor SAM is to a large part neutralized by the rSAM amidinium counterions.

To confirm the presence of the rSAM shell we then characterized the rSAM-NPs by mass and infrared spectroscopy. MALDI-TOF-MS was used as it features a mild ionization technique allowing analysis of neat samples of the NPs. Fig. S3 shows a representative MALDI-TOF-MS spectrum of rSAM-NPs and the masses of the amphiphile molecular ions. The identification of signals arising from E2 at 501 m/z, and E4-SA at 975 m/z provide evidence for the incorporation of both amphiphiles in the ligand shell.

FTIR-spectroscopy can be used to derive information regarding both the composition and molecular conformation of alkanethiol stabilized AuNPs.³⁹ Although the FTIR band positions of the SAMs on such curved surfaces are similar to the IRAS spectra of planar SAMs, the bands are typically broader and lack the orientation-sensitive intensity dependence seen in IRAS. This is confirmed by Fig. S4 showing the normalized attenuated total reflection infrared spectra (ATR-IR) of the SAM- and rSAM-NPs deposited on the ATR crystal from a pH 8 buffered solution. Noting here in brief that the rSAM modification results in the appearance of bands at 1620, 1273 and 838 cm^-1 (for assignements see Table S3) that reasonably well match the rSAM spectral signature and a conformational change reflected in a pronounced shift of the C-O-C ether stretch to lower frequencies (from 1186 to 1178 cm^-1 upon rSAM modification) we refer to the supporting information for a more detailed spectral interpretation.

To finally visualize the NP architectures we analysed the particles by transmission electron microscopy (TEM) (Fig. 2e,f). The samples were stained with PTA to determine whether the self assembled coatings could be observed and to estimate their thickness. The average thickness of the MUA-TEG shell (2.4±0.1 nm) and the rSAMs shell (3.2±0.2 nm) agreed with the data obtained by DLS (Table S2). Collectively, the above results support the presence of the rSAM on the MUA-TEG modified NPs.

rSAM-NPs interact strongly and selectively with viral proteins

To explore the affinity of the prepared rSAM-NPs for IAV, we first investigated their interactions with hemagglutinin (HA) from avian H5N1 (Table S4) by UV-vis spectroscopy and DLS. Assessing the plasmon band shift dependence of the AuNP core size and buffer medium showed that the use of smaller 50 nm AuNPs and phosphate buffer led to the strongest shifts (Table S2, Fig. S5). Hence, we fixed these parameters in all further experimentation. Selectivity was assessed using the galactose-binding lectin Concanavalin A (ConA) as reference with the anchor NPs functionalized with MUA-TEG used as control. The NPs were incubated with different concentrations of the proteins (0-12 nM), and the changes in the absorption spectra and the size of the nanoparticles were recorded 1h after protein addition (Fig. 3). The shift in the peak position was calculated by fitting the plasmon peak with the pseudoVoigt function using at least 40 data points.

As shown in Fig. 3a, the interaction with HA induced a gradual red shift in the SPR peak position of the rSAM-NPs with an estimated limit of detection of ca 1 nM. In contrast, only minor shifts were observed using the MUA-TEG based SAM-NPs. Moreover, no significant SPR and DLS shifts were observed for the reference protein ConA. These results were consistent with the data obtained by DLS (Fig. 3b) revealing a pronounced size increase of the rSAM-NPs after incubation with HA. The increase in NP size corresponds to a shell thickness increase of 15 nm which interestingly coincides with the estimated 14 nm upright dimension of the HA fusion protein. At higher HA concentrations, the rSAM-NPs exhibited an up to 30 nm increase in the hydrodynamic diameter suggesting the formation of multiple binding sites between the rSAM-NPs and the protein. Collectively these data show that the rSAM-NPs, as intended, interact tightly with the IAV lectin through the exposed SA ligands. Interestingly, the increase of the E4-SA concentration to 20 mol % (χ_E4-SA=0.20) led to a drop of the HA binding (Fig. 3a). This is consistent with our recently reported data obtained for planar substrates.²⁹

The affinity of the rSAM- modified NPs for HA was further investigated in a depletion experiment using highly sensitive ELISA-based detection (Fig. 3c-e). To assess the impact of particle shape on HA binding we compared the spherical NPs with nanorods (NRs) and nanocubes (NCs) and incubated the particles in a 30 pM HA solution. Already at the lowest particle concentration (< 10 pM) a more than 50% reduction in free HA was observed although with no obvious difference between the differently shaped AuNPs. This again indicates an exceptionally high affinity of the rSAM-NPs for the lectin.

rSAM-NPs interact strongly with deactivated H5N1

Next, we studied the interaction of the rSAM-NPs with the whole influenza virus in allantoic liquid, a protein and salt rich liquid originating from fetal membranes and used as medium for virus production. This would test whether the shell architecture would survive in more competitive environments. For this purpose, rSAM-NPs were incubated for 1h with the inactivated H5N1 (A/Anhui/01/2005 H5N1) and then the samples were characterized by UV-vis spectroscopy (Fig. 4) and DLS (Fig 5).

Increasing the virus concentration led to clear changes in both the visual appearance of the solutions (Fig. 4a) and their UV-Vis spectra (Fig. 4b). The plasmon band red-shifted, broadened and decreased in intensity while the absorbance at higher wavelengths increased. These changes indicate the presence of strong interactions between the virus and the particles and the formation of larger aggregates.⁴⁰ The shift of the plasmon band increased along with the virus concentration and appeared to level off at an IAV concentration of ca 3 HAU with significant shifts observed for IAV concentrations as low as 1 HAU (Fig. 4c). In view of the weak shifts observed for the SAM-NPs we conclude that these high-affinity interactions can be ascribed to the SA functionalized rSAM shell. For IAV concentrations exceeding 7.7 HAU, the plasmon band shift for the rSAM-NPs decreased whereas the apparent measurement error increased. We assigned this phenomenon to the hook effect, a known limitation of immunoassays manifested in a decrease in the measured signal at high analyte concentrations.⁴¹

At elevated IAV concentrations, the binding sites on the NPs become saturated which produces an increase in the interparticle distance and hence a decreased interparticle coupling. We therefore attribute the decreasing band shifts to this effect.⁴² The formation of agglomerates can be assessed by plotting the normalized A₆₀₀/A₅₃₉ ratio versus the concentration of the virus in suspension (Fig. 4d) ⁴⁰. These data showed a quasi-linear correlation between the degree of the rSAM-NPs aggregation and the virus concentration even at high HAU values. DLS confirmed this effect as shown by an additional population of particles of a size ranging from 200 to 450 nm appearing at virus concentrations above 5 HAU (Fig. 5). Increasing the IAV concentration led to a further increase of the cluster peak area reflecting a concomitant increase in the cluster concentration in the sample. In parallel, the area of the peak attributed to the individual rSAM-NPs decreased reflecting the transition from well-dispersed NPs to agglomerates. All in all, these data show that the rSAM shell imparts a very high affinity for the targeted virus which exceeds most previously published multivalent platforms.^18,20

rSAM-NPs effectively inhibit virus-cell interactions at picomolar concentrations

Having proven the ability of the rSAM-NPs to interact tightly with IAV we were curious to know whether this translated into an effective multivalent inhibition of IAV-cell interactions. We, therefore, focused on the well-established hemagglutination Inhibition Assay (HAI) (Fig. 6). For this purpose, different concentrations of rSAM- and SAM-NPs were preincubated with 8 HAU of the deactivated IAV for 1h. Thereafter the samples were exposed to glutaraldehyde-fixed turkey red blood cells (RBCs) and incubated for 30 min. In presence of the SAM- and rSAM-NPs formed from E2 only (χ_E4-SA=0), effective hemagglutination was observed in all samples indicating here a lack of inhibition (Fig. 6c, Fig. S7a,b). This contrasted with the samples treated with SA-decorated rSAM-NPs (χ_E4-SA=0.15). Here effective hemagglutination was absent even at sub-pM concentrations of the NPs. Hence, the rSAM-NPs can compete with the interaction of the viral particles with the RBCs and completely inhibit their hemagglutination at very low concentrations. As the hemagglutination of the RBCs proceeds via the interaction of the HA with sialic acid on the cell surface, the results again confirm that the binding of H5N1 by the rSAM-NPs indeed proceeds through the interaction between E4-SA and HA.

To relate these findings to previously reported multivalent SA-based inhibitors affinity comparisons have to be performed on a per SA basis, a measure that better reflects the enhanced binding due to multivalency. The number of SAs per NP has been estimated in Table S2 assuming a monodisperse distribution of particles, each particle being a sphere with a diameter of 50 nm and covered with densely packed SAM shells. For the rSAM-NP this leads to an estimated number of SA-ligands per particle of 5355 translating into a minimum inhibitory concentration of the NP per SA basis to ca 18 nM. Hence even per SA basis, our inhibitor appears exceptionally efficient showing complete inhibition at nanomolar concentrations per SA basis. This is in stark contrast with the micromolar concentrations required using similarly sized inhibitors based on covalently linked SA-ligands and the complete absence of inhibition using millimolar concentrations of free SA.^16,17

The principle of multivalent recognition is being increasingly exploited in the design of synthetic receptors, mostly bioinspired by the way cells, pathogens and proteins are interacting with each other. ^1,4,7 This has led drug discovery practice to move beyond small molecule inhibitor design based on the “lock and key” approach to now also comprise supramolecular multivalent inhibitors and drugs.^11,18,43 These typically rely on scaffolds with appended ligands with precisely tuned density and tether designs. There is now a growing realization of the crucial role of this fine tuning to achieve superselective binding and efficient tissue targeted medicines.⁴³ Also crucial in this context is to master the control of ligand mobility, a parameter known to strongly affect binding affinity, selectivity and cell signalling events.^5,9,10,44 Practical scaffolds in this regard are few, with most dynamic designs relying on labile lipid-based scaffolds. To fill this gap we have here introduced rSAMs as a versatile plug and play approach to design core shell nanoparticles featuring mobile ligands. We show that our previous demonstration of the crucial role of rSAM ligand lateral diffusivity on lectin affinity²⁹ also holds true for 3D designs, here in the form of sialic acid-rSAM coated AuNPs. This boosts affinity for lectins and AIV as reflected in highly sensitive optical sensing of hemagglutinin (LoD < 2 nM) and deactivated H5N1 (LoD < 1 HAU). As a logic extention of sensing applications we have tested rSAM-NPs as hemagluttination inhibitors i.e. probing their ability to inhibit virus cell interactions. In spite of a simple single sugar ligand design we noted a highly potent inhibition with a low nanomolar inhibitory concentration per ligand unit. This contrasts with most other reports on multivalent IAV inhibitors featuring significantly lower potencies.^18-20,45 A recurring question is whether such constructs exhibit sufficient selectivity for their targets. Apart from tuning the aforementioned design parameters, the inclusion of several orthogonally acting ligands is an efficient way to achieve this that we will report on in a forthcoming publication.

Preparation of SAM modified AuNPs (SAM-NPs)

The functionalization of the citrate stabilized nanoparticles was performed via a ligand-exchange reaction. A solution of the MUA-TEG (0.15 mM, 500 µL) in dry ethanol was added dropwise to 3 ml of AuNP suspensions (60 pM in deionized water, pH 11) under vigorous stirring. After thiol addition, the color of the suspension gradually changed from red to bordeaux. The reaction was allowed to proceed for 72h whereafter the colloidal solutions were centrifuged (8000 rpm, 10 min) and the pellet redispersed in (HEPES) buffer (HB) (10 mM, pH 8, 0.005% (w/w) Tween 20) followed by repeating the centrifugation and redispersion. The centrifugation-redispersion cycles were repeated five times.

Preparation of rSAM modified SAM-NPs (rSAM-NPs)

Stock solutions of the amidines E2 and E4-SA were prepared by dissolving them (c= 2.5 mM) in HB. E2 or the mixture of E2 and E4-SA (χ_E4-SA = 0.15 or χ_E4-SA = 0.20) stock solutions (2.4 µL) was added to 1 mL of the SAM-NP suspension (60 pM) in HB under shaking resulting in a final amidine concentration of 6 µM in the reaction mixture. Immediately after the addition of amidines, the suspension turned purple followed by sedimentation of the nanoparticle agglomerates. After 48h the rSAM-modified NP colloidal solutions were centrifuged (8000 rpm, 10 min) and the pellet redispersed in HB followed by repeating the centrifugation and redispersion at least three times.

Characterisation of rSAM-NPs by MALDI-ToF mass spectrometry

A matrix solution was prepared by dissolving 2,5-Dihydroxybenzoic acid (DHB) (25 mg) in 1 ml of MeCN/water (50/50 v/v) containing 1% phosphoric acid and 0.1% TFA. 0.5 uL of the matrix solution and 0.5 uL of the NP colloidal solution (8.4 nM in HB) were deposited on the target plate and mixed. Mass-spectrometric analysis of the samples was performed using a MALDI reflector time-of-flight mass spectrometer (Ultraflex mass spectrometer, Bruker-Daltonics GmbH, Bremen, Germany) equipped with a Scout-384 source in positive reflector mode. The spectra were collected by accumulating 5000 laser shots in the linear mode (relative laser focus: 80%) and further analyzed with the Flexanalysis 3.0 software (Bruker Daltonics).

Characterisation of rSAMs by in situ ellipsometry (ISE)

The adsorption process of amidines was monitored using in situ null ellipsometry. The instrument used was an automated Rudolph thin film ellipsometer (type 43603-200E, Rudolph Research, USA) using an angle of incidence of 68°. The light source was a xenon lamp, filtered to λ= 442.9 nm. The thiol-SAM modified gold substrates (vide supra) were immersed vertically into an ellipsometric quartz cuvette with ordinary microscopic cover glass windows containing 5 mL of HEPES buffer (10 mM, pH 8.0). The cuvette was thermostated to 25 °C and equipped with a magnetic stirrer at constant stirring rate of 350 rpm. Before each measurement, the refractive index of the SAM-modified gold substrate was determined by a 4-zone surface calibration. After a stable baseline was obtained, 100 µL of stock solution containing amidines E2 and E4-SA (χ_E4-SA = 0.15) (2.5 mM) was added to the curvette resulting in a final total amidine concentration of 50 µM. Kinetics data was collected until stabilization. The system was then rinsed with pH 8 HEPES buffer for 300 s (11 mL min^-1) in a continuous system. The thickness and the adsorbed mass were calculated using 3 layers substrate/ film model using the refractive index for the liquid: 1.335. The effective complex refractive index for the rSAMs was assumed 1.45. Refractive index increment, dn/dc, of 0.22 mg/mL was used to determine the amount of rSAMs adsorbed.

Interactions between proteins and NPs studied by UV-Vis spectroscopy and DLS

Incremental volumes of HA or ConA stock solutions (4.2 µM in PBS, pH 7.4) were added to the NP colloidal solutions (300 uL, 0.5 nM) in PB to make up protein concentrations of 1 nM, 2 nM, 4 nM, 6 nM, 8 nM, and 12 nM. After 1h, the samples were analyzed by UV-Vis spectroscopy and DLS. The position of the plasmon peak was determined as an average of three independent experiments and the peak shift calculated by fitting the plasmon peak with the pseudo Voigt function using at least 40 data points.

Interaction between inactivated H5N1 and NPs studied by UV-vis spectroscopy and DLS

Incremental amounts of BPL-inactivated influenza A virus, A/Anhui/01/2005(H5N1)-PR8-IBCDC-RG6 (128 HAU) was added to the NP colloidal solutions (30 pM) in PB to make up 300µL solutions containing 0.43-12.8 HAU of virus (Table S5). After 1h, the samples were analyzed by UV-vis spectroscopy and DLS. The shift in the LSPR peak position was calculated by fitting the plasmon peak with the pseudo Voigt function using at least 40 data points.

Interaction of HA with Au nanoparticles, nanorods and nanocubes

Interaction of HA with MUA-TEG modified spherical Au nanoparticles (NPs), nanorods (NRs) and nanocubes (NCs) functionalized with rSAM₁ (χ_E4-SA=0.15) was evaluated using a commercial ELISA kit for influenza H5N1 HA. 75 uL of an HA solution (30 pM in HB) was added to 75 µL particle suspensions of different concentrations (Fig. 2). After 1h of incubation at room temperature under shaking, the colloidal solutions were centrifuged (8000 rpm, 10 min) and the supernatants collected. The residual HA concentration in the supernatant was analyzed by H5N1-ELISA according to the protocol provided by manufacturer. The assay was based on duplicate measurements repeated three times.

Hemagglutination assay and hemagglutination inhibition assay

Hemagglutination (HA) and Inhibition of hemagglutination (HAI) assays were carried out as described in the WHO manual on influenza diagnosis and surveillance⁴⁶ with a few modifications. Both assays made use of glutaraldehyde stabilized red blood cells (RBCs) (5%). The cells were rinsed with PBS buffer supplemented with 1 mg/ml BSA (PBS-BSA) three times and diluted to 0.5% suspension with PBS-BSA buffer prior to use.

Hemagglutination assay: A serial dilution of virus (BPL-inactivated influenza A virus, A/Anhui/01/2005(H5N1)-PR8-IBCDC-RG6 (128 HAU) in allantoic liquid) was prepared across the rows 1-11 in a V-bottom shaped 96-well microtiter plate as follows. 50 μl of PBS (pH 7.4) was added to each of wells 1-12 and 50 uL of virus to well 1 leading to a two fold dilution. Then, 50 uL of virus was transferred from well 1 to well 2 leading to an overall four-fold dilution. The procedure was repeated eleven times in order to complete the dilution series. 50 ul of the RBCs (vide supra) were added to each well of the plate. The solutions were mixed and incubated at room temperature for 30 min. The hemagglutination titer was then estimated as the reciprocal of the dilution factor in the last well with complete haemagglutination. For the BPL-Inactivated Influenza, A H5N1Virus /Anhui/01/2005(H5N1)-PR8-IBCDC-RG6 the HA titer was found to equal 128 hemagglutination units (HAU) (Fig. S6). The assay was performed in duplicate.

Hemagglutination inhibition assay: NP solutions (0.113 – 30 pM) were first prepared by serial dilution of NP stock solutions (30 or 60 pM in PB) with PB. To 50 uL of each solution, 50 uL of virus (BPL-Inactivated Influenza A H5N1 Virus /Anhui/01/2005(H5N1)-PR8-IBCDC-RG6) in allantoic liquid diluted to 4 HAU with PBS was added and the suspensions incubated under shaking for 1 h. Thereafter, 50 ul of each solution was transferred to the wells of a V-bottom shaped 96-well microtiter plate, followed by the addition of 50 ul of 0.5% glutaraldehyde stabilized turkey red blood cells (RBCs) to each well of the plate. The RBCs treated with H5N1 (8 HAU) and RBCs alone were used as positive and negative controls, respectively. The solutions were mixed and incubated at room temperature for 30 min. The plate was thereafter tilted to 90 degrees for 30 s and then photographed. To distinguish between agglutinated and non-agglutinated cells, the shape of the RBC spots were recorded. Teardrop shaped spots were read as negative whereas absence of teardrop formation indicated positive agglutinated cells. Each assay was performed in duplicate.

To obtain the degree of hemagglutination, we analysed the path length of the RBCs after the plate was tilted using ImageJ software. We define the degree of hemagglutination as:

where Lc is the path length of the RBC negative control, Lv the path length of RBC incubated with H5N1 (4 HAU) as positive control, and L is the path length of the samples containing H5N1 (4 HAU) pre incubated with the NP solutions. The samples exhibiting hemagglutination exceeding 75% were assigned as fully agglutinated.

Author Contributions

Y.S. directed the project. B.S. conceived the idea and designed the initial strategy. Y.S. designed the experiments, evaluated the data and wrote the first version of the manuscript. B.S. wrote the final version of the manuscript. S.Y.Y assisted with amphiphile synthesis and development and characterisation of rSAMs on planar substrates. T.J. assisted with the initial optimization and characterization of the rSAM-NPs.

Funding

This work was supported by grants from Marie Skłodowska-Curie Actions (H2020-MSCA-IF-2017, 794999), the Swedish Knowledge Foundation (FO 2019/377), the Swedish Research Council (VR-NT 2018-04930 and VR-MH-2021-02118) and the Malmö University Foresight programme.

Acknowledgments

We are grateful to Dr. Nikolay Vinogradov, the MAX IV Laboratory, Lund University; Prof. Anette Gjörloff Wingren, Dr. Zahra El-Schich and Dr. Anil Incel, BRCB, Malmö University; Prof. Daniel Aili, IFM, Linköping University and Prof. Guoqing Pan, Jiangsu University, China for their assistance and fruitful discussions.

Notes

The authors declare the following competing financial interest(s): B.S., S.Y.Y. and Y.S. are co-inventors on a patent covering the concept presented in this report. The remaining authors declare no competing interests.

Supplementary Information. Reagents, apparatus and methods, SAM and rSAM-preparation and characterization and associated tables and figures.

Mammen, M., Choi, S.-K. & Whitesides, G. M. Polyvalent Interactions in Biological Systems: Implications for Design and Use of Multivalent Ligands and Inhibitors. Angewandte Chemie International Edition 37, 2754-2794 (1998). https://doi.org:10.1002/(SICI)1521-3773(19981102)37:20
Fasting, C. et al. Multivalency as a chemical organization and action principle. Angew Chem Int Ed Engl 51, 10472-10498 (2012). https://doi.org:10.1002/anie.201201114
Bhatia, S., Camacho, L. C. & Haag, R. Pathogen Inhibition by Multivalent Ligand Architectures. Journal of the American Chemical Society 138, 8654-8666 (2016). https://doi.org:10.1021/jacs.5b12950
Dubacheva, G. V., Curk, T., Auzély-Velty, R., Frenkel, D. & Richter, R. P. Designing multivalent probes for tunable superselective targeting. Proceedings of the National Academy of Sciences 112, 5579-5584 (2015). https://doi.org:10.1073/pnas.1500622112
Dubacheva, G. V., Curk, T., Frenkel, D. & Richter, R. P. Multivalent Recognition at Fluid Surfaces: The Interplay of Receptor Clustering and Superselectivity. Journal of the American Chemical Society 141, 2577-2588 (2019). https://doi.org:10.1021/jacs.8b12553
Bertozzi, C. R. & Kiessling, L. L. Chemical glycobiology. Science 291, 2357-2364 (2001). https://doi.org:10.1126/science.1059820
Ortiz Mellet, C., Nierengarten, J.-F. & García Fernández, J. M. Multivalency as an action principle in multimodal lectin recognition and glycosidase inhibition: a paradigm shift driven by carbon-based glyconanomaterials. Journal of Materials Chemistry B 5, 6428-6436 (2017). https://doi.org:10.1039/C7TB00860K
Overeem, N. J. et al. Hierarchical Multivalent Effects Control Influenza Host Specificity. ACS Central Science 6, 2311-2318 (2020). https://doi.org:10.1021/acscentsci.0c01175
Overeem, N. J., van der Vries, E. & Huskens, J. A Dynamic, Supramolecular View on the Multivalent Interaction between Influenza Virus and Host Cell. Small 17, 2007214 (2021). https://doi.org:https://doi.org/10.1002/smll.202007214
Di Iorio, D., Verheijden, M. L., van der Vries, E., Jonkheijm, P. & Huskens, J. Weak Multivalent Binding of Influenza Hemagglutinin Nanoparticles at a Sialoglycan-Functionalized Supported Lipid Bilayer. ACS Nano 13, 3413-3423 (2019). https://doi.org:10.1021/acsnano.8b09410
Zelikin, A. N. & Stellacci, F. Broad-Spectrum Antiviral Agents Based on Multivalent Inhibitors of Viral Infectivity. Advanced Healthcare Materials 10, 2001433 (2021). https://doi.org:https://doi.org/10.1002/adhm.202001433
Hassanzadeh, P. Nanotheranostics against COVID-19: From multivalent to immune-targeted materials. Journal of Controlled Release 328, 112-126 (2020). https://doi.org:https://doi.org/10.1016/j.jconrel.2020.08.060
Kim, Y., Hyun, J. Y. & Shin, I. Multivalent glycans for biological and biomedical applications. Chemical Society Reviews 50, 10567-10593 (2021). https://doi.org:10.1039/D0CS01606C
Sigal, G. B., Mammen, M., Dahmann, G. & Whitesides, G. M. Polyacrylamides Bearing Pendant α-Sialoside Groups Strongly Inhibit Agglutination of Erythrocytes by Influenza Virus: The Strong Inhibition Reflects Enhanced Binding through Cooperative Polyvalent Interactions. Journal of the American Chemical Society 118, 3789-3800 (1996). https://doi.org:10.1021/ja953729u
Tang, S. et al. Antiviral Agents from Multivalent Presentation of Sialyl Oligosaccharides on Brush Polymers. ACS Macro Letters 5, 413-418 (2016). https://doi.org:10.1021/acsmacrolett.5b00917
Papp, I. et al. Inhibition of Influenza Virus Activity by Multivalent Glycoarchitectures with Matched Sizes. ChemBioChem 12, 887-895 (2011). https://doi.org:https://doi.org/10.1002/cbic.201000776
Papp, I. et al. Inhibition of Influenza Virus Infection by Multivalent Sialic-Acid-Functionalized Gold Nanoparticles. Small 6, 2900-2906 (2010). https://doi.org:10.1002/smll.201001349
Budhadev, D. et al. Glycan-Gold Nanoparticles as Multifunctional Probes for Multivalent Lectin–Carbohydrate Binding: Implications for Blocking Virus Infection and Nanoparticle Assembly. Journal of the American Chemical Society 142, 18022-18034 (2020). https://doi.org:10.1021/jacs.0c06793
Lauster, D. et al. Phage capsid nanoparticles with defined ligand arrangement block influenza virus entry. Nature Nanotechnology 15, 373-379 (2020). https://doi.org:10.1038/s41565-020-0660-2
Ramos-Soriano, J., Ghirardello, M. & Galan, M. C. Carbon-based glyco-nanoplatforms: towards the next generation of glycan-based multivalent probes. Chemical Society Reviews 51, 9960-9985 (2022). https://doi.org:10.1039/D2CS00741J
Ribeiro-Viana, R. et al. Virus-like glycodendrinanoparticles displaying quasi-equivalent nested polyvalency upon glycoprotein platforms potently block viral infection. Nature Communications 3, 1303 (2012). https://doi.org:10.1038/ncomms2302
Waldmann, M. et al. A Nanomolar Multivalent Ligand as Entry Inhibitor of the Hemagglutinin of Avian Influenza. Journal of the American Chemical Society 136, 783-788 (2014). https://doi.org:10.1021/ja410918a
Johansson, E. et al. Pentavalent Sialic Acid Conjugates Block Coxsackievirus A24 Variant and Human Adenovirus Type 37–Viruses That Cause Highly Contagious Eye Infections. ACS Chemical Biology 15, 2683-2691 (2020). https://doi.org:10.1021/acschembio.0c00446
Reichert, A., Nagy, J. O., Spevak, W. & Charych, D. Polydiacetylene Liposomes Functionalized with Sialic Acid Bind and Colorimetrically Detect Influenza Virus. Journal of the American Chemical Society 117, 829-830 (1995). https://doi.org:10.1021/ja00107a032
Castellana, E. T. & Cremer, P. S. Solid supported lipid bilayers: From biophysical studies to sensor design. Surface Science Reports 61, 429-444 (2006). https://doi.org:https://doi.org/10.1016/j.surfrep.2006.06.001
Sut, T. N., Yoon, B. K., Jeon, W.-Y., Jackman, J. A. & Cho, N. J. Supported lipid bilayer coatings: Fabrication, bioconjugation, and diagnostic applications. Applied Materials Today (2021).
Jackman, J. A. & Cho, N.-J. Supported Lipid Bilayer Formation: Beyond Vesicle Fusion. Langmuir 36, 1387-1400 (2020). https://doi.org:10.1021/acs.langmuir.9b03706
Fonseka, N. M., Arce, F. T., Christie, H. S., Aspinwall, C. A. & Saavedra, S. S. Nanomechanical Properties of Artificial Lipid Bilayers Composed of Fluid and Polymerizable Lipids. Langmuir 38, 100-111 (2022). https://doi.org:10.1021/acs.langmuir.1c02098
Yeung, S. Y. et al. Lipid Bilayer-like Mixed Self-Assembled Monolayers with Strong Mobility and Clustering-Dependent Lectin Affinity. Langmuir 35, 8174-8181 (2019). https://doi.org:10.1021/acs.langmuir.9b01452
Yeung, S. Y. et al. Reversible Self-Assembled Monolayers (rSAMs): Adaptable Surfaces for Enhanced Multivalent Interactions and Ultrasensitive Virus Detection. ACS Central Science 3, 1198-1207 (2017). https://doi.org:10.1021/acscentsci.7b00412
Yeung, S. Y. et al. Reversible Self-Assembled Monolayers with Tunable Surface Dynamics for Controlling Cell Adhesion Behavior. ACS Applied Materials & Interfaces (2022). https://doi.org:10.1021/acsami.2c12029
Auer, F., Nelles, G. & Sellergren, B. Odd–Even Chain Length-Dependent Order in pH-Switchable Self-Assembled Layers. Chemistry – A European Journal 10, 3232-3240 (2004). https://doi.org:10.1002/chem.200305477
Sellergren, B., Auer, F. & Arnebrant, T. Selective binding of DNA oligonucleotides to switchable selfassembled molecular layers. Chem. Commun., 2001-2002 (1999).
Sellergren, B., Swietlow, A., Arnebrant, T. & Unger, K. Consecutive Selective Adsorption of Pentamidine and Phosphate Biomolecules on a Self - Assembled Layer: Reversible Formation of a Chemically Selective Coating. Anal. Chem. 68, 402-407 (1996).
Colangelo, E. et al. Characterizing Self-Assembled Monolayers on Gold Nanoparticles. Bioconjugate Chemistry 28, 11-22 (2017). https://doi.org:10.1021/acs.bioconjchem.6b00587
Sakthi Devi, R., Girigoswami, A., Siddharth, M. & Girigoswami, K. Applications of Gold and Silver Nanoparticles in Theranostics. Applied Biochemistry and Biotechnology 194, 4187-4219 (2022). https://doi.org:10.1007/s12010-022-03963-z
Auer, F. et al. Nanocomposites by Electrostatic Interactions: 1. Impact of Sublayer Quality on the Organization of Functionalized Nanoparticles on Charged Self-Assembled Layers. Langmuir 16, 7554-7557 (2000).
Yeung, S. Y. et al. Reversible Self-Assembled Monolayers (rSAMs): Adaptable Surfaces for Enhanced Multivalent Interactions and Ultrasensitive Virus Detection. ACS Cent Sci 3, 1198-1207 (2017). https://doi.org:10.1021/acscentsci.7b00412
Johnson, S. R., Evans, S. D., Mahon, S. W. & Ulman, A. Alkanethiol Molecules Containing an Aromatic Moiety Self-Assembled onto Gold Clusters. Langmuir 13, 51-57 (1997). https://doi.org:10.1021/la9607520
Shipway, A. N., Lahav, M., Gabai, R. & Willner, I. Investigations into the Electrostatically Induced Aggregation of Au Nanoparticles. Langmuir 16, 8789-8795 (2000). https://doi.org:10.1021/la000316k
Bazydlo, L. A. L., N.S. Harris, N. S. & Winter, W. E. in Accurate results in the clinical laboratory: A guide to error detection and correction (eds Amitava Dasgupta & Jorge L Sepulveda) Ch. 11, 155-175 (Elsevier, 2019).
Ghosh, S. K. & Pal, T. Interparticle Coupling Effect on the Surface Plasmon Resonance of Gold Nanoparticles: From Theory to Applications. Chemical Reviews 107, 4797-4862 (2007). https://doi.org:10.1021/cr0680282
Bila, H., Paloja, K., Caroprese, V., Kononenko, A. & Bastings, M. M. C. Multivalent Pattern Recognition through Control of Nano-Spacing in Low-Valency Super-Selective Materials. Journal of the American Chemical Society 144, 21576-21586 (2022). https://doi.org:10.1021/jacs.2c08529
Álvarez, Z. et al. Bioactive scaffolds with enhanced supramolecular motion promote recovery from spinal cord injury. Science 374, 848-856 (2021). https://doi.org:10.1126/science.abh3602
Wang, S.-K. & Cheng, C.-M. Glycan-based diagnostic devices: current progress, challenges and perspectives. Chemical Communications 51, 16750-16762 (2015). https://doi.org:10.1039/C5CC06876B
World Health Organization, Manual for the laboratory diagnosis and virological surveillance of influenza (Geneva, 2011).

Yes there is potential Competing Interest. The authors declare the following competing financial interest(s): B.S., S.Y.Y. and Y.S. are co-inventors on a patent covering the concept presented in this report. The remaining authors declare no competing interests.

ToCgraphics.docx
SergeevaetalNCommSuppInfo.pdf
Sergeeva et al NComm Supp Info

Download PDF

Version 1

posted

You are reading this latest preprint version

Nanoparticles with adaptable self-assembled shells sense and inhibit influenza virus at ultra-low concentrations

Status:

Version 1

Abstract

Figures

Introduction

Results and Discussion

Conclusions

Methods

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1