Specific and nondisruptive interaction of guanidium-functionalized gold nanoparticles with neutral phospholipid bilayers

Understanding and controlling the interaction between nanoparticles and biological entities is fundamental to the development of nanomedicine applications. In particular, the possibility to realize nanoparticles capable of directly targeting neutral lipid membranes would be advantageous to numerous applications aiming at delivering nanoparticles and their cargos into cells and biological vesicles. Here, we use experimental and computational methodologies to analyze the interaction between liposomes and gold nanoparticles (AuNPs) featuring cationic headgroups in their protecting monolayer. We find that in contrast to nanoparticles decorated with other positively charged headgroups, guanidinium-coated AuNPs can bind to neutral phosphatidylcholine liposomes, inducing nondisruptive membrane permeabilization. Atomistic molecular simulations reveal that this ability is due to the multivalent H-bonding interaction between the phosphate residues of the liposome’s phospholipids and the guanidinium groups. Our results demonstrate that the peculiar properties of arginine magic, an effect responsible for the membranotropic properties of some naturally occurring peptides, are also displayed by guanidinium-bearing functionalized AuNPs.


General
Chemical reagents were bought from Aldrich at highest commercial quality and used without further purification. Water was purified using a Milli-Q® and water purification system. Reactions were monitored by TLC developed on 0.25 mm Merck silica gel plates (60 F254) using UV light as visualizing agent and/or heating after spraying ninhydrin. Solvents were of analytical reagent grade, laboratory reagent grade or HPLC grade.
NMR spectra in the solution state were recorded on a AVIII 500 spectrometer (500 MHz for 1H frequency). UV-Vis absorption spectra were measured in water on a Varian Cary 50 spectrophotometer with 1 cm path length quartz cuvettes. Fluorescence spectra were measured in HEPES 10 mM or HEPES 10mM, NaCl 100 mM buffer at pH 7 on a Varian Cary Eclipse fluorescence spectrophotometer. Both the spectrophotometers were equipped with thermostatted cell holders. ESI-MS were recorded on Agilent Technologies 1100 Series system equipped with a binary pump (G1312A) and MSD SL Trap mass spectrometer (G2445D SL).
The hydrodynamic particle size (Dynamic Light Scattering, DLS) and Z-potential were measured with a Malvern Zetasizer Nano-S equipped with a HeNe laser (633nm) and a Peltier thermostatic system. Measurements were performed at 25 °C in water or HEPES 10 mM or HEPES 10mM, NaCl 100 mM buffer at pH 7. Transmission electron microscopy (TEM) was recorded on a FEI Tecnai G12 microscope operating at 100 kV. The images were registered with a OSIS Veleta 4K camera. Thermogravimetric analysis (TGA) was run on 0.4 mg nanoparticle samples using a Q5000 IR instrument from 25 to 1000 °C under a continuous air flow.
Confocal images were taken using a laser scanning confocal microscope (BX51WI-FV300, Olympus) coupled to an Argon laser (IMA-101040ALS, Melles Griot) emitting laser light at 488 nm. The laser beam was scanned on 512x512 px sample area using a 60x water immersion objective (UPLSAPO60xW-Olympus). Fluorescence emission was collected through the same objective, separated from excitation light through a 490 nm longpass dichroic mirror, and recorded by the PMT with a 510 nm longpass filter. For fluorescence lifetime experiments, the sample was excited using a frequency doubled Ti:Sapphire femtosecond laser at 440 nm, 76 MHz (VerdiV5-Mira900-F Coherent), coupled with the BX51WI-FV300 confocal microscope. The emission signal was sent to a single-photon avalanche photodiode (SPAD, MPD, Italy). Before the light enters the photodiode, it passes through a 525/50 bandpass filter. The laser sync and the output of the SPAD were fed to a time-correlated single photon counting (TCSPC) electronics (PicoHarp 300, PicoQuant, Germany) for the calculation of the emission decay curve. The fitting of decay curve was performed with the Symphotime software (PicoQuant, Germany), using a two-components exponential model. Vitrification of samples for cryo-EM was performed in liquid ethane cooled at liquid nitrogen temperature using the FEI Vitrobot Mark IV semiautomatic autoplunger. Bright field cryo-EM was run at -176 °C in a FEI Tecnai G2 F20 transmission electron microscope, working at an acceleration voltage of 200 kV and equipped, relevant for this project, with field emission gun and automatic cryo-box. The images have been acquired in low dose modality with a GATAN Ultrascan 1000 2kx2k CCD. Synthesis of 7-azidohept-1-ene. 7-bromohept-1-ene (516 mg, 3.14 mmol) and sodium azide (220 mg, 9.157 mmol) were dissolved in aqueous DMF (10 ml). After 10 hours stirring, the mixture was washed with H 2 O and extracted with DCM. The combined organic layer was concentrated in vacuo and used in the next step without any purification.

Synthesis and purification of gold nanoparticles
Tetraoctylammonium bromide (TOABr, 2.5 eq) was dissolved in toluene and the solution was interval, which well compare with typical values of alkylthiol protected gold nanoparticles.

Preparation of neutrally charged fluorogenic liposomes
Fluorogenic liposomes with calcein. 7 Then, 6 cycles of freeze/thaw were performed, followed by a 15 times extrusion filtration with a polycarbonate membrane (0.1 μm, 19 mm). Size extrusion chromatography (G75) was used to remove extravesicular fluorophore. The liposome samples were stored at 4ºC.

Characterization of the fluorogenic liposomes
Supplementary

Liposome experiments
Fluorescence recovery experiments were initiated by the addition of a nanoparticles stock solution to 2 ml buffered solution (HEPES 10 mM, NaCl 100 mM, pH 7.0 or HEPES 10 mM, glucose 200 mM, pH 7.0) containing liposomes (22 μM phospholipid concentration) in a quartz cell. Sample emission at 25 ºC was measured followed until no further variation were detected (usually within 5 minutes). The maximum fluorophore emission was measured after addition of Triton X100 in the cell.  liposomes after addition of 1 or 1'@AuNPs, which feature the same coating and different average sizes (see Table 1 and Table S1)   The release of calcein from liposomes was monitored with fluorescence confocal microscopy. Images recorded from samples of calcein-loaded PC liposomes showed the presence of scattered green dots, at the resolution of the microscope, these correspond to the residual emission of individual or small groups of liposomes ( Figure S29A). The addition of 4@AuNP ( Figure S29B) did not produce any significant changes in this pattern, consistent with the results of the previous experiments. However, for 1@AuNP ( Figure S29C), a relevant increase in the green background emission was clearly detected, while the dotted pattern of the emission from the liposomes remained substantially unaltered. We ascribed this effect to the release of calcein into the bulk solution. According to confocal images, this occurred without affecting the integrity of the liposomes. The unperturbed, equilibrated membrane displayed an equilibrium value of 3.81 ± 0.26 nm. Upon the binding of 1@AuNP in all four replica simulations, the lipids closer than 3.5 nm from the AuNP's COM on the XY plane showed a thickness of 3.84 ± 0.28 nm, 3.83 ± 0.28 nm, 3.85 ± 0.27 nm, and 3.85 ± 0.27 nm, for replicas 1, 2, 3, and 4, respectively.