Mechanisms of Uptake and Membrane Curvature Generation for the Internalization of Silica Nanoparticles by Cells

Nanosized drug carriers enter cells via active mechanisms of endocytosis but the pathways involved are often not clarified. Cells possess several mechanisms to generate membrane curvature during uptake. However, the mechanisms of membrane curvature generation for nanoparticle uptake have not been explored so far. Here, we combined different methods to characterize how silica nanoparticles with a human serum corona enter cells. In these conditions, silica nanoparticles are internalized via the LDL receptor (LDLR). We demonstrate that despite the interaction with LDLR, uptake is not clathrin-mediated, as usually observed for this receptor. Additionally, silencing the expression of different proteins involved in clathrin-independent mechanisms and several BAR-domain proteins known to generate membrane curvature strongly reduces nanoparticle uptake. Thus, nanosized objects targeted to specific receptors, such as here LDLR, can enter cells via different mechanisms than their endogenous ligands. Additionally, nanoparticles may trigger alternative mechanisms of membrane curvature generation for their internalization.

Controls were performed for each experiment, as reported elsewhere. 2 Samples were collected and prepared for flow cytometry as described below.

RNA interference
RNA interference was used to shut down the expression of different proteins involved in endocytosis and curvature sensing proteins. Briefly, 13000 cells/cm 2 were seeded in a 24-well plate (Greiner Bio-One BV). 24h after seeding, HeLa cells were washed in serum-free MEM for 20 minutes and then each well was incubated with 250 µl of a mix composed of 45 µl oligofectamine (Thermofisher), 10nmol of siRNA and Opti-MEM (Thermofisher). In order to exclude eventual effects of the transfection reagents and procedure on cells and on nanoparticle uptake, nanoparticle uptake in the silenced cells was compared to the uptake in control cells treated with oligofectamine and a scramble RNA using the same procedures (negative control). After 4 hours, 125 µl of MEM supplemented with 30% v/v Fetal Bovine Serum was added to cells. 72h after silencing, cells were incubated with 100 μg/ml nanoparticle-corona complexes dispersed in serum-free MEM and prepared as described above for the indicated times. Alternatively, cells were incubated for 4 hours with 1 or 2 μg/ml of BODIPYor Dil-labelled LDL (ThermoFisher Scientific), or for 10 minutes with 15 μg/ml of fluorescently labeled human transferrin (Alexa546-TF, ThermoFisher Scientific) in serum-free MEM. Samples were collected and analyzed by flow cytometry or fluorescence imaging as described below.

Plasmid transfection with AP180
The construct containing the GFP-tagged C-terminal of Ap180 expressed under a constitutive promoter was kindly provided by Yvonne Vallis and Harvey T McMahon (Cambridge University, UK). 3 HeLa cells were transfected with 0.2 ng of plasmid DNA, using 0.6 μl Fugene (Promega) as transfection reagent in cMEM. After 24h, cells were washed in serum-free MEM and incubated with nanoparticle-corona complexes or transferrin, as described above. Cells were then washed and samples prepared for fluorescence imaging as described below.

Flow Cytometry Analysis
Flow cytometry was used to measure the fluorescence intensity of cells incubated with fluorescent 50 nm SiO2 nanoparticles, LDL, or transferrin. After the required exposure time, cells were washed once with cMEM and twice with PBS in order to reduce the presence of nanoparticles and markers adhering to the cell membrane. HeLa cells were harvested using 0.05% trypsin-EDTA for 5 min at 37 °C, collected, centrifuged for 3 min at 300 rcf, and resuspended in 100 μl PBS for the measurement. Cell fluorescence was recorded using a BD FACSarray (BD Biosciences, Erembodegem, Belgium) using a 532 nm laser for SiO2 nanoparticles and Dil-LDL, or a Cytoflex Flow S Cytometer (Beckman Coulter, Woerden, Netherlands) with a 561 nm laser for SiO2 nanoparticles and for Alexa546 transferrin, and 488 nm for BODIPY LDL. Data were analyzed using Flowjo data analysis software (Flowjo, LLC). Cell debris and cell doublets were excluded by setting gates in the forward and side scattering double scatter plots. A total of at least 20000 cells were acquired per sample and each sample was performed in triplicate (unless specified).
Then the average and standard deviation of the median cell fluorescence intensity were calculated.
Experiments were repeated at least 3 times to confirm reproducibility. The results are the average and standard error of the average results obtained in 3 independent experiments (unless specified).

Immunohistochemistry
For immunohistochemistry, Hela cells were plated on glass coverslips inserted in 24-well plates and experiments were performed as described above. Cells were incubated with nanoparticlecorona complexes or transferrin. Lysosomal staining was performed by incubating cells with a mouse primary antibody against LAMP-1 (clone H4A3, BD Biosciences) for 1 h, followed by 1 h incubation with an Alexa Fluor 488 goat secondary anti-mouse antibody. After each step of antibody incubation, cells were washed 3 times with PBS. Nuclei were stained with 0.2 μg/ml DAPI (4',6-diamidino-2-phenylindole) and glass slides were mounted with Mowiol 4-88 mounting medium (EMD Chemical, Inc, CA, USA). Image acquisition was performed using a Leica TCS SP8 fluorescent confocal microscope (Leica Microsystems, Wetzlar, Germany) with a 405 nm laser for DAPI excitation, a 488 nm laser for Alexa Fluor®488 or GFP, and a 552 nm laser for 50 nm SiO2 nanoparticles or transferrin. Images were processed using ImageJ software (http://www.fiji.sc). In order to compare nanoparticle uptake in BIN1 silenced cells and control cells silenced with a scramble siRNA, the corrected total cell fluorescence (CTCF) was calculated.
Individual cells were selected manually from at least 4 different images for each condition and their integrated density and area were extracted. For each image, areas without cells were selected in order to obtain the mean fluorescence of the background. Thus, for each image, the CTCF was calculated as the integrated density of the selected cells minus their cell area multiplied by the mean fluorescence of the background. The results are the average corrected total cell fluorescence obtained from at least 4 frames for each condition for a total of 38 cells for control cells and 19 cells for BIN1 silenced cells.

mRNA expression
The expression levels of silenced proteins were determined by RT-qPCR using the primers listed in Table S1. Table S1. Primers used in this study for RT-qPCR. RT-qPCR was performed as described in the Methods to determine the expression levels of some cell receptors (LDL and transferrin receptor), and a series of targets involved in different endocytic pathways and curvature sensing proteins.    included as a reference (where 60% uptake is an indicative threshold on the effect of silencing on nanoparticle uptake). The results show increased uptake in cells silenced for FCHO2, which can be explained by the increased LDLR expression in these cells. Importantly, LDL uptake was lower in cells silenced for GRAF1, BIN1, and IST1 suggesting a potential role for these targets in LDL uptake. The reduced LDL uptake in IST1 silenced cells might be due to the reduction in LDLR expression in these cells. lines at 100% and 60% uptake, respectively, are included as a reference (where 60% uptake is an indicative threshold on the effect of silencing on nanoparticle uptake). The results show that also in these cells, silencing the LDLR strongly reduced nanoparticle uptake. Uptake reduction was observed also in cells silenced for BIN1 (and to a smaller extent its isoform BIN2), PACSIN2 and IST1, as observed in HeLa cells. However, for other targets for which an effect was observed in HeLa cells, no effects were observed in A549 cells. These results indicate that curvature sensing proteins do play a role in nanoparticle uptake also in these cells, but different types are involved depending on the cell type.