Ligand Density and Nanoparticle Clustering Cooperate in the Multivalent Amplification of Epidermal Growth Factor Receptor Activation

Abstract

Multivalent presentation of ligands on nanoparticles (NPs) is considered a general strategy for enhancing receptor binding and activation through amplification of ligand-receptor interactions within the footprint of the individual NPs. The spatial clustering of ligand-functionalized NPs represents an additional, less well understood mechanism for increasing local ligand-receptor interactions, especially for receptors that form higher-order assemblies, such as the epidermal growth factor (EGF) receptor (EGFR). To shed light on the interplay between ligand density ( i.e., multivalency) and NP clustering in signal amplification, we apply EGF-functionalized 72 ± 1 nm gold nanoparticles (NP-EGF) with known ligand loading (10-200 EGF/NP) as quantifiable and experimentally tractable units of EGFR activation and characterize the NP-mediated amplification of EGFR phosphorylation as a function of both EGF surface density and NP-EGF clustering for two cancer cell lines (HeLa and MDA-MB-468). The measurements confirm a strong multivalent amplification of EGFR phosphorylation through NP-EGF on the cellular level that results in EGF-loading-dependent maximum EGFR phosphorylation levels. A microscopic analysis of NP-EGF-induced EGFR phosphorylation reveals a heterogeneous spatial distribution of EGFR activation across the cell surface. Clustering of multivalent NP-EGF on sub-diffraction-limited length scales is found to result in a local enhancement of EGFR phosphorylation in signaling "hot spots" from where the signal can spread laterally in an EGF-independent fashion. Increasing EGF loadings of the NP enhances NP-EGF clustering and intensifies EGFR phosphorylation. These observations suggest that NP-EGF clustering and the associated local enhancement of ligand-receptor interactions are intrinsic components of the multivalent amplification of phosphorylation for the heterogeneously distributed EGFR through NP-EGF.

Keywords: cell signaling; epidermal growth factor receptor; gold nanoparticles; multivalency; nanoparticle−cell interactions.

Figure 1:

A. Schematic drawing of membrane-wrapped gold nanoparticle functionalized with EGF (NP-EGF) used in this work (left) and molecular structures and mol% of lipid components (right). The crystal structure of EGF is Protein Data Bank (pdb) structure 1JL9. B. TEM image of NP-EGF. C. Number of EGF molecules bound per NP as determined by ELISA for the particles used in this study. D. Geometrical model to estimate NP surface area O that is sufficiently close to the cell surface to facilitate EGF binding to EGFR. The length of the EGFR - EGF-PEG contact is approximated as 10 nm. E. Hydrodynamic diameter and F. zeta potential for citrate stabilized gold NP (Au NP), membrane wrapped NP (MWP), and membrane wrapped NP after EGF functionalization (NP-EGF). Data are provided for NP-EGF with 50, 100, 200 EGF / NP. G. UV-Vis spectra for NP-EGF with 50, 100, 200 EGF/NP (top to bottom). The data presented in C, E, F is the average of ten independent experiments.

Figure 2.

A. FIB-SEM was applied to mill MDA-MB-468 cells and to record SEM images at different depths. B. SEM image of an individual cell before milling. C. Cumulative depth distribution of NP-EGF after incubation times of t_inc = 15 min, 30 min, 60 min. D. SEM images of identical cellular regions for an exemplary cell recorded at milling depth of 30 nm (left) and 270 nm (right) for cells incubated with NP-EGF for t_inc = 15 min (top row), t_inc = 30 min (middle row), and t_inc = 60 min (bottom row). FIB milling was performed with a rate of approx. 6 nm/s, and the SEM images at depths of 30 nm and 270 nm were obtained after 5 s and 45 s of milling.

Figure 3:

A. Number of NP-EGF per MDA-MB-468 cell, as determined by ICP-MS, as function of input concentration and EGF loading (10, 50, 100, 200 EGF/NP). The NP-EGF were incubated with cells at 37°C for 10 min. The average cell number per ICP-MS run was 20,000. B. Same data for HeLa. C., D. EGFR phosphorylation determined by flow cytometry as function of NP-EGF input for nanoconjugates with specified EGF loading for MDA-MB-468 and HeLa. The individual data points represent geometric means. The average number of cells per run was 10,000. E., F. Phosphorylation signal for NP-EGF with different EGF surface loadings (50, 100, 200 EGF/NP) and free EGF at effective EGF concentrations of 1.5, 3.0, 6.0 nM for MDA-MB-468 and HeLa. G., H. Phosphorylation signal as function of the number of EGF peptides delivered per cell for MDA-MB-468 and HeLa. The data presented in A-H are averages of eight independent experiments.

Figure 4:

A. Immunoblots of pEGFR (Tyr 1068) and EGFR for (from left to right) blank; 15, 30, 60 pM of NP-EGF (50 EGF/NP); 15, 30, 60 pM of NP-EGF (100 EGF/NP); 15, 30, 60 pM of NP-EGF (200 EGF/NP); and supernatant of 60 pM of NP-EGF (200 EGF/NP) obtained with MDA-MB-468 (top) and HeLa (bottom). Cells were stimulated with NP-EGF and supernatant for 10 min. B. Relative phosphorylation for specified experimental conditions for MDA-MB-468. The phosphorylation was determined as ratio of pEGFR and EGFR signals. C. Same as in B. but for HeLa. The average number of cells per Western Blot was 20,000.

Figure 5:

A. Confocal image of immunolabelled pEGFR in the dorsal plasma membrane of MDA-MB-468 treated with 3 nM EGF. B. Confocal map of immunolabelled pEGFR in the plasma membrane of MDA-MB-468 after treatment with the supernatant of 30 pM NP-EGF (200 EGF/NP). C. Top row: Confocal backscattering image of NP-EGF (15pM, 50 EGF/NP) bound to the plasma membrane of an MDA-MB-468 cell (left); confocal scan of the immunolabeled pEGFR in the membrane (right). Bottom Row: merged image (left); computed colocalized pixel map (right). The effective EGF concentration is 0.75 nM. D. Same as in C. but with an EGF loading of 200 NP/EGF, corresponding to an effective EGF concentration of 3 nM. E. Mander’s Overlap Coefficients M1 and M2 for NP-EGF with 200 EGF/NP and 50 EGF/NP in MDA-MB-468 for three input concentrations: 5 pM, 15 pM, 30 pM. F. Same as in E. but for HeLa. Size bars are 10 μm. Data presented in E and F were collected from ten cells.

Figure 6:

Local pEGFR fluorescence intensity versus relative NP backscattering signal (normalized by monomer intensity) for A. HeLa with 50EGF/NP; B. HeLa with 200 EGF/NP; C. MDA-MB-468 with 50 EGF/NP; D. MDA-MB-468 with 200 EGF/NP. Three different NP-EGF concentration were used 5 pM (blue), 15 pM (red), 30 pM (green). The scattering intensity is normalized by average scattering intensity of an individual NP. E. The table summarizes the calculated Pearson Correlation Coefficients (PCC) for fluorescence intensity and scattering signal and associated Student’s t-test p-values for the individual conditions. The data in A-D were collected from ten cells per condition.

Figure 7:

A. Confocal map of pEGFR fluorescence (top) and map of peak LSPR wavelength of the same field of view obtained with hyperspectral darkfield imaging (bottom) in a single HeLa cell. NP-EGF with 200 EGF/NP and a concentration of 15 pM were used. The individual pixels are color-coded to identify their peak resonance wavelength in the spectral range 530 nm – 700 nm. Markings a-i indicate positions of colocalizing intensities in both channels. B. Histogram of the peak wavelength of cell-bound NP-EGF and NP-EGF bound to glass. C. pEGFR fluorescence as function of peak LSPR wavelength obtained from eight individual HeLa cells.