Effect of Nanoparticle Rigidity on the Interaction of Stromal Membrane Particles with Leukemia Cells

Abstract

In acute myeloid leukemia (AML), disease relapse is often observed because of therapy-resistant leukemic stem cells that re-initiate the disease. Leukemic stem cells can tightly associate with mesenchymal stromal cells inside the bone marrow, which is considered to further drive drug resistance. Here, the cell membrane of bone marrow stromal cells is used to prepare cell membrane nanoparticles and study their interactions with AML cells. Cell membrane liposomes (CM-Liposomes) of different charge are prepared and either used directly, or after deposition on silica cores to modulate nanoparticle mechanical properties. Nanoparticle size, zeta potential and coating efficiency are analyzed by dynamic light scattering (DLS) and cryo electron microscopy (Cryo-EM) imaging. Atomic force microscopy (AFM) is used to characterize the mechanical properties of CM-Liposomes and confirm bilayer deposition on silica cores. Finally, uptake by leukemic cells is determined. No difference in uptake is found between soft CM-Liposomes and liposomes of the same composition without membrane components. Instead, after deposition on a rigid core, uptake is higher for the cell membrane particles. Preliminary results on primary cells from leukemia patients confirm this observation. These results show that nanoparticle rigidity strongly affects the interaction between cell membrane nanoparticles and the targeted cells.

Keywords: cell membrane nanoparticles; leukemia; lipid‐coated nanoparticles; nanoparticle mechanical properties; nanoparticle targeting.

Figure 1

Schematic representation of the hypotheses tested in this study. The aim of this work was to study how nanoparticle rigidity and charge influence the interaction of cell membrane nanoparticles with the target cells: a) Softer and deformable CM‐Liposomes (CMLs) may not be able to bend the cell membrane and have low contact area.^{[ 22 ]} b) Rigid membrane coated nanoparticles may press into the target cell membrane and bend it.^{[ 22} , ^{24 ]} c) Agglomerates of partially coated rigid membrane nanoparticles may form and may press into the cell membrane, increasing the contact area.^{[ 25 ]} d) Schematic overview of soft and rigid model nanoparticles used in this study organized by charge (zeta potential, x‐axis) and mechanical properties (rigidity, y‐axis). Created in BioRender, de Weerd, S. (2024) https://BioRender.com/k44t679.

Figure 2

Characterization and cell uptake of liposomes and MS5‐CM‐Liposomes. a) Western blot of a whole cell RIPA lysate (MS5 wcl), crude membrane fraction (Crude), and purified membrane fraction (Purified, magenta) obtained with a cell membrane purification protocol optimized previously (n = 4, one representative blot is shown).^{[ 4} , ^{9 ]} Membranes were stained with markers for the plasma membrane, mitochondria, and cytosol in order to determine the composition of the different fractions (see Experimental Section for details). b) Size distribution of Pm‐Liposomes (black, n = 10, each measured in triplicate), Pm‐CM‐Liposomes (Pm‐CMLs, green, n = 10, each measured in triplicate) and purified membrane vesicles (magenta, one batch shown, measured in triplicate) in PBS (25 µg mL⁻¹ lipid for liposomes and CM‐Liposomes and ≈0.10 mg mL⁻¹ protein for membrane vesicles). c) Average median fluorescence intensity of K562 cells incubated with 10 µg mL⁻¹ Pm‐Liposomes (black‐triangles, mean ± SD, n = 3) and Pm‐CM‐Liposomes (green‐triangles, mean ± SD, n = 2) in 10% FBS over time, and to d) 50 µg mL⁻¹ Zw‐Liposomes (black‐circles) and Zw‐CM‐Liposomes (Zw‐CMLs, green‐triangles). Error bars are always included but too small to see in panel d. Figure legends created with Biorender.com.

Figure 3

Mechanical properties of liposomes and MS5‐CM‐Liposomes, as characterized by AFM imaging. a) Examples of AFM images of Pm‐Liposome and Pm‐CM‐Liposome (Pm‐CMLs). b,c) Heights of individual Pm‐Liposomes (n = 179, black‐triangles) and Pm‐CM‐Liposomes (n = 315, green‐triangles) in PBS from a representative batch at 80 pN, together with the calculated median (b), and the calculated mean and SEM (c). A Welch's‐t‐test was used to compare the mean values in panel c (p = 0.0009). d) Normalized height data for both nanoparticles at increasing imaging forces (Average ± SD). The comparable change in height suggests that both particles have similar mechanical properties and are equally soft. Figure legends created with Biorender.com.

Figure 4

Characterization of Liposome‐ and CM‐Liposome‐coated silica (LCS and MCS, respectively). a,b) Average size distribution by DLS of (a) all coated Zw batches (n = 6, each measured in triplicate), and (b) all coated Pm batches (n = 4, triplicate) in low ionic strength water (25 µg mL⁻¹) with SiO₂ as reference (n = 5, triplicate). c) Median and individual heights recorded by AFM at 80 pN for bare silica cores, Zw‐LCS, and Zw‐MCS. d) Mean of heights and calculated SEM for the same samples (SiO₂: n = 160; Zw‐LCS: n = 157, and Zw‐MCS: n = 190). A Welch and Brown‐Forsythe corrected ANOVA was used to compare the samples (adjusted p < 0.0001). e) Representative images of SiO₂, Zw‐LCS, Zw‐MCS by AFM. Arrows indicate some nanoparticle clusters observed by AFM (see Figure S8, Supporting Information for further images and cluster quantification). Figure legends created with Biorender.com.

Figure 5

Coating efficiency by cryo‐EM imaging and membrane fluidity of LCS and MCS. Representative Cryo‐EM images of a) Zw‐LCS, b) Zw‐MCS, c) Pm‐LCS and d) Pm‐MCS. Areas where coating is visible are highlighted by white arrows. e) Estimation of coating efficiency for LCS and MCS particles from Cryo‐EM images of a representative batch of particles (Zw‐LCS: n = 276, Zw‐MCS: n = 300, Pm‐LCS: n = 205, Pm‐MCS: n = 363). f) Laurdan GP ratio obtained at increasing temperature for the different formulations (Mean, n = 2, from 1 batch of particles). Pure DPPC was included as a control (DPPC). A lower GP indicates higher bilayer fluidity. Figure legends and inser ts created with Biorender.com.

Figure 6

Uptake of LCS and MCS in K562 cells. a,b) Median fluorescence intensity and SD of K562 cells incubated with 50 µg mL⁻¹ (based on silica concentration) of (a) Zw‐ and (b) Pm‐ LCS and MCS for increasing time. Uptake of bare silica was also included for comparison. Experiments were performed in duplicate and the average per experiment is shown (Mean ± SD, for a total of 6 and 3 independent kinetic experiments, in panel a and b, respectively). c) Staggered fluorescence distributions of K562 cells incubated with 50 µg mL⁻¹ silica, MCS, and LCS in K562 for 2 hours. d) Ratio (Fold change) of the uptake of MCS in respect to LCS of same composition at various incubation times (from kinetic data shown in panels a‐b with the addition of other experiments with fewer timepoints). The normalized median cell fluorescence obtained in 10 (Zw) and 7 (Pm) experiments are shown together with their mean, and SD (Mean ± SD, n = 10 (Zw), 7 (Pm), each with duplicate samples). Figure legends created with Biorender.com.

Figure 7

Schematic representation of the proposed mechanisms of interactions between soft and rigid cell membrane nanoparticles and targeted cells. a) A soft membrane particle has a lower capacity to bend the cell membrane for uptake. b) A rigid fully coated cell membrane particle may favor interaction with cell receptors and membrane bending for uptake. c) Agglomerates of partially coated cell membrane nanoparticles onto rigid cores may form and may promote membrane bending and interactions with multiple receptors on the targeted cells. d) Summary of the results reported in this work. Median cell fluorescence in K562 cells and zeta potential of the different particles studied (schematic). The blue line shows the expected trend in the results due to charge. Our results indicated that for deformable liposomes and CM‐Liposomes (CMLs, left side), uptake was higher for the more charged nanoparticles, but no difference was observed upon inclusion of cell membrane components. Instead for the more rigid LCS and MCS formed upon deposition of the bilayer onto silica cores, irrespective of particle charge, a higher uptake was observed upon inclusion of cell membrane components in the bilayer. These results suggested that a higher rigidity favors specific interactions between cell membrane nanoparticles and the targeted cells. Created in BioRender, de Weerd, S. (2024) https://BioRender.com/d93v796.