Lipid Phase Separation in Vesicles Enhances TRAIL-Mediated Cytotoxicity

Abstract

Ligand spatial presentation and density play important roles in signaling pathways mediated by cell receptors and are critical parameters when designing protein-conjugated therapeutic nanoparticles. Here, we harness lipid phase separation to spatially control the protein presentation on lipid vesicles. We use this system to improve the cytotoxicity of TNF-related apoptosis inducing ligand (TRAIL), a therapeutic anticancer protein. Vesicles with phase-separated TRAIL presentation induce more cell death in Jurkat cancer cells than vesicles with uniformly presented TRAIL, and cytotoxicity is dependent on TRAIL density. We assess this relationship in other cancer cell lines and demonstrate that phase-separated vesicles with TRAIL only enhance cytotoxicity through one TRAIL receptor, DR5, while another TRAIL receptor, DR4, is less sensitive to TRAIL density. This work demonstrates a rapid and accessible method to control protein conjugation and density on vesicles that can be adopted to other nanoparticle systems to improve receptor signaling by nanoparticles.

Keywords: Janus particles; TRAIL; lipid domains; liposomes; phase separation; vesicles.

Figure 1.

Characterization of TRAIL conjugation to lipid domain vesicles. (a) Microscopy images of GUVs show phase separation of saturated lipids (green) and unsaturated lipids (red). Scale bars are 10 μm. (b) Schematic of a FRET assay to determine the lipid domain presence in vesicles. (c) FRET analysis of vesicle lipid domains before TRAIL conjugation, where the increasing FRET ratio indicates the presence of domains. The FRET ratio is reported as F_donor/F_acceptor. (d) FRET analysis of vesicle lipid domains before and after TRAIL conjugation shows no change after TRAIL conjugation. (e) Size distribution of TRAIL-conjugated vesicles measured by dynamic light scattering (DLS). (f) Conjugation efficiency of TRAIL to vesicles, as determined by Western blot. Results were analyzed by ANOVA compared to 0:1 DSPC:DOPC as a control. **** p < 0.0001, *** p < 0.001, ** p < 0.01, * p < 0.05. (g) Cryo-electron microscopy images of TRAIL vesicles. Scale bar = 100 nm. Error bars represent SEM from n = 3 different vesicle preparations (panels c–f).

Figure 2.

Vesicle lipid domains enhance TRAIL activation in Jurkat cells. (a) Viability of Jurkat cells treated with different concentrations of soluble and vesicle TRAIL after 24 h. Error bars represent SEM from n = 6 using two different vesicle preparations. The concentration reported is the initial amount of TRAIL added to the vesicles during conjugation. Results are analyzed by two-way ANOVA with multiple comparisons compared to DOPC TRAIL. (b) Viability of Jurkat cells to 1 mM unconjugated lipid vesicles, which corresponds to the highest TRAIL-conjugated vesicle concentration tested. Error bars represent SEM from n = 8 from three different vesicle preparations. p values reflect an ANOVA with multiple comparisons compared to untreated Jurkat cells. (c) Flow cytometry analysis of Jurkat expression of TRAIL receptors DR4 and DR5. (d, e) Activation of caspase 3/7 (d) and caspase 8 (e) in Jurkat cells exposed to soluble and vesicle TRAIL (20 nM) after 3 h. Error bars represent SEM from n = 3. p values are generated using ANOVA with multiple comparisons. **** p < 0.0001, *** p < 0.001, ** p < 0.01, * p < 0.05.

Figure 3.

Cytotoxicity of TRAIL conjugated to vesicles with lipid domains depends on the cell type. Cell viability upon treatment with soluble and vesicle TRAIL after 24 h and expression of TRAIL receptors DR4/DR5 of U2-OS (a, f), U937 (b, g), MDA-MB-231 (c, h), K562 (d, i), and HCT-116 (e, j). Error bars represent SEM from n = 6 from two different vesicle preparations. The concentration reported is the initial amount of TRAIL added to the vesicles during conjugation. Only the 2:1 (*) and 3:1 (*) DSPC:DOPC TRAIL condition is statistically different from DOPC TRAIL for U2-OS and only the 3:1 DSPC:DOPC TRAIL condition is statistically different from DOPC TRAIL for U937 (*). Results are analyzed by two-way ANOVA with multiple comparisons compared to DOPC TRAIL as our control. * p < 0.05.

Figure 4.

Enhancement of TRAIL cytotoxicity with lipid domains is dependent on DR4 and DR5 expression levels in target cells. The DR5/DR4 ratio on target cells is shown relative to the resulting cytotoxicity difference between 3:1 DSPC:DOPC TRAIL vesicles and pure DOPC TRAIL vesicles. Cytotoxicity differences between vesicles containing domains and homogeneous DOPC vesicles are only seen when DR5 is expressed higher than DR4. The error bar represents the SEM from n = 6 (y-axis) and n = 3 (x-axis).

Scheme 1.

Design of TRAIL-Conjugated Nanoparticles Using Phase-Separated Lipid Vesicles^{a a}(Top) By varying the ratio of unsaturated lipid (DOPC), saturated lipid (DSPC), and cholesterol, lipid vesicles can be assembled containing homogenous or phase-segregated membranes. A nickel-conjugated lipid (DGS-NTA-Ni) that integrates into the liquid disordered phase of the membrane can bind histidine-tagged TRAIL proteins and generate nanoparticles with varying spatial densities of TRAIL. (Bottom) Nanoparticles with low unsaturated lipid content and increased density of TRAIL are hypothesized to enhance apoptosis in target cells.