Ligand-clustered "patchy" nanoparticles for modulated cellular uptake and in vivo tumor targeting

Abstract

Despite the evident success of using a multivalent approach to increase efficacy of targeted delivery, a clear understanding of how multiple ligands behave collectively to influence the uptake of nanoparticle cell-targeting agents has not been reached. Although when present in large quantity, multivalent ligands can increase binding avidities to cells, it is also conceivable, that the manner in which these ligands are presented to the cell may have a significant effect on uptake. Here we examine this parameter using a linear dendritic polymer construct that enabled us to pattern the surfaces of nanoparticles with variable sized ligand clusters in different spatial arrangements. We demonstrate for the first time the clear impact of folate presentation on intracellular uptake both in vitro and in vivo. The findings presented here suggest that the nature of ligand presentation on a nanoparticle surface may play an important role in drug targeting; the results suggest potential impact for other targeting moieties and provide a framework for further refinement of future multivalent targeting strategies.

Figure 1

Chemical structure and self-assembly of linear dendritic polymers (LDP). a) Detailed chemical structure of the LDP and illustration of its self-assembly (PBLA= poly(benzyl-L-aspartic acid), PED=polyester dendron, PEG=poly(ethylene glycol)). b) Labeling LDP used in our study. Labeling schemes are described in methods. c) Folate conjugated to dendron of LDP at various percentages. Conjugation chemistry is described in methods.

Figure 2

Characterization of LDP mixed micelle system. a) Theoretical vs experimental degree of conjugation to dendron. Experimental values for UV-Vis were calculated using at least 3 independent sets of measurements (mean±STD) with free folic acid as the calibration. ¹HNMR quantification is described in methods. b) TEM images of representative formulations are shown. Clusters of DTPA-Fe³⁺ on the micelle surface show the formation of mixed micelles. Scale bar = 100 nm. Formulations 10%F-100%mix (1), 20%F-60%mix (2), 30%F-40%mix (3) and 30%F-100%mix (4) are shown. c) Summary of the eight micelle formulations used in this study. 0%F-100%mix is the untargeted control, all other formulations present similar amounts (not statistically different by one way ANOVA analysis at the 95% CI) of folate but in different cluster size and arrangements. Total number of folate presented per micelle (mean±STD) is calculated based on at least 3 sets of data (UV-Vis) using an approximate micelle aggregation number of 1000 (based on measured diameters and estimates of LDP unimer dimensions from Materials Studio). Average micelle diameters and zeta potentials are given in mean±STD of the averages of at least 10 individual measurements per micelle. Additional information on diameter, zeta potential information and folate presentation are shown in Supplementary Figure 3 and 2b.

Figure 3

In vitro evaluation of patchy micelles on KB (folate receptor, FR+) and A375 (FR−) cells. a) KB cell associated fluorescence after 24 h of incubation with micelle formulations. Highest fluorescence was measured from KB cells incubated with 20%F-60%mix formulation (Supplementary 4a). b) Measured cell associated fluorescence of KB cells at 5 min, 1 h and 6 h post incubation. Increase in cell associated fluorescence between 1 h and 6 h are similar for all targeted formulations, indicating that rate of endocytosis is unaffected by folate presentation. c) The binding avidity (K_D) measured for targeted formulations show highest avidity with 20%F-60%mix. d) The measured rate of decrease in micelle fluorescence after fixed KB cells pre-incubated to equilibrium with micelles was re-suspended in PBS. Non-linear regression analysis using a one phase exponential decay model gives the apparent dissociation rate constant of micelles from membrane receptors (k_off). Calculated values for targeted formulations show longest dissociation rate constant with 20%F-60%mix. K_D, k_off and k_on values are shown in Table 1. All graphs shown are made with measurements on n>3 independent experiments and are given in mean±SEM values. Individual flow cytometry measurements were averaged out of 10000 events.

Figure 4

In vivo evaluation of patchy micelles on nude mice bearing two different tumors (Right flank: KB, Left flank: A375). a) Fluorescence 3-D optical imaging of tumored nude mice 48h after injection with different formulation micelles. Micelle fluorescence (VT680) from tumors on both flanks are indicated by the arrows. A side by side comparison of VT680 and AngioSence750 fluorescence is also shown in Supplementary Figure 7 to allow identification of the tumors. b) Normalized tumor fluorescence (VivoTag680(VT680)/AngioSense750(AS750)) of tumors (n = 4) showing the highest in vivo targeting with 20%F-60%mix micelles. Raw fluorescence data is an average of the fluorescence in the region of interest, given in units of efficiency. Ratios are 0.27±0.11, 0.30±0.06, 0.19±0.06 and 0.28±0.04 for A375 tumors and 0.25±0.04, 1.06±0.27, 0.66±0.11 and 0.38±0.16 for KB tumors (0%F-100%mix, 20%F-60%mix, 60%F-20%mix and 70%F-20%mix respectively). c) Average cell associated micelle fluorescence for cells gated in Q1+Q3 (see Supplemental Figure 7) are 160±16, 185±104, 140±17 and 126±15 for A375 tumors and 175±84, 1153±104, 611±71 and 380±126 for KB tumors (0%F-100%mix, 20%F-60%mix, 60%F-20%mix and 70%F-20%mix respectively). All data is given in mean±SEM.