Reaching the Tumor: Mobility of Polymeric Micelles Inside an In Vitro Tumor-on-a-Chip Model with Dual ECM

Abstract

Degradable polymeric micelles are promising drug delivery systems due to their hydrophobic core and responsive design. When applying micellar nanocarriers for tumor delivery, one of the bottlenecks encountered in vivo is the tumor tissue barrier: crossing the dense mesh of cells and the extracellular matrix (ECM). Sometimes overlooked, the extracellular matrix can trap nanoformulations based on charge, size, and hydrophobicity. Here, we used a simple design of a microfluidic chip with two types of ECM and MCF7 spheroids to allow "high-throughput" screening of the interactions between biological interfaces and polymeric micelles. To demonstrate the applicability of the chip, a small library of fluorescently labeled polymeric micelles varying in their hydrophilic shell and hydrophobic core forming blocks was studied. Three widely used hydrophilic shells were tested and compared, namely, poly(ethylene glycol), poly(2-ethyl-2-oxazoline), and poly(acrylic acid), along with two enzymatically degradable dendritic hydrophobic cores (based on hexyl or nonyl end groups). Using ratiometric imaging of unimer:micelle fluorescence and FRAP inside the chip model, we obtained the local assembly state and dynamics inside the chip. Notably, we observed different micelle behaviors in the basal lamina ECM, from avoidance of the ECM structure to binding of the poly(acrylic acid) formulations. Binding to the basal lamina correlated with higher uptake into MCF7 spheroids. Overall, we proposed a simple microfluidic chip containing dual ECM and spheroids for the assessment of the interactions of polymeric nanocarriers with biological interfaces and evaluating nanoformulations' capacity to cross the tumor tissue barrier.

Keywords: extracellular matrix; microfluidics; nanoparticle mobility; polymeric micelles; tumor-on-a-chip.

Figure 1

Experimental setup. A commercial microfluidic chip from AIM Biotech, with three channels separated by triangular pillars (A), is filled in the middle channel with a model of basal lamina ECM from Engelbreth–Holm–Swarm murine sarcoma cells at 5.25 mg/mL and in the right side channel with a gel mix of collagen type I (2.5 mg/mL, pH 7.4) and hyaluronic acid (0.8 mg/mL), representing the tumor ECM, in which are embedded spheroids of the MCF7 breast cancer cell line (B). The micelle sample is added to the flow channel as a 160 μM solution in full DMEM (10% FBS) and allowed to diffuse for 24 h before doing a functional readout in the confocal microscope, either as ratiometric imaging or as fluorescence recovery after photobleaching (FRAP) in different locations inside the chip (C).

Figure 2

Validation of ECM distribution and spheroid viability inside the chip. (A) Overview of ECM distribution inside the chip using Cy3-labeled basal lamina gel (ECM from Engelbreth–Holm–Swarm murine sarcoma cells) and Cy5-labeled tumor ECM model (mix of collagen type I and hyaluronic acid), labeled using EDC/NHS reaction. (B) Zoom into (a) basal lamina and (b) tumor ECM, showing the fluorescence in Cy3 (yellow) and Cy5 (red) channels. (C) Live/dead assay of MCF7 spheroid inside the tumor ECM channel, after 24 h inside the chip, stained with calcein (green) and propidium iodide (red) (image shown after log transformation) (n = 4). (D) MCF7 spheroid growth inside the tumor ECM channel. Scale bar is 100 μm for (A–D) and 50 μm for B.

Figure 3

Micelle structure and characterization. Chemical structure of the studied amphiphiles with three different hydrophilic groups (PEtOx, PEG, or PAA) and two lengths of the hydrophobic ends (“Hex” or “Non”), labeled with 7-DEAC (A). Zeta potential measurements plotted versus hydrodynamic size by DLS of the six micelle formulations in PBS, pH 7.4, are shown ([amphiphile] = 80 μM) (n = 3) (B).

Figure 4

Confocal imaging of micelles inside different chip compartments after 24 h of incubation at [amphiphile] = 160 μM in full DMEM (10% FBS). Total fluorescence (first row) or ratiometric images of the unimer/micelle pixel ratio after background removal (second row) are shown in the basal lamina compartment (A) or at the edge of an MCF7 spheroid, where a dotted line indicates the boundary between the ECM and spheroid (C). Scale bars are 20 μm. Intensity profiles (B) of 3 × 80 μm rectangles inside the basal lamina compartment are shown for total fluorescence images (blue line) and ratiometric images (pink line). The mean fluorescence intensity or unimer/micelle ratio is quantified for each chip compartment (D): flow channel (control), basal lamina, tumor ECM (outside spheroids), and inside spheroids (n ≥ 3). A horizontal gray line is drawn for visualization purposes, corresponding to a unimer:micelle ratio of 1. Asterisks indicate p values obtained from two-way ANOVA comparison, as follows: p ≥ 0.05 (ns), p ≤ 0.05 (*), p ≤ 0.01 (**), p ≤ 0.001 (***), and p ≤ 0.0001 (****).

Figure 5

FRAP recovery curves of unimer (blue) and micelle fluorescence (red) represented as mean and standard deviation (faint lines), with fitted one-component exponential curve (dark lines), shown in different parts of the chip (n ≥ 3). Control measurements represent micelles in solution (in full DMEM and 10% FBS).

Figure 6

Diffusion constants of unimers (blue) and micelles (red) calculated from FRAP measurements in different locations inside the chip. Control measurements (“C”) were performed in solution, on a glass slide. “BL”, “Ta”, and “Ts” represent basal lamina, tumor ECM away from spheroids, and tumor ECM next to spheroids, respectively (n ≥ 3). Asterisks indicate p values obtained from two-way ANOVA comparison, as follows: p ≥ 0.05 (ns), p ≤ 0.05 (*), p ≤ 0.01 (**), p ≤ 0.001 (***), and p ≤ 0.0001 (****).