Microfluidic co‐culture devices to assess penetration of nanoparticles into cancer cell mass

Abstract In vitro and in vivo assessment of safety and efficacy are the essential first steps in developing nanoparticle‐based therapeutic systems. However, it is often challenging to use the knowledge gained from in vitro studies to predict the outcome of in vivo studies since the complexity of the in vivo environment, including the existence of flow and a multicellular environment, is often lacking in traditional in vitro models. Here, we describe a microfluidic co‐culture model comprising 4T1 breast cancer cells and EA.hy926 endothelial cells under physiological flow conditions and its utilization to assess the penetration of therapeutic nanoparticles from the vascular compartment into a cancerous cell mass. Camptothecin nanocrystals (∼310 nm in length), surface‐functionalized with PEG or folic acid, were used as a test nanocarrier. Camptothecin nanocrystals exhibited only superficial penetration into the cancerous cell mass under fluidic conditions, but exhibited cytotoxicity throughout the cancerous cell mass. This likely suggests that superficially penetrated nanocrystals dissolve at the periphery and lead to diffusion of molecular camptothecin deep into the cancerous cell mass. The results indicate the potential of microfluidic co‐culture devices to assess nanoparticle‐cancerous cell interactions, which are otherwise difficult to study using standard in vitro cultures.


| I N TR ODU C TI ON
Canonical drug delivery research usually commences with the validation of a carrier or a drug using in vitro static cell cultures in which cells are grown in 2D monolayers and are subjected to the drug and subsequently tested through a variety of established methods for cellular uptake and cytotoxicity effects. If efficacy and toxicity outcomes in the static cultures are deemed satisfactory, then the carriers are advanced to in vivo studies. Currently, on average, five compounds from the initial pool of 5,000-10,000 enter clinical trials, and only one becomes a successful FDA approved drug. 1 Since carriers often alter the drug's efficacy and toxicity, drug-carrier combinations must also go through the same rigorous validation and approval process. This approach limits the likelihood and speed of translation of in vitro foundational research to in vivo outcomes. 2 The knowledge gap between the performance of the carriers in vitro and in vivo is often difficult to bridge due to the disparate nature of the two methods of studies. In vitro cell cultures are typically con- viscous forces and diffusive mixing within higher micro-and macroregime vessel sizes. 3 In addition, standard in vitro cultures lack the multicellular environment containing complex extracellular matrix, which is characteristic of tissues. These physical parameters strongly impact carrier performance in vivo, for example, the carrier's ability to extravasate and accumulate at the cancerous cell mass site.
Microfluidic devices offer the potential to bridge the gap between the standard in vitro and in vivo models for drug delivery and discovery because of their ability to integrate physiological processes which are often overlooked or not directly accounted for in traditional in vitro methods. 4 In comparison to traditional in vitro models, data from microfluidics devices can provide a more accurate and comprehensive prediction of how well a carrier will perform in vivo.
In this study, we utilized an idealized co-culture microfluidic device (ICD) with an inner tissue culture chamber and two flanking outer vascular channels connected to the tissue chamber via micron sized pores. 5 The inner tissue culture chamber of the ICD was cultured with murine breast cancer cell line, 4T1, and the outer vascular channels were cultured with the human umbilical vein endothelial cell line, Eahy.926. Cancerous and healthy cells were cultured in 3D in the tissue chamber and exposed to camptothecin (CPT) nanocrystals, with rod-shaped morphology, under physiologically relevant shear stresses found within micro-domain sized vessels. 6 Cells subjected to nontoxic and nonimmune reactive nanoparticles under physiologically relevant shear stresses are known to induce cell death due to the physical and mechanical interactions of particles and cell surfaces which are enhanced and impart a cytotoxic effect. 7 The devices were used to monitor penetration and efficacy of the nanocrystals within the cancerous cell mass site after short infusion time periods, akin to bolus injections.
The choice of therapeutics (camptothecin nanocrystal) was motivated by our previous studies, which demonstrated the benefits of rod-shaped nanoparticles over spheres. 8 Nanocrystals provide a unique ability to increase drug loading as well as control its release kinetics. 9 The crystalline nanorods used here comprise entirely of camptothecin, a Topo I inhibitor. Hydrophobic drugs have traditionally posed a challenge in drug delivery due to their poor solubility and dependence on amphiphilic carriers for their distribution. 10 Nanocrystals posit an alternative to the traditional hydrophobic drug carriers since they are entirely comprised of the hydrophobic drug; creating a high concentration of drug in a localized area. [11][12][13] Camptothecin nanocrystals were used either in their bare form or surface-modified to display PEG or PEG-folic acid. Folic acid was chosen for its ability to target the folic acid receptor on 4T1 cells. 14 To make CPT nanocrystals, 5 ml of 0.8 mg/ml of CPT (Sigma Aldrich) in DMSO solution was pipetted dropwise into a 120 ml water mixture containing 1% w/w alpha-tocopherol (Sigma). The mixture was stirred at 800 rpm under constant ultrasonication at room temperature (228C) for 1 hr. CPT-UM nanocrystals formed at the boundary where DMSO diffused into the water. The CPT-UM nanocrystals were then centrifuged three times at 208C with milliQ water (18.2 X) at 3,500 rpm. The concentration of CPT-UM nanocrystals was determined by dissolving the nanocrystals in DMSO and reading the absorbance at 366 nm using a spectrophotometer (Tecan M220 Infinite Pro) and a CPT calibration curve.
To prepare CPT-PEG nanocrystals, a mixture of 5 ml of 0.8 mg/ml CPT and 3.2 mg/ml DSPE PEG2K Amine (Avanti Polar Lipids) was added to the 1% alpha-tocopherol water mixture solution, all subsequent steps for preparation, purification, and quantification described above for CPT-UM nanocrystal preparation were followed. The suc- To incorporate folic acid into CPT crystals, 5 ml of 0.8 mg/ml of CPT (Sigma Aldrich) in DMSO solution was pipetted dropwise into a 120 ml water mixture containing 1% w/w alpha-tocopherol (Sigma).
The 20% acetonitrile-milliQ H 2 O (18.2 X) eluent containing PEG-FA conjugate was then added dropwise to this solution and the overall mixture containing CPT, DSPE PEG2K amine, and FA was stirred at 800 rpm with constant ultrasonification at room temperature (228C) for 1 hr. After 1 hr, the CPT-FA nanocrystals were then centrifuged three times at 208C with milliQ water (18.2 X) at 3,500 rpm. The presence of folic acid was quantified using absorbance at 290 and 370 nm, and CPT was quantified using fluorescence at 366/434 nm-both utilized a spectrophotometer (Tecan M220 Infinite Pro) and a CPT and FA calibration curve at all respective wavelengths. DSPE PEG2K Amine was quantified using P 31 NMR as described above in CPT-PEG constructs.
Morphologies of CPT, CPT-PEG, and CPT-FA nanocrystals were analyzed using a scanning electron microscope (SEM). Surface charges of all nanocrystalline scaffolds suspended in 1x PBS pH 7.4 were measured as zeta potential using a Nanoseries-Zetasizer (Malvern).

| Cell culture
All cell lines were commercially obtained from ATCC and were grown in a humidified incubator with 5% CO 2 at 378C. Endothelial cell line, EA.hy926 cells, were cultured using DMEM medium supplemented with 10% FBS and 1% penicillin-streptomycin (Pen-Strep). Murine mammary tissue cancerous cell line, 4T1 cells, were cultured using RPMI-1640 medium supplemented with 10% FBS and 1% Pen-Strep.

| Synthesis and characterization of camptothecin nanocrystals
Rod-shaped CPT nanocrystals were prepared using the solvent diffusion method and were visualized using SEM (Figure 2).
The dimensions of the nanocrystalline rods are reported in Table   1 All release data for CPT-UM and CPT-PEG were normalized to this positive control. The delayed effect of CPT release was attributed to the surface coating ( Figure 5).

| In vitro cell growth inhibition by CPT nanocrystal constructs
The

| Cell growth inhibition in idealized co-culture microfluidic devices
The effect of all CPT constructs on in vitro growth inhibition of 4T1

| D I SCUSSION
Nanoparticles developed for therapeutic applications typically advance through the canonical channels of drug delivery research, whereby the core platform is developed in vitro and then shunted into in vivo research routes. This serial assessment is not ideal and limits the likelihood of successful outcomes. [1][2][3][20][21][22][23][24][25][26] This study aims to couple conventional in vitro assays with microfluidics to assess nanoparticle Dissolution of nanocrystals appears to play a key role in its efficacy since the nanocrystals themselves were unable to penetrate deep within the cancerous cell mass. CPT-UM nanocrystals had the highest release rates of CPT. The release profile data suggested that CPT's increased solubility arose from its interactions with serum proteins. 36,37 Nanocrystalline constructs containing PEG or PEG-FA coating exhibited increased IC50 values compared to CPT-UM. Folate receptors (FR's) are known to be overexpressed in a variety of breast cancer cell lines. 16 For 4T1 cells, FR's are also slightly overexpressed on their cell surfaces. 14 Some decrease in IC50 value was seen for CPT-FA compared to CPT-PEG, thus suggesting the role of FA targeting. However, the effect was modest, thus indicating that the primary effect of CPT appears to be through drug dissolution.