Polyester Nanoparticles with Controlled Topography for Peroral Drug Delivery Using Insulin as a Model Protein

Abstract

Receptor-mediated polyester drug delivery systems have tremendous potential for improving the clinical performance of existing pharmaceutical drugs. Despite significant progress made in this area, it remains unclear how and to what extent the polyester nanoparticle surface topography would affect the in vitro, ex vivo and in vivo performance of a drug, and if there exists a correlation between in vitro and in vivo, as well as healthy versus pathophysiological states. Herein, we report a systematic investigation of the interactions between ligands and receptors as a function of the linker length, two-carbon (2C) versus four-carbon (4C). The in vitro, ex vivo and in vivo in healthy models validate the hypothesis that 4C has better reach and binding to the receptors. The results indicate that 4C offered better performance over 2C in vivo in improving the oral bioavailability of insulin (INS) by 1.1-fold (3.5-fold compared to unfunctionalized nanoparticles) in a healthy rat model. Similar observations were made in pathophysiological models; however, the effects were less prominent compared to those in healthy models. Throughout, ligand decorated nanoparticles outperformed unfunctionalized nanoparticles. Finally, a semimechanistic pharmacokinetic and pharmacodynamic (PKPD) model was developed using the experimental data sets to quantitatively evaluate the effect of P2Ns-GA on oral bioavailability and efficacy of insulin. The study presents a sophisticated oral delivery system for INS or hydrophilic therapeutic cargo, highlighting the significant impact on bioavailability that minor adjustments to the surface chemistry can have.

Keywords: nanobiointerface; nanotopography; oral insulin delivery; polyester nanoparticles; receptor-mediated transcytosis.

Figure 1.

Schematic of P2s, and synthesis of P2s-2C-GA or P2s-4C-GA and schematic of nanoparticles: i) EDC.HCl, DIEA, N-boc ethylenediamine/butanediamine, dimethylformamide, dichloromethane, 0 °C to r.t.; ii) trifluoroacetic acid, dichloromethane, 0 °C; iii) EDC.HCl, DIEA, gambogic acid, dichloromethane, 0 °C to r.t.; iv) w/o/w emulsion; INS-laden nanoparticle schematics with 2C and 4C chains coupled to GA (P2Ns-2C-GA and P2Ns-4C-GA).

Figure 2.

a) SEM micrographs of unfunctionalized and GA-functionalized nanoparticles with varying linker length. b) Corresponding particle size distribution plots determined via ImageJ.

Figure 3.

In vitro, ex vivo and in vivo performance of P2Ns/P2Ns-GA as a function of linker length, in healthy models. a) Confocal micrographs depicting cell uptake of nanoparticles in normal human intestine cells, FHs74; b) Mean fluorescence intensity of corresponding nanoparticles determined via flow cytometry; c) Fluorescent intensity penetrated through intestinal sacs into media in ex vivo study; d) INS plasma concentration profiles for P2Ns, P2Ns-2C-GA and P2Ns-4C-GA, in vivo, in healthy rat model. Data presented as mean ± standard error. *P < 0.05, ***P < 0.001, ****P < 0.0001; one-way ANOVA, with comparison to control group.

Figure 4.

Efficacy of oral INS-laden nanoformulations as a function of ligand, linker length and LA INS on blood glucose levels of STZ-induced Type-1 diabetes rat models. a) Daily dosing and blood glucose measurement strategy, b-c) Line graphs of blood glucose measurements assessed on day 1 and day 11 of study; d-e) Corresponding percentage reduction in blood glucose levels on day 1 and day 11; f-g) Fasting blood glucose levels assessed on both day 1 and day 11 following a 4-hour fasting period; h) Rat (endogenous) INS panel; and i) bovine INS vs human INS. Data presented as mean ± standard error, (n = 6). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; one-way ANOVA, with comparison to the diabetes group in d-g, and comparison with D+LA-INS in h-i.

Figure 5.

Efficacy of oral INS-laden nanoformulations as a function of ligand, linker length and LA INS on inflammatory and biochemistry panels in STZ-induced diabetes rat models. a) cf-DNA was assessed in the terminal plasma, monitoring the oxidized DNA fragment level; b-j) biochemistry panel which includes cholesterol, liver enzymes (amylase, ALT, AST, ALP, GGT), BUN, creatinine and CK in plasma. Data presented as mean ± standard error, (n = 6). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; one-way ANOVA, with comparison to the diabetes group.

Figure 6.

Superimposed model predicted a) pharmacokinetics (i.e., insulin concentrations), and b) pharmacodynamics (i.e., glucose levels) of insulin-laden nanoformulations.