Nanomedicine is a rapidly developing new field that utilizes nanoparticulate gene/drug delivery systems for therapeutic applications. The gene/drug delivery system’s size, size distribution, surface charge, morphology, and encapsulation efficiency play an extremely important role in the success and efficacy in vitro and in vivo. The current method of preparing these systems typically involve bulk mixing (BM) of the gene or drug with one or more condensing components such as polymer and/or lipids followed by vortexing. In addition, shielding and/or targeting components are also added during or post assembly of the systems. The BM process of multiple components often results in systems with heterogeneous populations. Therefore, an engineering approach to achieve controlled mixing of the different components was proposed.
Microfluidics is a versatile technology specialized for manipulation of liquid flows at the micrometer and picoliter scale and has been applied to proteomics, diagnostics, and therapeutics. Microfluidic devices generally operate at low Reynolds number so the flow is strictly laminar which allows well-defined mixing to be controlled solely by interfacial diffusion. Therefore, in this study, microfluidic devices with hydrodynamic focusing (MF) were developed for assembling polymer-DNA (polyplex) and lipid-polymer-DNA (lipopolyplex, LP) nanoparticles.
Polyethylenimine (PEI) can condense plasmid DNA (pDNA) into PEI/pDNA complexes for nonviral gene delivery. The conventional method for producing these complexes involves bulk mixing (BM) of PEI and DNA followed by vortexing, which at low N/P ratios results in large particle size distribution, low cytotoxicity, and poor gene transfection, whereas at high N/P ratios results in small particle size and better gene transfection but high cytotoxicity. To improve size control, gene transfection efficiency, and cytotoxicity, in this study, we used a 3-inlet MF device to synthesize PEI/pDNA complexes at N/P=3.3 and 6.7. Bulk mixing was used as a control, mouse NIH 3T3 fibroblast cells and mouse embryonic stem (mES) cells as model cell lines, plasmid encoding green fluorescent protein (pGFP) and secreted alkaline phosphatase (pSEAP) as the reporter gene, and commercially available Lipofectamine2000 as a positive control. The complexes were characterized by atomic force microscopy (AFM), dynamic light scattering goniometry, and zeta potential measurement. Confocal laser scanning microscopy (CLSM) and fluorescent labeling techniques were used to visualize the complex size distribution, complexation uniformity, and cellular distribution. The results showed that MF produced complexes were smaller, more uniformly complexed, had higher cell viability, and improved exogenous gene expression.
To further extend this technology, a multi-inlet MF system to prepare lipopolyplex (LP) nanoparticles containing Bcl-2 antisense deoxyoligonucleotide (ODN) was developed and evaluated. The lipopolyplexes (LPs) consist of ODN:protamine:lipids (1:0.3:12.5 wt/wt ratio) and the lipids included DC-Chol:egg PC:PEG-DSPE (40:58:2 mol/mol%). Using K562 human erythroleukemia cells, which contain an abundance of Bcl-2 and overexpression of transferrin receptors (TfR), and G3139 (oblimerson sodium or GenasenseTM) as a model cell line and drug, respectively, the Bcl-2 downregulation at the mRNA and protein levels were compared between the conventional bulk mixing (BM) method and the MF method, in addition to cellular uptake and apoptosis. The lipopolyplex size and surface charge were characterized by dynamic light scattering (DLS) and zeta potential (ζ) measurement, respectively, while the ODN encapsulation efficiency was determined by gel electrophoresis. Cryogenic transmission electron microscopy (Cryo-TEM) was used to determine the morphology of LPs. Our results demonstrated that MF produced LP nanoparticles had similar structures but smaller size and size distribution compared to BM LP nanoparticles. MF LP nanoparticles had higher level of Bcl-2 antisense uptake and showed more efficient downregulation of Bcl-2 protein level than BM LP nanoparticles.