Mathematical Modeling: A Tool for Optimization of Lipid Nanoparticle-Mediated Delivery of siRNA.

Lipid nanoparticles (LNPs) have been used to successfully deliver small interfering RNAs (siRNAs) to target cells in both preclinical and clinical studies and currently are the leading systems for in vivo delivery. Here, we propose the use of an ordinary differential equation (ODE)-based model as a tool for optimizing LNP-mediated delivery of siRNAs. As a first step, we have used a combination of experimental and computational approaches to develop and validate a mathematical model that captures the critical features for efficient siRNA-LNP delivery in vitro. This model accurately predicts mRNA knockdown resulting from novel combinations of siRNAs and LNPs in vitro. As demonstrated, this model can be effectively used as a screening tool to select the most efficacious LNPs, which can then further be evaluated in vivo. The model serves as a starting point for the future development of next generation models capable of capturing the additional complexity of in vivo delivery.

Challenges in the pharmaceutical development of lipid-based short interfering ribonucleic acid therapeutics.

INTRODUCTION: Harnessing RNA interference as a therapeutic approach has the potential to significantly expand the druggable target space, offering new hope for treatment of diseases that cannot be addressed with existing classes of drugs. A number of siRNA therapeutics have already progressed into preclinical and clinical development. Of these, lipid-based systems have emerged as one of the most mature classes of systemic delivery technologies. Despite tremendous advances in development, a number of significant challenges must still be addressed to enable commercialization of a lipid-based siRNA pharmaceutical product. AREAS COVERED: This review addresses specific challenges inherent to the pharmaceutical development of lipid-based siRNA therapeutics. Focus is placed on the development of a robust manufacturing process, the setting of appropriate product specifications and controls, development of strategies to assess and ensure product stability, and the evaluation of product comparability throughout development. EXPERT OPINION: Discovering and developing a lipid-based siRNA therapeutic that can be commercialized requires engineering a particle that selectively and efficiently delivers the cargo to the target tissue and cellular compartment. The particle assembly must be strictly controlled and physical properties thoroughly characterized to successfully develop an understanding of particle attributes that influence in vivo pharmaceutical properties. Correlation of particle physio-chemical properties to product performance is the foundation for advancements in discovery and assuring quality in a commercial drug product. Although difficult, we believe these development challenges can be addressed with appropriate scientific resources and that the industry will continue to progress siRNA therapeutic candidates through clinical development.

Lipid nanoparticle purification by spin centrifugation-dialysis (SCD): a facile and high-throughput approach for small scale preparation of siRNA-lipid complexes.

This paper describes the use of spin centrifugation-dialysis (SCD) for small-scale concentration/purification of siRNA-lipid complexes designed for use as therapeutic agents for gene silencing. SCD consists of a two-step method for concentration, filtration and buffer exchange of lipid nanoparticles (LNP) to provide a homogeneous preparation suitable for injection. Here, we compare SCD with the more traditionally used tangential flow filtration (TFF), and demonstrate the physicochemical and biological comparability of LNPs produced with both methods. TFF is a highly scalable method used in both developmental and production applications, but is limited in terms of miniaturization. In contrast to TFF, SCD is faster, less expensive, and requires less oversight for assembling LNPs for small-scale applications, such as target screening both in vitro and in vivo. The finding that SCD is a viable method for filtering LNPs in a manner similar to TFF, producing particles with comparable properties and biological activity, is significant given the complexity and sensitivity of LNPs to processing conditions.

External mechanical strain regulates membrane targeting of Rho GTPases by controlling microtubule assembly.

Transmission of externally applied mechanical forces to the interior of a cell requires coordination of biochemical signaling pathways with changes in cytoskeletal assembly and organization. In this study, we addressed one potential mechanism for this signal integration by applying uniform single external mechanical strains to aortic smooth muscle cells (SMCs) via their adhesion substrate. A tensile strain applied to the substrate for 15 min significantly increased microtubule (MT) assembly by 32 +/- 7%, with no apparent effect on the cells' focal adhesions as revealed by immunofluorescence and quantitative analysis of Triton X-100-insoluble vinculin levels. A compressive strain decreased MT mass by 24 +/- 9% but did not influence the level of vinculin in focal adhesions. To understand the decoupling of these two cell responses to mechanical strain, we examined a redistribution of the small GTPases RhoA and Rac. Tensile strain was found to decrease the amount of membrane-associated RhoA and Rac by 70 +/- 9% and 45 +/- 11%, respectively, compared with static controls. In contrast, compressive strain increased membrane-associated RhoA and Rac levels by 74 +/- 17% and 36 +/- 13%, respectively. Disruption of the MT network by prolonged treatments with low doses of either nocodazole or paclitaxel before the application of strain abolished the redistribution of RhoA and Rac in response to the applied forces. Combined, these results indicate that the effects of externally applied mechanical strain on the distribution and activation of the Rho family GTPases require changes in the state of MT polymerization.

Externally applied cyclic strain regulates localization of focal contact components in cultured smooth muscle cells.

Mechanical signals are critical regulators of cellular gene expression, yet little is understood of the mechanism whereby cells sense mechanical forces. In this study we have tested the hypothesis that mechanical strain applied to populations of cells via their adhesion substrate rapidly alters the cellular distribution of focal contact proteins. Focal contact-associated components (vinculin, a-actinin, paxillin) were assayed by immunofluorescence microscopy and quantitative western blotting. Application of a single step increase in strain in multiple experiments caused overall a small change in focal contact-associated vinculin. In contrast, cyclic strain induced a large and very reproducible increase in detergent-insoluble vinculin (52% relative to static) after just 1 min of strain. Insoluble paxillin was transiently enriched with a similar time course, whereas insoluble a-actinin did not change significantly in response to cyclic strain. Rhodamine-labeled chicken vinculin added to permeabilized cells preferentially localized to focal contacts in response to cyclic strain, but not a single step increase in strain. These findings establish that insoluble levels of focal contact components are altered rapidly following application of an appropriate number of mechanical perturbations, and suggest that at least one component of the mechanism does not involve soluble intermediates.

Quantification of fibronectin adsorption to silicone-rubber cell culture substrates.

As the role of mechanical force in cellular signaling gained recognition, investigators designed a number of devices to deliver controlled regimens of mechanical force to cultured cells. One type of device uses thin silicone-rubber membranes to support monolayer cell adhesion and to transmit mechanical force in the form of biaxial strain. We have observed that cell attachment and spreading are impaired on these membranes compared to polystyrene, even when both are passively coated with identical amounts of extracellular matrix. The purpose of these studies was to quantify the efficiency and stability of passive matrix adsorption onto commercially available elastic culture substrates. A theoretically saturating density (1 microg/cm2) of fibronectin was added to each well, and the initial efficiency of adsorption to the walls and elastic membranes was found to be 31 +/- 2% of the protein added. Strikingly, when the protein adsorbed specifically to the membranes was quantified after seven days, only 10-26 ng/cm2 fibronectin were present, revealing that most of the adsorption is to the sides of the wells. These results indicate that the adsorption of matrix proteins to silicone-rubber substrates is relatively inefficient and that investigators who use these systems must be aware of this fact and design their experiments accordingly.