(M1130-10-56) Screening of Significant Molecular Descriptors Affecting Particle Size: A Smart Exemplary to Developing Efficient Lipid-Based Delivery Systems in Pre-formulation Studies

Purpose: The phases of drug development are challenging in itself, ranging from series of events: screening of perfect molecules to delivering them the right way and finally making sure they are safe. The screening of active molecules in the initial stages is difficult, the moieties with high efficiency to risks ratio are certainly required. Therefore, computational screening methods are to rescue for larger databases screening, techniques like QSAR, molecular docking, and molecular dynamics are the best solutions to tedious failed experiments. Following this, the very crucial step in curing diseases is delivering the molecules in the right way! To exploit the potent activity of drugs, the physiochemical properties of drugs play a significant role. Therefore, physicochemical properties like Molecular weight, Partition coefficient (Log P), Melting point, Polar Surface area (PSA), and others should be taken into consideration while screening different formulations. More than 50% of drugs approved fall under BCS (Biopharmaceutics Classification System) Class 2, requiring solubility enhancement techniques. Recent publications highlight the effectiveness of SNEDDS (Self Nano Emulsifying Drug Delivery System) for increasing solubility of moieties. Earlier preliminary studies for SNEDDS development were saturation solubility of entities in excipients and construction of ternary phase diagrams. The primary purpose of our research flows around exploring the role of physicochemical properties of drugs in developing SNEDDS of different hydrophobic moieties. In this study, a set of drugs (n=40) was screened, wherein certain drugs emulsify to form systems with particle size than the absence of a drug. There was an indication that certain drugs tend to behave differently than the others in group, therefore, different studies were carried out to form SNEDDS system of these drugs and evaluated for certain properties.

Methods: The overall research was divided into a) experimental methodology and b) statistical methodology for the complete screening of drugs. The experimental studies were performed on set of 40 drug moieties with diverse physicochemical properties, i.e., poor water solubility and majorly log P > 2. Initially SNEDDS (Self nano-emulsifying drug delivery systems) were formulated employing Captex^® 300 (Medium Chain Triglyceride) as lipid phase, Kolliphor^® EL (a non-ionic surfactant) and DMA (N, N-Dimethyl Acetamide) as co-solvent. The procedure was adapted from our previous publications. Particle size (PS) and polydispersity index of nanoglobules prepared were determined using Malvern zetasizer Nano ZS, UK. Following, the effect of different solvents and lipids was also studied on the drugs with distinct behavior. The experimental results of distinct particle size differences were correlated to the computational results. Statistical methodology involved the series of steps: 1. Collection of physicochemical properties of drugs from literature 2. Calculation of molecular descriptors of drugs using the E-dragon software3. Correlation of molecular descriptors with particle size of SNEDDS of drugs (understanding correlation of which molecular descriptors affect particle size) using pearson’s correlation analysis 4. Further the correlation was depicted through two different graphical illustrations of principal component analysis (PCA) and Heat map.

Results: The experimental inference clearly showed 5 different moieties which drastically decrease particle size of SNEDDS in comparison to blank SNEDDS. Whereas incorporation of some drugs showed formation of destabilized nanoemulsion. These results were significantly correlated with the structural/ physicochemical properties of drugs. The computational analysis using the molecular descriptors depicts that the properties related to structure play a key role in depicting the final particle size of SNEDDS. This was confirmed when mostly topological molecular descriptors have a significant correlation with particle size. The significant topological descriptors are HNAr, GNar, PW4, Lop, JGI1, VRA², VRm² showed significant correlation to PS(p-value < 0.05). To further provide with proof of concept, some more experimental studies performed showed that this change in particle size was only evident in presence of DMA (N, N-Dimethyl Acetamide), the other solvents screened were N-methyl-2-pyrrolidone, Ethyl lactate, Isosorbide dimethyl ether and Diglyme. In the presence of DMA as a solvent, there was significant decrease in particle size of SNEDDS. Further, the evident interactions of drugs with DMA are to be confirmed by using NMR. Therefore, indicating structural properties play crucial role in understanding the effect on predicting formulation properties.

Conclusion: PS was successfully studied for correlation with different physicochemical properties of drugs. The significance of distinct behavior of drugs was confirmed through structural interactions of drugs with DMA as co-solvent, therefore leading to lower particle size systems. The study has immense importance in the development of LBDDS in the pre-clinical phase of delivery systems.


Figure 1. (A) Molecular descriptors as new aid in screening drug delivery systems(B)Drugs screening methodology: 1. experimental methodology for formulation of SNEDDS of 40 different drugs 2. Statistical methodology for calculation of different molecular descriptors for drugs.


Figure 2. (A) Dataset of drugs evaluated in the screening process(B)Correlation analysis results1. Heat map showing an evident correlation between topological descriptors and particle size 2. Principal component analysis(PCA) for understanding the most significant descriptors from the dataset of almost 1000 descriptors.