(291c) Cyto-Toxicity and Biocompatibility of a Family of Choline Phosphate Ionic Liquids

Recently a number of ionic liquids have been demonstrated to improve the thermostability and shelf life of proteins in liquid formulations, and thus show great promise as stabilizing excipients or solvents for protein therapeutics. Ionic liquids (ILs) consist of an anion and cation, but unlike typical salts, they are liquid at/near room temperature and can thus replace water and other solvents in a range of applications including protein-based pharmaceutical formulations intended for parenteral injection. The anion and cation choice can be tailored to provide desired solvent characteristics. When rationally designing new inactive compounds for use within the body, it can be beneficial to use starting materials that are already FDA approved for drug formulation or alternatively are naturally found within the body. Because of the prevalence of choline and phosphates in the human body, a family of choline phosphate salts was synthesized according to methods in the literature, and characteristics related to their biocompatibility were studied. Ionic liquids prepared for use in this work incorporate choline as a cation and the following anions dihydrogen phosphate (CDHP), dibutyl phosphate (CDBP), bis(2-ethyl hexyl) phosphate (CBEH), bis(2,4,4,-trimethyl pentyl) phosphinate (CTMP), and O,O′-diethyl dithiophosphate (CDEP). Simple salts and sugars, and an ammonium based ionic liquid, ethylammonium nitrate (EAN) were included for comparison. Both CDHP and EAN have been shown to dramatically improve the shelf life of several proteins.

To establish the relative toxicity of these compounds, mouse macrophages (J774) were used as a model cell type. Cell viability following exposure to ILs was measured using a resazurin-based fluorometric assay to estimate the number of metabolically active cells in 96-well plates. Viable cells reduce resazurin to fluorescent resorufin whereas nonviable cells do not generate a fluorescent signal. Stock solutions of ionic liquids were prepared in culture medium. Cells were exposed to 1:1 serial dilutions of each IL compound in 96-well plates for 48 hours. Each plate contained a media and an untreated cells control. Concentrations ranged from 0.2 to 113 mM, and were optimized for complete concentration-response curves including hormetic effects. The simple sugar (trehalose) showed incomplete inhibition in the concentration range tested and is reported with an EC₅₀ >100 mM, while the simple salts (sodium chloride, and choline chloride) had EC₅₀ values of 67 and 35 mM respectively. The EC₅₀ values (mM) of CDHP (19), CDBP (8.4), CDEP (7.4), and EAN (7.1) were lower than simple salts, suggesting that they were modestly more toxic, but still within the range of typical physiologic salts. The sulfanyl and sulfanylidene groups on CDEP did not appear to adversely affect toxicity compared to CDBP or EAN, but these three were approximately twofold more toxic than CDHP. The compounds CTMP and CBEH gave EC₅₀ values of 0.25 and 0.24 mM, respectively, lower than all the other compounds tested. This can be attributed to the longer and branched alkyl chains. In the case of both CDHP and CBEH, cell viability increased with increasing concentration before it was reduced with additional increasing ionic liquid concentration, i.e., both compounds induced a subtoxic stimulus or hormetic response. The hormetic effect of CDHP showed considerable variability. In the preparation of the stock CDHP in culture medium, the pH of the medium decreased indicating that the buffer capacity of the media had been exceeded. The acidity of the CDHP media may be the source of toxicity at higher CDHP concentrations and warrants further investigation.

To improve the biocompatibility of injected therapeutics the osmolality of these solutions are ideally formulated to be near the osmolality of extra- and intracellular fluids otherwise localized cell volumes can be altered, resulting in pain at the injection site. Because ionic liquids can range from molecular solution behavior to that of a fully dissociated ionic solution, it is important to understand this behavior in order to predict final solution osmolalities in formulations. Osmolality measurements of ILs formulated at 100 mM were determined using a vapor pressure osmometer. From these measurements, osmotic coefficients (φ) which account for the degree of non-ideality of a solution were calculated so that the degree of dissociation of each ionic liquid could be assessed. CDHP, CDBP, CDEP, and EAN have osmotic coefficients of φ = 0.8, 0.7, 0.5, and 0.7 respectively at 100 mM. Although the ionic liquids are composed of ionic salts, it appears that they have not dissociated completely or electrostatic effects are causing a depression in the coefficient. CTMP (φ = 0.3) and CBEH (φ = 0.1) have very low coefficients. Although most compounds were completely miscible with water, CTMP exhibited phase separation at 4 °C and room temperature, and appeared to be an emulsion or dispersion at 37 °C. With osmotic coefficients below 0.5, CTMP and CBEH appear to show more molecular versus ionic character. It is also possible that particle association is occurring, especially in the case of CTMP, which may not be fully soluble in aqueous based solutions.

This family of choline phosphate salts exhibit a range of toxicities from low (CDHP) to moderate (CTMP and CBEH) compared to other ILs that have been formulated for non-biological purposes, and many are in the range of basic physiologic salts. Some of the ILs exhibit molecular behavior, thus minimizing the contribution to osmolality compared to traditional salts. If a high IL concentration is required to achieve a stabilization effect these ILs might be best implemented as co-formulation stabilizers within slow-release nano- and micro-particles to maximize the biocompatibility.