Stability of monoclonal antibodies at high-concentration: head-to-head comparison of the IgG1 and IgG4 subclass.

Few studies have so far directly compared the impact of antibody subclass on protein stability. This case study investigates two mAbs (one IgG1 and one IgG4 ) with identical variable region. Investigations of mAbs that recognize similar epitopes are necessary to identify possible differences between the IgG subclasses. Both physical and chemical stability were evaluated by applying a range of methods to measure formation of protein aggregates [size-exclusion chromatography (SEC)-HPLC and UV340 nm], structural integrity (circular dichroism and FTIR), thermodynamic stability (differential scanning calorimetry), colloidal interactions (dynamic light scattering), and fragmentation and deamidation (SEC-HPLC and capillary isoelectric focusing). The impact of pH (4-9) and ionic strength (10 and 150 mM) was investigated using highly-concentrated (150 mg/mL) mAb formulations. Lower conformational stability was identified for the IgG4 resulting in increased levels of soluble aggregates. The IgG1 was chemically less stable as compared with the IgG4 , presumably because of the higher flexibility in the IgG1 hinge region. The thermodynamic stability of individual mAb domains was also addressed in detail. The stability of our mAb molecules is clearly affected by the IgG framework, and this study suggests that subclass switching may alter aggregation propensity and aggregation pathway and thus potentially improve the overall formulation stability while retaining antigen specificity.

Viscosity of high concentration protein formulations of monoclonal antibodies of the IgG1 and IgG4 subclass - prediction of viscosity through protein-protein interaction measurements.

The purpose of this work was to explore the relation between protein-protein interactions (PPIs) and solution viscosity at high protein concentration using three monoclonal antibodies (mAbs), two of the IgG4 subclass and one of the IgG1 subclass. A range of methods was used to quantify the PPI either at low concentration (interaction parameter (kD) obtained from dynamic light scattering, DLS) or at high concentration (solution storage modulus (G') from ultrasonic shear rheology). We also developed a novel method for the determination of PPI using the apparent radius of the protein at either low or high protein concentration determined using DLS. The PPI measurements were correlated with solution viscosity (measured by DLS using polystyrene nanospheres and ultrasonic shear rheology) as a function of pH (4-9) and ionic strength (10, 50 and 150 mM). Our measurements showed that the highest solution viscosity was observed under conditions with the most negative kD, the highest apparent radius and the lowest net charge. An increase in ionic strength resulted in a change in the nature of the PPI at low pH from repulsive to attractive. In the neutral to alkaline pH region the mAbs behaved differently with respect to increase in ionic strength. Two mAbs (A and B) showed little or no effect of increasing ionic strength, whereas mAb-C showed a remarkable decrease in attractive PPI and viscosity. Previous studies have mainly investigated mAbs of the IgG1 and IgG2 subclass. We show here, for the first time, that mAbs of the IgG4 subclass behave similar as the other subclasses. By comparison of the three tested mAbs with mAbs investigated in other studies a clear linear trend emerges between the pH of strongest attractive PPI and highest solution viscosity. The determination of PPI using either kD or apparent radius is thus a useful prediction tool in the determination of solution conditions that favors low solution viscosity at high protein concentration of therapeutically used mAb molecules. The novel methodology using apparent radius is a simple and rapid alternative to determine relative PPI directly under formulation conditions. The method can potentially serve as a high-throughput screening tool in formulation development.

Bupivacaine salts of diflunisal and other aromatic hydroxycarboxylic acids: aqueous solubility and release characteristics from solutions and suspensions using a rotating dialysis cell model.

In the search for poorly soluble bupivacaine salts potentially enabling prolonged postoperative pain relief after local joint administration in the form of suspensions the solubility of bupivacaine salts of diflunisal and other aromatic hydroxycarboxylic acids were investigated together with the release characteristics of selected 1:1 salts from solutions and suspensions using a rotating dialysis cell model. The poorest soluble bupivacaine salts were obtained from the aromatic ortho-hydroxycarboxylic acids diflunisal, 5-iodosalicylic acid, and salicylic acid (aqueous solubilities: 0.6-1.9 mM at 37 degrees C). Diffusant appearance rates in the acceptor phase upon instillation of solutions of various salts in the donor cell applied to first-order kinetics. Calculated permeability coefficients for bupivacaine and the counterions diflunisal, 5-iodosalicylic acid, and mandelic acid were found to be correlated with the molecular size of the diffusants. Release experiments at physiological pH involving suspensions of the bupivacaine-diflunisal salt revealed that at each sampling point the diflunisal concentration exceeded that of bupivacaine in the acceptor phase. However, after an initial lag period, a steady state situation was attained resulting in equal and constant fluxes of the two diffusants controlled by the permeability coefficients in combination with the solubility product of the salt. Due to the fact that the saturation solubility of the bupivacaine-salicylic acid salt in water exceeded that of bupivacaine at pH 7.4, suspensions of the latter salt were unable to provide simultaneous release of the cationic and anionic species at pH 7.4. The release profiles were characterised by a rapid release of salicylate accompanied by a much slower appearance of bupivacaine in the acceptor phase caused by precipitation of bupivacaine base from the solution upon dissolution of the salt in the donor cell.

Aqueous solubility study of salts of benzylamine derivatives and p-substituted benzoic acid derivatives using X-ray crystallographic analysis.

Twenty two p-substituted benzoic acid derivates were used to prepare salts of N-methylbenzylamine (II) and N,N-dimethylbenzylamine (III), respectively. Only five salts of (II) and two salts of (III) were obtained in a crystalline state. The solubility of these salts was orders of magnitude higher than those reported for the corresponding salts of benzylamine (I). Thermal analysis indicated that the increased solubility was caused by reduced crystal lattice energy, which was most likely due to the reduced number of strong hydrogen bonds of the salt of (II) and (III). X-ray crystallographic analysis of p-hydroxybenzoic acid salt of (I), (II) and (III) suggested that the reduced number of hydrogen bonds caused the apparent higher solubility. Further analyses of seven salts of (I) were performed. It was not possible to identify any relationship between the number of hydrogen bonds and the corresponding solubility of the salts.

Assessment of drug salt release from solutions, suspensions and in situ suspensions using a rotating dialysis cell.

A rotating dialysis cell consisting of a small (10 ml) and a large compartment (1000 ml) was used to study the release of drug salt (bupivacaine 9-anthracene carboxylate) from (i). solutions, (ii). suspensions and (iii). in situ formed suspensions. Initial release experiments from suspensions indicated that the release of drug salt in deionized water was predominantly limited by the diffusion across the membrane whereas it is essentially dissolution rate controlled in 0.05 M phosphate buffer (pH 7.40). Thus, the in vitro model appears to have a potential in formulation screening when phosphate buffer is used as release media. Generally, the initial release of the drug salt from in situ suspensions occurred faster as compared to conventional suspensions, probably due to incomplete precipitation of the drug salt, and hence formation of supersaturated solutions where the rate of release is predominantly determined by the concentration gradient. However, when an adequately concentrated solution of the drug salt was used to prepare the in situ suspension, the initial fast release was followed by a substantial sustained release indicating that the release had become dissolution rate limited.

Correlation of aqueous solubility of salts of benzylamine with experimentally and theoretically derived parameters. A multivariate data analysis approach.

Twenty two salts of benzylamine and p-substituted benzoic acids were prepared and characterized. The p-substituent was varied with regard to electronic, hydrophobic, and steric effects as well as hydrogen bonding potential. A multivariate data analysis was used to describe the relationship between the aqueous solubility of the salts and experimentally determined physicochemical parameters and theoretically derived molecular descriptors. The model, based on all descriptors, gave R(2)=0.86 and Q(2)=0.72. The most significant descriptors exhibiting VIP (variance of importance) values above 1.0 were intrinsic dissolution rate, intrinsic solubility of the unionized acids (S(0)), Hansch's hydrophobic parameter, Charton's steric parameter and molecular weight (MW). Statistically good models for predicting solubility of a selected test set were obtained by using simple models consisting of a few descriptors only: (i) Charton, Hansch and MW (R(2)=0.73; Q(2)=0.70), and (ii) Charton and S(0) (R(2)=0.74; Q(2)=0.72).

Characteristics of drug substances in oily solutions. Drug release rate, partitioning and solubility.

In vitro rate of drug release from oil solutions was investigated in a rotating dialysis cell. A log linear correlation was established between the rate constant (k(obs)) for attainment of equilibrium and apparent partition coefficient (P(app)) between oil vehicle and release media using various weak acids and bases and non-electrolytes. Collander like linear free energy relationships were observed allowing various oil-aqueous buffer partition coefficients to be calculated from known octanol-aqueous buffer partition coefficients. Solubility of the various drug substances in oil vehicles were investigated. A linear correlation was observed between log molar solubility and melting point of the solutes. Release profiles obtained for release of two local anaesthetics dissolved in the same oil vehicle exhibited an unexpected behavior involving an initial delayed release of the most lipophilic local anaesthetic.