Investigating the impact of aromatic ring substitutions on selectivity for a multimodal anion exchange prototype library

The increasing prevalence of low pI non-mAb therapeutics as well as current challenges in mAb-aggregate separations and low recoveries motivate further development in the multimodal anion exchange (MM AEX) space. In this work, linear salt gradient experiments at pH 7 were used to evaluate the retention of model proteins (with pI from 3.4 to 6.8) in 17 novel MM AEX prototype systems. The ligands were organized into three series. Series 1 extended previous work in multimodal ligand design and included a hydroxyl variant and linker length variants. Series 2 and 3 investigated the nature of hydrophobicity in MM AEX systems by adding hydrophobic (series 2) or fluorine (series 3) substituents to a solvent exposed phenyl ring. Compared to the commercial resin Capto Adhere, the series 1 and 3 ligands exhibited weaker binding, while some of the series 2 aliphatic prototypes showed dramatically increased retention and unique selectivities. Within series 1, the model proteins eluted earlier in the gradient as the charge-hydrophobic group distance on the ligand was increased from 4.9 Å to 8.5 Å. For the aliphatic variants in series 2, proteins that eluted early in the salt gradient were not affected by the increase in ligand hydrophobicity, while the later eluting proteins bound stronger as the length of the aliphatic substituent increased. The series 3 variants indicated that phenyl ring fluorination created subtle changes in protein elution in these MM AEX systems. Retention data from the three series was used to generate a partial least squares QSAR model based on both protein and ligand descriptors which accurately predicted protein retention with a training R2 of 0.81 and a test R2 of 0.76. The retention characteristics of some prototypes such as the earlier elution and unique selectivities compared to Capto Adhere suggest that they could potentially provide unique selectivities and increased recovery for the downstream processing of both mAb and non-mAb biotherapeutics.