Developability profiling of a panel of Fc engineered SARS-CoV-2 neutralizing antibodies.

To combat the COVID-19 pandemic, potential therapies have been developed and moved into clinical trials at an unprecedented pace. Some of the most promising therapies are neutralizing antibodies against SARS-CoV-2. In order to maximize the therapeutic effectiveness of such neutralizing antibodies, Fc engineering to modulate effector functions and to extend half-life is desirable. However, it is critical that Fc engineering does not negatively impact the developability properties of the antibodies, as these properties play a key role in ensuring rapid development, successful manufacturing, and improved overall chances of clinical success. In this study, we describe the biophysical characterization of a panel of Fc engineered ("TM-YTE") SARS-CoV-2 neutralizing antibodies, the same Fc modifications as those found in AstraZeneca's Evusheld (AZD7442; tixagevimab and cilgavimab), in which the TM modification (L234F/L235E/P331S) reduce binding to FcγR and C1q and the YTE modification (M252Y/S254T/T256E) extends serum half-life. We have previously shown that combining both the TM and YTE Fc modifications can reduce the thermal stability of the CH2 domain and possibly lead to developability challenges. Here we show, using a diverse panel of TM-YTE SARS-CoV-2 neutralizing antibodies, that despite lowering the thermal stability of the Fc CH2 domain, the TM-YTE platform does not have any inherent developability liabilities and shows an in vivo pharmacokinetic profile in human FcRn transgenic mice similar to the well-characterized YTE platform. The TM-YTE is therefore a developable, effector function reduced, half-life extended antibody platform.

Machine learning prediction of antibody aggregation and viscosity for high concentration formulation development of protein therapeutics.

Machine learning has been recently used to predict therapeutic antibody aggregation rates and viscosity at high concentrations (150 mg/ml). These works focused on commercially available antibodies, which may have been optimized for stability. In this study, we measured accelerated aggregation rates at 45°C and viscosity at 150 mg/ml for 20 preclinical and clinical-stage antibodies. Features obtained from molecular dynamics simulations of the full-length antibody and sequences were used for machine learning model construction. We found a k-nearest neighbors regression model with two features, spatial positive charge map on the CDRH2 and solvent-accessible surface area of hydrophobic residues on the variable fragment, gives the best performance for predicting antibody aggregation rates (r = 0.89). For the viscosity classification model, the model with the highest accuracy is a logistic regression model with two features, spatial negative charge map on the heavy chain variable region and spatial negative charge map on the light chain variable region. The accuracy and the area under precision recall curve of the classification model from validation tests are 0.86 and 0.70, respectively. In addition, we combined data from another 27 commercial mAbs to develop a viscosity predictive model. The best model is a logistic regression model with two features, number of hydrophobic residues on the light chain variable region and net charges on the light chain variable region. The accuracy and the area under precision recall curve of the classification model are 0.85 and 0.6, respectively. The aggregation rates and viscosity models can be used to predict antibody stability to facilitate pharmaceutical development.

Investigation of amino acid specificity in the CydX small protein shows sequence plasticity at the functional level.

Small proteins are a new and expanding area of research. Many characterized small proteins are composed of a single hydrophobic α-helix, and the functional requirements of their limited amino acid sequence are not well understood. One hydrophobic small protein, CydX, has been shown to be a component of the cytochrome bd oxidase complex in Escherichia coli, and is required for enzyme function. To investigate small protein sequence specificity, an alanine scanning mutagenesis on the small protein CydX was conducted using mutant alleles expressed from the E. coli chromosome at the wild-type locus. The resulting mutant strains were assayed for CydX function. No single amino acid was required to maintain wild-type resistance to ß-mercaptoethanol. However, substitutions of 10-amino acid blocks indicated that the N-terminus of the protein was required for wild-type CydX activity. A series of double mutants showed that multiple mutations at the N-terminus led to ß-mercaptoethanol sensitivity in vivo. Triple mutants showed both in vivo and in vitro phenotypes. Together, these data provide evidence suggesting a high level of functional plasticity in CydX, in which multiple amino acids may work cooperatively to facilitate CydX function.

In vitro reconstitution, functional dissection, and mutational analysis of metal ion transport by mitoferrin-1.

Iron is universally important to cellular metabolism, and mitoferrin-1 and -2 have been proposed to be the iron importers of mitochondria, the cell's assembly plant of heme and iron-sulfur clusters. These iron-containing prosthetic groups are critical for a host of physiological processes ranging from oxygen transport and energy consumption to maintaining protein structural integrity. Mitoferrin-1 (Mfrn1) belongs to the mitochondrial carrier (MC) family and is atypical given its putative metallic cargo; most MCs transport nucleotides, amino acids, or other small- to medium-size metabolites. Despite the clear importance of Mfrn1 in iron utilization, its transport activity has not been demonstrated unambiguously. To bridge this knowledge gap, we have purified recombinant Mfrn1 under non-denaturing conditions and probed its metal ion-binding and transport functions. Isothermal titration calorimetry indicates that Mfrn1 has micromolar affinity for Fe(II), Mn(II), Co(II), and Ni(II). Mfrn1 was incorporated into defined liposomes, and iron transport was reconstituted in vitro, demonstrating that Mfrn1 can transport iron. Mfrn1 can also transport manganese, cobalt, copper, and zinc but discriminates against nickel. Experiments with candidate ligands for cellular labile iron reveal that Mfrn1 transports free iron and not a chelated iron complex and selects against alkali divalent ions. Extensive mutagenesis identified multiple residues that are crucial for metal binding, transport activity, or both. There is a clear abundance of residues with side chains that can coordinate first-row transition metal ions, suggesting that these could form primary or auxiliary metal-binding sites during the transport process.