Hinge Truncation to Improve Aggregation Kinetics and Thermal Stability of an Antibody Fab Fragment

Abstract

The hinge region of antibody fragments plays a crucial role in their stability and aggregation properties. In this study, we investigated the effects of hinge truncations on the thermal stability and aggregation propensity of the A33 Fab antibody fragment. Eight Fab variants were engineered by introducing stop codons to truncate 1-8 residues at the hinge region (heavy chain residues 221-228). These variants were then expressed, purified, and characterized in terms of stability and aggregation propensity using SDS-PAGE, SEC-HPLC, LC-MS, and thermal stability assays. Our findings demonstrate that truncating the hinge region can enhance the thermal stability and reduce the aggregation of Fab fragments, and that progressive truncations identified an optimal hinge length for stability. Notably, the 227TGA variant exhibited a significant 14.5% reduction in aggregation rate compared to the wild type, without compromising thermal stability. By contrast, 221TGA removed all of the hinge and reduced the aggregation rate by 13%, but also decreased the thermal stability. These results suggest that hinge truncation is a promising strategy for improving the developability of therapeutic antibody Fab fragments by mitigating some of the stability issues associated with aggregation.

Keywords: aggregation; engineering; hinge; stability; therapeutic; truncation.

1

Structure of A33 Fab with modifications at the hinge region. Light chain (green) and heavy chain (cyan) are depicted in cartoon format. The five disulfide bonds are represented in sticks and colored yellow, with the interchain disulfide bond between LC214 and HC220 indicated by an arrow. Residues 221–228 of the heavy chain, highlighted in red and also indicated by an arrow, are progressively truncated from 1 to 8 residues to create eight variant structures with a shortened hinge region.

2

Comparative SDS-PAGE analysis under nonreducing (A) and reducing (B) conditions. Protein ladders are loaded in the outermost lanes of both gels. Eight Fab variants of 221TGA to 228TGA along with the wild type samples are loaded in the center wells, progressing from left to right.

3

HPLC size exclusion chromatograms of the wild type (WT) and eight hinge-truncated variants from calibration. Samples were prepared at a concentration of 2 mg/mL, with injection volumes of 2 μL. Each variant is distinguished by a different color. The dimer species predominantly elutes at approximately 2.2–2.3 min, whereas the monomer elutes at around 2.7 min. Panel (B) provides a zoomed-in view of Panel (A) focusing on the elution time between 2.0 and 2.6 min.

4

Mass spectroscopy analysis of the variants 221TGA (A), 222TGA (B) and wild type (C). Background signals were subtracted prior to deconvolution of the data. The complete data set for all the variants can be found in the Supporting Information.

5

Barycentric mean (BCM) of intrinsic fluorescence spectra across temperature gradients for variants 221TGA (A&C) and 222TGA (B&D), fitted to van’t Hoff thermal transitions. The black squares represent actual measurements while the red lines indicate model fittings using two-state (top, A,B) and three-state (bottom, C,D) equations. The transition temperatures are marked with arrows. Comments on the fitting quality are included to assess the adequacy of the model fit for each scenario.

6

Melting temperature (T _m, orange bars) and van’t Hoff entropy change (ΔS _vh, green bars) for the Fab wild type (WT) and its variants labeled as 221TGA to 228TGA, fitted with a two-state thermal transition equation. For the 221TGA and 222TGA variants, values from a three-state fit are included. Dashed reference lines indicate the baseline values for WT, facilitating direct comparison of each variant’s stability relative to the WT.

7

Static light scattering analysis of Fab variants. Panels (A) and (B) display the static light scattering (SLS) intensity profiles for the Fab wild type (WT) and its variants at 266 and 473 nm wavelengths, respectively. Each variant is represented by a distinct color with three repeat measurements in the same color. The corresponding aggregation onset temperatures (T _agg) are summarized in panel (C), indicating T _agg values at 266 nm (orange bars) and 473 nm (green bars), with standard error of the mean (SEM) represented by error bars for each data point. Dashed reference lines indicate the baseline values for WT, facilitating direct comparison of each variant’s stability relative to the WT.

8

Aggregation kinetics as measured by SEC-HPLC. Panel (A) the monomer retention over time for the wild type (WT) and eight hinge-truncated variants. All three replicas are shown for each variant at each time-point. For each variant, a single monomer retention curve is fitted to the combined replica data (24 data points) using eq to derive the kinetic constant, k, modeled reaction coefficient, A, and associated standard fitting errors. Unique colors and symbols represent each variant. Panel (B) aggregation rate constant k and initial aggregation rate derived from eq and A × k respectively. Reference lines for WT are included for comparison. Error bars represent standard errors based on parameter fitting errors.

9

Linear regression analysis of the initial aggregation rate versus van’t Hoff entropy change (ΔS _vh) for various Fab variants. The red line represents the linear fit, excluding outliers 221TGA and 222TGA.