3.1. Antibody-epitope interface
Success factors for discovery of biopharmaceuticals include
understanding the epitope-efficacy relationship in addition to other
parameters such as affinity and manufacturability (Puchades et al.,
2019; Sun et al., 2021; Zhu et al., 2019). HDMS is robust, high in
throughput (Qi et al., 2021), ensuring quick and efficient determination
of binding regions in a near-native physicochemical condition — a
feature very attractive to the fast-paced pharmaceutical R&D sector. In
this regard, HDMS has made significant contributions to probing binding
interfaces of antibody-antigen complexes, i.e., epitope mapping, in
industry as well as academia (Figure 3A) (Masson et al., 2017).
In particular, exceptional applicability of HDMS to biomacromolecules is
useful for mapping an epitope for an antigen-antibody complex which is
usually as large-sized as more than 100 kDa — far beyond a measurable
limit of NMR analysis. Huang et al. subjected a homotrimeric cytokine
(90 kDa), TL1A, complexed with three antibodies (450 kDa) at an
equimolar binding stoichiometry to HDMS for the epitope mapping. The
analysis allowed residue-level determination of a discontinuous epitope
within a predicted binding interface of TL1A and its cognate receptor
DR3, rationalizing a strong potency of the antibody as a TL1A antagonist
(Huang et al., 2018). The epitope mapping by HDMS is less idiosyncratic
and resource-intensive, permitting comparative epitope analyses for a
high-molecular-weight antigen targeted by two distinct antibodies. For
instance, the epitopes of a large homotrimeric food allergen, cashew 11S
globulin allergen (286 kDa), bound to either of two monoclonal
antibodies, were mapped by the solution-phase HDMS (Zhang et al., 2011).
Due to the large size and complexity of the antigen-antibody complex,
the epitope mapping was inaccessible by NMR or X-ray crystallography.
Desirably, HDMS was readily applicable to the complex, yielding
comparative epitope analyses between two different antibodies in
parallel, and required less overall effort than the conventional epitope
mapping strategies such as mutagenesis and peptide scanning (Gershoni et
al., 2007; Ozohanics and Ambrus 2020; Ständer et al., 2021).
One of the novel characteristics of immunoglobulin G (IgG) is the
FcRn-mediated recycling that is responsible for long circulation
half-lives of IgG. Understanding their structural mechanisms of
interaction is crucially important for optimizing pharmacokinetics of
therapeutic antibodies (Liu 2018). As revealed in the X-ray
crystallography (Oganesyan et al., 2014), the Fc region was mainly
involved in FcRn interaction. However, a lack of structural information
on a full-length IgG (150 kDa) complexed with FcRn (50 kDa,
extracellular domain) in solution raised a possibility of potential
binding interfaces located at the Fab region. Jensen et
al. studied the interaction between a full-length human IgG1 and human
FcRn via HDMS, and identified several loci at the Fabregion which were substantially protected from hydrogen/deuterium
exchange in the presence of FcRn. Some of the loci were found even in
VH and VL domain remote from the Fc
region, implying a landscape of the FcRn-IgG binding interface extended
throughout the whole IgG structure or conformational dynamics of the
Fab region relevant to the FcRn binding (Jensen et al.,
2015). Future studies might provide supplementary CLMS datasets that
include defined distance information between FcRn and the protected loci
of IgG detected by HDMS, and therefore should scrutinize either
possibility unambiguously.
A unique analytic feature of HDMS is its ability to directly monitor the
interfacial dynamics at various protein concentrations in solution.
Antibodies are prone to reversible self-association when highly
concentrated (Yadav et al., 2011). Arora et al. investigated two
antibody samples at low and high concentrations by HDMS in parallel, and
could define major interfacial hot spots in CDRs that are related to the
concentration-dependent reversible self-association (Arora et al.,
2015). Furthermore, the HDMS results demonstrated that
CH1-CH2 interface at the hinge region,
distal to the self-association interface, exhibited significant local
backbone flexibility in a concentration-dependent manner, presumably
attributed by long-range, distant dynamic coupling effects.
Abovementioned and related studies are summarized in Table 2.