Predictive tools for stabilization of therapeutic proteins

Monoclonal antibodies represent the fastest growing class of pharmaceuticals. A major problem, however, is that the proteins are susceptible to aggregation at the high concentration commonly used during manufacturing and storage. Our recent publication describes a technology based on molecular simulations to identify aggregation-prone regions of proteins in silico. The technology, called spatial aggregation propensity (SAP), identifies hot-spots for aggregation based on the dynamic exposure of spatially-adjacent hydrophobic amino acids. Monoclonal antibodies (mAbs) in which patches with high-SAP scores are changed to patches with significantly reduced SAP scores via a single mutation are more stable than wild type, thus validating the SAP method for mapping aggregation-prone regions on proteins. We propose that the SAP technology will be useful for protein stabilization, and as a screening tool to bridge discovery and development of protein-based therapeutics by a rational assessment of the developability of candidate protein drugs.     

Elucidation of acid-induced unfolding and aggregation of human immunoglobulin IgG1 and IgG2 Fc

Understanding the underlying mechanisms of Fc aggregation is an important prerequisite for developing stable and efficacious antibody-based therapeutics. In our study, high resolution two-dimensional nuclear magnetic resonance (NMR) was employed to probe structural changes in the IgG1 Fc. A series of (1)H-(15)N heteronuclear single-quantum correlation NMR spectra were collected between pH 2.5 and 4.7 to assess whether unfolding of C(H)2 domains precedes that of C(H)3 domains. The same pH range was subsequently screened in Fc aggregation experiments that utilized molecules of IgG1 and IgG2 subclasses with varying levels of C(H)2 glycosylation. In addition, differential scanning calorimetry data were collected over a pH range of 3-7 to assess changes in C(H)2 and C(H)3 thermostability. As a result, compelling evidence was gathered that emphasizes the importance of C(H)2 stability in determining the rate and extent of Fc aggregation. In particular, we found that Fc domains of the IgG1 subclass have a lower propensity to aggregate compared with those of the IgG2 subclass. Our data for glycosylated, partially deglycosylated, and fully deglycosylated molecules further revealed the criticality of C(H)2 glycans in modulating Fc aggregation. These findings provide important insights into the stability of Fc-based therapeutics and promote better understanding of their acid-induced aggregation process.     

Competing interactions stabilize pro- and anti-aggregant conformations of human Tau

Aggregation of Tau into amyloid-like fibrils is a key process in neurodegenerative diseases such as Alzheimer. To understand how natively disordered Tau stabilizes conformations that favor pathological aggregation, we applied single-molecule force spectroscopy. Intramolecular interactions that fold polypeptide stretches of ~19 and ~42 amino acids in the functionally important repeat domain of full-length human Tau (hTau40) support aggregation. In contrast, the unstructured N terminus randomly folds long polypeptide stretches >100 amino acids that prevent aggregation. The pro-aggregant mutant hTau40ΔK280 observed in frontotemporal dementia favored the folding of short polypeptide stretches and suppressed the folding of long ones. This trend was reversed in the anti-aggregant mutant hTau40ΔK280/PP. The aggregation inducer heparin introduced strong interactions in hTau40 and hTau40ΔK280 that stabilized aggregation-prone conformations. We show that the conformation and aggregation of Tau are regulated through a complex balance of different intra- and intermolecular interactions.     

Isolation and characterization of IgG1 with asymmetrical Fc glycosylation

N-glycosylation of immunoglobulin G (IgG) at asparigine residue 297 plays a critical role in antibody stability and immune cell-mediated Fc effector function. Current understanding pertaining to Fc glycosylation is based on studies with IgGs that are either fully glycosylated [both heavy chain (HC) glycosylated] or aglycosylated (neither HC glycosylated). No study has been reported on the properties of hemi-glycosylated IgGs, antibodies with asymmetrical glycosylation in the Fc region such that one HC is glycosylated and the other is aglycosylated. We report here for the first time a detailed study of how hemi-glycosylation affects the stability and functional activities of an IgG1 antibody, mAb-X, in comparison to its fully glycosylated counterpart. Our results show that hemi-glycosylation does not impact Fab-mediated antigen binding, nor does it impact neonatal Fc receptor binding. Hemi-glycosylated mAb-X has slightly decreased thermal stability in the CH2 domain and a moderate decrease (～20%) in C1q binding. More importantly, the hemi-glycosylated form shows significantly decreased binding affinities toward all Fc gamma receptors (FcγRs) including the high-affinity FcγRI, and the low-affinity FcγRIIA, FcγRIIB, FcγRIIIA and FcγRIIIB. The decreased binding affinities to FcγRs result in a 3.5-fold decrease in antibody-dependent cell cytotoxicity (ADCC). As ADCC often plays an important role in therapeutic antibody efficacy, glycosylation status will not only affect the antibody quality but also may impact the biological function of the product.     

Site of binding of IgG2b and IgG2a by mouse macrophage Fc receptors by using cyanogen bromide fragments

Cyanogen bromide fragments of murine IgG2b and IgG2a immunoglobulins were used to localize the sequences that are bound by specific IgG2b and IgG2a Fc receptors on murine macrophages. One fragment from the CH2 domain of IgG2b bound to the gamma 2b Fc receptor. Two fragments from IgG2a--one one from the CH2 domain, differing by only four amino acids from the homologous IgG2b fragment, and the other from the CH3 domain--specifically bound to the gamma 2a Fc receptor. In both a rosetting assay and a radioactive binding assay, these two fragments from IgG2a competed with intact IgG2a: however, they did not compete with each other. Rather, binding of the fragment from the CH3 domain of IgG2a augmented the binding of the fragment from the CH2 domain of IgG2a but not that of the homologous fragment from IgG2b. The binding of both IgG2a fragments was abolished by trypsin treatment of macrophages. These data suggest that 1) a sequence in the CH2 domain of IgG2b is sufficient for binding to the gamma 2b Fc receptor, 2) sequences from both the CH2 and CH3 domains of IgG2a bind to the gamma 2a Fc receptor, and 3) the binding of sequences from the CH3 domain of IgG2a may induce a conformational change in the gamma 2a Fc receptor that leads to enhanced binding of sequences from the CH2 domain.     

Characterization of domains of herpes simplex virus type 1 glycoprotein E involved in Fc binding activity for immunoglobulin G aggregates.

Herpes simplex virus type 1 glycoproteins gE and gI form receptors for the Fc domain of immunoglobulin G (IgG) which are expressed on the surface of infected cells and on the virion envelope and which protect the virus from immune attack. Glycoprotein gE-1 is a low-affinity Fc receptor (FcR) that binds IgG aggregates, while gE-1 and gI-1 form a complex which serves as a higher-affinity FcR capable of binding IgG monomers. In this study, we describe two approaches used to map an Fc binding domain on gE-1 for IgG aggregates. First, we constructed nine plasmids encoding gE-1/gD-1 fusions proteins, each containing a large gE-1 peptide inserted into the ectodomain of gD-1. Fusion proteins were tested for FcR activity with IgG-sensitized erythrocytes in a rosetting assay. Three of the fusion proteins containing overlapping gE-1 peptides demonstrated FcR activity; the smallest peptide that retained Fc binding activity includes gE-1 amino acids 183 to 402. These results indicate that an Fc binding domain is located between gE-1 amino acids 183 and 402. To more precisely map the Fc binding domain, we tested a panel of 21 gE-1 linker insertion mutants. Ten mutants with insertions between gE-1 amino acids 235 and 380 failed to bind IgG-sensitized erythrocytes, while each of the remaining mutants demonstrated wild-type Fc binding activity. Taken together, these results indicate that the region of gE-1 between amino acids 235 and 380 forms an FcR domain. A computer-assisted analysis of the amino acid sequence of gE-1 demonstrates an immunoglobulin-like domain contained within this region (residues 322 to 359) which shares homology with mammalian FcRs.     

Evidence for a modular structure of the homologous repetitive C-terminal carbohydrate-binding sites of Clostridium difficile toxins and Streptococcus mutans glucosyltransferases.

The homologous C-terminal repeats of Clostridium difficile toxins (ToxA and ToxB) and streptococcal glucosyltransferases appear to mediate protein-carbohydrate interactions at cellular binding sites with sugar moieties as substrates. A consensus sequence of 134 repeating units from gram-positive bacteria indicates that these repeats have a modular design with (i) a stretch of aromatic amino acids proposed to be involved in the primary carbohydrate-protein interaction, (ii) an amplification of this interaction by repetition of the respective sequences, and (iii) a second domain, not characterized, that is responsible for carbohydrate specificity.     

A stable engineered human IgG3 antibody with decreased aggregation during antibody expression and low pH stress

Human IgG comprises four subclasses with different biological functions. The IgG3 subclass has a unique character, exhibiting high effector function and Fab arm flexibility. However, it is not used as a therapeutic drug owing to an enhanced susceptibility to proteolysis. Antibody aggregation control is also important for therapeutic antibody development. To date, there have been few reports of IgG3 aggregation during protein expression and the low pH conditions needed for purification and virus inactivation. This study explored the potential of IgG3 antibody for therapeutics using anti‐CD20 IgG3 as a model to investigate aggregate formation. Initially, anti‐CD20 IgG3 antibody showed substantial aggregate formation during expression and low pH treatment. To circumvent this phenomenon, we systematically exchanged IgG3 constant domains with those of IgG1, a stable IgG. IgG3 antibody with the IgG1 CH3 domain exhibited reduced aggregate formation during expression. Differential scanning calorimetric analysis of individual amino acid substitutions revealed that two amino acid mutations in the CH3 domain, N392K and M397V, reduced aggregation and increased CH3 transition temperature. The engineered human IgG3 antibody was further improved by additional mutations of R435H to obtain IgG3KVH to achieve protein A binding and showed similar antigen binding as wild‐type IgG3. IgG3KVH also exhibited high binding activity for FcγRIIIa and C1q. In summary, we have successfully established an engineered human IgG3 antibody with reduced aggregation during bioprocessing, which will contribute to the better design of therapeutic antibodies with high effector function and Fab arm flexibility.     

Glycosylation in the Fc domain of IgG increases resistance to proteolytic cleavage by papain

IgG antibodies (Abs) and fragments of IgG Abs are becoming major biotherapeutics to treat an assortment of human diseases. Commonly prepared fragments of IgGs include Fc, Fab, and F(ab')2 fragments, all of which can be made using the sulfhydryl protease papain, although prolonged digestion times and/or excessive amounts of papain typically result in further cleavage of the Fc domain into smaller fragments. During our attempts to use papain to isolate Fc fragments from different IgG monoclonal Abs, it was observed that prior removal of Fc glycans resulted in a faster rate of papain-mediated degradation of the Fc domain. Subsequent time-course experiments comparing glycosylated and deglycosylated versions of IgG antibodies showed that the majority of molecules in a deglycosylated IgG sample were converted into Fab, Fc, and smaller Fc fragments in less than one hour, whereas the original glycosylated IgG required more than two hours to convert into a comparable amount of Fab and Fc fragments. Furthermore, whereas papain digestion converted almost all of a deglycosylated Fc fragment into smaller fragments of approximately 10 and approximately 12 kDa within 4 h, more than 40% of a glycosylated Fc fragment remained intact even after 24 h of digestion. These results indicate that the presence of CH(2) domain glycans in either IgGs or purified Fc fragments increases resistance to papain digestion. Increased sensitivity of non-glycosylated Fc domains to papain is consistent with the Fc domains lacking a defined structure, as exemplified by their inability to bind Fcgamma receptors, since misfolded proteins are often degraded by proteases because of increased accessibility of their proteolytic cleavage sites. Based on these observations it is possible to use papain sensitivity as a means of assessing proper Fc structure of IgG molecules.     

Monovalent IgG4 molecules

Antibodies have become the fastest growing class of biological therapeutics, in part due to their exquisite specificity and ability to modulate protein-protein interactions with a high biological potency. The relatively large size and bivalency of antibodies, however, limits their use as therapeutics in certain circumstances. Antibody fragments, such as single-chain variable fragments and antigen binding-fragments, have emerged as viable alternatives, but without further modifications these monovalent formats have reduced terminal serum half-lives because of their small size and lack of an Fc domain, which is required for FcRn-mediated recycling. Using rational engineering of the IgG4 Fc domain to disrupt key interactions at the CH3-CH3 interface, we identified a number of point mutations that abolish Fc dimerization and created half-antibodies, a novel monovalent antibody format that retains a monomeric Fc domain. Introduction of these mutations into an IgG1 framework also led to the creation of half-antibodies. These half-antibodies were shown to be soluble, thermodynamically stable and monomeric, characteristics that are favorable for use as therapeutic proteins. Despite significantly reduced FcRn binding in vitro, which suggests that avidity gains in a dimeric Fc are critical to optimal FcRn binding, this format demonstrated an increased terminal serum half-life compared with that expected for most alternative antibody fragments.     

Engineering a human IgG2 antibody stable at low pH

IgG2 subclass antibodies have unique properties that include low effector function and a rigid hinge region. Although some IgG2 subclasses have been clinically tested and approved for therapeutic use, they have a higher propensity than IgG1 for aggregation, which can curtail or abolish their biological activity and enhance their immunogenicity. In this regard, acid‐induced aggregation of monoclonal antibodies during purification and virus inactivation must be prevented. In the present study, we replaced the constant domain of IgG2 with that of IgG1, using anti‐2,4‐dinitrophenol (DNP) IgG2 as a model antibody, and investigated whether that would confer greater stability. While the anti‐DNP IgG2 antibody showed significant aggregation at low pH, this was reduced for the IgG2 antibody containing the IgG1 CH2 domain. Substituting three amino acids within the CH2 domain—namely, F300Y, V309L, and T339A (IgG2_YLA)—reduced aggregation at low pH and increased CH2 transition temperature, as determined by differential scanning calorimetric analysis. IgG2_YLA exhibited similar antigen‐binding capacity to IgG2, low affinity for FcγRIIIa, and low binding ability to C1q. The same YLA substitution also reduced the aggregation of panitumumab, another IgG2 antibody, at low pH. Our engineered human IgG2 antibody showed reduced aggregation during bioprocessing and provides a basis for designing improved IgG2 antibodies for therapeutic applications.     

Structural basis for cyclodextrins' suppression of human growth hormone aggregation

Many therapeutic proteins require storage at room temperature for extended periods of time. This can lead to aggregation and loss of function. Cyclodextrins (CDs) have been shown to function as aggregation suppressors for a wide range of proteins. Their potency is often ascribed to their affinity for aromatic amino acids, whose surface exposure would otherwise lead to protein association. However, no detailed structural studies are available. Here we investigate the interactions between human growth hormone (hGH) and different CDs at low pH. Although hGH aggregates readily at pH 2.5 in 1 M NaCl to form amorphous aggregates, the presence of 25 to 50 mM of various beta-CD derivatives is sufficient to completely avoid this. alpha- and gamma-CD are considerably less effective. Stopped-flow data on the aggregation reaction in the presence of beta-CD are analyzed according to a minimalist association model to yield an apparent hGH-beta-CD dissociation constant of approximately 6 mM. This value is very similar to that obtained by simple fluorescence-based titration of hGH with beta-CD. Nuclear magnetic resonance studies indicate that beta-CD leads to a more unfolded conformation of hGH at low pH and predominantly binds to the aromatic side-chains. This indicates that aromatic amino acids are important components of regions of residual structure that may form nuclei for aggregation.     

Correlation between the exposure of aromatic chromophores at the surface of the Fc domains of immunoglobulin G and their ability to bind complement.

The recognition that certain biological effector functions associated with the Fc region of human IgG are mediated exclusively by either the Cgamma2 or Cgamma3 domains prompted a study of some of the physical properties of the isolated domains in an attempt to correlate these with functional differentiation. The degree of aromatic chromophore exposure of intact Fc and fragments corresponding to the Cgamma2 and Cgamma3 domains were determined by solvent perturbation difference spectroscopy using 20% ethylene glycol. For the monomeric Cgamma2 fragment one of the two tryptophans and all four of the tyrosines were exposed to solvent. In the pFc' fragment, which represented a dimer of two intact Cgamma3 domains, an average of 0.4 of the two tryptophans of 3.3 of the five tyrosines per chain were exposed. These data were consistent with the suggested involvement of tryptophan in complement fixation since Cgamma2 binds C1q but pFc' does not. Several fragments derived from the Cgamma3 region had previously been shown to have differing environments for their aromatic side chains from circular dichroism studies. These fragments have now been shown to exhibit different degrees of chromophore exposure to solvent. Removal of the carboxy-termimal heptapeptide from the intact, Cgamma3 domain resulted in a fragment not only showing a greater exposure of aromatic residues but also having the ability to bind Clq. Our data suggest that the structural requirements for C1Q binding may be quite commonplace within Fc, but tertiary folding limits their expression except in Cgamma2 in the native molecule. The solvent perturbation observed with Fc was somewhat lower than would have been expected from the results with the isolated domains, suggesting that interdomain interactions may result in burial of aromatic residues.     

Impact of SPR biosensor assay configuration on antibody: Neonatal Fc receptor binding data

Binding interactions with the neonatal Fc receptor (FcRn) are one determinant of pharmacokinetic properties of recombinant human monoclonal antibody (rhumAb) therapeutics, and a conserved binding motif in the crystallizable fragment (Fc) region of IgG molecules interacts with FcRn. Surface plasmon resonance (SPR) biosensor assays are often used to characterize interactions between FcRn and rhumAb therapeutics. In such assays, generally either the rhumAb (format 1) or the FcRn protein (format 2) is immobilized on a biosensor chip. However, because evidence suggests that, in some cases, the variable domains of a rhumAb may also affect FcRn binding, we evaluated the effect of SPR assay configuration on binding data. We sought to assess FcRn binding properties of 2 rhumAbs (rhumAb1 and rhumAb2) to FcRn proteins using these 2 biosensor assay formats. The two rhumAbs have greater than 99% sequence identity in the Fc domain but differ in their Fab regions. rhumAb2 contains a positively charged patch in the variable domain that is absent in rhumAb1. Our results showed that binding of rhumAb1 to FcRn was independent of biosensor assay configuration, while binding of rhumAb2 to FcRn was highly SPR assay configuration dependent. Further investigations revealed that the format dependency of rhumAb2-FcRn binding is linked to the basic residues that form a positively charged patch in the variable domain of rhumAb2. Our work highlights the importance of analyzing rhumAb-FcRn binding interactions using 2 alternate SPR biosensor assay configurations. This approach may also provide a simple way to identify the potential for non-Fc-driven FcRn binding interactions in otherwise typical IgGs.     

Rational design of therapeutic mAbs against aggregation through protein engineering and incorporation of glycosylation motifs applied to bevacizumab

The aggregation of biotherapeutics is a major hindrance to the development of successful drug candidates; however, the propensity to aggregate is often identified too late in the development phase to permit modification to the protein's sequence. Incorporating rational design for the stability of proteins in early discovery has numerous benefits. We engineered out aggregation-prone regions on the Fab domain of a therapeutic monoclonal antibody, bevacizumab, to rationally design a biobetter drug candidate. With the purpose of stabilizing bevacizumab with respect to aggregation, 2 strategies were undertaken: single point mutations of aggregation-prone residues and engineering a glycosylation site near aggregation-prone residues to mask these residues with a carbohydrate moiety. Both of these approaches lead to comparable decreases in aggregation, with an up to 4-fold reduction in monomer loss. These single mutations and the new glycosylation pattern of the Fab domain do not modify binding to the target. Biobetters with increased stability against aggregation can therefore be generated in a rational manner, by either removing or masking the aggregation-prone region or crowding out protein-protein interactions.     

Influence of N-glycosylation on effector functions and thermal stability of glycoengineered IgG1 monoclonal antibody with homogeneous glycoforms

Glycosylation of the conserved asparagine residue in each heavy chain of IgG in the CH2 domain is known as N-glycosylation. It is one of the most common post-translational modifications and important critical quality attributes of monoclonal antibody (mAb) therapeutics. Various studies have demonstrated the effects of the Fc N-glycosylation on safety, Fc effector functions, and pharmacokinetics, both dependent and independent of neonatal Fc receptor (FcRn) pathway. However, separation of various glycoforms to investigate the biological and functional relevance of glycosylation is a major challenge, and existing studies often discuss the overall impact of N-glycans, without considering the individual contributions of each glycoform when evaluating mAbs with highly heterogeneous distributions. In this study, chemoenzymatic glycoengineering incorporating an endo-β-N-acetylglucosaminidase (ENGase) EndoS2 and its mutant with transglycosylation activity was used to generate mAb glycoforms with highly homogeneous and well-defined N-glycans to better understand and precisely evaluate the effect of each N-glycan structure on Fc effector functions and protein stability. We demonstrated that the core fucosylation, non-reducing terminal galactosylation, sialylation, and mannosylation of IgG1 mAb N-glycans impact not only on FcγRIIIa binding, antibody-dependent cell-mediated cytotoxicity, and C1q binding, but also FcRn binding, thermal stability and propensity for protein aggregation.     

N-glycosylation of enhanced aromatic sequons to increase glycoprotein stability

N-glycosylation can increase the rate of protein folding, enhance thermodynamic stability, and slow protein unfolding; however, the molecular basis for these effects is incompletely understood. Without clear engineering guidelines, attempts to use N-glycosylation as an approach for stabilizing proteins have resulted in unpredictable energetic consequences. Here, we review the recent development of three "enhanced aromatic sequons," which appear to facilitate stabilizing native-state interactions between Phe, Asn-GlcNAc and Thr when placed in an appropriate reverse turn context. It has proven to be straightforward to engineer a stabilizing enhanced aromatic sequon into glycosylation-na?ve proteins that have not evolved to optimize specific protein-carbohydrate interactions. Incorporating these enhanced aromatic sequons into appropriate reverse turn types within proteins should enhance the well-known pharmacokinetic benefits of N-glycosylation-based stabilization by lowering the population of protease-susceptible unfolded and aggregation-prone misfolded states, thereby making such proteins more useful in research and pharmaceutical applications.     

Molecular recognition of Fc‐specific ligands binding onto the consensus binding site of IgG: insights from molecular simulation

Immunoglobulin G (IgG) plays an important role in clinical diagnosis and therapeutics. Meanwhile, the consensus binding site (CBS) on the Fc domain of IgG is responsible for ligand recognition, especially for Fc‐specific ligands. In this study, molecular simulation methods were used to investigate molecular interactions between the CBS of the Fc domain and seven natural Fc‐specific ligands. The analysis on the binding energy of the Fc–ligand complex indicated that hydrophobic interactions provide the main driving force for the Fc–ligand binding processes. The hot spots on the ligands and Fc were identified with the computational alanine scanning approach. It was found that the residues of tryptophan and tyrosine on the ligands have significant contributions for the Fc–ligand binding, while Met252, Ile253, Asn434, His435, and Tyr436 are the key residues of Fc. Moreover, two binding modes based on tryptophan or tyrosine were summarized and constructed according to the pairwise interaction analysis. Guidelines for the rational design of CBS‐specific ligands with high affinity and specificity were proposed. Copyright © 2014 John Wiley & Sons, Ltd.     

Accounting for protein-solvent contacts facilitates design of nonaggregating lattice proteins

The folding specificity of proteins can be simulated using simplified structural models and knowledge-based pair-potentials. However, when the same models are used to simulate systems that contain many proteins, large aggregates tend to form. In other words, these models cannot account for the fact that folded, globular proteins are soluble. Here we show that knowledge-based pair-potentials, which include explicitly calculated energy terms between the solvent and each amino acid, enable the simulation of proteins that are much less aggregation-prone in the folded state. Our analysis clarifies why including a solvent term improves the foldability. The aggregation for potentials without water is due to the unrealistically attractive interactions between polar residues, causing artificial clustering. When a water-based potential is used instead, polar residues prefer to interact with water; this leads to designed protein surfaces rich in polar residues and well-defined hydrophobic cores, as observed in real protein structures. We developed a simple knowledge-based method to calculate interactions between the solvent and amino acids. The method provides a starting point for modeling the folding and aggregation of soluble proteins. Analysis of our simple model suggests that inclusion of these solvent terms may also improve off-lattice potentials for protein simulation, design, and structure prediction.