Prediction of primary structure deamidation rates of asparaginyl and glutaminyl peptides through steric and catalytic effects.

The primary sequence dependence of deamidation has been quantitatively explained on the basis of a simple steric and catalytic model. Application to the known deamidation rates of peptides produces a table of coefficients that permits calculation of the known deamidation rates and prediction of deamidation rates for peptide sequences that have not yet been measured. This work permits a better understanding of deamidation, provides a prediction procedure for protein engineering, and facilitates improved computation of peptide and protein primary, secondary, tertiary, and quaternary structure deamidation rates.     

Mass spectrometric evaluation of synthetic peptides as primary structure models for peptide and protein deamidation.

Solid-phase peptide synthesis and deamidation measurements using a novel mass spectrometric technique were carried out for 94 model asparaginyl peptides from 3 to 13 residues in length. Deamidation rates of these peptides in pH 7.4, 37.0 degrees C, 0.15 M Tris-HCl buffer were measured and evaluated. It was found that they validate the use of pentapeptide models as surrogates for the primary sequence dependence of peptide and protein deamidation rates and the discovery by difference of secondary, tertiary and quaternary structure effects. Deamidation of the pentapeptide models, compared with that of longer peptides of more intricate structure, is discussed, and the application of this technique to deamidation measurement of intact proteins is demonstrated.     

Structure-dependent nonenzymatic deamidation of glutaminyl and asparaginyl pentapeptides.

Nonenzymatic deamidation rates for 52 glutaminyl and 52 asparaginyl pentapeptides in pH 7.4, 37.0 degrees C. 0.15 m Tris-HCl buffer have been determined by direct injection mass spectrometry. These and the previously reported 306 asparginyl rates have been combined in a self-consistent model for peptide deamidation. This model depends quantitatively upon peptide structure and involves succinimide, glutarimide and hydrolysis mechanisms. The experimental values and suitable interpolated values have been combined to provide deamidation rate values in pH 7.4, 37.0 degrees C. 0.15 m Tris-HCl buffer for the entire set of 648 single-amide permutations of ordinary amino acid residues in GlyXxxAsnYyyGly and GlyXxxGlnYyyGly. Thus, knowledge about sequence-dependent deamidation in peptides is extended to include very long deamidation half-times in the range of 2-50 years.     

The propensity for deamidation and transamidation of peptides by transglutaminase 2 is dependent on substrate affinity and reaction conditions

Transglutaminase 2 (TG2) catalyzes cross-linking or deamidation of glutamine residues in peptides and proteins. The in vivo deamidation of gliadin peptides plays an important role in the immunopathogenesis of celiac disease (CD). Although deamidation is considered to be a side-reaction occurring in the absence of suitable amines or at a low pH, a recent paper reported the selective deamidation of the small heat shock protein 20 (Hsp20), suggesting that deamidation could be a substrate dependent event. Here we have measured peptide deamidation and transamidation in the same reaction to reveal factors that affect the relative propensity for the two possible products. We report that the propensity for deamidation by TG2 is both substrate dependent and influenced by the reaction conditions. Direct deamidation is favored for poor substrates and at low concentrations of active TG2, while indirect deamidation (i.e. hydrolysis of transamidated product) can significantly contribute to the deamidation of good peptide substrates at higher enzyme concentrations. Further, we report for the first time that TG2 can hydrolyze iso-peptide bonds between two peptide substrates. This was observed also for gliadin peptides introducing a novel route for the generation of deamidated T cell epitopes in celiac disease.     

Evidence for the in vivo deamidation and isomerization of an asparaginyl residue in cytosolic serine hydroxymethyltransferase

Rabbit liver cytosolic serine hydroxymethyltransferase exists in several subforms which have different isoelectric points. Incubation of the purified enzyme with chymotrypsin cleaves the enzyme at Trp14. The released amino-terminal 14-mer peptide was shown to exist in three forms of equal concentration. The peptides differ in structure only at the asparaginyl residue at position 5. In addition to asparagine at this position we found both aspartyl and isoaspartyl residues. The deamidation of Asn5 does not appear to occur during the purification of the enzyme. The in vitro rate of deamidation of Asn5 in the enzyme is more than 5-fold slower than the rate of deamidation of this residue in the free 14-mer peptide. The isoaspartyl residue at position 5 serves as a substrate for protein carboxyl methyltransferase both in the free 14-mer peptide and the native enzyme. The enzyme which has had the amino-terminal 14 residues removed by digestion with chymotrypsin still exists in several forms with different isoelectric points. Reaction of peptides from this enzyme with carboxyl methyltransferase suggests that there is at least one more asparaginyl residue in this enzyme other than Asn5 which has undergone deamidation with the formation of isoaspartyl bonds.     

Effects of amino acid sequence, buffers, and ionic strength on the rate and mechanism of deamidation of asparagine residues in small peptides

The nonenzymatic rates of deamidation of Asn residues in a series of pentapeptides with the sequences VSNXV and VXNSV, where X is one of 10 different amino acids, were determined at neutral, alkaline, and acid pH values. The results demonstrate that in neutral and alkaline solutions the amino acid residue on the amino side of the Asn had little or no effect on the rate of deamidation regardless of its charge or size. The group on the carboxyl side of Asn affected the rate of deamidation significantly. Increasing size and branching in the side chain of this residue decreased the rate of deamidation by as much as 70-fold compared to glycine in the N-G sequence, which had the greatest rate of deamidation. In acidic solution, the rate of deamidation of the Asn residue was not affected by the amino acid sequence of the peptide. The products for each deamidation reaction were tested for the formation of isoAsp residues. In neutral and alkaline solutions, all products showed that the isoAsp:Asp peptide products were formed in about a 3:1 ratio. In acidic solution, the Asp peptide was the only deamidation product formed. All peptides in which a Ser residue follows the Asn residue were found to undergo a peptide cleavage reaction in neutral and alkaline solutions, yielding a tripeptide and a dipeptide. The rate of the cleavage reaction was about 10% of the rate of the deamidation pathway at neutral and alkaline pH values. The rates of deamidation of Asn residues in the peptides studied were not affected by ionic strength, and were not specific base catalyzed. General base catalysis was observed for small bases like ammonia. A model for the deamidation reaction is proposed to account for the observed effects.     

Study of the mechanism of action of anoplin, a helical antimicrobial decapeptide with ion channel-like activity, and the role of the amidated C-terminus.

Anoplin, an antimicrobial, helical decapeptide from wasp venom, looses its biological activities by mere deamidation of its C-terminus. Secondary structure determination, by circular dichroism spectroscopy in amphipathic environments, and lytic activity in zwitterionic and anionic vesicles showed quite similar results for the amidated and the carboxylated forms of the peptide. The deamidation of the C-terminus introduced a negative charge at an all-positive charged peptide, causing a loss of amphipathicity, as indicated by molecular dynamics simulations in TFE/water mixtures and this subtle modification in a peptide's primary structure disturbed the interaction with bilayers and biological membranes. Although being poorly lytic, the amidated form, but not the carboxylated, presented ion channel-like activity on anionic bilayers with a well-defined conductance step; at approximately the same concentration it showed antimicrobial activity. The pores remain open at trans-negative potentials, preferentially conducting cations, and this situation is equivalent to the interaction of the peptide with bacterial membranes that also maintain a high negative potential inside.     

Role of peptide conformation in the rate and mechanism of deamidation of asparaginyl residues

The tetrapeptides Val-Asn-Gly-Ala and N-acetyl-Val-Asn-Gly-Ala undergo deamidation of the asparaginyl residue at pH 7.0 at similar rates. However, they form different products. The N-acetyl peptide gave a 3:1 ratio of N-acetyl-Val-isoAsp-Gly-Ala and N-acetyl-Val-Asp-Gly-Ala, respectively. The nonacetylated peptide gave no detectable amounts of these products but rather gave a cyclic peptide formed from the nucleophilic displacement of the asparaginyl side chain amide by the amino terminus of valine. This compound was slowly inverted at carbon 2 of the asparaginyl residue. At pH values above 7.5, the nonacetylated peptide also underwent deamidation to form Val-isoAsp-Gly-Ala and Val-Asp-Gly-Ala in the 3:1 ratio. Proton NMR spectra of the acetylated and nonacetylated tetrapeptides show that below pH 7.5 they have very different preferred conformations, and it is these different conformations which result in the different mechanisms of deamidation. Above pH 9.0, both peptides have similar conformations and deamidate by the same mechanism to give equivalent products. Neither mechanism of deamidation was subject to general base catalysis by the buffer. These results suggest that deamidation rates of the asparaginyl-glycyl sequence in proteins will vary according to the conformation of the peptide backbone of each respective protein. The results also show that asparaginyl residues which are penultimate to the amino terminus can react to form an N-terminal-blocked seven-membered ring.     

Effect of viscosity on the deamidation rate of a model Asn-hexapeptide.

The effect of viscosity on the deamidation rate of a model Asn-containing hexapeptide (l-Val-l-Tyr-Pro-l-Asn-Gly-l-Ala) was assessed in aqueous solution and in solids containing varying amounts of poly(vinyl pyrrolidone) (PVP) and water. Stability studies were conducted at 0.1 mg/mL peptide and 0-50% PVP (w/w) in aqueous solution, and at 5% (w/w) peptide and different relative humidities (31.6, 53.1, 74.4 and 96%) in the solid state. The parent peptide and its deamidation products were analysed by reverse-phase high-performance liquid chromatography. Deamidation rates decreased with increasing solvent viscosity in a manner described by a semi-empirical mathematical model developed to describe this relationship. The results suggest that the motion of the Asn side-chain along the reaction coordinate is a function of the macroscopic solvent viscosity. However, the apparent energy barrier for the diffusive movement of the side-chain appears to be less than the energy barrier for that associated with macroscopic viscosity. The dependence of the deamidation rate on viscosity in both viscous solution and hydrated solids further demonstrates the importance of mobility in peptide deamidation.     

Sites of deamidation and methylation in Tsr, a bacterial chemotaxis sensory transducer

The sensory transducer proteins in bacterial chemotaxis undergo two covalent modifications, deamidation and reversible methylation, in response to attractants and repellents. Oligonucleotide-directed mutagenesis was used to alter putative methylation and deamidation sites in one of the transducers to further define these sites and their role in chemotaxis. The mutations, in combination with peptide maps and Edman analysis, have clarified the sites of covalent modification in Tsr. Tsr contains six specific glutamates and glutamines that serve as methyl-accepting sites. An arginine-containing tryptic peptide (R1) has two sites, one at glutamate 493 and a newly located site at glutamate 502. A lysine-containing peptide (K1) has four methyl-accepting sites. Two of the lysine peptide sites are glutamates and can accept methyl groups without deamidation. The other two sites are glutamines and two methyl-accepting sites are created by two distinct deamidations. Both deamidations can occur on the same polypeptide chain. Single glutamate mutants have shown that one deamidation (at glutamine 311) proceeds rapidly, while the other deamidation (at glutamine 297) has a half-life of approximately 60 min under our experimental conditions.     

The effects of alpha-helix on the stability of Asn residues: deamidation rates in peptides of varying helicity

Asn deamidation was monitored in Ala-based octadecapeptides of varying alpha-helicity. Gly was substituted for Ala residues at positions 6 and 16 to create a peptide with less helicity. Ala --> Gly substitutions were made at three or more residues from the Asn to negate known primary sequence effects on deamidation rates. The extent of helicity and rate of Asn deamidation for alkaline aqueous solutions of each peptide was measured as a function of temperature by circular dichroism and reversed-phase high-performance liquid chromatography, respectively. The rate of deamidation in the peptides was inversely proportional to the extent of alpha-helicity. The results support the conclusion that Asn deamidation only occurs in the nonhelical population of conformers.     

Effect of protein structure on deamidation rate in the Fc fragment of an IgG1 monoclonal antibody

The effects of secondary structure on asparagine (N) deamidation in a 22 amino acid sequence (369‐GFYPSDIAVEWESNGQPENNYK‐390) of the crystallizable (Fc) fragment of a human monoclonal antibody (Fc IgG1) were investigated using high‐resolution ultra performance liquid chromatography with tandem mass spectrometry (UPLC/MS). Samples containing either the intact Fc IgG (～50 kD) (“intact protein”), or corresponding synthetic peptides (“peptide”) were stored in Tris buffer at 37°C and pH 7.5 for up to forty days, then subjected to UPLC/MS analysis with high energy MS1 fragmentation. The peptide deamidated only at N₃₈₂ to form the isoaspartate (isoD₃₈₂) and aspartate (D₃₈₂) products in the ratio of ～4:1, with a half‐life of ～3.4 days. The succinimide intermediate (Su₃₈₂) was also detected; deamidation was not observed for the other two sites (N₃₈₇ and N₃₈₈) in peptide samples. The intact protein showed a 30‐fold slower overall deamidation half‐life of ～108 days to produce the isoD₃₈₂ and D₃₈₇ products, together with minor amounts of D₃₈₂. Surprisingly, the D₃₈₂ and isoD₃₈₇ products were not detected in intact protein samples and, as in the peptide samples, deamidation was not detected at N₃₈₈. The results indicate that higher order structure influences both the rate of N‐deamidation and the product distribution.     

Low levels of asparagine deamidation can have a dramatic effect on aggregation of amyloidogenic peptides: Implications for the study of amyloid formation

The polypeptide hormone amylin forms amyloid deposits in Type 2 diabetes mellitus and a 10-residue fragment of amylin (amylin(20-29)) is commonly used as a model system to study this process. Studies of amylin(20-29) and several variant peptides revealed that low levels of deamidation can have a significant effect on the secondary structure and aggregation behavior of these molecules. Results obtained with a variant of amylin(20-29), which has the primary sequence SNNFPAILSS, are highlighted. This peptide is particularly interesting from a technical standpoint. In the absence of impurities the peptide does not spontaneously aggregate and is not amyloidogenic. This peptide can spontaneously deamidate, and the presence of less than 5% of deamidation impurities leads to the formation of aggregates that have the hallmarks of amyloid. In addition, small amounts of deamidated material can induce amyloid formation by the purified peptide. These results have fundamental implications for the definition of an amyloidogenic sequence and for the standards of purity of peptides and proteins used for studies of amyloid formation.     

Deamidation, isomerization, and racemization at asparaginyl and aspartyl residues in peptides. Succinimide-linked reactions that contribute to protein degradation

Aspartyl and asparaginyl deamidation, isomerization, and racemization reactions have been studied in synthetic peptides to model these spontaneous processes that alter protein structure and function. We show here that the peptide L-Val-L-Tyr-L-Pro-L-Asn-Gly-L-Ala undergoes a rapid deamidation reaction with a half-life of only 1.4 days at 37 degrees C, pH 7.4, to give an aspartyl succinimide product. Under these conditions, the succinimide product can further react by hydrolysis (half-time, 2.3h) and by racemization (half-time, 19.5 h). The net product of the deamidation reaction is a mixture of L- and D-normal aspartyl and beta-transpeptidation (isoaspartyl) hexapeptides. Replacement of the asparagine residue by an aspartic acid residue results in a 34-fold decrease in the rate of succinimide formation. Significant racemization was found to accompany the deamidation and isomerization reactions, and most of this could be accounted for by the rapid racemization of the succinimide intermediate. Replacement of the glycyl residue in the asparagine-containing peptide with a bulky leucyl or prolyl residue results in a 33-50-fold decrease in the rate of degradation. Peptide cleavage products are observed when these Asn-Leu and Asn-Pro-containing peptides are incubated. Our studies indicate that both aspartic acid and asparagine residues may be hot spots for the nonenzymatic degradation of proteins, especially in cells such as erythrocytes and eye lens, where these macromolecules must function for periods of about 120 days and 80 years, respectively.     

Engineering an anti-CD52 antibody for enhanced deamidation stability

ABSTRACT

Deamidation evaluation and mitigation is an important aspect of therapeutic antibody developability assessment. We investigated the structure and function of the Asn-Gly deamidation in a human anti-CD52 IgG1 antibody light chain complementarity-determining region 1, and risk mitigation through protein engineering. Antigen binding affinity was found to decrease about 400-fold when Asn³³ was replaced with an Asp residue to mimic the deamidation product, suggesting significant impacts on antibody function. Other variants made at Asn³³ (N33H, N33Q, N33H, N33R) were also found to result in significant loss of antigen binding affinity. The co-crystal structure of the antigen-binding fragment bound to a CD52 peptide mimetic was solved at 2.2Å (PDB code 6OBD), which revealed that Asn³³ directly interacts with the CD52 phosphate group via a hydrogen bond. Gly³⁴, but sits away from the binding interface, rendering it more amendable to mutagenesis without affecting affinity. Saturation mutants at Gly³⁴ were prepared and subjected to forced deamidation by incubation at elevated pH and temperature. Three mutants (G34R, G34K and G34Q) showed increased resistance to deamidation by LC-MS peptide mapping, while maintaining high binding affinity to CD52 antigen measured by Biacore. A complement -dependent cytotoxicity assay indicated that these mutants function by triggering antibody effector function. This study illustrates the importance of structure-based design and extensive mutagenesis to mitigate antibody developability issues.     

Effect of lysine residues on the deamidation reaction of asparagine side chains

The effect of lysine residues on the deamidation reaction of the asparagine side chain has been studied on the peptide and on its lysine-acetylated derivative in a wide range of pH values. The amino acid sequence of these peptides is similar to the local sequence flanking the labile Asn-67 in RNAse A. The experimental data show that Lys influences both the deamidation rate and the relative yield of the two reaction products, i.e., the aspartic acid and beta-aspartic acid containing peptide. These effects are pH dependent and can be rationalized based on the mechanism previously proposed for the deamidation reaction via succinimide derivative.     

Identification of deamidation and isomerization sites on pharmaceutical recombinant antibody using H(2)(18)O

Asparagine (Asn) deamidation and aspartic acid (Asp) isomerization are spontaneous and common alterations occurring in pharmaceutical protein drugs in solution. Because those reactions may cause functional changes, it is important to identify the product-related substances, especially when biopharmaceuticals are under development. In this study, we used H(2)(18)O to identify Asn deamidation and Asp isomerization sites on a recombinant humanized monoclonal antibody (mAb) by using high-performance liquid chromatography-mass spectrometry (HPLC-MS). This strategy takes advantage of reactions whereby (18)O is incorporated into the protein molecule. The mAb was lyophilized and reconstituted in H(2)O or H(2)(18)O, followed by incubation at 50 degrees C for 1 month. Samples were reduced/carboxymethylated and digested by trypsin and then subjected to HPLC-MS and HPLC-tandem mass spectrometry (MS/MS) analysis. Among all of the peptide fragments analyzed, there were two in which deamidation and/or isomerization was observed. In one peptide fragment, an obvious mass shift ( approximately 3Da) at Asn was observed in the newly produced peptide when the mAb was incubated in H(2)(18)O, whereas it was barely feasible to identify this mass shift in H(2)O. In the other peptide fragment, isomerization of Asp was identified after incubation in H(2)(18)O, although it was impossible to distinguish when using H(2)O. By means of this procedure, identification of deamidation and isomerization sites can be accomplished easily even when they are difficult or impossible to detect by the usual peptide mapping.     

Enzymatic methyl esterification of a deamidated form of mouse epidermal growth factor

The enzyme S-adenosylmethionine:protein carboxyl-O-methyl-transferase, type II (EC 2.1.1.77; PCMT) from eukaryotes methyl esterifies peptides containing isoAsp residues, which can arise from spontaneous deamidation of labile Asn residues. We report here a study on in vitro methyl esterification of mouse EGF by bovine brain PCMT. This peptide contains two Asn in the sequences Asn1-Ser2 and Asn16-Gly17. It is known from the literature that the presence of a small residue on the carboxyl side of asparaginyl makes this residue susceptible to deamidation through the spontaneous formation of a succinimide intermediate. Therefore EGF was incubated under deamidating conditions (pH 9.0, 37 degrees for 48 h) and the extent of deamidation monitored by enzymatically measuring the NH3 produced during the alkali treatment: a release of 0.80 mol NH3/mol EGF was calculated. The alkali-treated EGF, analyzed by anion-exchange chromatography, shows two major components identified as native EGF (nEGF) and its deamidated form (dEGF). When incubated in the presence of purified PCMT neither nEGF nor dEGF showed any methyl accepting capability. Since it is known that the three-dimensional structure of a protein may hinder the methyl esterification of a potential ethyl accepting site, dEGF was unfolded by reducing and alkylating the intrachain disulfide bridges. Only a slight increase in the methyl accepting capability could be observed. Conversely, when EGF was deamidated after its unfolding, the resulting protein was stoichiometrically methylated by PCMT, presumably at level of isoAsp16. Our findings strongly suggest that the three-dimensional structure of a protein is a major specificity determinant for both deamidation and methyl esterification processes.     

Site-specific transamidation and deamidation of the small heat-shock protein Hsp20 by tissue transglutaminase

Crosslinking of small heat-shock proteins (sHsps) by tissue transglutaminase (tTG) is enhanced by stress and under pathological conditions. We here used hexapeptide probes to determine the amine donor (K) and acceptor (Q) sites for tTG in Hsp20. Mass spectrometric peptide mass fingerprinting and peptide fragmentation established that Q31 and the C-terminal K162 are involved in inter- and intramolecular crosslinking (transamidation). Q31 is a conserved glutamine in sHsps where the neighboring residue determines its reactivity. Moreover, we detected highly efficient simultaneous deamidation of Q66, which suggests that tTG-catalyzed transamidation and deamidation is specific for different glutamine residues.     

Structure Based Prediction of Asparagine Deamidation Propensity in Monoclonal Antibodies

Identification of asparagine (Asn) sites that are prone to deamidation is critical for the development of therapeutic monoclonal antibodies (mAbs). Despite a common chemical degradation pathway, the rates of Asn deamidation can vary dramatically among different sites, and prediction of the sensitive deamidation sites is still challenging. In this study, characterization of Asn deamidation for five IgG1 and five IgG4 mAbs under both normal and stressed conditions revealed dramatic differences in the Asn deamidation rates. A comprehensive analysis of the deamidation sites indicated that the deamidation rate differences could be explained by differences in the local structure conformation, structure flexibility and solvent accessibility. A decision tree was developed to predict the deamidation propensity for all Asn sites in IgG mAbs based on the analysis of these three structural parameters. This decision tree will allow potential Asn deamidation hot spots to be identified early in development.