Residue-Specific Peptide Modification: A Chemist’s Guide

Abstract

Advances in bioconjugation and native protein modification are appearing at a blistering pace, making it increasingly time consuming for practitioners to identify the best chemical method for modifying a specific amino acid residue in a complex setting. The purpose of this perspective is to provide an informative, graphically rich manual highlighting significant advances in the field over the past decade. This guide will help triage candidate methods for peptide alteration and will serve as a starting point for those seeking to solve long-standing challenges.

Graphical abstract

Just as some of the most pivotal advances in total synthesis can be tied to overcoming obstacles in chemoselectivity,¹ advances in bioconjugation chemistry are intimately linked to the development of targeted amino acid modifications for the precise engineering of proteins. Given the sheer number of functional groups present and the general requirement for aqueous reaction media, protein modification is perhaps the ultimate expression of a chemoselective reaction. A growing appreciation for the therapeutic potential of peptides^2,3 has led to numerous initiatives that necessitate an equivalent degree of chemoselectivity in diverse, late-stage peptide functionalizations. This challenge has mobilized peptide and organic chemists alike, a point illustrated by the upsurge of interdisciplinary collaborations over the past decade. Such endeavors have uniquely altered the landscape of traditional amino acid modification strategies for both peptide functionalization and protein bioconjugation, and will be the focus of this perspective.

This survey is not intended to provide a comprehensive history of bioconjugation and related chemistry;^4–7 rather, it should serve as a graphically rich catalog for the practitioner interested in recently developed transformations for the discrete modification of polypeptide and protein substrates. Ligation,⁸ stapling,⁹ macrocyclization,¹⁰ and cross-linking¹¹ strategies will not be covered, nor will methods that rely on enzymatic transformations¹² or the incorporation of designer residues.¹³ Instead, chemical modifications specific to proteinogenic residues, select noncanonical but naturally occurring residues [e.g., dehydroalanine (Dha)], and peptide C- and N-termini will be highlighted. The format is pedagogically similar to The Portable Chemist’s Consultant,¹⁴ an e-book recently published for those engaged in chemical synthesis.

A model protein bioconjugation should be site-selective and robust and should proceed under mild conditions (physiological pH, ambient temperature and pressure, and aqueous solvent preferred). Moreover, preservation of the protein structure and conservation of activity are among the most important considerations. While shorter peptide sequences allow for some deviation, the operational simplicity epitomized by these standards is an ideal goal for the development of versatile peptide modification strategies. With these (and other) factors in mind, the aim of this user guide is to provide an etic assessment of both the advantages offered and limitations imposed by the strategies featured herein. Organized by amino acid, this schematic manual will detail key elements of each reaction, including demonstrated or implied compatibility (indicted by brackets) with common peptide functional groups, attributes unique to the method described, and challenges that may arise in reagent preparation or reaction protocol. Specific examples and general trends are highlighted in the text to guide the reader and provide essential information in a succinct manner.

Finally, this guide is supplemented with a high-level view of the field [see Table 3 (Amino Acid Side-Chain Modification Report Card) and the Supporting Information] that offers a snapshot of the available literature in the area. These resources have been compiled with an eye toward gaps in current methodology and opportunities for innovation.

Table 3.

^a

^aSee the Supporting Information for details.

ALIPHATIC SIDE CHAINS

Without an apparent functional handle, there are relatively few methods available (vide infra) for derivatization of the aliphatic residues. However, fundamental advances in the direct functionalization of C–H bonds have ushered in new opportunities for targeted modifications.¹⁵ In perhaps the most notable recent example, Yu and co-workers disclosed a palladium-catalyzed C–H arylation of N-terminal alanine (Ala) residues facilitated by coordination of the metal catalyst to the peptide backbone.¹⁶ Although currently limited to di-, tri-, and tetrapeptide substrates, this seminal report illustrates the vast potential of postassembly C(sp³)–H functionalization as an enabling tool for hydrocarbon modifications. A promising metal-mediated approach to the γ-arylation of valine (Val) and isoleucine (Ile) within dipeptide substrates has also been reported¹⁷ (Table 1).

Table 1.

POLAR, NONIONIZABLE SIDE CHAINS

Similar to aliphatic residues, the primary amides of asparagine (Asn) and glutamine (Gln) are all but immovable and have proven to be difficult targets for selective modification. One notable exception is the molecular recognition strategy employed by Popp and Ball, which enabled a targeted, proximity-driven modification of Asn and Gln using dirhodium metallopeptides.¹⁹ In contrast, creative reactions seeking to modify methionine (Met) are far more common. With a relatively high oxidation potential, the reversible oxidation of the thioether to the corresponding sulfoxide or sulfone is a well-defined Met reaction pathway, both synthetically and biochemically.⁹⁶ Strikingly, it is the only native residue that can be alkylated under acidic conditions.⁹⁷ In fact, early approaches to protein degradation for sequencing exploited these unique properties; cyanogen bromide in formic or hydrochloric acid buffer, for example, is a reliable method for Met-specific peptide bond hydrolysis that is still in use today.⁹⁸ The methylthioether side chain can also be used as a homocysteine proxy, as in entry 2.²¹ Alkylation and subsequent demethylation provides the homologated cysteine (Cys) analogue.

In a recent joint effort, Toste, Chang, and co-workers reported the use of redox-activated chemical tagging (ReACT) for site-selective Met modification (entry 4).²³ This bioconjugation strategy harnesses the previously underexplored nitrogen-transfer properties of oxaziridines to form the functionalized sulfimide. The resulting conjugates demonstrate remarkable stability when exposed to a wide pH range, disulfide reducing agents, and bioorthogonal reaction conditions. Furthermore, the ReACT approach exhibits exclusive selectivity for Met in both protein and antibody substrates.

AROMATIC SIDE CHAINS

Among the most abundant targets for new synthetic transformations, residues with aromatic side chains can be further divided into those that are readily ionized (histidine, His, and tyrosine, Tyr) and those that are not (tryptophan, Trp, and phenylalanine, Phe). It should be noted, however, that side-chain pK_a is not an exclusive predictor of reactivity in this case, as indicated by the wealth of transformations specific to the indole of Trp. Importantly, the nucleophilicity of both His and Tyr can be modulated by the protonation state of the side chain. As such, the standard modes of Tyr reactivity can be effectively tuned by pH control: in acidic or neutral environments, the aromatic carbons ortho to the hydroxyl group may undergo ene-type reactions (entry 5),^26,27 diazonium couplings (entry 6),²⁸ and Mannich-type condensation reactions (entry 9),^31,32 while at a pH near the pK_a of phenol (∼10), alkylation or acylation of the oxygen is observed (entries 8 and 10).^30,33,34 In 2009, Barbas and co-workers reported a robust aqueous ene-type reaction that enables click-like Tyr bioconjugation (entry 5).^26,27 Further investigation has expanded the scope and utility of this method, allowing for the installation of diverse handles, including PEG chains and bifunctional linkers.²⁷ In the case of Trp, modifications involving the indole C-2 position dominate the reaction landscape. A recent, metal-free organoradical conjugation strategy (entry 12)³⁶ supplements a variety of powerful metal-based methods (entries 13–15).^19,37–43 Metal-catalyzed C–H functionalization is a particularly useful strategy that enables alkynylation (entry 13)^37,38 and arylation (entry 14)^39–41 of the indole side chain. While not yet fully realized, emerging trends in C–H functionalization are expected to facilitate the development of robust modification protocols for Phe and His.

POLAR, IONIZABLE SIDE CHAINS

As the most readily modified of amino acids, residues with polar, ionizable side chains have been widely examined. Cys, undoubtedly the most well-studied residue, occupies a distinguished place in the bioconjugation literature.^4,99 The inherently low pK_a (∼8.3) and substantial nucleophilic character provide a convenient handle for site-selective derivatization. Historically, cysteine–maleimide conjugation has been the most common manner of thiol modification, but the low stability of the adducts has rendered this particular reaction outmoded.¹⁰⁰ However, as a mild and selective modification, conjugate addition remains a popular approach for functionalization. With recent improvements in conjugate stability (entries 22–25),^52–56 including the design of self-hydrolyzing maleimides,¹⁰¹ as well as the development of bifunctional Michael acceptors (entries 22 and 24),^52,54,55 it is expected that this venerable reaction will retain prominence for years to come.

In an effort to improve the delivery of compact, high-value bioisosteres—namely, cyclobutanes, cyclopentanes, and related analogues—our lab reported the development of a variety of designer, spring-loaded reagents for strain-release functionalization.^45,46 Remarkably, a number of bicyclobutane sulfones and chiral “housanes” were shown to rapidly and selectively functionalize Cys (entry 17)^45,46 in the presence of several unprotected, nucleophilic peptide functional groups.

The use of transition metals in bioconjugation chemistry has gained recent traction, with the combined efforts of Pentelute, Buchwald, and co-workers at the forefront of these endeavors.^47,60 In 2015, the researchers reported a palladium(II)-promoted arylation of Cys under mild conditions (entry 18).⁴⁷ Of note, the palladium(II) complexes used are readily synthesized and easy to handle. In addition, the arylation reaction proved to be an effective tool in peptide stapling and antibody–drug conjugation, leading to stable bioconjugates. The site-guided palladium-mediated arylation of metal-binding proteins¹⁰² and expansion of transition metal functionalizations to the alkynylation (entry 21) and alkenylation of Cys by Messaoudi and co-workers⁵⁰ demonstrate the adaptability of this approach.

While a myriad of acid-specific reactions exists, differentiation between peptide α- and side-chain carboxylic acids (e.g., aspartic acid, Asp, and glutamic acid, Glu) has proven to be a formidable challenge in bioconjugation chemistry. Initially reported by our group in 2016,¹⁰³ the use of “redox-active esters” (RAEs) as a means to activate and engage carboxylic acids in a nickel-catalyzed cross-coupling event has since been parlayed into a broad approach for the construction of carbon–carbon (entries 27, 28, and 44)^58,59,79 and carbon–heteroatom (entry 46) bonds.⁸² The robust procedure, which repurposes activating reagents typically employed in amide bond formation (e.g., HOAt, N-hydroxyphthalimides), has been successfully applied to both resin-bound^58,79 and solution-phase peptide substrates,^59,79,82 allowing for the introduction of discrete cross-coupled products at either side-chain positions or α-positions.

Approaches to lysine (Lys) modification encounter similar obstacles in α- versus ε-amine differentiation, but selective acylation or alkylation can be achieved through careful pH control. Recent advances in arylation (entry 29)⁶⁰ and condensation (entry 30)⁶¹ reactions have also expanded the diversity of functionalizations. The potential for reversible labeling (entries 31 and 32)^62,63 of Lys is particularly intriguing, with envisioned applications toward the development of antibody–drug conjugates (ADCs). Finally, modifications to arginine (Arg) are currently limited in scope; acylation and condensation are common approaches (entries 35 and 36),^66–68 with the most provocative adduct (entry 37)⁶⁹ being a serendipitous byproduct formed during a copper-catalyzed azide–alkyne cycloaddition. To date, chemoselective side-chain modifications at serine (Ser) and threonine (Thr) are exceedingly rare, generally relying on proximity-driven¹⁹ or sequence-specific modifications; more broadly applicable methods represent an opportunity for invention.

NONCANONICAL AMINO ACIDS

Biosynthetically, Dha is often incorporated into peptides through the enzymatic dehydration of Ser;¹⁰⁴ targeted chemical approaches, in contrast, typically proceed through activation–elimination of the Cys thiol¹⁰⁵ or an analogous, and similarly facile, transformation at modified selenocysteine (Sec) residues.¹⁰⁶ The electrophilic, unsaturated side chain is frequently utilized as an acceptor in 1,4-conjugate additions (entries 39 and 40).^71–75 In a recent example, both the Davis laboratory⁷⁵ and the collaborative team of Söll, Lee, and Park⁷⁴ independently reported a radical-based conjugate addition (entry 40) to Dha for the site-selective introduction of diverse non-native residues into several target proteins.

In addition to its role as a Dha precursor, Sec is often purposed as a handle for native chemical ligation.¹⁰⁷ Relatively few examples of covalent derivatization exist, primarily because of the low reduction potential and high polarizability of the selenol functional group.¹⁰⁸ Capitalizing on these otherwise unwieldy characteristics, Pentelute and Buchwald reported a copper-promoted umpolung approach to the arylation of Sec, which takes advantage of the electrophilicity of selenyl sulfides (entry 41).⁷⁶ While the reaction conditions were conclusively determined to favor Sec arylation, oxidation of unprotected Cys was unavoidable.

C-TERMINUS AND N-TERMINUS

Modification of peptide termini is an attractive approach, particularly in cases in which strict bioconjugate stoichiometry is required. The use of specialized linkers to reveal unique functionality on resin cleavage is a common approach to the alteration of C-terminal acids and has been the subject of prior reviews.^109,110 Focusing instead on postassembly synthetic transformations, it is unsurprising that the aforementioned chemoselectivity challenges (see the discussion of Asp and Glu in Polar, Ionizable Side Chains) apply to α-carboxylic acids as well. Nevertheless, orthogonal protecting group strategies and judicious sequence choice have been employed to enable several decarboxylative functionalizations of α-acids, including 1,4-additions (entry 44),⁷⁹ heteroarylations (entry 45),^80,81 and borylations (entry 46).⁸²

Bioconjugation reactions that target α-amines have been the subject of much investigation for use in ligation and macrocyclization; for further insight into these invaluable processes, we direct readers to the many extensive reviews available.^4–10 Alternative modifications at the N-terminus¹¹¹ often rely on participation from the associated side-chain functional group; reactions that engage ionizable residues (e.g., Cys) are common (entries 47 and 48),^83–85 as are examples that employ aromatic (entry 56)^94,95 and other nucleophilic side chains (e.g., Ser and Thr).⁸⁹ The formation of heterocycles is a common approach (entries 47–49, 51, 53, and 56)^{83–86,89,91,94,95} as is N-terminal oxidation and further diversification (entries 50 and 51).^87–89

BACKBONE MODIFICATION

As in small molecule chemistries, transition metals have enabled previously unimaginable bioconjugate transformations,^112,113 an assertion particularly true in the burgeoning area of backbone modifications (Scheme 1).^114–117 In 2016, White and co-workers reported an iron-catalyzed oxidative functionalization inspired by known nonribosomal peptide synthetase (NRPS) pathways.¹⁸ In an impressive display of versatility, the approach was used to synthesize 21 non-natural amino acids from four native residues, with no erosion of chirality observed. Notably, C–H oxidation of the tertiary carbon center of Leu and Val was also demonstrated.¹⁸

Scheme 1.

In another intriguing example, Ball and co-workers detailed a copper-mediated, His-directed amide N-functionalization in a rare example of selective backbone modification in the context of proteins.¹¹⁷ While relatively few examples of these methods currently exist, the profound effects such modifications have on peptide and protein structure and function^118,119 will undoubtedly drive further innovation.

FUTURE OUTLOOK

The biological importance of modified peptides and proteins is a powerful impetus for the continued exploration of diverse amino acid-specific modifications. To this end, the operational simplicity and inherent flexibility of direct modifications of proteinogenic amino acids are undeniable. A high-level view of progress toward residue-specific peptide and protein modifications is outlined in Table 3 (Amino Acid Side-Chain Modification Report Card) and further detailed in the Supporting Information. Categorized by amino acid residue and type of chemical transformation, this graphical depiction illustrates recent and historical successes in the field, with the mild reaction conditions and exquisite chemoselectivity required for protein modifications serving as the ultimate evaluation parameters. Framed by these ideals, the amino acid landscape is marked by several remaining challenges, or rather, opportunities for invention: (1) robust aliphatic functionalization, (2) the transition from “proof of concept” (e.g., single amino acids or short, unadorned peptide sequences) to “widely applicable” (e.g., larger peptides, proteins), and (3) well-defined, site-selective modification in the presence of chemically equivalent residues. We anticipate that advances in metal-catalyzed C–H functionalization will address a number of current limitations and continue to enlist both peptide and small molecule chemists, leading to an expanding library of interdisciplinary methods. Indeed, the requisite functional group selectivity and mild reaction conditions are enviable benchmarks for all new synthetic methods. Translational approaches to peptide and protein modifications therefore have the capacity to revolutionize the capabilities of modern organic synthesis while also delivering high-value biological targets to meet urgent societal needs.

Table 2.

ABBREVIATIONS

PEG

polyethylene glycol

Ab

antibody

Ar

aryl

Boc

tert-butyloxycarbonyl

Alk

alkyl

2-pym

2-pyrimidine

PG

protecting group

Conj

conjugate

HetAr

heteroaryl

ee

enantiomeric excess

NHS

N-hydroxysuccinimide

HPLC

high-performance liquid chromatography

DHA

dehydroascorbate

PTM

post-translational modification

Fmoc

fluorenylmethyloxycarbonyl

SPPS

solid-phase peptide synthesis

Bpin

boronic acid pinacol ester

dr

diastereomeric ratio

AA

amino acid

HOAt

1-hydroxy-7-azabenzotriazole

Nu

nucleophile

LDA

lithium diisopropyl amide

PMP

p-methoxyphenyl

TBHP

tert-butyl hydroperoxide

PDP

2-({(R)-2-[(R)-1-(pyridin-2-ylmethyl)-pyrrolidin-2-yl]pyrrolidin-1-yl}methyl)pyridine

HEPES

4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid