Co-translational Installation of Posttranslational Modifications by Non-canonical Amino Acid Mutagenesis

Abstract

Protein posttranslational modifications (PTMs) play critical roles in regulating cellular activities. Here we provide a survey of genetic code expansion (GCE) methods that were applied in the co-translational installation and studies of PTMs through noncanonical amino acid (ncAA) mutagenesis. We begin by reviewing types of PTM that have been installed by GCE with a focus on modifications of tyrosine, serine, threonine, lysine, and arginine residues. We also discuss examples of applying these methods in biological studies. Finally, we end the piece with a short discussion on the challenges and the opportunities of the field.

Graphical Abstract

Protein post translational modifications (PTMs) regulate every aspect of cellular function with profound implications in biology and health. To facilitate study of PTMs, genetic code expansion (GCE) has emerged as a powerful tool to synthesize proteins with homogeneous and stoichiometric PTMs or their analogs. This Concept paper summarizes several common PTMs that are successfully installed through GCE and discusses their applications in biological studies.

Introduction

Posttranslational modification (PTM) of proteins greatly expands the diversity of proteome and is essential for proteins’ biological functions. The most common types of PTM include phosphorylation, glycosylation, lipodation, methylation, acylation, ubiquitination, SUMOylation, sulfation et al.^[1] Given their importance in biology and medicine, it is necessary to investigate and understand the function of these PTMs. Such studies often require the synthesis of proteins with homogeneous and stoichiometric modifications at specific site(s). Although one can use native enzymes to install desirable PTMs, general limitations exist. On one hand, some modification enzymes have broad substrate specificity, a desirable PTM cannot be obtained at a specific site but not affect other site(s). On the other hand, some modification enzymes have relatively rigid substrate specificity, a non-native PTM cannot be installed at positions beyond the naturally occurring sites. Furthermore, the dedicated PTM enzymes are largely unknown in many cases, which limits a broad application of the enzymatic method. So, can we directly synthesize proteins with desirable PTMs without relying on native enzymes? Recent advances in chemical biology methods, including native chemical ligation,^[2] expressed protein ligation,^[3] and other enzyme-catalyzed peptide/protein ligations, have made great strides towards this goal by enabling the syntheses of numerous biologically important proteins with selective PTMs.^[4–6] While these methods are generally applicable to the installation of many PTMs, the sites of modification are usually limited to the N- or C-terminus. Installation of a desirable PTM at an internal site of a large protein frequently requires ligations of multiple fragments.^[5] Furthermore, because these methods are typically carried out in vitro, the synthesized proteins are mostly used for studies outside of live cells. As an alternative approach, a PTM of interest can be site-specifically installed into any permissive sites of target proteins in live cells through the genetic code expansion (GCE). This concept paper describes fundamentals and applications of GCE to the investigation of PTMs.

Genetic Code Expansion (GCE)

The genetic incorporation of noncanonical amino acids (ncAAs) with defined chemical and physical properties into proteins is a useful tool for biological studies. A general approach for the site-specific incorporation of ncAAs in live cells has been developed (Figure 1).^[7–18] In this system, orthogonal aminoacyl-tRNA synthetase (aaRS)/tRNA pairs, which do not cross-react with any of the endogenous tRNAs and aaRSs of the host cell, are generated to recognize a blank codon (codon that does not encode natural proteinogenic amino acids, e.g., amber UAG codon or a quadruplet codon). The orthogonal aaRS is then modified to charge its cognate tRNA with only the desired ncAA. This methodology enabled the incorporation of over 150 ncAAs with a variety of side chain structures and functions in live cells, including spectroscopic probes,^[19–24] metal chelators,^[25–30] posttranslational modification analogs,^[31–33] redox-active groups,^[34–36] and probes with unique chemical reactivities.^[37–41]

Figure 1.

General scheme of Genetic Code Expansion (GCE). A UAG codon is used as a representative blank codon. aaRS, aminoacyl-tRNA synthetase, ncAA, noncanonical amino acid.

Installation of PTMs with GCE

The use of GCE for site-specific installation of PTMs is mainly focused on three approaches: (1) direct incorporation of a ncAA that contains the modification of interest. Once the target protein is synthesized, the desirable PTM is properly installed, and biological investigations can be conducted without any further manipulations; (2) incorporation of a ncAA that contains a masked PTM. After translation, the masking group can be (photo)chemically removed and the resulting protein is ready for subsequent studies; and (3) incorporation of a ncAA that contains a chemical handle which enables the introduction of a desirable PTM through subsequent bioconjugation reactions. For most cases, this approach yields a mimicry of the desirable PTM.

There are over 400 different types of PTMs, targeting a range of canonical amino acids that contain chemically reactive side chains.^[42] This Concept paper does not intend to cover every report of using GCE to install PTMs, but rather focus on several representative types of modification to discuss the approach. Overall, GCE has mainly been used for adding PTMs to lysine, tyrosine, serine, threonine, and arginine residues (Figure 2). There are two reasons for this. Firstly, these are among the most frequently modified residues with significant biological implications.^[43] Secondly, the GCE has extensively explored incorporation systems targeting derivatives of these amino acids, including the tyrosyl-tRNA synthetase (TyrRS)/tRNA^Tyl, pyrrolysyl-tRNA synthetase (PylRS)/tRNA^Pyl, and phosphoseryl-tRNA synthetase (SepRS)/tRNA^Sep pairs.

Figure 2.

Posttranslational modifications (PTMs) and their analogs installed by GCE and discussed in this Concept paper. Compound abbreviations are included in the main text. R* represents formyl, propionyl, lactyl, butyryl, 2-hydroxybutyryl, 2-hydroxyisobutyryl, crotonyl, benzoyl, lipoyl, amino acid acyl groups, ubiquitin and SUMO proteins.

Tyrosine phosphorylation

Tyrosine phosphorylation participates in many physiological and pathological processes, and therefore it is of great significance to study how particular tyrosine phosphorylation events are relayed through the cellular signaling network to regulate important biological phenomena. Genetic incorporation of phosphotyrosine (pTyr) is challenging since it has low permeability through cell membranes and is susceptible to hydrolysis by intracellular phosphatases. To overcome these challenges, several approaches have been developed.

The first strategy aims to elevate the intracellular concentration of free pTyr by including high concentration of pTyr in the growth media and decreasing cellular phosphatase activities in parallel.^[44] In this work, five phosphatases that displayed relatively high activity to pTyr were knocked out from the genome of the E. coli host. After screening a collection of mutants that were derived from a previously reported M. jannaschii TyrRS variant for sulfotyrosine (sTyr) incorporation^[33] and engineering the EF-Tu to better bind phosphotyrosylate tRNA, genetic encoding of pTyr in live E. coli cells was achieved for the first time albeit with low efficiency. In an alternate approach, a dipeptide that contained a C-terminal pTyr residue was used to increase the cellular uptake.^[45] It has been reported that the E. coli dipeptide transporter DppA mainly recognizes the N-terminal amino acid. Once inside cells, the dipeptide was hydrolyzed by nonspecific peptidase(s) to yield free pTyr. Its subsequent incorporation in E. coli utilized a previously reported M. jannaschii TyrRS mutant^[32] that recognizes p-carboxymethyl-L-phenylalanine (p-CMF), a pTyr analog, as the substrate. While incorporation of pTyr has been achieved in live cells, its practical utilities may be limited due to dephosphorylation events that readily happens even in the presence of a phosphatase inhibitor during protein expression.^[45]

The second strategy explored a two-step approach, in which a stable and masked pTyr precursor was incorporated into the target protein, followed by deprotection to yield phosphorylated tyrosine.^[46] In this work, a phosphoramidate-containing pTyr precursor, which does not have a negatively charged side chain and is resistant to hydrolysis by phosphatases, was used. This strategy can largely avoid dephosphorylation of target proteins inside cells. However, a relatively harsh acidic condition is required for the deprotection step, and the re-folding of target proteins may be needed following the acid hydrolysis. Furthermore, the strategy is mainly used to obtain homogeneously phosphorylated proteins for in vitro investigations and is not designed for live cell studies.

The third strategy is to incorporate cell-permeable and sable pTyr analogs. In addition to potentially overcoming the uptake and hydrolysis issues associated with pTyr, non-hydrolyzable pTyr analogs are useful to study specific phosphorylation events. Indeed, the dynamic inter-conversion of phosphorylated and dephosphorylated isoforms can complicate interpretations of experimental data and further hinder deciphering the roles of individual phosphorylation sites. The first demonstration of a co-translational incorporation of a non-hydrolyzable pTyr analog, p-carboxymethyl-L-phenylalanine (pCMF), was achieved in E. coli by engineering a M. jannaschii TyrRS/tRNA pair.^[32] pCMF is a pTyr mimic albeit bearing only a single negative charge at physiological pH. Later, the genetic incorporation of another pTyr analog, 4-phosphomethyl-L-phenylalanine (Pmp), into proteins in E. coli was reported.^[45] In comparison to other analogs, such as pCMF, Pmp is both structurally and electrostatically more similar to pTyr. As another step forward, direct incorporation of pCMF was recently achieved in mammalian cells, reported independently by two groups.^{[47, 48]} This will facilitate real-time studies of pTyr-associated cellular events in the native mammalian host cells.

Tyrosine O-sulfation

Protein tyrosine O-sulfation (PTS) plays a crucial role in many physiological and pathological processes.^[49–53] To advance the study of PTS, several GCE approaches have been developed to genetically encode sTyr in E. coli and mammalian cells.^{[33, 54–56]} In 2006, a M. jannaschii TyrRS/tRNA pair was successfully engineered to incorporate sTyr into recombinant proteins in E. coli.^[33] The efficiency of this system can be improved by combining six copies of the orthogonal tRNA and two copies of sTyrRS.^[54] Although this system has been proven efficient in bacteria cells, it cannot be applied in mammalian cells because the M. jannaschii TyrRS/tRNA pair is not orthogonal to the mammalian TyrRS/tRNA. To overcome this limitation, two groups independently demonstrated the directed evolution of the E. coli TyrRS/tRNA pair for the site-specific incorporation of sTyr into recombinant proteins in mammalian cells.^{[55, 56]}

Serine/Threonine phosphorylation

Serine/threonine phosphorylation is more abundant in the phosphoproteome than that of tyrosine phosphorylation. Significant efforts and progresses have been made towards the genetic encoding of phosphoserine (Sep), phosphothreonine (pThr) and their analogs in live cells.

The first demonstration of a co-translational incorporation of Sep in E. coli cells was achieved by engineering a phosphoseryl-tRNA synthetase (SepRS)/tRNA^Sep pair that was derived from Methanococcus maripaludis SepRS and M. jannaschii tRNA^Cys.^[57] An engineered EF-Tu (EF-Sep), which bound the negatively charged Sep-tRNA more productively than the wild-type EF-Tu, was required for a successful incorporation of Sep. To further improve incorporation efficiency of Sep, additional efforts were made: (1) to knock out release factor 1 and/or to use genomically recoded E. coli hosts with large-scale or selective (leading to healthier cells) removal of endogenous TAG codon^[58–61] in order to enhance amber suppression efficiency; (2) to further engineer SepRS, tRNA^Sep, and EF-Tu, which led to a significant improvement of efficiency and fidelity in Sep incorporation;^{[62, 63]} and (3) to use serB knockout strains or cell lines,^{[63, 64]} which minimize the dephosphorylation of Sep. In addition to Sep, the genetic incorporation of a Sep analog, phosphonomethylene alanine (Pma), that is resistant to hydrolysis by serine phosphatases was achieved in a metabolically engineered E. coli using an optimized SepRS/tRNA^Sep pair.^[63] A recent study further demonstrated the successful genetic encoding of Sep and Pma in mammalian cells using an engineered SepRS/tRNA^Sep pair.^[65]

Because of the structural similarity between pThr and Sep, the genetic encoding of pThr was achieved through further evolving a reported SepRS/tRNA^Sep pair.^[66] Remarkably, the engineered pThr tRNA synthetase (pThrRS) exclusively recognizes pThr but not Sep. The efficiency of this system was further increased with an elevated intracellular concentration of pThr through its in situ synthesis in an E. coli host with a heterologously expressed Salmonella threonine kinase.^{[67, 68]} Future efforts for the genetic encoding of pThr calls for the development of an incorporation system in mammalian cells, where threonine phosphorylation plays crucial roles in regulating diverse cellular processes. Furthermore, incorporation systems that can recognize nonhydrolyzable analog(s) of pThr is expected to have universal utility for the study of threonine phosphorylation in live cells.

PTMs of lysine

Lysine residues are primary targets for PTM and can be modified with many different chemical groups.^[69] Among them, GCE has been successful in the incorporation of acylated lysines. Indeed, several PylRS/tRNA pairs have been identified to site-specifically incorporate acetyl lysine (AcK) into proteins in microbes, mammalian cells, and animals.^[70–73] To simplify biological studies of reversible lysine acetylation events, non-deacetylatable AcK analogs have been genetically incorporated into proteins using engineered PylRS/tRNA pairs as well.^[74–77] In addition to acetylation, engineered PylRS variants also enabled the genetic encoding of other acylated lysines, including propionyl lysine,^{[78, 79]} butyryl lysine,^{[78, 79]} 2-hydroxyisobutyryl lysine,^[80] crotonyl lysine, ^{[78, 79, 81]} formyl lysine,^[82] benzoyl lysine variants,^[83–85] lactyl lysine, β-hydroxybutyryl lysine, lipoyl lysine,^[86] and amino acid acylated lysine.^{[87, 88]}

Methylation of lysine residues is another hallmark PTM event in epigenetics,^[89] but the genetic encoding of methylated Lys is currently underdeveloped. A direct incorporation of methylated lysine variants has not been achieved in live cells. It is possible that engineering PylRS to specifically recognize methylated lysine variants but not lysine is challenging. To overcome this potential hurdle, several caged monomethylated lysine analogs have been genetically incorporated into proteins in E. coli and mammalian cells using engineered PylRS/tRNA pairs.^[90–92] Removal of protecting groups usually requires acidic conditions, ruthenium catalysis, or UV irradiation. More complex schemes that require a series of chemical manipulation was devised for installing dimethylated and trimethylated lysines.^[93–95] While additional step(s) complicates the experimental procedure and therefore may not be ideal for certain biological studies, the GCE methods do provide a means for controlling lysine methylation events.

PTM with ubiquitin functions as critical regulators of many cellular processes. Ubiquitin is typically attached to lysine residues of substrate proteins, or to one of the seven lysine residues or the N-terminal of ubiquitin itself to form polyubiquitin chains. Similarly, proteins can also be modified by ubiquitin-like proteins (e.g., SUMO), which are a family of proteins that are structurally and evolutionarily close to ubiquitin. Due to their large size, direct contranslational incorporation of these PTMs is not possible. Therefore, step-wise approaches were developed to couple GCE with either chemical reactions^{[96, 97]} or native chemical ligation followed by desulfurization.^[98] GCE approach, in combination with Staudinger reduction and sortase-mediated transpeptidation, has also been developed to site-specifically modify target proteins with ubiquitylation and SUMOylation.^[99]

Citrullination of arginine

The citrullination of arginine entails the hydrolysis of the guanidium side chain under the catalysis of protein arginine deiminases.^[100] This PTM converts a positively charged arginine into a neutral citrulline residue, which affects the structure and function of a protein. Indeed, citrullination of arginine is essential for many physiological processes and is implicated in the development of certain diseases.^[101–105] Since citrulline is structurally similar to arginine, it is difficult to engineer an aaRS that specifically recognizes citrulline but not arginine. To meet this challenge, a step-wise strategy was developed to first install a photocaged-citrulline into the target protein using an engineered E. coli leucyl-tRNA synthetase followed by decaging at 365 nm to form citrulline.^[106] The utility of this GCE approach was demonstrated by studying the role of citrullination at Arg372 and Arg374 in the activity of the protein arginine deiminase 4.^[106]

Biological Investigations

Majority of PTMs installed by the GCE method are limited to feasibility demonstration with broad applications to real biological studies lagging behind. Followings are a few representative examples for which the installation contributed to the mechanistic investigation of particular PTMs in biological processes.

Acetyl-lysine (AcK) and analogs

Lysine acetylation is a universal protein PTM in both prokaryotes and eukaryotes.^[107]. It entails the addition of an acetyl group at the N^ε-position of the side chain, which leads to changes of nucleophilicity, hydrophobicity, and local structure. The acetylation of lysine is a reversible process that is catalyzed by lysine acetyltransferases and lysine deacetylases.^{[108, 109]} Since its discovery as an epigenetic marker in histone proteins, lysine acetylation has been reported in numerous functional and regulatory proteins. In human proteome, over 38,000 sites have been mapped for this PTM.^[110] It plays crucial roles in regulating many cellular processes, including transcription, chromatin remodeling, DNA repairs, metabolism, and cell proliferation.^[111–113]

A detailed biological investigation of lysine acetylation is often hindered by the inaccessibility to a large quantity of high-purity proteins with homogeneous and well-defined modifications. The challenge is addressed with GCE-enabled site-specific incorporation of AcK using engineered PylRS/tRNA^Pyl systems.^[70–72] Because the important role played by lysine acetylation in chromatin remodeling, AcK has been installed at defined positions in multiple human histone proteins to investigate the effect of specific modification events. In these efforts, acetylated histones were heterologously expressed in E. coli cells and purified for biophysical and biochemical characterizations.^{[79, 114–117]} An early example demonstrated the feasibility of in vitro reconstitution of histone octamer and nucleosome using purified protein with defined lysine acetylation.^[114] In this study, the effects of H3 K56 acetylation on chromatin remodeling were examined. The availability of uniformly acetylated histone protein also enabled activity studies of human histone deacetylases (HDACs) on substrates that better resemble the natural chemical context than short peptides. Research that focused on the deacetylation of H3 observed higher catalytic activities on the full-length protein than on peptide substrates when SIRT1, SIRT2, and HDAC8 were examined.^{[115, 117]} Besides the core histone proteins, GCE-enabled AcK incorporation was also applied to study the acetylation-dependent interactome of linker histone H1.^[118] Using purified none, mono-, and di-acetylated H1.2 as baits in cell lysates, affinity purification mass spectrometry analysis revealed interacting partners, including transcriptional regulators and translational initiators, that distinctly recognized acetylated proteins. Structural biology studies of modified nucleosomes also benefited from the site-specific installation of AcK through GCE.^[119]

Besides histone proteins, the GCE-enabled site-specific incorporation of AcK was broadly applied to investigate how lysine acetylation affects the functions of other structural, regulatory, and metabolic proteins.^[120–131] Cyclophilin A (CypA) is a peptidylprolyl cis-trans isomerase (PPIase) with important implications in human health^[132]. Its acetylation on lysine residues was detected in both intracellular and secreted forms.^{[120, 133, 134]} A detailed structural and biophysical study enabled by AcK incorporation at the K125 position revealed changes in the binding affinity and catalytic activity of CypA as a result of altered local electrostatic environment rather than large structural reshaping caused by the acetyl group.^[120] With acetylation at K125, the catalytic efficiency of CypA was reduced 35 folds, which correlated to a switch in the preferred binding mode of HIV-1 capsid from the trans to cis conformation at the Gly89-Pro90 peptide bond. The results clearly showed that the K125 acetylation event plays a key role in regulating biological functions of CypA. This report is also the first structural biology study of a protein with biologically relevant lysine acetylation.

Understanding the biological role of specific acetylation event(s) is important when multiple lysine residues in a protein are subjected to the modification. A well-accepted approach was to use arginine or glutamine to mimic the lack or presence of an AcK residue at the site of interest, respectively. However, a few examples showed that glutamine does not faithfully represent the property changes following the acetylation.^{[127, 135–137]} By taking advantage of the high-fidelity and site-specificity of GCE-enabled incorporation, five lysine acetylation sites in the small GTP-binding protein Ran were studied using monoacetylated full-length proteins.^[123] Results from the study revealed how the acetylation at a specific position in Ran impacts its interaction with the guanine exchange factor RCC1 and its essential functions, such as nucleotide hydrolysis, binding to import and export receptors. The GCE-enabled method was also adapted in the first comprehensive report on lysine acetylation in ubiquitin (Ub).^[127] As one major type of PTM, how ubiquitination is regulated through lysine acetylation of Ub was not well understood. In this study, a collection of Ub proteins with AcK at each one of the seven lysine sites was generated using GCE. Initial NMR analysis revealed that each variant had unique structural changes, which were translated into substrate preference by E3 ligases and distinct sets of binding partners that interacted with modified Ubs in an acetylation site-specific manner. Meanwhile, in vitro assay results showed that the HDAC6 deacetylates ubiquitin, and its activity is position independent. Above studies were made possible due to the single-residue resolution of the GCE-enabled AcK incorporation.

As part of the intricate regulatory network in mammalian cells, a single protein is often subjected to multiple types of PTM for accurate spatial and temporal controls. In an effort to investigate the chromosomal segregation during mitosis, the interplay between lysine acetylation and threonine phosphorylation in Aurora B kinase was studied using proteins generated from GCE-enabled AcK incorporation.^[124] It was shown that the K215AcK variant underwent comparable autophosphorylation at Thr232 as the wild-type protein, but the pT232 modification was only largely preserved in the acetylated protein following the treatment with PP2A phosphatase. Together with results from inhibition study and affinity assay, the research supports the conclusion that acetylation at Lys215 ensured the Aurora B kinase to maintain an activated form that is necessary for the robust chromosomal segregation.

Rapid turnover caused by deacetylase activities presents another challenge in studying the biological effect of lysine acetylation. Nonhydrolyzable AcK analogs have been broadly applied in studies using peptide substrates. GCE-based methods to incorporate these analogs, including trifluoroacetyl-lysine (TfAcK)^{[74, 138]} and thioacetyl-lysine (ThioAcK)^[75], open new research avenues. When utilized as a probe in ¹⁹F NMR analysis of acetylated proteins, TfAcK at the K164 position of p53 tumor suppressor protein was shown to be resistant to the NAD⁺-dependent SIRT1 and SIRT2 deacetylases in vitro and in vivo ^[74]. Similar observation was made in forkhead domain of FOXO4 transcription factor with TfAcK at position K189.^[139] Meanwhile, ThioAcK incorporated at K140 position of the E. coli malate dehydrogenase was also resistant to the activity of a bacterial sirtuin deacetylase, CobB.^[75] On the other hand, both analogs are still susceptible to zinc-dependent lysine deacetylases.^{[140, 141]} Interestingly, the deacetylation of TfAcK proceeded at an elevated rate than AcK,^{[139, 141]} probably due to the strong electron-withdrawing effect of the trifluoro substituent. When tested within histone H3 peptides, TfAcK and ThioAcK interacted with bromodomains at a similar to slightly reduced level in comparison to AcK.^[139] Although concerns over the downstream effects of covalent intermediate formation with the ThioAcK remains^[142], the results paved the way for cellular studies with these nonhydrolyzable AcK analogs following GCE.

In comparison to in vitro studies, cell-based investigation of lysine acetylation using the GCE approach is less explored. In a recent example, acetylation of the trans-activation response DNA-binding protein of 43 kDa (TDP-43) was examined using genetically incorporated AcK in HEK293 cells.^[143] The study showed that the K136 acetylation led to phase separation of TDP-43 and identified SIRT1 as the cognate deacetylase of this site. While the first observation correlated with results from studies using glutamine as the surrogate of AcK, the second observation can only be made using the authentic AcK modification. In a similar study, AcK incorporation was applied to study IFN regulatory factors 3 and 7, i.e., IRF3 and IRF7, which play important roles in the innate immune system’s response to virial infections. The cell-based studies showed that when AcK was site-specifically installed into IRF3 or IRF7, both proteins failed to phase separate and to form discrete nuclear puncta. The results confirmed that lysine acetylated IRF3 and IRF7 have reduced affinity towards IFN-stimulated response elements and therefore are not transcriptionally active forms.^[130] To facilitate cellular studies, stable mouse embryonic stem cell lines with GCE machinery for AcK were generated using PiggyBac transposon.^[73] Incorporation of AcK into H3.3 protein resulted in gene transcription level changes that are likely regulated by acetylation. A similar study was carried out in yeast cells using plasmid-supplied genetic components to compare the effects of acetylation and crotylation at K56 in the H3 protein.^[144] The encoded AcK provided better protection against genotoxic reagents and led to the upregulation of genes mainly functioning in cell wall and sexual reproduction pathways when treated with methyl methanesulfonate. The acetylation at K56 also changes the transcription level of PTM enzymes.

Sulfotyrosine (sTyr)

Protein tyrosine sulfation is a PTM that is identified in mammalian cell surface and secreted proteins following their transit through the trans-Golgi.^{[52, 145]} Under the catalysis of tyrosylprotein sulfotransferase 1 or 2 (TPST1 or TPST2), a sulfate group from 3’-phosphoadenosine 5’-phosphosulfate (PAPS) is transferred to a protein tyrosine residue. The introduction of a negatively charged sulfate group changes a protein’s electrostatic property and local structure, which contribute to the binding affinity and specificity between protein partners. Therefore, tyrosine sulfation plays crucial roles in extracellular biomolecular interactions that dictate various biological processes including cell adhesion, hemostasis, hormone activities, inflammation, and immunity^[146–148]. Unlike most other PTMs, the sulfation of protein tyrosine is considered as an irreversible modification. It is also a type of PTM that is currently understudied.

Due to the instability of the sulfate ester bond under mass spectrometry conditions, a global profiling of tyrosine sulfation sites in human proteome has not been achieved. Computational algorithms were developed to predict potential sites of the modification.^[149–152] GCE-based incorporation of sTyr in E. coli cells led to the production of recombinant sulfated hirudin^[54] and a cell surface immune receptor in rice (RaxX)^[153] with enhanced biological functions in in vitro studies. Following the development of sTyr GCE system in mammalian cells, the site-specific installation enabled live-cell studies under biologically relevant conditions.^[55] An axis formed by the C-X-C chemokine receptor type 4 (CXCR4) and its cognate ligand, i.e., chemokine CXCL12, has emerged as a potential tumor therapeutic target because of its roles in activating multiple pathways of tumor initiation and progression.^[154] When sTyr was introduced at Tyr21 of CXCR4 by GCE, a functional CXCR4 protein led to calcium flux in cells treated with CXCL12. The report validated results from studies using chemically synthesized peptides.^[155] Similar to lysine acetylation, multiple sites of a protein are often subjected to sulfation. Understanding the biological effects of each sulfation event often relies on loss-of-function method which entails mutating the tyrosine into a phenylalanine residue. The developed GCE-enabled sTyr incorporation is a gain-of-function method for the investigation of multi-site sulfation in live cells. As another application, the effect of sulfation on distinct Tyr residues of human heparin cofactor II (HCII) was investigated using purified protein.^[56] It was found that sTyr73 likely contributed more to the binding of HCII’s N-terminal domain to its glycosoaminoglycan binding domain, and the sTyr60 could be more important for thrombin recruitment. A recent report on in situ biosynthesis of sTyr further streamline the GCE approach.^[156]

New discovery

Besides applications to study known types of PTM, a recent report discovered a previously unknown type of PTM by studying GCE-enabled PTM events. The reversible aminoacylation of lysine by endogenous amino acids was first reported in proteomics studies of liver cancer cells as a regulatory mechanism.^[157] Mass spectrometry study of GCE-enabled incorporation of methionyl lysine revealed that the free amino group of the methionine moiety can be further ubiquitinated.^[88] Further investigation of lysine that are aminoacylated by other amino acids showed similar ubiquitination and the reactions were catalyzed by the ubiquitin-conjugating enzyme UBE2W.

Future Directions

Tremendous progresses have been made to install various types of PTMs or their analogs into proteins through GCE. Meanwhile, this burgeoning field in chemical biology still bears many challenges and opportunities.

The slow adaptation of the GCE-enabled method in cellular studies reflects a general concern over the impact of GCE machinery on normal biological functions, albeit previous studies have shown minimal fitness changes.^{[73, 158–160]} Methods that allow stringent control over the activity of the GCE machinery can minimize undesirable effects. One possible solution emerges from studies into how short nucleotide fragments that flank the site of blank codon affected the decoding efficiency.^{[161, 162]} Identified enhancer sequences can be embedded as signals for high level of decoding at the designated site while maintaining an overall low level of readthroughs. Accessibility to ncAAs presents another general obstacle. Recent efforts towards in situ biosyntheses present a viable solution to certain cases but may not be applicable to all. Robust and easy-to-follow (bio)synthetic methods will further empower (chemical) biologists to apply GCE.

Many engineered orthogonal aaRS/tRNA pairs are less efficient than native ones. While an immediate solution is to increase the expression level of aaRS/tRNA in host cells, further engineering of aaRS/tRNA pairs themselves to improve their intrinsic efficiency are also desirable. In addition, the engineered orthogonal aaRS/tRNA pairs are foreign to host cells and their interactions with endogenous translation machinery need to be optimized. Such efforts include modifications of tRNA,^[163–165] EF-Tu,^{[57, 62, 166]} ribosome,^[167–171] and genome (codon replacement and release factor knockout).^[172–177] While these strategies are currently used mostly for GCE in bacteria, some of them are potentially applicable to mammalian systems as well.

The precise choreography of various protein PTM events is essential for the survival and propagation of living cells. A thorough investigation of such regulatory mechanism by GCE calls for the ability to install multiple types of PTM simultaneously, which requires mutually orthogonal aaRS/tRNA pairs and blank codons. This is also one of the current focuses of the GCE field. Meanwhile, bacterial systems are better developed for certain types of PTM (e.g., tyrosine phosphorylation), their direct installation and study using GCE in the native mammalian hosts are highly desirable. Furthermore, general GCE methods to directly install certain types of PTM or to target certain amino acid residues in any cell types still require further development.

In summary, we discussed recent advances and representative examples in applying GCE to the study of PTMs. Besides basic biological investigations, GCE can also be used to generate proteins with PTMs at natural or non-natural sites at a large scale for biotechnological and therapeutical applications. Readers are directed to recent reviews that cover additional aspects of applying GCE to PTM research.^{[178, 179]}

Biographies

Dr. Jiantao Guo obtained his B.S. and M.S. degrees in Organic Chemistry from Nankai University with Professor Jin-Pei Cheng, and his Ph.D. degree from Michigan State University with Professor John W. Frost. He pursued postdoctoral studies with Professor Peter G. Schultz at The Scripps Research Institute before joining the University of Nebraska – Lincoln as an assistant Professor. Dr. Guo is currently a Professor of Chemistry and the Director of the Nebraska Center for Integrated Biomolecular Communication. Dr. Guo’s research is mainly at the interface of Chemistry and Biology, focusing on chemical biology tools, genetic code engineering, and protein posttranslational modifications.

Dr. Wei Niu obtained her B.S. degree in Organic Chemistry and M.S. degree in Biophysics with Professor Senfang Sui from Tsinghua University in China. She conducted her Ph.D. study of biocatalysis under the supervision of Professor John W. Frost in the Chemistry Department of Michigan State University. After spending five years working in biotech companies, Dr. Niu returned to academia and currently is an Associate Professor in the Department of Chemical and Biomolecular Engineering of University of Nebraska - Lincoln. Her research focused on synthetic biology with special interests in protein engineering, microbial metabolic engineering, and genetic code expansion.