2.1. L-Lysine.

L-Lysine (Lys or K) is an essential α-amino acid building block of proteins with distinct biochemical andbiophysical properties. As one of the three basic natural amino acids (Lys, Arg and His), Lys contains a primary amine(ε-amine) appended to its α-carbon via a linear four-carbon linker (Figure 2).The lone-pair electrons of the ε-amine are protonated at physiological pH of 7.4 to a degree of > 99.9%, asestimated by 3 orders of magnitude higher pK_a of the ε-amine (~10.5 versus 7.4).(43,44) The protonated ε-amine of Lys thuscarries a formal charge of +1.(44) For Lys residues in proteins, this polar andpositively-charged ε-amine can be positioned either within catalytic sites or more often on protein surface with exposureto bulk solvent. The C4 hydrocarbon chain of Lys, unlike its ε-amine, maintains hydrophobic properties. This amphiphiliccharacter thus enables Lys to engage in molecular recognition through diverse modes such as cation-π interaction (acolumbic attraction between a protonated Lys cation and π-electron-rich surface of an aromatic amino acid residue), anion-pairing salt bridge, hydrogen bonds (the ε-amine as a donor and an acceptor), and the hydrophobic interaction viaentropy-driven desolvation of its C4 hydrocarbon. In aqueous media, the cation-π interaction of Lys is energeticallypreferred in comparison with the formation of a salt bridge with a carboxylate moiety.(44,45) Such preference is attributed to higher free-energy penalty of desolvation upon theformation of a salt bridge, which involves a negatively charge carboxylate ion such as Glu and Asp. In contrast, thecation–π complex between a protonated Lys and an aromatic residue gains comparable electrostatic attraction but paysless energy penalty of desolvation for a hydrophobic aromatic residue.

To illustrate Lys-participating interactions, the ε-amine moiety of Lys (its protonated state) is often shown as apositively charged nitrogen surrounded by a methylene and three neutral hydrogen atoms.(27)However, it has been less appreciated that the +1 charge on the ε-nitrogen is merely a formal charge, which should not beviewed as the positive electrostatic density localized on an electronegative nitrogen atom (Figure2).(44,46) Instead, the nitrogen stillcarries significant negative charge with the net +1 charge delocalized around nearby atoms (three hydrogen atoms and theε-amine-adjacent methylene group) (Figure 2). Whereas it is convenient to depict the+1 formal charge of a protonated amine moiety for its formation of hydrogen bonds and a salt bridge, it is more relevant toenvision the actual electronic density on the ε-nitrogen and its neighboring atoms for cation-π interactions (seediscussion later). Additionally, cation-π interactions are strongest when a cation is situated perpendicular to thearomatic plane of a Phe, Tyr or Trp residue.(46) This spatial orientation is preferred toengage in the maximal cation-π interaction between an ammonium cation and the side chain of an aromatic amino acid.(46)

2.2. Epsilon-N-monomethyl lysine (Kme1).

The ε-amine moiety of Lys can be methylated up to three times from unmodified lysine to mono-, di- andtri-methylated Lys (Kme1, Kme2 and Kme3) (Figure 2). The progression of Lys methylationgradually alters biophysical properties of Lys, which determine how these Lys modifications (Kme1/2/3) engage in specificinteractions and are selectively recognized by their biological effectors.(30,31) The pK_a value of the secondary amine of Kme1 is ~ 10.7(estimated on the basis of pK_a of dimethylamine), which is comparable with thepK_a 10.5 of free Lys (Figure 2). At physiological pH of 7.4,the secondary amine of Kme1 is protonated to a degree of > 99.9% with a formal charge of +1 on the electronegativeε-nitrogen (Figure 2). As a more relevant view, the +1 charge is spread aroundadjacent hydrogen and hydrocarbon moieties (two hydrogen atoms, one methyl and one methylene) (Figure 2). Despite dispersion of the +1 charge into two neighboring hydrocarbons, positively charged Kme1 maintainsthe ability to form a salt bridge with negatively charged amino acids such as Glu and Asp (Figure2). In contrast to a protonated Lys as the donor and acceptor of three hydrogen bonds, Kme1 can only serve as the donorand acceptor of two hydrogen bonds (Figure 2). Addition of one methyl group onto free Lysalso increases the overall size and hydrophobicity of its side chain (Figure 2). Such adifference can play an important role in distinguishing Lys monomethylation from other states of Lys methylation by structurallymatched biological effectors as detailed later.(30,31)

2.3. Epsilon-N-dimethyl lysine (Kme2).

Similar to Lys and Kme1, Kme2 contains two lone-pair electrons on its ε-amine. The pK_avalue of the tertiary ε-amine of Kme2 is 10.2 (estimated on the basis ofN,N-dimethylethylamine), which is comparable to those of the ε-amines of Lys and Kme1(Figure 2). Given its pK_a of 10.2, the ε-nitrogen ofKme2 is largely protonated at physiological pH (>99%). In this protonation state, Kme2 can serve as the donor or acceptorof one hydrogen bond, and form a salt bridge. The positively charged Kme2 after protonation can also engage in cation-πinteractions (Figure 2). In comparison with the net +1 formal charge of Lys and Kme1, thepositive charge of Kme2 is more dispersed among three neighboring hydrocarbon moieties (two methyl groups and one methylenemoiety) and one hydrogen atom (Figure 2). Addition of two methyl groups significantlyincreases the overall size and hydrophobicity of the side chain of Lys (Figure 2). Thecombined effects are associated with the higher spatial requirement for Kme2 to engage in cation-π and hydrogen bondinteractions for its molecular recognition.(30,31)

2.4. Epsilon-N-trimethyl lysine (Kme3).

Kme3 is the extreme state of lysine methylation (Figure 2). The side chain of Kme3 isan obligatory cation with a permanent +1 formal charge at physiological pH (Figure 2). LikeLys, Kme1 and Kme2, the positively charged Kme3 can form a salt bridge with negatively charged amino acid residues such as Asp andGlu (Figure 2). Unlike free Lys, Kme1 and Kme2, Kme3 cannot form a hydrogen bond either as adonor or acceptor because of the quaternary ammonium moiety (Figure 2). Here the permanent +1formal charge of Kme3 is broadly dispersed on the four neighboring hydrocarbon moieties (three methyl groups and one methylenemoiety) (Figure 2). Addition of three methyl groups substantially enhances the overall sizeand hydrophobicity of the side chain of Lys (Figure 2). Compared with free Lys, Kme1 andKme2, Kme3 is characterized with the largest van der Waals radius and thus the highest spatial demand upon engaging incation-π interactions. Collectively, it is more accurate to depict the quaternary ε-ammonium of Kme3 as ahydrophobic cloud with an electronegative nitrogen atom enshrouded centrally and the overall +1 charge evenly dispersed around thefour surrounding hydrocarbons (Figure 2). This description will become more apparent uponelucidating the underlying mechanism for recognition of Kme3 in a biological setting.

3.1. Overview of protein methylation.

Polar lysine residues are often positioned at solvent-exposed protein surface regions. The lone-pair electrons of theε-amine of Lys, together with its preferred localization on protein surface, make lysine susceptible to diverseposttranslational modifications (Figure 3). Among characterized lysine modifications ineukaryotic cells are methylation, acetylation, propionylation, malonylation, butyrylation, succinylation, glutarylation,myristoylation, biotinylation, ubiquitination, SUMOylation, and neddylation (Figure 3).(27,47–52) There are at least two factors that make protein lysine methylation distinct from other lysine posttranslationalmodifications (Figure 3): (a) each addition of a methyl group onto Lys does not alter the +1formal charge of the ε-amine moiety at physiological pH, whereas other Lys modifications are acylation and convert theprotonated, positively charged ε-amine into a neutral amide; (b) lysine methylation is the smallest posttranslationalmodification and thus minimally alters overall size of the side chain in comparison with other lysine modifications.(21)

Lys, Kme1 and Kme2 contain lone-pair electrons on their ε-amine moieties and are expected to be readily alkylated(e.g. methylated) under basic conditions. However, high pK_a of Lys, Kme1 and Kme2(10.2 ~ 10.7) makes their ε-amines mainly protonated under physiological conditions (pH = 7.4) and thus inert asnucleophiles. To circumvent the activation barrier of deprotonation for the subsequent nucleophilic substitution reaction, proteinlysine methylation is catalyzed by various PKMT enzymes in biological settings (Figure4).(53–57) The human genomeencodes more than 60 characterized PKMTs, which can be classified as SET domain-containing PKMTs (Class V methyltransferases, SETforSuppressor of variegation 3-9,Enhancer ofzeste andTrithorax, three genetic phenotypes of Drosophila) and non-SET-domain PKMTs(Class I methyltransferases) (Figure 4,Figure 5 andTable 1).(27–29,40) More than 90% of PKMTs belong to the family of SETdomain-containing PKMTs (Table 1 andFigure 5).DOT1L,(54) METTL10,(58) METTL20,(59,60) METTL21A,(61)METTL21B,(62,63) METTL21C,(64) METTL21D (VCP-KMT),(65,66) METTL22,(67) eEF1A-KMT1 (N6AMT2 as named previously, the homolog of yeastEfm5),(68) eEF2-KMT,(69) and CaM-KMT(70) are among well-characterized members of non-SET domain human PKMTs (Figure 5 andTable 1). Although > 60 human proteins havebeen characterized as PKMTs, they are often multiple-domain macromolecules consisting of hundreds to thousands of amino acidsembedding small methyltransferase domains (130-aa SET or non-SET domain).(55,71,72) Their methyltransferase activities may onlyaccount for part of their biological functions. Caution should thus be taken to dissect the functions of the methyltransferasedomains of PKMTs from their other roles as full-length proteins (e.g. as effectors for recognition of otherposttranslational modifications or as structural scaffolds for formation of protein complexes).

3.2. SET domain-containing PKMTs.

The SET domain of PKMTs (Class V methyltransferases) consists of approximately 130 amino acids, often flanked by pre-SETand post-SET domains (Figure 4).(71) Its structuraltopology is distinct from other types of methyltransferases by a characteristic ‘pseudoknot’ fold of the SET domain(Figure 4).(71) On the basis of phylogeneticsequences of SET domains, human PKMTs can be further divided into subfamilies with respective structural topology (Table 1 andFigure 5).(73)For instance, G9a and GLP1 belong to a subfamily of classical PKMTs in which their SET domains alone are sufficient for enzymecatalysis (Table 1 andFigure 5).(74) In contrast, ASH1L, SETD2 and NSD1/2/3 are within a subfamily of PKMTs containing anauto-inhibitory SET domain, whose apo-isomer is expected to be catalytically inactive and must go through dramatic conformationalchanges for substrate binding and enzyme catalysis (Table 1 andFigure 5).(55,75–78) The SET domains of the MLL subfamily alone are inert butcatalytically active in the presence of their binding partners such as WDR5, RbBP5, Ash2L and Dpy30 (referred as WRAD) (Table 1,Figure 5 andFigure 6).(79–81) A similarrequirement for the formation of a protein complex for enzyme catalysis has also been documented for EZH1 and EZH2 (EZH1/2, EEDand Suz12 referred as PRC2 complexes) (Table 1,Figure5 andFigure 6).(82,83) A 5-member SMYD subfamily and a 17-member PRDM subfamily of human PKMTs are characterized byinsertion of the MYND domain (myeloid translocation protein 8, Nervy, and DEAF-1)(84–88) and the PR (PRDI-BF1 and RIZ) domain within their SETdomains,(89) respectively (Table 1 andFigure 5).

3.3. Non-SET domain PKMTs.

In comparison with SET domain-containing PKMTs harboring well-defined methyltransferase activities, DOT1L,(54) METTL10,(58) METTL20 (ETFβ-KMT),(59,60) METTL21A,(61)METTL21B (eEF1A-KMT3),(62,63) METTL21C,(64) METTL21D (VCP-KMT),(65,66) METTL22,(67) eEF1A-KMT1 (N6AMT2 as named previously, thehomolog of yeast Efm5),(68) eEF2-KMT,(69) andCaM-KMT(70) are among human non-SET domain PKMTs demonstrated with methyltransferaseactivities (Figure 5). These PKMTs are structurally homologous to protein argininemethyltransferases (PRMTs) (Figure 5).(28)Non-SET-domain PKMTs and PRMTs belong to the canonical Rossmann-fold-like family of methyltransferases (Class I) with theirstructural topology featuring a seven-stranded β-sheet connected by α-helices. Besides protein Lys and Arg,small-molecule, DNA and RNA are among the substrates of the class-I methyltransferases (Figures4,5). The human genome encodes 200 putative methyltransferases, most of which haveyet been characterized.(28) It will be intriguing to explore whether some of these putativemethyltransferases are PKMTs. Recently, after examining a catalytically-dead SET-domain mutant of MLL1, the Cosgrove groupreported that the MLL1-WRAD (WDR5, RbBP5, Ash2L and Dpy30) complex contains a noncanonical site located at its Ash2L subunit tocatalyze H3K4 methylation.(95–97) BecauseAsh2L has no homology with SET domain-containing Class-V methyltransferases and non-SET-domain Class-I methyltransferases, it isof great interest to explore the structural features of the cryptic PKMT site, which can be informative to identify novelPKMTs.

4.1. Overview of PKMT-catalyzed methylation.

PKMTs are classified as transferase enzymes (EC 2 in enzyme nomenclature), which catalyze the transfer of specificmoieties (e.g., a methyl group, EC 2.1.1 subclass) from one molecule (co-substrate or cofactor such asS-adenosyl-l-methionine or SAM) to a substrate (e.g., a lysine or arginine residuefor protein methylation). PKMT-catalyzed methylation reactions are expected to go through an approximately linear S_N2transition state with the ε-nitrogen of Lys, Kme1 and Kme2 as a nucleophile and SAM’s sulfonium as a leaving group(Figure 7).(98) In order to assembleS_N2 transition states for effective catalysis, PKMTs, in particular those with promiscuous substrate profiles (seediscussion below), need to bind various substrates with their target Lys(Kme1/2) residues positioned within their catalyticpockets; align the sulfonium methyl group of the SAM cofactor on a nearly linear path required for a S_N2 reaction;deprotonate the ε-nitrogen of Lys(Kme1/2) to free its lone pair of electrons (Figure7).(98,99) Upon poising the resultingactivated lysine substrate and the SAM cofactor along a nearly linear reaction trajectory, PKMTs may engage in other dynamicmotions and interactions to further stabilize the S_N2 transition state and facilitate the subsequent chemical step ofmethylation (Figure 7).(98,99) Upon the completion of the transfer reaction, PKMTs release the methylated product and a byproductS-adenosylhomocysteine (SAH) for another catalytic cycle.(19,100) As detailed below, the structures of many PKMTs have been elegantly tuned to promotethe stepwise catalysis.

4.2. SAM-binding modes of PKMTs.

3-D structures of PKMTs reveal that these enzymes bind the SAM cofactor in two distinct modes (Figure 4).(101,102) TheSAM cofactor in SET domain-containing PKMTs is oriented in a relatively packed configuration with theCβ-Cγ-Sδ-C5′ dihedral angle around 120° (e.g., 114° for SET7/9 and131° for SETD8).(103–109) Incontrast, the SAM cofactor in non-SET-domain PKMTs, as exemplified with DOT1L,(54) adoptsan extended conformation with the Cβ-Cγ-Sδ-C5′ dihedral angle around 180° (Figure 8). Upon further analyzing the modes of interaction (MOI) of the SAM cofactor in SETdomain-containing PKMTs and the non-SET-domain DOT1L, the Schapira group generalized that PKMTs interact with the α-aminoacid moiety of SAM through salt bridges, hydrogen bonds or their combination; the adenine moiety of SAM through combined hydrogenbond and hydrophobic interactions (Figure 8).(102)However, MOIs of SET domain-containing PKMTs and the non-SET-domain DOT1L are different at least in two aspects.(102) First, the 2′, 3′-hydroxyl moiety of SAM, as well as its mimics SAH and sinefungin,in SET domain-containing PKMTs are solvent exposed and does not involve any essential interaction with neighboring residues, whilethe comparable 2′,3′-hydroxyl moiety in DOT1L is deeply embedded in a cofactor-binding pocket and forms two stronghydrogen bonds with the carboxylic side chain of the nearby E186 residue (Figure 8).(54,110) Second, SAM’s sulfonium methyl groupin SET domain-containing PKMTs resides in a highly compacted, electronegative environment, whereas the comparable sulfonium methylmoiety in DOT1L is localized in a relatively vacant space (Figure 8). (54,71,110) TheseSAM-interacting networks are expected to be essential for PKMTs to position the SAM cofactor in a ready orientation for enzymecatalysis.

A combination of experimental and computational methods is often required to elucidate molecular details of how PKMTsinteract with the SAM cofactor.(98,111,112) In the case of SET7/9, the characteristic¹H-NMR chemical shifts, quantummechanical (QM) calculation, crystallographic data and kinetics of SET7/9 mutants implicated that the sulfonium methyl moiety ofthe cofactor SAM engages in noncanonical carbon-oxygen (CH•••O) hydrogen-bonding interactions with thehydroxyl group of Y335 as well as the two amide oxygen atoms of H293 and G264 (Figure9).(113–116) These interactionswere shown to be stronger than typical CH•••O hydrogen-bonding interactions, likely because the positivelycharged sulfonium acidifies the protons on the adjacent carbon atoms (Figure 9). Forinstance, the¹H chemical shift of the methyl group of SET7/9-bound SAM is 3.8 ppm, which is 0.8 ppm downfield relativeto that of SAM in solution and thus reflects a more electron-withdrawing environment surrounding SAM’s methyl group inSET7/9 (Figure 9).(113) In comparison with Y335 innative SET7/9, the Y335F substitution, which does not significantly alter the overall structure and stability of SET7/9,diminishes SAM’s affinity by 3 orders of magnitude.(113) Consistent with a lesselectron-withdrawing environment, the Y335F substitution shifts the¹H-NMR signal of SAM’s methyl group toupfield.(113) These observations can be partially rationalized by the lostCH•••O interaction of the methyl group with SET7/9’s Y335 residue.

Binding modes of PKMTs for the SAM cofactor in PKMTs can also be probed experimentally with binding isotope effects(BIEs).(98,111,112) In this approach, the changes of bond orders and vibrational modes of the SAM cofactor uponbinding PKMTs can be reported by relevant BIEs when the corresponding atoms of SAM are replaced with heavy isotopes (Figure 10).(98,112) Upon measuring the BIE of [CT₃]-SAM in complex with NSD2, the Schramm laboratory observed an inverseBIE of 0.65.(112) This large inverse BIE strongly argues that the SAM cofactor, from asolution state to the NSD2-bound state, experiences extreme steric impingement and thus restricts the stretching vibrational modesof the BIE-associated C-H bonds.(112) Interestingly, upon the formation of the pseudoternary Michaelis-Menten complex of NSD2 with the SAM cofactor and H3K36M nucleosome (H3K36MNuc, an inactive substrate mimic), thesteric impingement is relaxed and the BIE of [CT₃]-SAM was reported to be 0.990.(111,112) While this [CT₃]-BIE is significantly less inverse thanthat of the binary SAM-NSD2 complex (0.990 of the ternary SAM-NSD2-H3K36MNuc complex versus 0.65 of the binary SAM-NSD2 complex),this small inverse BIE of the ternary complex is comparable with [CD₃] and [CT₃] BIEs of SAM in complex withSETD8 (0.990 versus 0.959 and 0.979).(98,111,112) As further supported by computational modeling, these slightly inverse BIEs areconsistent with the noncanonical carbon-oxygen (CH•••O) hydrogen-binding interactions of the methyl group ofSAM with the PKMTs through a conversed phenolic hydroxyl group and amide oxygen atoms as implicated in several PKMTs (Y1179, R1135and F1117 of NSD2; Y336, R295 and C270 of SETD8; Y335, H293 and G264 of SET7/9) (Figures 9,10).(98,111,112) In comparison, the binary SAM-NSD2 complex and the ternarySAM-NSD2-H3K36MNuc complex show comparable [5′-³H₂]-SAM BIEs of 0.973 and 0.982, respectively.(111,112) These different BIEs argue that SAMbinding is different for the two PKMTs (Figures 9,10).

Upon probing molecular rationales for PKMTs to recognize the SAM cofactor, the Trievel laboratory also examined potentialnoncovalent interactions between SAM’s sulfonium and its adjacent oxygen atoms (S•••O chalcogen bonds)around the active site of SET7/9.(117) The sulfur-oxygen (S•••O)chalcogen interaction originates from partial bond order through a σ-antibonding orbital of SAM’s sulfur atom and alone pair of electrons of the backbone amide oxygen of the nearby N265 residue (Figure 11).In comparison with N265 in native SET7/9, the N265A mutation does not alter the overall structure and stability of SET7/9 butdecreases its affinity to SAM by 8-fold.(117) In contrast, the N265A variant diminishesthe affinity to SAH, a SAM mimic with SAM’s sulfonium replaced by the cognate neutral thioether, only by 2-fold. Consistentwith this observation, quantum mechanical (QM) calculation suggests that sulfonium cations form a strongerS•••O chalcogen bond than cognate neutral thioethers. The noncanonical sulfur-oxygen(S•••O) chalcogen interaction is thus essential for SET7/9 to engage in the maximal binding to the SAMcofactor.(117) It remains to be investigated experimentally whether unconventionalcarbon-oxygen (CH•••O) and sulfur-oxygen (S•••O) chalcogen interactions play roles inrecognizing the SAM cofactor by other SET domain-containing PKMTs in a broad manner.

4.3. Histone and nonhistone targets of PKMTs.

SET domain-containing PKMTs were first characterized to methylate histones and originally named as histonemethyltransferases.(27) However, recent advancement in identifying PKMT substratesindicated that SET domain-containing PKMTs can act on diverse nonhistone substrates (e.g., full-length G9a andGLP1 can methylate hundreds to thousands of nonhistone substrates, see discussion later).(21) Using the nomenclature of PKMTs rather than histone methyltransferases is more consistent with diverse substrateprofiles of these enzymes. In contrast, only a small set of proteins were identified as the substrates of non-SET domain PKMTs asexemplified by DOT1L for K79 of histone H3,(54,118) METTL10 for K318 of eEF1A,(58) METTL20 for K199 and K202 of theβ-subunit of the electron transfer flavoprotein (ETFβ),(59,60) METTL21A for HSP70s (K561 in HSPA1, K585 in HSPA5, and K561 in HSPA8),(61) METTL21B for K165 of eukaryotic elongation factor 1 alpha (eEF1A),(62,63) METTL21C for K390 of eEF1A as well as K35 automethylation,(64,119) METTL21D (VCP-KMT) for K315 ofvalosin-containing protein (VCP),(65) METTL22 for K135 of Kin17,(67) eEF1A-KMT1 for K79 of eEF1A,(68) eEF2-KMT for K525 ofeukaryotic elongation factor 2 (eEF2),(69) and CaM-KMT for K115 of calmodulin.(70) Given that many known Lys methylation sites have not been assigned to specific PKMTactivities, it remains to be determined whether non-SET domain PKMTs, like SET domain-containing PKMTs, can act on diversesubstrates.

4.4. Sequence motifs and binding modules of PKMT substrates.

While some PKMTs recognize substrates promiscuously and others are more specific, there are no general rules to defineunambiguously PKMT substrates through their primary sequences. For instance, SET7/9 was shown to methylate histone H3, p53 andTAF10 by recognizing a common peptide motif [R/K][S/T]K (K as the modificationsite).(106,120) After further structural andbiochemical characterization of the enzyme-substrate complex, the consensus sequence of potential SET7/9 substrates was extendedto [R/K][S/T/A]K[DNQK].(121) Upon searching SET7/9 targets with thesequence lead, several new substrates including TAF7 and E2F1 were then identified.(106,122) Nevertheless, the consensus sequence of SET7/9 substrates was then expanded moresystematically with arrayed peptides as substrate candidates and led to identification of 91 potential substrates.(123) However, no consensus sequence was defined for many newly-identified SET7/9 substratesincluding STAT3,(124) androgen receptor,(125,126) Yap,(127) SIRT1,(128) FOXO3,(129,130)DNMT1,(121) Rb(131) and NF-kappa B.(132) A similar situation was also encountered upon identifying substrates of G9a and GLP1,which were first characterized to methylate histone H3 and DNMT3A through a consensus sequence TARK(K as the modification site).(133–135) The structural characterization of GLP1-substrate complexes revealed that Asp1135 and Asp1145 ofGLP1 directly interact with R at the −1 position and T at the −3 position through salt bridges and hydrogen bonds,respectively.(135) However, this consensus sequence cannot be found innewly-identified G9a/GLP1 substrates such as p53,(136) C/EBP-β,(137) Reptin,(138) and MyoD.(139) More importantly, recently advanced technologies allowed revealing hundreds of substrates of G9a and GLP1,(99,140–144) which do not share well-defined consensus sequences. Similarly, no consensus sequence can be defined uponexamining SETD8 substrates histone H4 (RHRK),(107) PCNA(GHIK)(145) and Numb (LERK);(146) EZH2 substrates histone H3 (AARKS),(147) RORα (SARKS)(148) and STAT3(KTLKS).(124,149,150) Notably, citrullination of H3R26 in H3 (AARKS) suppressesthe EZH2-mediated methylation of H3K27 (AARKS) by 30,000 fold. This result suggests that the presence of acitrulline residue rather than the absence of an arginine residue at the “-1” position diminishes the reactivity ofH3K27 as EZH2’s substrates.(151)

Accumulated evidence strongly argues that many SET domain-containing PKMTs such as SET7/9,(152) G9a,(153) GLP1,(99,140–143) SETD8(146,154) and EZH2(124,149,150) can bind multiple substrates in apromiscuous manner. A possible working model is that their substrate-binding pockets are structurally flexible and thus can adoptmultiple configurations to accommodate different substrates. An alternative but nonexclusive explanation for substrate promiscuityis that a binding pocket recognizes many substrates mainly through their backbones rather than side chains. In addition,structurally flexible substrates may also adopt multiple conformations for optimal binding by PKMTs. Computational modeling showedthat SMYD3 can engage in substrate interactions through multiple conformations of substrate peptides.(155,156) On the basis of this model, a few proteins wereidentified as novel substrates of SMYD3.(155,156)While these hypotheses could be readily tested by examining 3-D structures of the corresponding PKMT-substrate complexes, theso-far solved structures of PKMT-substrate complexes do not cover a broad scope of PKMT substrates. As indirect evidence, CARM1,an arginine protein methyltransferase with a structural topology similar to that of DOT1L, recognizes its substrates mainlythrough amide backbones and with minimal interactions with the side chains of these substrates.(157) In contrast to SET domain-containing PKMTs, no 3-D structure of an enzyme-substrate complex has been solved fornon-SET domain PKMTs. Many of non-SET domain PKMTs (e.g., DOT1L)(54,118,158,159) recognize full-length proteins rather than truncated peptides as active substrates.

4.5. Lysine binding pockets of PKMTs.

Although it is difficult to predict consensus sequences and binding modes for PKMTs to recognize their substrates, SETdomain-containing PKMTs do follow some general rules to position the side chains of their substrates (Lys, Kme1 and Kme2) intocatalytic sites. All SET domain-containing PKMTs bind the side chains of Lys, Kme1 and Kme2 through a hydrophobic channel mainlyconsisting of aromatic amino acids (Tyr, Phe and occasionally Trp).(71) These structuralmotifs are expected to achieve maximal engagement with lysine, Kme1 and Kme2 through both desolvation of their C4 hydrocarbon(hydrophobic effect) and cation-π interactions through their positively charged ε-amine moieties. The hydroxylresidue of a highly-conserved Tyr (e.g., Y1154 in G9a, seeFigure 12),which is located at the interface of the SAM cofactor and the ε-amine substrate, also forms a strong hydrogen bond with theε-amine nitrogen. Such an interaction locks this ε-amine moiety and positions its lone-pair electrons in a readyconfiguration for subsequent methylation. Lysine-binding pockets of SET domain-containing PKMTs are thus well tuned to engagesubstrates for catalysis. It has also been noted that certain PKMTs such as SUV420H1/2(160) and EZH2 Y641F/N/S/H/C mutants(161–165) prefer pre-methylated lysine residues as substrates. Given that no similar aromatic-rich pocket can be mapped forClass-I PKMTs such as DOT1L, METTL10, METTL20, METTL21A, METTL21B, METTL21C, METTL21D (VCP-KMT), METTL22, eEF1A-KMT1, eEF2-KMT,and CaM-KMT, these PKMTs likely adopt different strategies to engage their substrates for catalysis.

4.6. Lysine deprotonation in PKMT-catalyzed methylation reaction.

Besides efficient engagement of substrates, catalytic sites of SET domain-containing PKMTs are also elegantly structuredto remove one proton from Lys, Kme1 and Kme2 substrates and thus activate them for subsequent methylation. With SET7/9 as anexample, a rudimentary mechanism of substrate deprotonation was speculated to involve SET7/9’s Y335 as a generalbase.(166) However, this mechanism was soon ruled out because of the comparablepK_a between the phenolic hydroxyl group of Tyr (pKa = 10.5) and the ε-amines of Lys, Kme1and Kme2 (pK_a = 10.2 ~ 10.7) (Figure 2). Cation-πinteractions between the ε-amine of Lys (free or Kme1/2) and nearby aromatic residues in lysine-binding pockets furtherincreases the pKa of Lys, Kme1 and Kme2 to favor their protonated states. The phenolic hydroxyl group withpKa around 10 is thus not sufficient to remove the proton from the ε-amine group at physiological pHof 7.4. Despite the lack of a general base at the catalytic site of SET7/9 (other SET-domain PKMTs as well), computationalmodeling revealed that deprotonation of the Lys ε-amine can occur through transient formation of dynamic water channelsaround SET7/9’s catalytic site.(167) Molecular dynamics simulations and quantummechanical/molecular mechanics (QM/MM) calculations further suggest that the SAM binding by SET7/9 at its catalytic site induceselectronic repulsion between the positive sulfonium moiety of the SAM cofactor and the protonated ε-amine of Lys, and thusfurther decreases the pK_a of the latter to 8.2.(167) Theformation of the dynamic water channels and the proximity between SAM’s sulfonium and the substrate lysine were proposed tobe essential for efficient catalytic turnover of SET7/9. Computational modeling also showed that similar deprotonation activationalso occurs for SETD8 to methylate its H4K20 substrate,(168) likely for other SETdomain-containing PKMTs in general. It remains to be investigated whether the formation of dynamic water channels is also requiredfor non-SET-domain PKMTs to catalyze their methylation reactions, given similar lack of a general base for lysine deprotonation atcatalytic sites.

4.7. Transition state stabilization of SET domain-containing PKMTs.

Stepwise progression of protein lysine methylation is accompanied by changes of bond order and vibrational modes of thesulfonium methyl moiety of the SAM cofactor and the ε-amine moiety of the substrate Lys.(98) Multiple computational and experimental methods have been implemented to elucidate molecular features of thetransition states of PKMT-catalyzed methylation reactions.(98,111,167–175)The structures of the transition states of PKMT-catalyzed methylation can be probed experimentally with kinetic isotope effects(KIEs) in combination with computational modeling.(98,111) Computational modeling can afford an unbiased set of candidate structures of transition states. KIEs(e.g., the methyl group of SAM and the ε-amine of the substrate Lys) of individual PKMTs are then usedas electrostatic and geometric constraints to define the best matched transition state structure(s) with atomic resolution. WhileKIEs are conventionally determined by ratios of steady-state kinetic parameters (k_cat andK_m) with a pair of isotopic substrates, more precise approaches with a mixed isotopic pair in acompetition format are required for PKMT-catalyzed methylation, for which KIEs are often within a range of a few percent fromunity.(98,111,176) Among successful approaches to determine precisely KIEs of PKMT-catalyzed methylation are remoteradioactive labeling and isotope ratio mass spectrometry (MS).(98,111)

For NSD2-catalyzed H3K36 methylation, the Schramm laboratory relied on the remote radioactive labeling method to determineintrinsic [¹⁴CH₃]-SAM, [³⁶S]-SAM, [CT₃]-SAM and [CD₃]-SAM KIEs (Figure 13).(111) A large primary[¹⁴CH₃] KIE of 1.113 and a nearly extreme [³⁶S] KIE of 1.018 are characteristic for aS_N2 transition state with the methyl transfer reaction as a rate-limiting step.(111) [CT₃]-SAM and [CD₃]-SAM of NSD2-catalyzed H3K36 methylation show inverse KIEs of 0.77 and0.83, respectively (Figure 13). To measure KIEs of SETD8-catalyzed H4K20 methylation, wedeveloped a set of mathematic algorithm, in combination with matrix-assisted laser desorption ionization time-of-flight(MALDI-TOF) mass spectrometry (MS), to determine isotopic ratios of monomethylated H4K20.(98) SETD8 showed a characteristic primary intrinsic [¹³CH₃]-SAM KIE of 1.04 for a S_N2transition state and an inverse intrinsic α-secondary [CD₃]-SAM KIE of 0.90 (Figure 13).(98) Here the different KIEs between SETD8 and NSD2 argue twodistinct transition states with certain commonly shared S_N2 characters.(98)Using these KIEs as experimental constraints, in combination with computational modeling, NSD2 was revealed to adopt a late,asymmetric S_N2 transition state with its N-C distance of 2.10 Å and S-C distance of 2.53 Å (Figure 13).(111) In comparison, the transition stateof SETD8 was characterized with an early, asymmetric S_N2 character with its N-C distance of 2.38 Å and S-Cdistance of 2.05 Å (Figure 13).(98)

Besides the transition state structures of NSD2 and SETD8 solved with KIEs as experimental constraints, unconstrainedcomputational modeling also revealed the transition state structures of several SET domain-containing PKMTs including SETD8 andSET7/9 (Figure 14).(167–175) The computationally modeled transition state structure of SETD8 is consistent with an early,asymmetric S_N2 transition state with the KIEs as experimental constraints.(168)Interestingly, the modeled transition state structures of SET7/9 are substrate-dependent with symmetric S_N2 charactersfor histone 3 lysine 4 peptide substrate (comparable N-C and S-C distances of 2.2~2.4 Å) and with late, asymmetricS_N2 characters for p53 lysine 372 peptide substrate (Figure 14).(167,170) Remarkably, all of the known transitionstate structures of PKMTs feature a relatively fixed distance of 4.4~4.6 Å between the nitrogen nucleophile and thesulfonium leaving group but differ in the position of the transferred methyl group (Figure14).

Computational modeling further revealed that the reaction path of SET7/9-catalyzed lysine methylation toward itstransition state features less degree of electrostatic change in comparison with that of a solution-phase S_N2methylation reaction.(171,173) SET7/9’smethylation reaction is accelerated by pre-organization of an electrostatic environment at its catalytic site.(171,173) Several pieces of nonexclusive evidence support thepreorganized electrostatic environment in SET7/9 and likely many other SET domain-containing PKMTs for enzyme catalysis. SET7/9binds SAM in the ground state partially through leveraging three essential non-canonical CH•••Ohydrogen-bonding interactions of SAM’s methyl group with the two amide oxygens of H293 and G264 and the phenolic oxygen ofTyr336 (Figure 9).(113–116) On the path toward the transition state, these CH•••O hydrogen-bondinginteractions are expected to be maintained and thus construct an electrostatic pore to confine the motion of the sulfonium methylat a catalytically ready conformation. In the course of examining methylation reactions catalyzed by catechol-O-methyltransferase,glycine-N-methyltransferase and their mutants, the Klinman group obtained inverse 2° kinetic isotope effects (KIEs) of[CT₃]-SAM and found that the magnitude of these inverse KIEs positively correlates with their catalyticefficiency.(177,178) The origin of theseinverse KIEs was attributed to the compaction of active site residues toward SAM’s methyl group at the transitionstates.(179) The compaction model proposed for catechol-O-methyltransferase andglycine-N-methyltransferase can be also applied to rationalizing how SET7/9 and other PKMTs pre-organize their electrostaticenvironment to align SAM’s methyl group along the nearly linear trajector of a S_N2 transition state for enzymecatalysis. This rationale is consistent with the inverse α-secondary [CD₃]/[CT₃]-SAM KIEs of 0.77 and0.83 for NSD2 and 0.90 for SETD8 (Figure 13), whose magnitude is too large to berationalized solely by noncanonical carbon-oxygen (CH•••O) hydrogen-binding interactions of SAM’smethyl group upon binding PKMTs.(98,111) Whereasthe large inverse KIEs can be readily attributed to the axial compression of the N-S distance along the linear trajectory of thetransition state, computational modeling suggests that other modes such as equatorial compression can also lead to comparableKIEs.(179) It is likely that the equatorial compression through non-canonicalCH•••O hydrogen-bonding interactions and the N-S axial compression act together for assembling transitionstates of diverse PKMTs.(98,179)

4.8. Product specificity of PKMTs.

A lysine side chain of PKMT substrates can be subjected to methylation for up to three times to generate mono-, di- andtri-methylated products (Kme1/2/3), respectively. This product specificity is controlled by multiple factors including thetopology of catalytic sites of PKMTs, amino acid sequences and prior methylation states of substrates, as well as their biologicalcontexts such as the presence of coactivators. The molecular mechanism underlying the product specificity can be summarized by thefollowing general principle: a methylation reaction can only occur if PKMTs can assemble a lysine substrate (Lys, Kme1 and Kme2)and the SAM cofactor into an active transition-state configuration. In the context of the reaction path discussed above (Figure 7), the whole process of catalysis demands PKMT to bind the Lys(Kme1/2) substrate andSAM in a manner so that a water channel for deprotonation can be formed at the catalytic site and thus free the lone pair ofelectrons of the ε-amine as a nucleophile; the activated ε-amine needs to be aligned in proximity with SAM’ssulfonium methyl group and in a nearly linear trajectory for a S_N2 reaction. This general principle can then be appliedto rationalize the product specificity of many PKMTs.

It has been well characterized that most SET domain-containing PKMTs can leverage a Phe/Tyr switch to control productspecificity with a smaller Phe residue preferred for higher methylation states (di, tri) and a bulkier Try residue for lowermethylation states (mono, di) (Figure 15).(180,181) For instance, human G9a, which has F1152 as the Phe/Tyr switch site, can di- andtri-methylate histone H3K9, whereas its F1152Y mutant can only monomethylate this substrate (Figure15).(182) In contrast, SET7/9 and SETD8, which have a Tyr residue(SETD8’s Y335 and SET7/9’s Y305) as their Phe/Tyr switch, are in general characterized as the PKMTs formonomethylaitoin (Figure 15).(181,183) Structural characterization of SET7/9, SETD8 and their Tyr-to-Phe switch mutants indicates thatnative SET7/9 and SETD8 rely on the phenolic oxygen of the Try residue to immobilize a water molecule at their catalytic sites(Figure 15).(181,183) As further revealed by the structures of the Y335F mutant of SETD8 and the Y305F mutant of SET7/9,the absence of the phenolic oxygen will liberate the water molecule and thus spare a cavity to accommodate the methyl group ofmethyllysine substrates. This mode of interaction thus allows positing the lone-pair electrons of methyllysine’sε-amine adjacent to SAM’s methyl moiety for catalysis. (183)

It is worth noting the exception for the rule of the Phe/Tyr switch. For instance, although SET7/9 and SETD8 have beenclassified as the PKMTs harboring monomethylation activity because of their characteristic Tyr switch, several exceptions havebeen reported that the two PKMTs can also carry out dimethylation reactions for certain substratesin vitro andunder cellular settings. Among the characterized dimethylated products are K2076 of MINT(123) and K140 of STAT3 for SET7/9(124) and K158 of Numb for SETD8.(146) Although alternative interpretation of these observations could be that SET7/9 andSETD8 carried out monomethylation and other PKMTs added the second methyl group under a cellular setting, computational modelingindeed suggests that SET7/9 can carry out dimethylation by releasing the Tyr-bound water molecule via a channel formed by theG292, A295, Y305 and Y335 residues around the catalytic site of SET7/9.(184)

Set1/COMPASS complex is the yeast homolog of human MLL and SET1 complexes.(185)Although yeast Set1, human MLL and human SET1 all contain a Tyr residue (Y1052 for yeast Set1) at the Phe/Tyr switch site, theycan mono, di-, and trimethylate H3K4in vivo.(81,186) With the yeast Set1 as a model, Cps40/Ypl138 was shown to be essential for the activity of H3K4trimethylation and thus proposed to induce the conformational change of Y1052 upon interacting with Set1 for a higher degree ofmethylation.(187–189) This observationthus presents the feasibility to modulate product specificity of SET domain-containing PKMTs via their binding partners. Besidesmodulating the conformation of the Phe/Tyr switch, other regions of SET domain-containing PKMTs can also be tailored to alterproduct specificity. For instance, the Y641 residue of EZH2, which is located in the opposite side of the switch residue F724 ofEZH2, is often mutated to F/N/S/H/C in human B-cell lymphomas.(161–165) The Y641F/N/S/H/C mutants elevate the catalytic activity of EZH2 toward trimethylation productslikely through expanding its catalytic site to accommodate bulkier mono/di-methylated histone H3K27 substrates. For non-SET-domainPKMTs, the full-length DOT1L mainly monomethylates histone H3K79 in the absence of its binding partner AF10, whereas the presenceof AF10 alters the specificity of DOT1L toward the dimethylated product.(190)Collectively, product specificity of PKMTs can be determined not only by their primary sequences such as the canonical Phe/Tyrswitch but also by the features of substrates and biological contexts of individual PKMTs.

4.9. Potential product inhibition of PKMTs.

Methylated proteins andS-adenosylhomocysteine (SAH) are products and the byproduct of PKMT-catalyzedmethylation, respectively. While only limited kinetic studies were carried out to explore potential product inhibition forPKMT-catalyzed methylation, the inhibitory effect of methylated products was shown to be minimal. An unmethylated substrate andthe methylated product are structurally similar except at the site of lysine modification. Methylated products at catalytic sitesof PKMTs can often be competed readily by substrates for multiple turnovers.

In contrast, the reaction byproduct SAH shows a broad range of inhibition against PKMTs (e.g.,K_d = 23 μM for SET7/9 versusK_d = 0.36 μM for SYMD2).(28) Here no general rule can be found to predictK_d values of SAHagainst even closely-related PKMTs. For PKMTs, SAM generally shows higher affinity than SAH, likely because of the formation ofstronger S•••O chalcogen bonds via SAM’s sulfonium moiety.(117) In mammalian cells, SAH-mediated byproduct inhibition of PKMTs is less concerned because intracellular SAH can beefficiently degraded into adenine and homocysteine by SAH hydrolase (SAHH).(191,192) Indeed, a common practice to increase intracellular concentrations of SAH is to treatcells with SAHH inhibitors such as adenosine dialdehyde.(193) As an extreme case, thelevel of intracellular SAH can be elevated by dysregulating a SAM-SAH-associated metabolic pathway. For instance, certain cancers(e.g., lung, liver, kidney, bladder and colon cancers) can upregulate intracellular SAH by rapid consumptionof SAM by overexpressing nicotinamideN-methyltransferase.(194,195) The resulting elevation of intracellular SAH and thus inhibition of a panel of PKMTshave been linked to an altered epigenetic state and cancer malignancy.(195) Recent studiesalso showed that intracellular concentrations of SAM are modulated in a context-dependent manner.(1) For instance, depletion of methionine (a precursor for SAM’s biosynthesis) via diet restriction can rapidlyreduce the level of intracellular SAM.(196) Meanwhile, the amount of intracellular SAM canbe indirectly affected by signaling enzymes.(1,196)It is thus interesting to investigate how SAH-mediated inhibition affects the activities of PKMTs in the context of variedconcentrations of intracellular SAM.

5.1. Classification of reader domains that recognize protein methyllysine(s).

As protein lysine methylation progresses for mono-, di-, and to tri-methylated products, the size of lysine side chaingradually increases without altering the overall +1 formal charge on the ε-amine at physiological pH (Figure 2). However, these methylation events can slightly alter the ability of the target Lys to engagein cation-π interactions (increased dispersion of +1 charge around neighboring hydrocarbons and decreased desolvationenergy penalty upon binding reader motifs). The progressive lysine methylation also alters the ability to form hydrogen bonds as adonor and acceptor (Figure 2). The collective difference between Lys, Kme1, Kme2 and Kme3,though subtle, can be distinguished by the proteins containing methyllysine reader motifs (Table2). A wealth of Lys(Kme1/2/3)-specific reader modules have been discovered over the past decade and can be classifiedinto at least 15 classes (Table 2): ADD (e.g., a domain in DNMT3L),(197) ankyrin (e.g., domains in G9a and GLP1),(198) BAH (e.g., a domain in ORC1),(199) chromobarrel (e.g., a domain in MSL3),(200) chromodomain(e.g., a domain in HP1),(201) double chromodomain(e.g., a domain in CHD1),(202) HEAT (e.g., a domainin Condensin II),(203) MBT (e.g., domain in L3MBT1),(204,205) PHD (e.g., domains in TAF3, BPTF,ING2 and BHC80),(206–209) PWWP(e.g., domains in hMSH6 and NSD2: PDB 2GFU, 5VC8), SAWADEE (e.g., a domain in SHH1),(210) tandem tudor domain (e.g., a domain in 53BP1),(205) Tudor (e.g., a domain in PHF1),(211) WD40(e.g., a domain in EED)(212,213) and zf-CW (e.g., a domain in ZCWPW1).(214,215) It is worth noting that, while the list above contains all methyllysine reader modulescharacterized so far, it does not mean that a protein containing such module(s) can always interact with a methyllysine-containingprotein. For instance, ADD (e.g., DNMT3a),(216) Tudor(e.g., TDRD3, SMN and SPF30)(217–220) and WD40 (e.g., WDR5),(221) thoughgenerally classified as methyllysine readers on the basis of sequence homology, can also be reader proteins of methylarginine. Inaddition, cognate targets of many methyllysine-specific reader modules remain to be uncovered; no homology model can predictmethyllysine-flanking sequences recognized by potential readers. For instance, L3MBTL1 harbors three MBT repeats with eachcontaining a characteristic aromatic pocket (see discussion later) for potential binding of methyllysine.(204,222–224)However, the binding assays and crystallographic data of native L3MBTL1 and its binding pocket mutants revealed that only thesecond MBT repeat involves the interaction with its binding partner H4K20me1/2.(204,222,223) Given that most well-characterized bindingpartners of methyllysine reader modules are restricted to histones and that nonhistone substrates of PKMTs are rapidly emerging,it is of great interest to examine whether methyllysine reader modules such as the first and third MBT repeats of L3MBTL1 canrecognize nonhistone targets. It is likely that methyllysine reader modules can recognize cognate histone and nonhistone peptidescontaining similar but not identical methyllysine sequence motifs.

5.2. Mechanisms for reader modules to recognize methyllysine(s).

Selective recognition of Lys(Kme1/2/3) by their matched reader modules is essential for downstream functions of proteinlysine methylation. Although it remains challenging to predict binding partners of potential readers solely on the basis ofsequence homology, the accumulated structures of reader domains in complex with their methyllysine ligands have shed light onseveral general mechanistic rules underlying such recognition. One common trait for most methyllysine-specific readers is torecognize methyllysine(s) through a hydrophobic pocket containing aromatic residues (e.g., Phe, Tyr and Trp).Given that Lys, Kme1, Kme2 and Kme3 all carry the overall +1 formal charge at physiological pH, the aromatic pocket serves as apreferred docking site for cation-π interactions. For an optimal cation-π interaction, an ammonium cation (Lys,Kme1, Kme2 and Kme3) is expected to be poised perpendicular to an aromatic ring within its van der Waals contact.(44,45) In contrast ot the cation-π interaction, thehydrophobic effect of the aromatic pocket only plays a secondary role in recognizing methyllysine.(225) It has been estimated that a favorable cation-π interaction between an ammonium ion and anaromatic ring can gain 2.6 kcal/mol net binding free energy.(44) This value is even higherthan that gained via the formation of an ammonium-carboxylate salt bridge, because of higher desolvation penalty for the coulombinteraction, or via a modest hydrogen bond.(44)

While free energy gained through cation-π interactions is the main driver in binding methyllysine targets, thestructures of these reader modules have also been tuned to discriminate distinct states of lysine methylation by exploiting subtledifferences of the biophysical properties of Lys, Kme1, Kme2 and Kme3. As a general rule with few exceptions, the reader domainsthat prefer Kme3 over Lys, Kme1, Kme2 often have their binding pockets purely consisting of aromatic residues.(30,31,215) The sizeand geometry of these aromatic pockets are well suited to maximize cation-π interactions with partially chargedδ-methylene and three ε-N methyl groups. For the reader domains with the preference of Kme3, replacing any of thethree methyl groups with a hydrogen atom causes the loss of the corresponding cation-π and van der Waals interactions. Sucha loss is gradually magnified when the second and third methyl groups are removed. Among the known examples (Figure 16) are the interactions of the PHD finger domain of BPTF with H3K4me3 (PDB: 2F6J);(207) the chromodomain of Polycomb and the WD40 domain of EED with H3K27me3 (PDB: 1PFB;3IIW);(212,213) the Tudor domain of PHF andthe PWWP domain of ZMYND11 with H3K36me3 (PDB: 4HCZ; 4N4H);(211,226) the tandem chromodomains of human CHD1 (PDB: 2B2W).(202)One exception is ATRX’s ADD motif, whose methyllysine-recognizing pocket consists of only one aromatic residue andotherwise polar residues (PDB: 3QLA) (Figure 16).(227) The ATRX ADD domain slightly prefers H3K9me3 over H3K9me2/1 through forming nonconventional carbon-oxygen(CH•••O) hydrogen-bonding interactions between the pocket residues Y203, Q219 and A224 and the ε-Nmethyl groups (Figure 16).(227)

Unlike Kme3, Kme2 can form a hydrogen bond or salt bridge with an acidic residue such as Glu and Asp. This difference isoften exploited by reader modules that prefer Kme2 over Kme3. The Kme2-binding pockets of these reader domains are mainly composedof aromatic amino acids in combination with an acidic amino acid such as Glu or Asp (Figure17).(198,199,204,205) The aromatic residues engage in cation-πinteractions through the partially charged δ-methylene and two ε-N methyl groups of Kme2, while Glu or Asp forms asalt bridge or hydrogen bond with the tertiary amine of Kme2 (Figure 17). Among theseexamples are the interactions of the BAH domain of ORC1, the second MBT domain of L3MBTL1, the tandem Tudor domain of 53BP1 withH4K20me2 (E93, D355, D1521 in PDB 4DOW, 2PQW and 3LGF, respectively),(199,204,205,224) and the ankyrin repeat of GLP1 with H3K9me2 (E851 in PDB 3B95) (Figure17).(198) Here it is worth noting that the different modes of interactionbetween Kme2 and Kme3 often cause only a modest difference inK_d (within 10-fold), likely because theloss of a salt bridge or hydrogen bond for Kme3 can be compensated by hydrophobic and cation-π interactions of the extraε-amine methyl group in Kme3. Similarly, the slight selectivity of the reader modules for Kme2 over Kme1 largely arisesfrom the former’s ability to engage in cation-π and hydrophobic interactions with aromatic residues. Thecomplimentary contributions of cation-π and hydrophobic interactions and the formation of a salt bridge or hydrogen bondmay rationalize the promiscuous recognition of H3K4me3/2/1 by the double chromodomain of CHD1,(202) the PHD finger domain of Pygo-HD1-BCL9 complex (PDB 2B2W; 2VPE);(202)H3K4me2/3 by the tandem tudor domain of SGF29 (PDB 3MEA);(228) H3K9me3/2/1 by thePHD-tandem Tudor domain of UHRF1 (PDB 3ASK);(229) H3K9me3/2 by the chromodomain of HP1(PDB 1KNE);(201) and likely H3K36me3/2 by the PWWP domain of BRPF1 (PDB 2X4W) (Figure 18).(230) One exception is the recognition modeof H3K9me2 by the SAWADEE domain of SHH1 (PDB 4IUT), for which no obvious polar interaction can be identified within itsmethyllysine-binding pocket (Figure 18).(210)

To solely recognize lower methylation states such as free Lys and Kme1, reader modules engage in more salt bridges andhydrogen bonds rather than cation-π interactions. For instance, the ADD domain of DNMT3L recognizes unmodified H3K4 throughtwo salt bridges and one hydrogen bond between its Glu88/Glu90/Asn93 residues and the ε-amine of H3K4 (PDB 2PVC) (Figure 19).(197) Given the limited space of this polarpocket, adding one methyl group onto the free Lys was shown to significantly decrease the affinity to the ADD domain and addingtwo or three methyl groups is expected to completely abolish this interaction. As another example, the PHD finger domain of BHC80recognizes unmodified H3K4 by forming a salt bridge with D489, and hydrogen bonds with the backbone carbonyl group of E488 and onewater molecule (PDB 2PUY) (Figure 19).(209) Thesmall polar pocket likely can only accommodate one extra methyl group (Kme1). Interestingly, the chromo barrel domain of MSL3 wasuncovered as a methyllysine reader module to recognize H4K20me through an aromatic-rich pocket (PDB 3OA6) (Figure 19), whose spatial arrangement is similar to those of the reader domains of Kme3.(200) It remains to be determined how the chromo barrel of MSL3 distinguishes H4K20me fromH4K20me2/3.

5.3. Strategies to amplify affinity and selectivity of individual methyllysine reader modules.

While many methyllysine reader domains have been documented to recognize methyllysine flanked by specific neighboringresidues, the affinity and selectivity of a single reader domain for its target is often modest. The majority of methyllysinereader domains interact with methyllysine peptides with dissociation constants (K_d) in a range of highnanomolar to low micromolar affinity.(30,31,215,231) Additionally, many of these domains canonly modestly distinguish the targets that differ by one methyl group (e.g., Kme3 versus Kme2). For instance, thePHD finger of human BPTF (Figure 20), the largest subunit of the ATP-dependentchromatin-remodelling complex NURF (nucleosome remodelling factor), bind H3K4me2/3 peptides with comparableK_d values of 5.0 μM and 2.7 μM (a difference of 0.4 kcal/mol of binding freeenergy).(207) The extremely modest gain (~2-fold) of the affinity from H3K4me2to H3K4me3 likely arises from the additional cation-π interaction of Kme3 with aromatic residue(s) (Tyr10, Tyr17, Tyr23 andTrp32) in the methyllysine binding pocket of BPTF (Figure 16).

Given limited affinity and selectivity of a single methyllysine reader module for its target, one can envision that thereare additional mechanisms to enhance recognition and binding to these Lys marks in biological settings. One of these mechanisms isto increase local concentrations of methyllysine proteins above thresholdK_d values. If methyllysinetargets can be clustered to reach an extremely high local concentration (>10-fold aboveK_d), itis likely that such an effect will facilitate recruitment of methyllysine readers. For instance, HP1 (heterochromatin-associatedprotein, a small protein containing around 200 amino acids) harbors a chromodomain to recognize H3K9me2/3 with modestK_d values of 2.5~7 μM (Figure 18).(201) In a cellular context, the efficient recruitment of HP1 to specific chromatin loci ispartially benefited from high local concentrations of H3Kme3 (Figure 21).(232)

An alternative way to enhance target recognition of methyllysine reader proteins is to cluster multiple cognate readermodules for a specific methyllysine motif within a single protein or protein complex (Figure21). This microenvironment is also expected to lower apparentK_d values for target binding.Among many examples are the ankryin repeats of G9a and GLP1, which recognize H3K9me1/2.(198) G9a and GLP1 have been shown to form a heteromeric complex and depend on each other to execute their fullfunctions.(198) Both of the proteins harbor an ankryin repeat with theK_d values of 14 μM and 5 μM for H3K9me, and 6 μM and 7 μM for H3K9me2,respectively.(198) The local-concentration effect can be further enforced bymultiple-module reader proteins (or protein complexes) in combination with regionally clustered cognate methyllysine proteins (amultivalent binding mode) (Figure 21).(233) Itappears that ZMET2 and likely its homologs leverage such a mode to recognize H3K9me2 marks and then deposit DNA methylation.(234) ZMET2 contains BAH and chromo domains (PDB 4FT2), both of which interact with H3K9me2(Figure 22).(234) The proposed model ofH3K9me2-directed DNA methylation is that ZMET2 recognizes H3K9me2-containing nucleosomes through simultaneous binding of twoH3K9me2-containing tails with the BAH domain and chromo domain (Figure 22).(234) This multivalent interaction then positions the DNA methyltransferase domain of ZMET2 forsubsequent methylation of nearby nucleosome DNA.

The multivalent binding mode can also rationalize enhanced affinity and selectivity of a multiple-module reader protein(or protein complex) against a methyllysine in the context of other matched binding motifs (Figure21). In an ideal setting, the apparentK_d value of a multiple-module reader protein in aprefect complex with its binding partners is expected to be the product of the individualK_dvalues.(235) The selectivity and affinity of a multiple-module reader protein (proteincomplex) to a specific methyllysine can thus be fulfilled in the presence of other matched binding partners. One example is thePHD-bromodomain cassette of BPTF (Figure 20).(207)Here the PHD domain and the bromodomain are separated by an α-helical linker, and recognize H3K4me3 and acetylated lysine,respectively. The recognition of H3K4me3 is enhanced in the presence of H4K16 acetylation (H4K16ac) but not other H4 acetylationmarks such as H4K12ac and H3K20ac because of the preferred distance between H3K4me3 and H4K16ac.(207)

On the other hand, the affinity for a specific methyllysine can also be suppressed in the presence of an unmatched bindingmotif (Figure 21). For instance, the PHD finger of ING2 (PDB: 2G6Q) preferentiallyrecognizes H3K4me3 with the K4me3 and the H3R2 separated by Thr3 (Figure 23).(208) This was shown to form a preferred hydrogen bond between Thr3 and ING2’s K249.In contrast, the same PHD finger disfavors H3K9me3 and H3K27me3, which contain an Arg (R8/R26) instead of Thr adjacent to Kme3 andlikely abolish the T3-involved hydrogen bond (Figure 23). Collectively, even thoughindividual methyllysine reader modules may recognize their binding partners with only modest affinity and specificity, theseeffects can be significantly amplified via multivalent interactions associated with a target methyllysine and its adjacent bindingmotifs (Figure 21). Interestingly, many PKMTs or their complexes (e.g.NSD1-3 and MLL1-4) contain an array of reader modules for methyllysine(s), as well as other posttranslational modificationmarks.(55,72,236) Such an observation strongly argues that the enzymatic activities of PKMTs can be regulated reciprocally bymultivalent interactions of methyllysine reader domains.

6.1. Overview of human KDMs.

The chemically inert nature of methyllysine allows few options for enzymes to remove a methyl group from theε-amine of a protein lysine residue. The two so-far well-characterized mechanisms of enzymatic demethylation of proteinmethyllysine involve amino-oxidation and hydroxylation.(33–35) The former is catalyzed by lysine-specific demethylases (LSD).(237,238) The human genome encodes 2 LSD enzymes (LSD1 and LSD2), which arecharacterized by the presence of an amine oxidase-like (AOL) domain, known for several metabolic enzymes, and a SWIRM domain,which is characteristic for chromatin-associated proteins (Figure 24).(237,238) The hydroxylation reaction of protein methyllysine iscarried out by KDMs bearing characteristic JmjC domains (Figure 25).(239) These JmjC domains are expected to adopt a double-stranded β-helix topology, as observedfor a superfamily of 2-oxoglutarate (2-OG) dioxygenases, with the catalytic sites of KDMs harbored at one end (see KDM4A as anexample, PDF 2OQ6,Figure 26).(240) The human genomeencodes around 30 JmjC-domain-containing proteins, > 50% of which have been demonstrated as active KDMs.(239) Human KDMs are further diverged into 7 subfamilies on the basis of the sequence homology of theirJmjC domains (Figure 25).(239)

6.2. Catalytic motifs and reaction mechanisms of KDMs.

Although the net reaction of all KDMs is to remove a methyl group from the ε-amine of a protein lysine residue,they belong to two families of oxidoreductases (EC 1 in enzyme nomenclature) according to their different catalytic mechanisms.LDS1 and LDS2 are classified as EC 1.5.8 subfamily oxidoreductases with the CH-NH moiety of substrates as an electron donor and aflavin cofactor as an electron acceptor (Figure 27). The crystal structures of LDS1 and LSD2revealed that their catalytic AOL domains fold into two well-separated, intercalated subdomains with one binding to the FADcofactor in an extended conformation and the other binding to a methyllysine substrate on the opposite side (Figure 24).(237,238)The center at the interface of the two domains harbors a spacious cavity as the enzyme active site (Figure 24). This binding mode and the conformation of the FAD cofactor in LDS1 and LSD2 closely resemble those ofFAD-dependent oxidases.(241) After recruiting their substrates, the side chain ofmethyllysine is poised adjacent to the flavin moiety of the cofactor (Figure 27).(237,238) Such proximity together with matched redoxpotential promotes oxidation of the ε-amine methyl group, coupled with reduction of FAD into FADH2, to generate thecorresponding imine intermediate (Figure 27).(241)The catalytic cycle is then completed upon hydrolysis of the imine intermediate into formaldehyde as a byproduct.Dioxygen-mediated oxidation of FADH2 restores FAD with H₂O₂ as a byproduct (Figure 27). The involvement of the key imine intermediate prevents LDS1 and LDS2 from oxidizing Kme3 as a substratebecause of the incapability of Kme3 to be oxidized into an imine.(241)

JmjC-domain-containing KDMs, classified as EC 1.13 subfamily oxidoreductases, are dioxygenases with an ε-aminemethyl moiety and 2-OG as electron donors. Structural characterization of the JmjC catalytic core of KDM4A provided the firstmechanistic insight of KDMs (Figure 26, PDB 2OQ6).(240) Sequence alignment and structural comparison revealed that JmjC-domain-containing KDMs resemble cupinmetalloenzymes in relation to their double-stranded β-helix and jellyroll-like topology (Figure 26).(243) KDM-catalyzed lysine demethylation involves two cofactors,Fe(II) and 2-oxoglutarate, which are located at one end of the jellyroll fold (Figure 28).In general, the catalytic Fe is coordinated with two histidines and one acidic residue (HXX[D/E]···H) via ahighly-conserved 2-His-1-carboxylate facial triad (Figure 28).(239,244) Other binding sites of the Fe cofactor are eithervacant or occupied by waters in the absence of 2-OG. The 2-OG cofactor chelates the Fe cofactor through its 2-oxo carboxylatemoiety (Figure 28). Interestingly, not all JmjC KDMs bear the characteristicHXX[D/E]···H motif. One exception is PHF2 (Figure 29), which wascharacterized as an H3K9me2 demethylase.(245) This enzyme appears to use a Tyr (Y321)instead of the conserved His of the facial triad in other KDMs to coordinate with the Fe cofactor (Figure 29).(245)

While a stepwise mechanism of KDM-catalyze demethylation remains to be elaborated, dioxygen is anticipated to enter thecatalytic cycle of KDMs by occupying a vacant Fe-coordinate site at the 2-OG-bound stage (Figure28).(239) Subsequent formation of an Fe(IV) oxo species is facilitated by theelectron transfer from Fe coupled with the oxidation of 2OG into succinate and CO₂. In the proximity of the highlyreactive Fe(IV)=O species is positioned the methyl group of a methyllysine substrate so that the Fe(IV)=O intermediate can extracta hydrogen atom from the methyl group to yield Fe(III)-OH and a N-methylene radical. The following rebound of the N-methyleneradical with Fe(III)-OH leads to a hemi-aminal lysine intermediate (Figure 28).Alternatively, the hemi-aminal intermediate can be the product of Fe(IV)=O-mediated oxygen insertion (Figure 28). Between the two mechanisms, the oxygen-radical rebound has been favored given a similarobservation in a model system.(242) The catalytic cycle is then completed upon hydrolysisof the hemi-aminal lysine intermediate to release the byproduct formaldehyde (Figure 28). Incomparison with LDS1/2, which rely on the production of an imine intermediate and thus can only act on Kme1/2 as substrates,(241) JmjC-domain-containing KDMs catalyze a direct hydroxylation of methyllysine and thuscan act on Kme1/2/3 as substrates. Recently, an unconventional mechanism was reported for lysyl oxidase-like 2 protein (LOXL2) toremove H3K4me3 through an oxidative deamination at the δ-carbon of methyllysine (Figure30).(246) For this deamination reaction, though removing a methyllysine mark,installs a novel allysine residue rather than restores an intact lysine residue as the final product.

6.3. Substrate specificity of KDMs.

KDMs recognize substrates through their methylation states and the sequences flanking methyllysine residues. WhileJmjC-domain-containing KDMs can remove methyl groups from Kme1/2/3 and LSD1/2 can only act on Kme1/2 as substrates,(239,241) it remains to be determined whether LSD1/2indeed could not bind Kme3 substrates or can bind them but fail to promote catalysis. A general observation is that KDMs,harboring demethylase activities towards the substrates with high degrees of methylation (e.g. Kme3/2), can acton the cognate methyllysine substrates with lower degrees of methylation (e.g. Kme1/2). The structures of KDMs incomplex with substrates reveal that these enzymes contain a cavity to accommodate the methyllysine side chain of substrates (Figure 26,31).(237,238,240) The methyllysine sidechain, which is buried deeply near the Fe-(II) cofactor, engages in few specific interactions with KDMs (Figure 26).(240) It is thus expected that Kme2-binding pocketsof some KDMs provide certain steric constraints and thus prevent them from binding Kme3-containing substrates. In contrast, it isready for KDMs to accommodate the cognate substrates with lower degrees of methylation.(244) Collectively, at least two factors, the demethylation mechanism (JmjC enzymes versus LSD1/2) and the spatialrestriction of methyllysine binding pockets, determine whether a KDM can act on Kme3 as a substrate. In contrast, demethylaseactivities of most KDMs have been demonstrated with Kme1/2 as substrates.

Besides a specific methylation state of lysine substrates, sequences flanking a methyllysine residue also play roles onsubstrate recognition of KDMs. The structures of KDMs (LSD1/2 and JmjC enzymes) in complex with substrates reveal long-rangeinteractions for substrate recognition (Figure 31).(237,238,240) For instance, the sidechains of Arg2, Gln5, Thr6, Arg8 and Ser10 of histone H3 involve intermolecular or intramolecular interactions for recognition ofLSD1’s substrate (Figure 31).(237) So far,most histone lysine methylation marks are shown to be substrates of KDMs.(247) Similar toPKMTs that can act on diverse nonhistone substrates, KDMs have been shown to demethylate methyllysine in nonhistone proteins suchas E2F1,(122) DNMT1,(248) STAT3,(124) ERα(249) and HSP90α.(250) Most of these nonhistone targets were characterized as LSD1’s substrates withoutwell-defined sequence homology at their demethylation sites (Figure 31). This observation isremarkable because LSD1 was shown to bind its H3K4me2 substrate in a highly sequence-specific manner (Figure 31). The paradox between the promiscuous sequences and the specific recognition of LSD1substrates can be rationalized by at least two nonexclusive mechanisms. First, the substrate-recognizing pocket of LSD1 can beflexible and thus adopt multiple conformations to accommodate different substrates. Second, LSD1’s substrate specificity ismodulated through recruitment of other regulatory partners. For instance, the SWIRM domain of LSD1 can associate with androgenreceptors and this interaction alters LSD1’s substrate preference from H3K4me2 to H3K9me2.(251) It is also of interest to explore whether other KDMs like LSD1 act on diverse nonhistone substrates.

7.1. Uncovering lysine methylomes through chemical labeling.

PKMT-catalyzed protein lysine methylation is characterized by the transfer of SAM’s sulfonium methyl group toprotein lysine substrates through a S_N2 transition state.(98) Given thechallenge to profile proteome-wide lysine methylation events, an attractive alternative is to uncover lysine methylomes throughchemical labeling.(21) In this approach, the sulfonium methyl group of the SAM cofactor canbe replaced with its isotopic or chemical isosteres (e.g., CD₃,¹³CD₃,CT₃,¹⁴CH₃, allyl and propargyl), which are transferred by native or engineered PKMTs totheir substrates.(141,256–260) The distinct properties of these methyl isosteres then allow more robust signal readouts or targetenrichment.

7.2. Enzymatic labeling lysine methylomes with isotopic SAM cofactor.

For chemical labeling of a methylome with isotopic isosteres of the SAM cofactor (heavy methyl labeling),S-[methyl-¹³C/²H]-SAM and radioactiveS-[methyl-¹⁴C/³H]-SAM have been broadly used.(257,258) Because of SAM’s poor membrane permeability, the direct useof the isotopic SAM cofactors is largely restricted for its use inin vitro settings such as cell lysates.(21) SAM is synthesized within living cells by highly-conserved methionineadenosyltransferases (MATs) with endogenous ATP and methionine as co-substrates (Figure32).(19,21) To label PKMT targets insideliving cells with isotopically labeled SAM cofactors,S-[methyl-²H/³H/¹³C/¹⁴C]-methionine can be used as the precursorsfor the MAT-catalyzed biosynthesis (Figure 32). The isotopically labeled methionine can bereadily internalized by living cells, likely through amino acid transporter(s), and processed into the corresponding SAM cofactorsby MATs for target labeling (Figure 32).(257,258) In this labeling process, protein translation inhibitors such as cycloheximide areadded to prevent direct incorporation of the isotopic methionine into protein through RNA translation machinery.(258)

ForS-[methyl-²H/¹³C]-SAM, which contain stable isotopes,[CH₃]/[methyl-²H/¹³C]-SAM are often used in a combination (Figure 32). The resultant [CH₃]/[methyl-²H/¹³C]-labeled PKMT targets can then becharacterized unambiguously upon detecting the pair of [CH₃]/[methyl-²H/¹³C]-labeled products orin a more quantitaive manner by analyzing their isotopic ratios with MS (Figure 32).(257) Ong and Mann pioneered chemical labeling of a methylome withS-[¹³CD₃]-methionine as the biosynthetic precursor ofS-[¹³CD₃]-SAM and uncovered around 100 methylation events from the proteome of HeLacells. The existence of a pair of mass shifts of 14 Da for [CH₃]-labeling and 18 Da for[¹³CD₃]-labeling is characteristic upon revealing bona fide methylation events (Figure 32).(257) With the aid of a recently developed mathematicalgorithm, the isotopic ratios can be determined precisely by MS with statistical errors within only a few percent.(98)

For chemical labeling of a methylome withS-[methyl-³H/¹⁴C]-SAM, these radioactivecofactors are often used as tracers in a mixture containing a large amount of unlabeled SAM as a cold carrier.(258) The limited use of radioactive materials can be due to the presence of unlabeled cognates inavailable radioactive materials (e.g., for a high percentage of¹²C-SAM in¹⁴C-containingSAM), cost of excessive radioactive materials and potential environmental contamination. Therefore, usingS-[methyl-³H/¹⁴C]-SAM as PKMT cofactors is often restricted in well-definedinvitro settings because of their poor membrane permeability.(27) In contrast,S-[methyl-³H/¹⁴C]-methionine can be readily uptaken by cells and processed byendogenous MATs intoS-[methyl-³H/¹⁴C]-SAM for target labeling inside living cells. Theresultant [methyl-³H/¹⁴C]-labeled targets can be probed by autoradiography for characteristic radioemission. Detection thresholds and signal-to-noise ratios ofS-[methyl-³H/¹⁴C]-SAM-labeledtargets can be further improved in combination with advanced sample enrichment strategies as discussed later.(141)

Recent work on KIEs of PKMTs with [CD₃]-SAM cofactors and [CT₃]-SAM further revealed that[CD₃]-SAM and [CT₃]-SAM are more reactive than SAM on the basis of their inverse KIEs and 20~30%largerk_cat/K_m values in comparison with [CH₃]-SAM.(98,111) This difference can originate from theincreased steric impingement of vibrational modes of the methyl group of the SAM cofactor at the transition state in comparisonwith those in the ground state (Figures 9,10). Such aneffect allows heavy SAM cofactors such as [CD₃]-SAM and [CT₃]-SAM to react faster than native SAM. Incontrast, [¹⁴CH₃]-SAM is slightly less reactive than SAM as reflected by its normal KIEs and 3~5%smallerk_cat/K_m values.(98,111) This rate decrease is due to the loss of the overall bond order of the transferredmethyl group at the transition state. Given low radiospecificity of¹⁴C, [CT₃]-SAM is preferred over[¹⁴CH₃]-SAM as a cofactor surrogate for target labeling.

7.3. Further chemical derivatization of lysine methylomes for MS analysis.

Besides enzymatic labeling of methylomes with isotopic SAM cofactors, proteome-wide Kme1/2 sites as well as unmodified Lyscan be subjected to further chemical derivatization (Figure 33).(256,261–263)For instance, Lys and Kme1 can be modified by anhydrides (e.g., propionic anhydride) to generate thecorresponding acylated products (Figure 33).(256,261) Free Lys and Kme1/2 can be modified by formaldehyde followed byNaBH₄-mediated reduction to yield the corresponding Kme3 products (Figure33).(263) These chemical modifications make the processed protein samples morehomogenous for trypsin digestion and MS analysis. In the process of methylome profiling, methyllysine-containing proteins areoften subjected to proteolytic digestion by trypsin alone or in combination with other proteases followed by MS analysis.(263) While trypsin primarily cleaves the amide bonds after basic amino acid residues such asLys and Arg, my laboratory noted that there is residual activity of trypsin to cleave the amide bonds after a Kme1 site. Incontrast, the amide bonds of acylated Lys and Kme1/2 are inert for trypsin cleavage (an observation in the Luo laboratory). As aresult, the pattern of MS-revealed peptide sequences after trypsin digestion can alter according to the sites and degrees (Lys,Kme1, Kme2 and Kme3) of protein lysine methylation. This chemical derivatization step essentially spares all Lys sites fromtrypsin digestion (Figure 33) and thus allow a homogenous, ArgC-like proteolytic digestionpattern regardless of the prior status of Lys methylation. For the formaldehyde-involving reductive amination, isotopicallylabeled formaldehyde can be used to further distinguish the methylation events associated with chemical derivatization from thosecatalyzed by PKMTs.(263) In addition, the processed peptides have Kme3 at all Lys sites.The prior states of lysine methylation thus have no effect on their ionization efficiency for quantitive MS analysis.Collectively, the distinct chemical properties and thus selective derivatization of Lys, Kme1, Kme2 and Kme3 allow the productionof more homogenous lysine methylome samples for trypsin digestion and MS quantification (Figure33).

7.4. Chemical labeling by native PKMTs with clickable SAM analog cofactors.

Given the merit of terminal-alkyne/azide-containing moieties for copper-catalyzed cyclization addition (click reaction),terminal-alkyne/azide-containing (clickable), activity-based probes have been developed to examine protein posttranslationalmodifications.(27) While the utility of clickable SAM analogs was documented formethyltransferases in 2006, it was until 2010 that there was the first report of an active clickable SAM analog to label PKMTtargets (Figure 34).(27) The Weinhold group firstshowed that (E)-pent-2-en-4-ynyl SAM analog (EnYn-SAM) has the detectable activity as a SAM surrogate toward afungal PKMT (Dim-5) and two human PKMTs (MLL1 and MLL4).(264) Here EnYn-SAM featuresSAM’s sulfonium methyl group replaced by a sterically bulky, clickable (E)-pent-2-en-4-ynyl moiety (Figure 34). To explore general applicability of sulfonium-alkyl SAM analogs as cofactors ofwild-type PKMTs, the Luo laboratory evaluated five SAM analogs (allyl-SAM, propargyl-SAM, EnYn-SAM,(E)-hex-2-en-5-ynyl-SAM (Hey-SAM) and 4-propargyloxy-but-2-enyl-SAM (Pob-SAM) (Figure 34) against a panel of human PKMTs.(191) Among the examined 5×5pairs of PKMTs and SAM analogs, only native SUV39H2, G9a and GLP1 are active towards allyl-SAM. In contrast, the bulky SAM analogssuch as EnYn-SAM, Hey-SAM and Pob-SAM showed undetectable activities as cofactors of the 5 examined native PKMTs.

The lack of activity of propargyl-SAM particularly caught our attention given that the overall structure of this SAManalog is comparable to that of allyl-SAM, which is an active cofactor for several PKMTs.(21,27,143,191,259,265)Propargyl-SAM was envisioned as the smallest SAM surrogate containing a clickable moiety for the transfer reaction and subsequenttarget characterization with the well-established alkyne-azide click chemistry (Figure34).(259) Unfortunately, progargyl-SAM, though was characterized to be stableat an acidic pH, is unstable at physiological pH with a half-life time shorter than 1 min, a time scale that is too short to labelPKMT targets in an efficient manner.(259) At physiological pH, progargyl-SAM can besubjected to deprotonation at the double-activated alkyne-sulfonium carbon and decompose into keto-SAM via a putative alleneintermediate (Figure 34).(259) Our laboratory aswell as the Weinhold laboratory independently overcome the stability issue by replacing the sulfonium with selenium to lower theacidicity at the double-activated carbon center (Figure 34).(266) The resultant ProSeAM (propargylicSe-adenosyl-L-selenomethionine) has ahalf-life time of 1–2 hours at physiological pH with its decomposition mechanism different from that ofpropargyl-SAM.(259) Consistent with the activity profile of allyl-SAM, ProSeAM isactive toward human SUV39H2, G9a and GLP1, which harbor di-/tri-methylation activities (Figure34).(259,260) Under a steady-state,GLP1 processes ProSeAM as a cofactor withk_cat of 0.375 min⁻¹ andK_m of 45.4 μM, which are only 5- to 15-fold different fromk_catof 1.97 min⁻¹ andK_m of 3.1 μM of the native cofactor SAM.²⁴⁷ Although the work of the Weinhold group showed that ProSeAM is also activetoward SET7/9, which was known for its monomethylation activity,(266) the weak activity ofSET7/9 on ProSeAM likely arises from a highly sensitive assay used there. Collectively, the activity profile of ProSeAM towardsPKMTs strongly argues that ProSeAM is an active SAM surrogate cofactor for the PKMTs (e.g., SUV39H2, G9a andGLP1) whose enzymatic active sites are spacious enough for di-/tri-methylation.

7.5. Chemical labeling with allele-specific PKMT-cofactor pairs.

Although ProSeAM and EnYn-SAM are active cofactors for native PKMTs and have demonstrated their use for targetcharacterization through their clickable moieties, the labeling efficiency of these SAM analogs is modest and their generalapplicability towards a broad family of PKMTs remains to be examined.(143,265) More importantly, it is challenging to correlate the revealed targets to the activitiesof specific PKMTs because ProSeAM and EnYn-SAM are active towards multiple endogenous PKMTs.(259,260,264) It is also challengingto associate an altered lysine methylome of a designated PKMT activity through specific perturbation, given the potential cascadeeffect of such perturbation. To address these challenges as well as low activities of wild-type PKMTs toward sulfonium-alky SAManalogs, we envisioned a Bioorthogonal Profiling Protein Methylation (BPPM) technology (Figure35).(27,99,143,265,267) In BPPM,highly conserved SAM-binding pockets of PKMTs are engineered to process bulky sulfonium-alky SAM analogs (e.g.,Hey-SAM or Ab-SAM), which are otherwise too steric to be accommodated by wild-type PKMTs.(99,143)

Upon searching engineered PKMT-cofactor pairs for BPPM, we first focused on G9a and GLP1 given their well-characterizedin vitro methyltransferase activities.(99,143) The Y1154A mutant of G9a and its equivalent Y1211A mutant of GLP1 but not wild-type G9a and GLP1were shown to be active towards Hey-SAM and 4-azidobut-2-enyl SAM (Ab-SAM). Remarkably, the apparentk_cat/K_m,cofactor values of the two PKMT variants with Hey-SAM as acofactor are comparable to those of wild-type G9a and GLP1 with the native cofactor SAM.(99,143) Benefited from the clickable terminal-alkyne/azido moieties of these SAManalog cofactors, the resultant labeled substrates can be readily visualized through conjugation with fluorescent dyes or enrichedwith biotin reporters (Figure 36).(268,269) Since these engineered PKMTs contain a single-site mutation, which is remote from theirsubstrate-binding pockets, these mutations are not expected to affect substrate recognition. This argument is further supported byready validation of many BPPM-revealed substrate candidates.(268,269)

Allyl-SAM, ProSeAM, EnYn-SAM, Hey-SAM, Pob-SAM and Ad-SAM have demonstrated cofactor activities towards native orengineered PKMTs.(99,143) One common structuralfeature of these active SAM analog cofactors is that the sp³ carbon center is double-activated by its immediateadjacent sulfonium and sulfonium-β sp¹/sp² carbons for a S_N2 transfer reaction (Figure 34).(99,143) The sulfonium-β sp¹/sp² carbons are indispensable forS-alkyl SAManalogs to be active towards PKMTs (native or engineered), whereas the equivalentS-alkyl SAM analogs lacking thesulfonium-β sp¹/sp² carbons are typically inert as methyltransferase cofactor surrogates.(99) A direct rationale for this observation is that the sulfonium-βsp¹/sp² carbons conjugate and thus stabilize S_N2 transition states of the PKMT-catalyzedtransalkylation reaction (Figure 37).(99) Consistentwith partial S-C bond breaking at S_N2 transition states of PKMTs revealed by KIEs, the replacement of the sulfonium inS-alkyl SAM analogs with selenium only slightly accelerates the rates of the transalkylation reaction of asmall set of PKMTs (native or engineered).(260)

Structural analysis on native PKMTs suggests that the highly conserved Tyr residue in SET domain-containing PKMTs(e.g., Y1154 of G9a and Y1211 of GLP1) is essential for their transmethylation reactions.(99) Small but significant inverse BIEs upon binding [CD₃]-SAM and [CT₃]-SAM byPKMTs suggest the steric impingement of the surrounding environment of SAM’s methyl group (Figure 13).(98,111,112) The observed inverse BIEs could partially originate from noncanonical CH•••Ohydrogen-bonding interaction.(98,111,112) Such interaction is expected to construct an electrostatic pore to confine the motionof the sulfonium methyl group to assemble a S_N2 transition state (Figure 7).(98,111,112)Disruption of this interaction with Ala mutation in G9a and GLP1 leads to a 300-fold decrease of the catalytic efficiency with theSAM cofactor (Figure 37).(99) We envision that theloss of the noncanonical CH•••O hydrogen-bonding interaction in G9a Y1154 mutant and GLP1 Y1211 mutant withthe native SAM cofactor can be partially compensated by the transition-state stabilization through the sulfonium-βsp² carbons of these SAM analogs. This model is consistent with the 10-fold higherk_catvalue of the allyl-SAM than that of the native SAM as the cofactors of G9a Y1154 mutant and GLP1 Y1211 mutant (Figure 37).(99) Remarkably, double-activatedsulfonium-β-alkenyl SAM analogs show comparable affinity to the Y1211A/Y1154A mutants as reflected by less than 5-folddifference of theirK_m values (12–80 μM for Y1154 and 8–40 μM for Y1211A),despite the dramatic difference in the size of their sulfonium-δ-substituents. In contrast, the correspondingk_cat values can alter by 100-fold (Figure 37).(99) This observation suggests that the two PKMT mutants are spacious and flexible enough toaccommodate structurally diverseS-alkyl SAM analogs.(99) However, theoptimal binding of structurally matchedS-alkyl SAM analogs is essential to correctly position thedouble-activated sp³ carbon center for a linear S_N2 transition state (Figure37). Collectively, two essential structural features for allele specificS-alkyl SAM analog cofactorsare (i) sterically matched δ-substituents for optimal cofactor binding to assemble S_N2 transition states; (ii)the sulfonium-β-sp² moiety to lower the energy barrier of S_N2 transition states.

7.6. Bioorthogonal profiling of protein lysine methylation inside living cells.

Similar to the native SAM cofactor,S-alkyl SAM analog cofactors show poor membrane permeability and thusare not suitable for target labeling in living cells.(21,270) Inspired by the success of BPPM in uncovering the methylome of designated PKMTs with cell lysates, we advancedthe BPPM technology for living cells (Figure 38).(21,270) The key step of implementing BPPM inside living cells is to hijack thebiosynthetic pathway of SAM with engineered MATs and membrane-permeableS-alkyl methionine analogs forinsitu production of the correspondingS-alkyl SAM analogs. The three-step BPPM within living cellsconsist of the biosynthesis of SAM analogs from methionine analog precursors by MAT2A I117A mutant,in situtarget labeling by engineered PKMTs, and subsequent enrichment of the distinct modified targets via the click chemistry (Figure 38).(21,270) Human MAT2A I117A variant was shown to be the so-far most efficient enzyme to process bulkyS-alkyl methionine analogs into the corresponding SAM analogs.(19,270) With G9a and GLP1 as examples, we showed that (E)-hex-2-en-5-ynylhomocysteine (Hey-methionine analog) can be processed by the I117A variant of human MAT2A into the corresponding Hey-SAM.(270) In the presence of the BPPP-feasible G9a and GLP1 mutants (Y1154A of G9a and Y1211A ofGLP1), histone H3 and other chromatin targets of G9a and GLP1 can be efficiently labeled. The subsequent chromatin enrichment witha biotin-azide probe followed by genome-wide sequencing revealed the preferred chromatin loci harboring the methyltransferaseactivities of G9a and GLP1.(270) Besides robust enrichment through the clickable moietyand unambiguous assignment of the targets to specific PKMTs, the BPPM method in living cells has the merit to capture dynamicmethylation events that are subjected to rapid demethylation by KDMs. Bulky lysine modifications by alkyl-SAM cofactors areexpected to be more inert for the removal by KDMs.

7.7. Recognizing lysine methylomes through physical interaction.

Because methyllysine-containing proteins often account for a small fraction of a cellular proteome, efficient enrichmentof methyllysine-containing proteins is essential to increase signal-to-noise ratios for subsequent characterization.(21,27) Given the similarity of the overall physicalproperties between free lysine and methyllysine, a current hurdle for methylome profiling is to access reagents that candistinguish methyllysine versus free lysine of a broad spectrum of otherwise cognate proteins. While many high-qualityanti-methyllysine antibodies were developed to recognize specific methyllysine modifications within well-defined peptidesequences, broad recognition of these antibodies for methyllysine-containing proteomes remains to be improved.(41) In parallel with the effort to develop high-quality pan-anti-methyllysine antibodies, methyllysinereader domains can also be utilized to recognize and enrich lysine methylomes.(140–142,255) These reagents,in combination with advanced MS technologies, have become valuable additions to elucidating methylomes.

8.1. General criteria of PKMT inhibitors as chemical probes.

PKMT inhibitors can be developed either as chemical probes to perturb the methyltransferase activities of specific PKMT(s)or as drug candidates for therapeutic use.(38,273)On the basis of these respective uses, different principles should be applied upon evaluating overall quality of PKMTinhibitors.(274) To develop chemical probes, Frye introduced five general principlesin his article of “The Art of the Chemical Probe”.(274) Upon rephrasingthese principles in the current setting, high-quality PKMT chemical probes should: (1) show sufficientin vitropotency and selectivity against one or a set of designated PKMT(s); (2) show decentin vivo or at least cellularpotency and selectivity, which correlate with the correspondingin vitro data (in vivo orcellular EC₅₀ versusin vitro IC₅₀); (3) be well characterized in term of MOAinvitro andin vivo or in a cellular context (e.g., SAM-competitive,substrate-competitive, allosteric or covalent inhibitors); (4) demonstrate at least one utilization (e.g., thetreatment of a PKMT chemical probe recapitulates certain biological readouts implicated by genetic perturbation of the PKMT); (5)be accessible through well-described chemical methods or from commercial vendors. Additional criteria required for drug candidatesinclude proper pharmacokinetic, pharmacodynamic and toxicity profiles.

While a well-defined spectrum of potential off-target effects is an essential parameter for a PKMT chemical probe, atherapeutic index of the dose of efficacy versus toxicity is more concerned for a drug candidate. Great caution should be madewhen PKMT inhibitors are used under unprecedented biological settings (e.g., across different cell types,cellular versusin vivo use, oral, intraperitoneal injection versus intravenous dose). Even for the bestcharacterized PKMT inhibitors, it is essential to reevaluate vigorously their target engagement and inhibitory efficiency underdifferent biological settings. Current efforts have been mainly focused on developing small-molecule inhibitors against humanPKMTs. While the sequences of human PKMTs can be homologous across species, certain caution should be made by assuming that thepotency and selectivity of a PKMT inhibitor could be fully maintained for sequence-related PKMT homologues. There are not short ofexamples that single-point mutations are sufficient to alter potent inhibitors into completely inactive compounds.(93) A PKMT inhibitor that is characterized for its use in human cell lines needs to be reevaluated uponperturbing related PKMTs across species.

8.2. Inhibition of PKMTs through SAM-competitive MOA.

A common strategy for developing PKMT inhibitors is to engage small molecules to occupy SAM-binding pockets of PKMTs andthus inhibit their catalysis.(38) Among well-characterized SAM-competitive SAM analoginhibitors are EPZ004777 against DOT1L,(275) Pr-SNF/Bn-SNF against SETD2 and Pr-SNFagainst NSD2 (Figure 39).(75,77) A key character of these SAM-competitive inhibitors is that their IC₅₀ values show alinear increase in the presence of increased concentrations of the SAM cofactor. A SAM-competitive MOA can be further supported bystructural data of these PKMTs in complex with corresponding inhibitors, for which SAM-binding pockets are occupied by theinhibitors.(75,77,276) As revealed by structural data, most of these PKMTs show dramatic conformational changes uponbinding these SAM-competitive inhibitors. For instance, several SAM-binding loop regions of DOT1L, which used to interact with theα-amino carboxylic acid moiety of the SAM cofactor, open up to accommodate the bulkier phenyl urea moiety of EPZ004777; theautoinhibitory loop of SETD2, which used to be in a close configuration upon binding SAM or SAM analogs, makes nearby a 180-degreeflipping to accommodate the bulky benzyl moiety of Bn-SNF.(75,276) While Bn-SNF was characterized as a SAM-competitive inhibitor, its parent IC₅₀ alsodepends on the presence of a H3K36 peptide substrate (Figure 39).(75) This inhibitor shows higher affinity toward SETD2 by forming a ternary inhibitor-SETD2-substratecomplex. As a result, Bn-SNF is better defined as a SAM-competitive, substrate-dependent inhibitor (Figure 39).(75)

The pyridone-based GSK126(277,278) andUNC1999 (EZH2 or dual EZH1/2 inhibitors),(279) and several fragment-based DOT1L inhibitorsrecently developed by Novartis(280–282) arethe SAM-competitive PKMT inhibitors with no structural similarity to the SAM cofactor (Figure39). Like canonical SAM-competitive cofactor analog inhibitors (Figure 39), thesenon-nucleotide-based compounds also show linearly increased IC₅₀ values in the presence of increased concentrations ofthe SAM cofactor. The SAM-competitive MOA is also supported by their structural data with the SAM-binding pockets fully orpartially occupied by these compounds.(93,280–282) Interestingly, upon binding most of these compounds, both EZH2and DOT1L show dramatic conformational changes around their SAM-binding sites.(93,276,280–282) For instance, several newly released structures of EZH2 or its homolog enzyme in complex with pyridone-basedinhibitors showed that a loop region adjacent to the adenine-binding pocket adopts a new conformation upon replacing SAH withthese inhibitors as ligands.(93) In the case of fragment-based DOT1L inhibitors, thebinding of these compounds by DOT1L (Figure 39) also involves a distinct pocket adjacent toits SAM-binding site and then induces the conformational changes that are unfavorable for SAM binding.(280–282) The inhibitor-induced conformational changescould also rationalize how these inhibitors selectively interact with specific PKMTs but not others even if they all sharehighly-conversed domains for SAM recognition. In most of these scenarios, PKMTs are subjected to dramatic conformational changesupon binding these SAM-competitive inhibitors. Similar to the situation of protein kinases,(283) the strong preference of certain PKMTs through the exploration of their structurally matched alternativeconformations could be essential for the high potency and selectivity of these PKMT inhibitors. Exploring and targetingstructurally distinct conformers of PKMTs may present a general strategy to developing potent and selective SAM-competitiveinhibitors.

Multiple factors can affect the potency of SAM-competitive, substrate-noncompetitive PKMT inhibitors inside cells. Ingeneral, canonical SAM-competitive PKMT inhibitors act better in a cellular setting with a low concentration of endogenous SAM(Figure 39). Nevertheless, as long as concentrations of these PKMT inhibitors issufficient to compete with SAM, the methyltransferase activities of the engaged PKMTs should be inhibited regardless of thepresence of peptide substrates (Figure 39). Given that intracellular concentrations of SAMmay be dramatically affected by the level of other metabolites such as methionine and vary by several folds across cell or tissuetypes,(1,196) it is thus important to evaluatethe efficiency of SAM-competitive inhibitors in different cellular contexts. The situation can be further complicated in thepresence of PKMT complexes. For instance, EZH2 requires the binding of EED and SUZ12 to form a core complex to efficientlycatalyze H3K27 methylationin vivo.(82,213,284) The methyltransferase activity of the core complex can be furtherenhanced allosterically through interaction of the EED subunit with an H3K27me3 peptide.(82,213,284) Here, the SAM-competitiveEZH2 inhibitor GSK126 showed around 10-fold increase of its affinity to the EED-SUZ12-EZH2 complex with the additional presence ofan H3K27me3 peptide.(277) Given that DOT1L-catalyzed H3K79 methylation can be acceleratedupon forming the DOT1L-AF10 complex,(190) it will be interesting to explore whether theSAM-competitive DOT1L inhibitor EPZ004777 shows different affinity to DOT1L versus the DOT1L-AF10 complex. Target engagement canbe distinct for SAM-competitive, substrate-dependent inhibitors such as Bn-SNF for SETD2.(75) In these cases, the optimal cellular setting for the target engagement is the presence of an substrate and theabsence of the SAM cofactor to facilitate the maximal formation of the ternary inhibitor-PKMT-substrate complex (Figure 39).(75)

8.3. Inhibition of PKMTs through substrate-competitive MOA.

Another MOA of PKMT inhibitors is to engage small molecules to occupy peptide-binding pockets of PKMTs(substrate-competitive inhibitors) (Figure 40).(33,38) Given different PKMTs may contain distinct peptide-binding pockets to recognizetheir substrates, the substrate-competitive MOA is expected to be more feasible for developing selective PKMT inhibitors.UNC0638/UNC0642 (G9a/GLP1 inhibitors),(285) (R)-PFI-2 (an inhibitor ofSETD7/9),(127) LLY-507(272) and A-893 (twoinhibitors of SMYD2),(286) and A-196 (an inhibitor of SUV4-20H1/2)(287) are examples of well-characterized substrate-competitive PKMT inhibitors with cellular activities.Target inhibition of these compounds is characterized by a linear increase of their apparent IC₅₀ values with increasedconcentrations of substrates. The substrate-competitive MOA of these inhibitors is also consistent with the binary PKMT structuresin which the small-molecule ligands occupy the substrate-binding pockets of PKMTs (Figure40). Substrate-competitive PKMT inhibitors can be further classified according to whether the SAM cofactor can be involvedto form more stable inhibitor-PKMT-SAM ternary complexes (Figure 40). The G9a/GLP1inhibitors UNC0638 and UNC0642 and the SUV4-20H1/2 inhibitor A-196 are canonical substrate-competitive inhibitors with theirapparent IC₅₀ independent upon the presence of the SAM cofactor.(285,287) In contrast, (R)-PFI-2 is a substrate-competitive, SAM-dependentinhibitor with the strong preference to form the inhibitor-SET7/9-SAM ternary complex and thus strengthen the interaction of(R)-PFI-2 with SET7/9 in the presence of the SAM cofactor (decreased apparentK_dvalues with increased concentrations of SAM).(127) AZ505, an analog of the SMYD2 inhibitorA-893, also showed a similar substrate-competitive, SAM-dependent character.(86) It remainsto be determined whether the presence of the SAM cofactor increases the affinity of substrate-competitive inhibitors A-893 andLLY-507 to SMYD2.

Cellular target engagement of canonical substrate-competitive PKMT inhibitors negatively correlates with the affinity ofPKMTs to inhibitor-competitive substrates as well as their intracellular concentrations. Because PKMTs may methylate diversehistone and nonhistone targets with a broad range ofK_m,substrate values and intracellularconcentrations, EC₅₀ values of substrate-competitive PKMT inhibitors will likely vary on different substrates. It isthus likely to apply modest doses of substrate-competitive PKMT inhibitors to perturb the methylation events associated with asubset of substrates with low affinity (highK_m,substrate) and intracellular concentrations, but sparethose substrates with high affinity (lowK_m,substrate) and high intracellular concentrations frominhibition.(38)

It is worth noting that some PKMT inhibitors may partially occupy substrate-binding pockets but do not show the expectedsubstrate-competitive character (the increase of IC₅₀ versus increased concentrations of substrates). For instance,EPZ030456 is a selective and potent inhibitor of SMYD3 with a non-competitive character for its substrate MAP3K2.(288) However, the structure of SMYD3 in complex with EPZ030456 shows that the inhibitor residues in thepocket that is otherwise occupied by MAP3K2, a SMYD3 substrate, in the structure of the SMYD3-MAP3K2 complex (Figure 41).(288) The Lys-binding site of the MAP3K2 substratecan also be occupied by the propyl dimethylamino moiety of the SMYD3 inhibitor GSK2807 (Figure41).(289) However, GSK2807 also does not show the noncompetitive characteragainst the substrate MAP3K2. To rationalize these paradoxical observations, it was proposed that the optimal binding of MAP3K2 toSMYD3 may mainly occur outside of the substrate pocket with minimal engagement of the lysine substrate and its neighboringresidue(s). As a result, the presence of the MAP3K2 substrate has no inference in SMYD3’s binding to EPZ030456 and GSK2807.It therefore should be cautious to conclude substrate-competitive MOA solely on the basis of potential steric clash betweeninhibitors and substrates revealed by their PKMT complexes. The clashed occupancy of inhibitors and substrates does not alwaysgrant a substrate-competitive MOA because they may accommodate each other through alternative conformations without significantenergy penalty. In contrast, because of the well-defined binding mode of SAM in PKMTs, it is more confident to conclude aSAM-competitive MOA upon observing even partial occupancy of inhibitors in SAM-binding pockets of PKMTs.

8.4. Inhibition of PKMTs by allosteric MOA.

Besides SAM-competitive and substrate-competitive MOAs, a distinct set of PKMT inhibitors were developed by targeting theallosteric sites essential for catalysis (Figure 42). It has been documented that themethyltransferase activities of certain PKMTs depend upon the formation of higher-order complexes and thus participation of theresidues remote from their catalytic sites. For instance, the H3K4 methyltransferase activity of MLL1 depends on the formation ofa core complex with at least three binding partners WDR5, ASH2L and RbBP5 with the MLL1-WDR5 interaction essential for thecatalysis.(79–81) MM-401(81) and OICR-9429(79) were developed as the allostericinhibitors of MLL1 by their competitive occupancy of the central channel of WDR5, which is otherwise occupied by MLL1 to form thecore complex for catalysis (Figure 6 andFigure 41).Similar strategies have also been applied to develop MI-2/3/463(80,290) and SAH-EZH2(83) as disruptors of MLL1-Menin andEZH2-EED-SUZ12 complexes for allosteric inhibition of MLL1 and EZH2, respectively (Figure 6andFigure 41). The methyltransferase activity of the core EED-SUZ12-EZH2 complex can bealloterically modulated through the interaction of the aromatic cage of EED subunit consisting of F97, Y148 and Y365 with anH3K27me3 peptide. EED226, A-395 and their derivatives were developed by occupying this aromatic cage.(82,284) The induced conformational arrangement of the aromaticcage upon binding these inhibitors is expected to be rendered to remote site(s) to perturb catalysis (Figure 6 andFigure 41). The PKMT inhibitors characterized asallosteric MOA often engage their targets at the sites remote from the catalytic pockets and thus show noncompetitive characterstowards the SAM cofactor and substrates (EC₅₀ independent upon the presence of substrate and SAM) (Figure 42). PKMT inhibitors with the allosteric MOA are therefore complimentary with SAM-competitive andsubstrate-competitive PKMT inhibitors for engagement of PKMTs in a cellular context.

8.5. Inhibition of PKMTs through covalent or suicide MOA.

While the development of cysteine-targeted covalent inhibitors has been well documented for kinases, few PKMTs inhibitorswere developed through covalent MOA. With substrate-competitive SETD8 inhibitors UNC0379 and MS2177 as lead scaffolds, the Jinlaboratory developed its analog MS453 by installing an electrophilic acrylamide group to selectively target the C311 residue ofSETD8 (Figure 43).(291,292) MS453 showed time-dependent target engagement and thus decreased IC₅₀ values asextending preincubation time with SETD8. This covalent MOA is also supported by the structure of SETD8 in complex withMS453.(292) Upon screening > 5,000 commercial compounds, SPS8I1 (NSC663284),SPS8I2 (ryuvidine) and SPS8I3 (BVT948) were identified as SETD8 inhibitors with modest selectivity (Figure 43).(293,294) These compoundsas well as their derivatives such as SGSS05-NS and SPECS21(295) feature a common quinonemotif with anticipated covalent modification of a Cys residue. SPS8I1 (NSC663284) and SPS8I2 (ryuvidine) were characterized tocovalently modify C311 of SETD8, while SPS8I3 (BVT948) may target multiple Cys residues of SETD8.(293) To develop inhibitors against DOT1L, Yao and coworkers described SAM-based suicide inhibitors (Figure 43).(110) These compounds are expected to undergo anintracellular cyclization to generate a highly reactive aziridinium (Figure 43). Thisintermediate then reacts with the K79 of the substrate H3 to generate a binary substrate-inhibitor adduct for DOT1L inhibition(Figure 43).(110) A similar suicide strategy wasalso reported to develop potent inhibitors of protein arginine methyltransferases.(296) Itremains to investigate whether covalent or suicide MOAs can be generally applicable for PKMT inhibition.

8.6. Evaluation of target engagement of PKMT inhibitors.

Because complicated factors contribute to the potency of PKMT inhibitors, their cellular target engagement should beevaluated. PKMT inhibitors are often appended with a functional anchor (e.g., a terminal alkyne, azide or biotinmoiety) at an inert position of inhibition. This anchor can then be used to pull down engaged PKMT(s) from cell lysates.(38) Alternatively, a cellular thermal shift assay (CTSA) can be implemented.(297) Here the binding of inhibitors to PKMTs is expected to increase their thermal stability. Targetengagement of PKMT inhibitors can also be evaluated indirectly through their efficiency to block relevant methylation marks.(38) For canonical SAM-competitive and allosteric PKMT inhibitors, complete target engagementsimply requires to show a dose-dependent decrease of relevant methylation marks (e.g. H3K27me3 for EZH1/2).(38) For canonical SAM-competitive or allosteric MOAs, lack of PKMT-dependent methylationoften indicates that the current dose is sufficient for these inhibitors to fully occupy the SAM-binding pocket or allosteric siteof the targeted PKMT to prevent catalysis. In contrast, complete target engagement of a substrate-competitive PKMT inhibitorshould be evaluated with the most robust substrate (the highest intracellular concentration and the lowestK_m,substrate). In addition, the depletion of specific methylation marks also depends on thelifetime of the examined protein substrates and antagonistic activities of demethylases. A longer period of target engagement isrequired for more stable methylation marks to be suppressed.(38)

It is worth noting that, besides the inhibition of methyltransferase activities, the engagement of PKMT inhibitors withtheir targets can lead to conformational changes of PKMTs and thus alter other binding events. While substrate-competitive PKMTinhibitors perturb the catalysis through preventing PKMTs from binding their substrates, such perturbation is also expected toalter the integrity of the whole PKMT complex (e.g. MM-401 disrupts the interaction of WDR5 with MLL1 in theMLL1–WDR5–ASH2L–RbBP5 complex).(81) Therefore, the treatment of PKMTinhibitors can alter components of PKMT complexes and have profound effects beyond the only inhibition of substratemethylation.

9.1. Cysteine-specific chemical conjugation.

The free-thiol of cysteine in cysteine-containing proteins is often explored for site-specific chemical installation oflysine, methyllysine or their analogs. With the free-thiol of cysteine as a nucleophile, the Shokat laboratory first reported asemisynthetic approach to conjugate anN-methyl aminoethyl moiety to proteins (Figure 44).(298) The resultingN-methylatedaminoethylcysteines are generated as methyllysine analogs (MLA) (Figure 44). The overallsimilarity between methyllysine and MLA has been confirmed by their equivalent recognition as epitopes by anti-methyllysineantibodies and methyllysine binders, and as substrates of multiple PKMTs.(298) Forinstance, Margueronet. al. relied on this MLA approach to prepare a series of MLA-containing histones and usedthem as substrates to examine the crosstalk between the PRC2 complex (EED-Suz12-EZH2) and various histone methylation marks(e.g., H3K27, H3K36, and H3K9).(213) This work demonstrated that theEED subunit of the PRC2 complex enhances EZH2’s catalysis by interacting with nucleosomes containing the MLAs of H3K27me3and H3K9me1/2/3 but not H3K36me1/2/3.(213) In a more systemic manner, the Zhu laboratoryevaluated biochemical compatibility of MLA-containing histones including their interactions with the reader proteins 53BP1 and theG9a ankyrin repeats, and their reactivities as the substrates of 6 PKMTs and one KDM.(299)In comparison with natural methyllysine, the replacement of the γ methylene moiety with the sulfide results in a slightincrease of the corresponding bond length by 0.28 Å and a small decrease of pKa by 1.1 unit.(299) Such difference seems to have more impact upon quantitative evaluation of MLAs asmethyllysine surrogates, in which natural methyllysine is often better than MLAs to engage binding partners.(300)

The Davis laboratory recently developed a cysteine-specific, radical-based semisynthetic method to incorporatemethyllysine as well as other side chains into proteins.(301) On the basis of the priorwork of 2,5-dibromohexanediamide (DBHDA) as a double-activated electrophile, the coworkers in the Davis laboratory converted acysteine residue in proteins into dehydroalanine (Dha) (Figure 44).(301,302) The resultant Dha-containing protein was thensubjected to NaBH₄-mediated radical coupling with organic iodides to afford the corresponding methyllysine-containingproducts (Figure 44). With the newly-developed semisynthetic method, they generated a widerange of methyllysine-containing proteins. With H3K9me3-containing H3 as a demonstration, they showed that the semisynthetichistone can serve as a substrate of KDM4A. This semisynthetic method to access diverse methyllysine-containing proteins featuresits readiness, robustness and scalability. Yang and coworkers independently reported a method to convert Dha into methyllysinethrough a Zn/Cu-mediated radical coupling reaction (Figure 44).(303) Here a Dha residue is generated from Cys byO-mesitylenesulfonylhydroxylamine(MSH), which can be accessed chemically or introduced genetically in a site-specific manner through unnatural amino acidprecursors (see discussion below) (Figures 44,45).(303) In comparison with the DBHDA/MSH-mediated production of Dha from anatural cysteine, site-specific incorporation of unnatural amino acid precursors allows installing methyllysine without influenceof other existing cysteine residues.(303) One limitation for this radical-basedsemisynthetic method lies in the introduction of aD,L-methyllysine mixture rather thannaturally occurringL-methyllysine. In addition, this method has not demonstrated its utility in living cellsgiven relatively harsh conditions for the radical reactions.

9.2. Incorporation of methyllysine precursors through nonsense suppressors.

Unnatural amino acids can be incorporated into proteins in a site-specific manner with orthogonally engineered tRNA/tRNAsynthetase pairs.(27) This method has been well documented for its utility in a cell-freetranslational system, within bacterial, fungal and mammalian cells, and in living animals.(304,305) Given the challenge of recognizing the small difference betweenmethyllysine and free lysine by cellular translational machinery, site-specific incorporation of methyllysine into proteins isoften achieved by installing methyllysine precursors or caged methyllysine, followed by additional chemical modifications to yieldmethyllysine-containing proteins (Figure 45). For instance, the Schultz laboratory reportedthe work to prepare proteins containing MLA through site-specific phenylselenocysteine (SecPh) chemistry (Figure 45).(306) Here SecPh was incorporated into proteinsthrough a nonsense suppressor coupled with orthogonal tRNA/tRNA synthetase pairs. The SecPh was then subjected toH₂O₂-mediated oxidation to afford Dha.(306) In this situation,the authors reacted the Dha-containing protein withN-methylated 2-mercaptoethylamine to generate thecorresponding MLA-containing product (Figure 45). Yang and coworkers recently reported a newsynthetic path to access Dha-containing proteins.(303) Distinct from the DBHDA-mediatedproduction of Dha from cysteine (the Davis laboratory)(301) and theH₂O₂-mediated production of Dha from SecPh (the Schultz laboratory),(306) Yang and coworkers site-specifically introduced anO-phosphoserine (Sep) with orthogonallyengineered tRNA/tRNA synthetase pairs (Figure 45).(303) The Sep residue was then chemically eliminated to afford Dha under basic conditions. Theoretically, theseDha-containing protein intermediates can be subjected to eitherN-methylated 2-mercaptoethylamine to affordMLA-containing protein products or the transition-metal-mediated radical coupling with the corresponding organic iodides to affordmethyllysine-containing protein products (Figures 44,45). It is worth noting that the Dha-mediated chemistry always affords theD,Lisomers of methyllysine or MLAs and needs to be performedin vitro with denatured proteins.

To access proteins containing enantiomerically pure methylysine, the Liu laboratory recently reported theamber-suppressor-mediated incorporation ofN^ε-(4-azidobenzoxycarbonyl)-δ,ε-dehydrolysine (AcdK) as a methyllysineprecursor.(307) AcdK contains a liable δ,ε-dehydrolysine side chain,which is protected by an azidobenzoxycarbonyl moiety (Figure 45). This moiety is readilysubjected to phosphine-triggered self-cleavage to afford allysine (AlK) through an unstable intermediateδ,ε-dehydrolysine (Figure 45).(307)The subsequent NaCNBH₃-mediated reductive amination of AlK with monomethylamine or dimethylamine yields correspondingmethyllysine-containing proteins. In comparison with the Dha-mediated incorporation of methyllysine and MLAs, the AcdK-AlK-basedmethod is distinct for its feature to install enantiomerically pure methyllysine (Figure45). Given the requirement of the reductive amination upon the conversion of AlK into methyllysine, this approach can onlybe implemented for installation of Kme1 and Kme2 but not Kme3 into proteins (Figure45).(307) Because of involvement of multiple chemical steps under biologicallyincompatible conditions, the AcdK-AlK-based semisynthetic method, similar to Dha-mediated incorporation of methyllysine, wasdeveloped forin vitro biochemical settings.

Another broadly used method to access methyllysine-containing proteins relies on amber-suppressor-mediated incorporationof caged methyllysine precursors, followed by chemical removal of the caging moieties (Figure45). To prepare recombinant proteins containing site-specific mononmethyllysine, the Schultz laboratory and the Liulaboratory documentedN-allyloxycarbonyl(308) andN-(ortho-nitro)benzyl methyllysine,(309)respectively, as unnatural amino acid building blocks (Figure 45). With the aid of apyrrolysine(Pyl)-based genetic code expansion system, the two caged methyllysine building blocks can be incorporated into proteinsthrough site-specific nonsense suppressors. Methyllysine can also be masked by a carboxybenzyl (Cbz)(310) or atert-butyloxycarbonyl (Boc) group.(311,312) The resultant methyllysine is readily unmasked upon transition-metalor acid-triggered decaging reactions (Figure 45). In conjunction of developing natural aminoacid caging strategies, propargyloxycarboxyl,(313) derivatized(ortho-nitro)benzyl,(314) ortho-azidobenzyloxycarbonyl(315) and 1-(trans-cyclooctene)oxycarbonyl(316) moieties have been documented as lysine caging reagents.(317) Theresultant lysine derivatives can be readily recognized by corresponding engineered Pyl-tRNA synthetase (PylRS)-tRNA systems asnatural amino acids for protein biosynthesis. The lysine decaging can be achieved through transition-metal-, UV light-, phosphine-and tetrazine-based bioorthogonal cleavage chemistry, respectively. These lysine-caging strategies are expected to betransferrable for the site-specific incorporation of caged monomethyllysine (Kme1) into proteins (Figure 45). More importantly, the metal-, photo- and small-molecule-induced lysine decaging chemistry is bioorthogonaland biocompatible, and can be combined with amber-suppressor-mediated translation to afford Kme1-containing proteins inside livingcells.(317) The bioorthogonal cleavage chemistry also provides the feasibility torelease the caged lysine residues of functional importance and interrogate the associated biology with a temporal control.However, the lysine caging strategies can only be applied to install Kme1. To prepare proteins containing dimethyllysine with theamber-suppressor system, the site-specific dimethylation of target proteins has to be carried out in the context of globallyprotected lysine residues.(312) Recently, the Li laboratory reported a lysine analog witha photo-labile diazirine appended at the γ-position of lysine (Figure 45). Given thesimilarity between the diazirine lysine analog and free lysine, this lysine analog is caged as an unnatural amino acid and thencan be processed by orthogonally engineered tRNA/tRNA synthetase pairs for amber-suppressor-mediated protein synthesis.(318) The peptide containing the diazirine methyllysine analog can render photo-inducedcross-linking reactivity to covalently capture their binding partnersin vitro and inside living cells.(318,319)

9.3. Chemical ligation.

In comparison with cysteine-specific chemical conjugation and nonsense suppressor-mediated protein biosynthesis, chemicalligation features its ability to assemble a target protein containing multiple types of lysine methylation.(27) Native chemical ligation (NCL) and expressed protein ligation (EPL) are two chemical ligationmethods to study protein post-translational modifications including methylation (Figure46).(320) A key transformation of chemical ligation involves the couplingreaction between the C-terminal thioester of a peptide and an N-terminal cysteine-containing peptide (Figure 46). The residual cysteine, if unnecessary, can be converted into alanine throughdesulfurization.(320) A key difference between NCL and EPL is that the latter involvesan intein to generate a recombinant thioester intermediate and then catalyze the subsequent ligation reaction.

With the aid of NCL, the Muir laboratory is able to semi-synthesize nucleosomes containing more than 50 combinations ofmethylation with other posttranslational modifications.(321,322) These nucleosomes can be individually DNA-barcoded and used as baits to dissect their recognitionpartners by their effector proteins. With NCL strategies, the Danishefsky laboratory chemically synthesized several biologicallyimportant glycosylated proteins including erythropoietin and granulocyte-macrophage colony-stimulating factor.(323,324) Such efforts can be also readily transferred tochemically synthesize methyllysine-containing proteins with modest complexity. To apply the expressed protein ligation (EPL)inside living cells, the Muir laboratory recently reported protein semisynthesis using ultrafasttrans-splicinginteins (Figure 46).(325–327) Here they genetically fused the carboxylate end of the N-terminal fragment of a target proteinproduct with the N-terminal fragment of ultrafast split intein. The complementary C-terminal fragment of the ultrafast splitintein can be chemically synthesized and fused with the C-terminal fragment of the target protein product as well as acell-penetrating peptide. The cell-penetrating peptide drives the cellular uptake of the fused C-terminal intein and is cleavedintracellularly. The N-terminal intein and the C-terminal intein then form the functional splicing complex to ligate theN-terminal and C-terminal fragment to afford the target protein within living cells. Similar to NCL, installation of methyllysinecan be envisioned at the N-terminal fragment through a nonsense suppressor or at the C-terminal fragment through chemicalsynthesis. One limitation of NCL, EPL andtrans-splicing intein ligation is that these reactions occur at acysteine site or at sites with a cysteine as the precursor for desulfurization. Such a requirement may not be fulfilled for allprotein targets but can be avoided potentially with subtiligase as reported recently by the Cole laboratory (Figure 46).(328) Subtiligase is an engineered peptide ligasederived from the protease subtilisin through mutagenesis, which renders peptidase activity toward aminolysis and thus facilitatesthe coupling of a peptide containing activated C-terminal esters with a peptide containing an N-terminal α-amine moiety. Inaddition, it is more convenient to implement NCL, EPL andtrans-splicing intein ligation to install methyllysineat N-terminal or C-terminal region rather than in the central region of target proteins.

• (1).Reid MA, Dai Z, and Locasale JWThe Impact of Cellular Metabolism on Chromatin Dynamics and Epigenetics. Nat. CellBiol. 2017, 19, 1298–1306. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (2).Allis CD, and Jenuwein TThe Molecular Hallmarks of Epigenetic Control. Nat. Rev. Genet2016, 17, 487–500. [DOI] [PubMed] [Google Scholar]
• (3).Jakovcevski M, and Akbarian SEpigenetic Mechanisms in Neurological Disease. Nat. Med2012, 18, 1194–1204. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (4).Han S, and Brunet AHistone Methylation Makes its Mark on Longevity. Trends Cell Biol.2012, 22, 42–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (5).Flavahan WA, Gaskell E, and Bernstein BEEpigenetic Plasticity and the Hallmarks of Cancer. Science2017, 357, eaal2380. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (6).Kouzarides TChromatin Modifications and their Function. Cell2007, 128, 693–705. [DOI] [PubMed] [Google Scholar]
• (7).Lyko FThe DNA Methyltransferase Family: a Versatile Toolkit for Epigenetic Regulation. Nat. Rev.Genet2017, 19, 81–92. [DOI] [PubMed] [Google Scholar]
• (8).Tolstykh T, Lee J, Vafai S, and Stock JBCarboxyl Methylation Regulates Phosphoprotein Phosphatase 2A by Controlling the Association of Regulatory BSubunits. EMBO J. 2000, 19,5682–5691. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (9).Clarke SAging as War between Chemical and Biochemical Processes: Protein Methylation and the Recognition of Age-damagedProteins for Repair. Ageing Res. Rev. 2003, 2,263–285. [DOI] [PubMed] [Google Scholar]
• (10).Winter-Vann AM, and Casey PJPost-prenylation-processing Enzymes as New Targets in Oncogenesis. Nat. Rev.Cancer2005, 5, 405–412. [DOI] [PubMed] [Google Scholar]
• (11).Clarke SGProtein Methylation at the Surface and Buried Deep: Thinking Outside the Histone Box.Trends Biochem. Sci. 2013, 38,243–252. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (12).Biggar KK, and Li SSNon-histone Protein Methylation as a Regulator of Cellular Signalling and Function. Nat.Rev. Mol. Cell Biol. 2015, 16,5–17. [DOI] [PubMed] [Google Scholar]
• (13).Klotz AV, Thomas BA, Glazer AN, and Blacher RWDetection of Methylated Asparagine and Glutamine Residues in Polypeptides. Anal.Biochem1990, 186, 95–100. [DOI] [PubMed] [Google Scholar]
• (14).Sprung R, Chen Y, Zhang K, Cheng D, Zhang T, Peng J, and Zhao YIdentification and Validation of Eukaryotic Aspartate and Glutamate Methylation in Proteins.J. Proteome Res. 2008, 7,1001–1006. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (15).Blanc RS, and Richard SArginine Methylation: the Coming of Age. Mol. Cell2017, 65, 8–24. [DOI] [PubMed] [Google Scholar]
• (16).Nimura K, Ura K, and Kaneda YHistone Methyltransferases: Regulation of Transcription and Contribution to Human Disease.J. Mol. Med2010, 88, 1213–1220. [DOI] [PubMed] [Google Scholar]
• (17).Zhao BS, Roundtree IA, and He CPost-transcriptional Gene Regulation by mRNA Modifications. Nat. Rev. Mol. CellBiol. 2017, 18, 31–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (18).Fontecave M, Atta M, and Mulliez ES-Adenosylmethionine: Nothing Goes to Waste. Trends Biochem. Sci.2004, 29, 243–249. [DOI] [PubMed] [Google Scholar]
• (19).Wang R, Zheng W, and Luo MA Sensitive Mass Spectrum Assay to Characterize Engineered Methionine Adenosyltransferases withS-Alkyl Methionine Analogues as Substrates. Anal. Biochem2014, 450, 11–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (20).Klimasauskas S, and Weinhold EA New Tool for Biotechnology: AdoMet-dependent Methyltransferases. TrendsBiotechnol. 2007, 25, 99–104. [DOI] [PubMed] [Google Scholar]
• (21).Wang R, and Luo MA Journey toward Bioorthogonal Profiling of Protein Methylation inside Living Cells. Curr.Opin. Chem. Biol2013, 17, 729–737. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (22).Thomsen M, Vogensen SB, Buchardt J, Burkart MD, and Clausen RPChemoenzymatic Synthesis andin situ Application ofS-Adenosyl-L-MethionineAnalogs. Org. Biomol. Chem2013, 11, 7606–7610. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (23).Booker SJRadical SAM Enzymes and Radical Enzymology. Biochim. Biophys. Acta2012, 1824, 1151–1153. [DOI] [PubMed] [Google Scholar]
• (24).Lin HS3-Adenosylmethionine-Dependent Alkylation Reactions: When are Radical ReactionsUsed?Bioorg. Chem2011, 39, 161–170. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (25).Geske GD, O’Neill JC, and Blackwell HEExpanding Dialogues: From Natural Autoinducers to Non-Natural Analogues that Modulate Quorum Sensing inGram-negative Bacteria. Chem. Soc. Rev2008, 37, 1432–1447. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (26).Gutierrez JA, Luo M, Singh V, Li L, Brown RL, Norris GE, Evans GB, Furneaux RH, Tyler PC, Painter GF, et al. Picomolar Inhibitors as Transition-State Probes Of 5′-Methylthioadenosine Nucleosidases.ACS Chem. Biol. 2007, 2,725–734. [DOI] [PubMed] [Google Scholar]
• (27).Luo MCurrent Chemical Biology Approaches to Interrogate Protein Methyltransferases. ACS Chem.Biol. 2012, 7, 443–463. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (28).Richon VM, Johnston D, Sneeringer CJ, Jin L, Majer CR, Elliston K, Jerva LF, Scott MP, and Copeland RAChemogenetic Analysis of Human Protein Methyltransferases. Chem. Biol. Drug. Des2011, 78, 199–210. [DOI] [PubMed] [Google Scholar]
• (29).Copeland RA, Solomon ME, and Richon VMProtein Methyltransferases as a Target Class for Drug Discovery. Nat. Rev. DrugDiscov. 2009, 8, 724–732. [DOI] [PubMed] [Google Scholar]
• (30).Andrews FH, Strahl BD, and Kutateladze TGInsights into Newly Discovered Marks and Readers of Epigenetic Information. Nat. Chem.Biol2016, 12, 662–668. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (31).Wagner T, Robaa D, Sippl W, and Jung MMind the Methyl: Methyllysine Binding Proteins in Epigenetic Regulation.Chemmedchem2014, 9, 466–483. [DOI] [PubMed] [Google Scholar]
• (32).Musselman CA, Lalonde ME, Cote J, and Kutateladze TGPerceiving the Epigenetic Landscape through Histone Readers. Nat. Struct. Mol.Biol2012, 19, 1218–1227. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (33).Kaniskan HU, Martini ML, and Jin JInhibitors of Protein Methyltransferases and Demethylases. Chem. Rev2018, 118, 989–1068. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (34).Mcallister TE, England KS, Hopkinson RJ, Brennan PE, Kawamura A, and Schofield CJRecent Progress in Histone Demethylase Inhibitors. J. Med. Chem2016, 59, 1308–13029. [DOI] [PubMed] [Google Scholar]
• (35).Thinnes CC, England KS, Kawamura A, Chowdhury R, Schofield CJ, and Hopkinson RJTargeting Histone Lysine Demethylases - Progress, Challenges, and the Future. Biochim.Biophys. Acta2014, 1839, 1416–14132. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (36).Helin K, And Dhanak DChromatin Proteins and Modifications as Drug Targets. Nature2013, 502, 480–488. [DOI] [PubMed] [Google Scholar]
• (37).Kudithipudi S, and Jeltsch AApproaches and Guidelines for the Identification of Novel Substrates of Protein LysineMethyltransferases. Cell Chem. Biol. 2016, 23,1049–1055. [DOI] [PubMed] [Google Scholar]
• (38).Luo MInhibitors of Protein Methyltransferases as Chemical Tools. Epigenomics2015, 7, 1327–1338. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (39).Schneider R, Bannister AJ, and Kouzarides TUnsafe Sets: Histone Lysine Methyltransferases and Cancer. Trends Biochem. Sci.2002, 27, 396–402. [DOI] [PubMed] [Google Scholar]
• (40).Kaniskan HU, and Jin JRecent Progress in Developing Selective Inhibitors of Protein Methyltransferases. Curr.Opin. Chem. Biol2017, 39, 100–108. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (41).Egelhofer TA, Minoda A, Klugman S, Lee K, Kolasinska-Zwierz P, Alekseyenko AA, Cheung M-S, Day DS, Gadel S, Gorchakov AA, et al. An Assessment of Histone-modification Antibody Quality. Nat. Struct. Mol. Biol2011, 18, 91–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (42).Rothbart SB, Dickson BM, Raab JR, Grzybowski AT, Krajewski K, Guo AH, Shanle EK, Josefowicz SZ, Fuchs SM, Allis CD, et al. An Interactive Database for the Assessment of Histone Antibody Specificity. Mol.Cell2015, 59, 502–511. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (43).Andre I, Linse S, and Mulder FAResidue-specific pKa Determination of Lysine and Arginine Side Chains by Indirect 15N and 13C NMR Spectroscopy:Application to apo Calmodulin. J. Am. Chem. Soc. 2007, 129,15805–15813. [DOI] [PubMed] [Google Scholar]
• (44).Beaver JE, and Waters MLMolecular Recognition of Lys and Arg Methylation. ACS Chem. Biol.2016, 11, 643–653. [DOI] [PubMed] [Google Scholar]
• (45).Gallivan JP, and Dougherty DAA Computational Study of Cation-pi Interactions vs Salt Bridges in Aqueous Media: Implications for ProteinEngineering. J. Am. Chem. Soc2000, 122, 870–874. [Google Scholar]
• (46).Gallivan JP, and Dougherty DACation-pi Interactions in Structural Biology. Proc. Natl. Acad. Sci. USA1999, 96, 9459–9464. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (47).Seeler JS, and Dejean ASUMO and the Robustness of Cancer. Nat. Rev. Cancer2017, 17, 184–197. [DOI] [PubMed] [Google Scholar]
• (48).Sabari BR, Zhang D, Allis CD, and Zhao YMetabolic Regulation of Gene Expression Through Histone Acylations. Nat. Rev. Mol. CellBiol. 2017, 18, 90–101. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (49).Enchev RI, Schulman BA, and Peter MProtein Neddylation: Beyond Cullin-RING Ligases. Nat. Rev. Mol. Cell Biol.2015, 16, 30–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (50).Mattiroli F, and Sixma TKLysine-targeting Specificity in Ubiquitin and Ubiquitin-like Modification Pathways. Nat.Struct. Mol. Biol2014, 21, 308–316. [DOI] [PubMed] [Google Scholar]
• (51).Choudhary C, Weinert BT, Nishida Y, Verdin E, and Mann MThe Growing Landscape of Lysine Acetylation Links Metabolism and Cell Signalling. Nat. Rev.Mol. Cell Biol. 2014, 15,536–550. [DOI] [PubMed] [Google Scholar]
• (52).Kothapalli N, Camporeale G, Kueh A, Chew YC, Oommen AM, Griffin JB, and Zempleni JBiological Functions of Biotinylated Histones. J. Nutr. Biochem2005, 16, 446–448. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (53).Melcher M, Schmid M, Aagaard L, Selenko P, Laible G, and Jenuwein TStructure-function Analysis of SUV39H1 Reveals a Dominant Role in Heterochromatin Organization, ChromosomeSegregation, and Mitotic Progression. Mol. Cell. Biol2000, 20, 3728–3741. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (54).Min J, Feng Q, Li Z, Zhang Y, and Xu RMStructure of the Catalytic Domain of Human DOT1L, a Non-SET Domain Nucleosomal HistoneMethyltransferase. Cell2003, 112, 711–723. [DOI] [PubMed] [Google Scholar]
• (55).Bennett RL, Swaroop A, Troche C, and Licht JDThe Role of Nuclear Receptor-binding SET Domain Family Histone Lysine Methyltransferases in Cancer.Cold Spring Harb. Perspect Med. 2017, 7. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (56).Milite C, Feoli A, Viviano M, Rescigno D, Cianciulli A, Balzano AL, Mai A, Castellano S, and Sbardella GThe Emerging Role of Lysine Methyltransferase SETD8 in Human Diseases. Clin.Epigenetics2016, 8, 102. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (57).Liu Y, Liu K, Qin S, Xu C, and Min JEpigenetic Targets and Drug Discovery: Part 1: Histone Methylation. Pharmacol.Ther2014, 143, 275–294. [DOI] [PubMed] [Google Scholar]
• (58).Shimazu T, Barjau J, Sohtome Y, Sodeoka M, and Shinkai YSelenium-based S-Adenosylmethionine Analog Reveals the Mammalian Seven-β-strand Methyltransferase METTL10 tobe an EF1A1 Lysine Methyltransferase. Plos One2014, 9, E105394. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (59).Malecki J, Ho AY, Moen A, Dahl HA, and Falnes POHuman METTL20 is a Mitochondrial Lysine Methyltransferase that Targets the β-Subunit of Electron TransferFlavoprotein (ETFβ) and Modulates its Activity. J. Biol. Chem2015, 290, 423–434. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (60).Rhein VF, Carroll J, He J, Ding S, Fearnley IM, and Walker JEHuman METTL20 Methylates Lysine Residues adjacent to the Recognition Loop of the Electron Transfer Flavoprotein inMitochondria. J. Biol. Chem2014, 289, 24640–24651. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (61).Jakobsson ME, Moen A, Bousset L, Egge-Jacobsen W, Kernstock S, Melki R, and Falnes POIdentification and Characterization of a Novel Human Methyltransferase Modulating Hsp70 Protein Function throughLysine Methylation. J. Biol. Chem2013, 288, 27752–27763. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (62).Malecki J, Aileni VK, Ho AYY, Schwarz J, Moen A, Sorensen V, Nilges BS, Jakobsson ME, Leidel SA, and Falnes POThe Novel Lysine Specific Methyltransferase METTL21B Affects mRNA Translation through Inducible and DynamicMethylation of Lys-165 in Human Eukaryotic Elongation Factor 1 Alpha (Eef1a). Nucleic AcidsRes. 2017, 45, 4370–4389. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (63).Hamey JJ, Wienert B, Quinlan KGR, and Wilkins MRMETTL21B is a Novel Human Lysine Methyltransferase of Translation Elongation Factor 1A: Discovery by CRISPR/Cas9Knockout. Mol. Cell. Proteomics2017, 16, 2229–2242. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (64).Jakobsson ME, Davydova E, Malecki J, Moen A, and Falnes POSaccharomyces Cerevisiae Eukaryotic Elongation Factor 1A (Eef1a) is Methylated at Lys-390 by a METTL21-LikeMethyltransferase. Plos One2015, 10, E0131426. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (65).Kernstock S, Davydova E, Jakobsson M, Moen A, Pettersen S, Maelandsmo GM, Egge-Jacobsen W, and Falnes POLysine Methylation of VCP by a Member of a Novel Human Protein Methyltransferase Family.Nat. Commun2012, 3, 1038. [DOI] [PubMed] [Google Scholar]
• (66).Cloutier P, Lavallee-Adam M, Faubert D, Blanchette M, and Coulombe BA Newly Uncovered Group of Distantly Related Lysine Methyltransferases Preferentially Interact with MolecularChaperones to Regulate their Activity. Plos Genet. 2013, 9,E1003210. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (67).Cloutier P, Lavallee-Adam M, Faubert D, Blanchette M, and Coulombe BMethylation of the DNA/RNA-binding Protein Kin17 By METTL22 Affects its Association with Chromatin.J. Proteomics2014, 100, 115–124. [DOI] [PubMed] [Google Scholar]
• (68).Hamey JJ, Winter DL, Yagoub D, Overall CM, Hart-Smith G, and Wilkins MRNovel N-terminal and Lysine Methyltransferases that Target Translation Elongation Factor 1A in Yeast andHuman. Mol. Cell. Proteomics2016, 15, 164–176. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (69).Davydova E, Ho AY, Malecki J, Moen A, Enserink JM, Jakobsson ME, Loenarz C, and Falnes POIdentification and Characterization of a Novel Evolutionarily Conserved Lysine-specific Methyltransferase TargetingEukaryotic Translation Elongation Factor 2 (Eef2). J. Biol. Chem2014, 289, 30499–30510. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (70).Magen S, Magnani R, Haziza S, Hershkovitz E, Houtz R, Cambi F, and Parvari RHuman Calmodulin Methyltransferase: Expression, Activity on Calmodulin, and Hsp90 Dependence.Plos. One2012, 7, E52425. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (71).Qian C, and Zhou MMSET Domain Protein Lysine Methyltransferases: Structure, Specificity and Catalysis. Cell.Mol. Life Sci. 2006, 63,2755–2763. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (72).Vougiouklakis T, Hamamoto R, Nakamura Y, and Saloura VThe NSD Family of Protein Methyltransferases in Human Cancer. Epigenomics2015, 7, 863–874. [DOI] [PubMed] [Google Scholar]
• (73).Wu H, Min JR, Lunin VV, Antoshenko T, Dombrovski L, Zeng H, Allali-Hassani A, Campagna-Slater V, Vedadi M, Arrowsmith CH, et al. Structural Biology of Human H3K9 Methyltransferases. Plos One2010, 5, E8570. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (74).Tachibana M, Sugimoto K, Fukushima T, and Shinkai YSET Domain-containing Protein, G9a, is a Novel Lysine-preferring Mammalian Histone Methyltransferase withHyperactivity and Specific Selectivity to Lysines 9 and 27 of Histone H3. J. Biol. Chem.2001, 276, 25309–25317. [DOI] [PubMed] [Google Scholar]
• (75).Zheng W, Ibanez G, Wu H, Blum G, Zeng H, Dong A, Li F, Hajian T, Allali-Hassani A, Amaya MF, et al. Sinefungin Derivatives as Inhibitors and Structure Probes of Protein Lysine MethyltransferaseSETD2. J. Am. Chem. Soc2012, 134, 18004–18014. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (76).An S, Yeo KJ, Jeon YH, and Song JJCrystal Structure of the Human Histone Methyltransferase ASH1L Catalytic Domain and its Implications for theRegulatory Mechanism. J. Biol. Chem2011, 286, 8369–8374. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (77).Tisi D, Chiarparin E, Tamanini E, Pathuri P, Coyle JE, Hold A, Holding FP, Amin N, Martin AC, Rich SJ, et al. Structure of the Epigenetic Oncogene MMSET and Inhibition by N-Alkyl Sinefungin Derivatives.ACS Chem. Biol. 2016, 11,3093–3105. [DOI] [PubMed] [Google Scholar]
• (78).Yang S, Zheng X, Lu C, Li GM, Allis CD, and Li HMolecular Basis for Oncohistone H3 Recognition by SETD2 Methyltransferase. GenesDev. 2016, 30, 1611–1616. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (79).Grebien F, Vedadi M, Getlik M, Giambruno R, Grover A, Avellino R, Skucha A, Vittori S, Kuznetsova E, Smil D, et al. Pharmacological Targeting of the WDR5-MLL Interaction in C/EBPα N-terminal Leukemia.Nat. Chem. Biol2015, 11, 571–578. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (80).Borkin D, He S, Miao H, Kempinska K, Pollock J, Chase J, Purohit T, Malik B, Zhao T, Wang J, et al. Pharmacologic Inhibition of the Menin-MLL Interaction Blocks Progression of MLL Leukemia in vivo.Cancer Cell2015, 27, 589–602. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (81).Cao F, Townsend EC, Karatas H, Xu J, Li L, Lee S, Liu L, Chen Y, Ouillette P, Zhu J, et al. Targeting MLL1 H3K4 Methyltransferase Activity in Mixed-Lineage Leukemia. Mol.Cell2014, 53, 247–261. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (82).He Y, Selvaraju S, Curtin ML, Jakob CG, Zhu H, Comess KM, Shaw B, The J, Lima-Fernandes E, Szewczyk MM, et al. The EED Protein-protein Interaction Inhibitor A-395 Inactivates the PRC2 Complex. Nat.Chem. Biol2017, 13, 389–395. [DOI] [PubMed] [Google Scholar]
• (83).Kim W, Bird GH, Neff T, Guo G, Kerenyi MA, Walensky LD, And Orkin SHTargeted Disruption of the EZH2-EED Complex Inhibits EZH2-dependent Cancer. Nat. Chem.Biol2013, 9, 643–650. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (84).Mazur PK, Reynoird N, Khatri P, Jansen PW, Wilkinson AW, Liu S, Barbash O, Van Aller GS, Huddleston M, Dhanak D, et al. SMYD3 Links Lysine Methylation of MAP3K2 to Ras-driven Cancer. Nature2014, 510, 283–287. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (85).Sirinupong N, Brunzelle J, Doko E, and Yang ZStructural Insights into the Autoinhibition and Posttranslational Activation of Histone MethyltransferaseSmyd3. J. Mol. Biol2011, 406, 149–159. [DOI] [PubMed] [Google Scholar]
• (86).Ferguson AD, Larsen NA, Howard T, Pollard H, Green I, Grande C, Cheung T, Garcia-Arenas R, Cowen S, Wu J, et al. Structural Basis of Substrate Methylation and Inhibition of SMYD2. Structure2011, 19, 1262–1273. [DOI] [PubMed] [Google Scholar]
• (87).Sirinupong N, Brunzelle J, Ye J, Pirzada A, Nico L, and Yang ZCrystal Structure of Cardiac-specific Histone Methyltransferase SMYD1 Reveals Unusual Active SiteArchitecture. J. Biol. Chem2010, 285. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (88).Saddic LA, West LE, Aslanian A, Yates JR, Rubin SM, Gozani O, and Sage JMethylation of the Retinoblastoma Tumor Suppressor by SMYD2. J. Biol. Chem2010, 285, 37733–37740. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (89).Mzoughi S, Tan YX, Low D, and Guccione EThe Role of PRDMs in Cancer: One Family, Two Sides. Curr. Opin. Genet. Dev2016, 36, 83–91. [DOI] [PubMed] [Google Scholar]
• (90).Justin N, Zhang Y, Tarricone C, Martin SR, Chen S, Underwood E, De Marco V, Haire LF, Walker PA, Reinberg D, et al. Structural Basis of Oncogenic Histone H3K27M Inhibition of Human Polycomb Repressive Complex 2.Nat. Commun2016, 7, 11316. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (91).Li Y, Han J, Zhang Y, Cao F, Liu Z, Li S, Wu J, Hu C, Wang Y, Shuai J, et al. Structural Basis for Activity Regulation of MLL Family Methyltransferases. Nature2016, 530, 447–452. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (92).Hsieh JJ, Cheng EH, and Korsmeyer SJTaspase1: A Threonine Aspartase Required for Cleavage of MLL and Proper HOX Gene Expression.Cell2003, 115, 293–303. [DOI] [PubMed] [Google Scholar]
• (93).Brooun A, Gajiwala KS, Deng YL, Liu W, Bolanos B, Bingham P, He YA, Diehl W, Grable N, Kung PP, et al. Polycomb Repressive Complex 2 Structure with Inhibitor Reveals a Mechanism of Activation and DrugResistance. Nat. Commun2016, 7, 11384. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (94).Jiao L, and Liu XStructural Basis of Histone H3K27 Trimethylation by an Active Polycomb Repressive Complex 2.Science2015, 350, Aac4383. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (95).Shinsky SA, Hu M, Vought VE, Ng SB, Bamshad MJ, Shendure J, and Cosgrove MSA Non-active-site SET Domain Surface Crucial for the Interaction of MLL1 and the Rbbp5/Ash2L Heterodimer within MLLFamily Core Complexes. J. Mol. Biol2014, 426, 2283–2299. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (96).Patel A, Vought VE, Swatkoski S, Viggiano S, Howard B, Dharmarajan V, Monteith KE, Kupakuwana G, Namitz KE, Shinsky SA, et al. Automethylation Activities within the Mixed Lineage Leukemia-1 (MLL1) Core Complex Reveal Evidence Supporting a“Two-Active Site” Model for Multiple Histone H3 Lysine 4 Methylation. J. Biol.Chem. 2014, 289, 868–884. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (97).Patel A, Vought VE, Dharmarajan V, and Cosgrove MSA Novel Non-SET Domain Multi-subunit Methyltransferase Required for Sequential Nucleosomal Histone H3 Methylationby the Mixed Lineage Leukemia Protein-1 (MLL1) Core Complex. J. Biol. Chem2011, 286, 3359–3369. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (98).Linscott JA, Kapilashrami K, Wang Z, Senevirathne C, Bothwell IR, Blum G, And Luo MKinetic Isotope Effects Reveal Early Transition State of Protein Lysine Methyltransferase SET8.Proc. Natl. Acad. Sci. USA2016, 113, 8369–8378. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (99).Islam K, Chen Y, Wu H, Bothwell IR, Blum GJ, Zeng H, Dong A, Zheng W, Min J, Deng H, et al. Defining Efficient Enzyme-cofactor Pairs for Bioorthogonal Profiling of Protein Methylation.Proc. Natl. Acad. Sci. USA2013, 110, 16778–16783. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (100).Ibáñez G, Mcbean JL, Astudillo YM, and Luo MAn Enzyme-coupled Ultrasensitive Luminescence Assay for Protein Methyltransferases. Anal.Biochem. 2010, 401, 203–210. [DOI] [PubMed] [Google Scholar]
• (101).Schapira MStructural Chemistry of Human SET Domain Protein Methyltransferases. Curr. Chem.Genomics2011, 5, 85–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (102).Campagna-Slater V, Mok MW, Nguyen KT, Feher M, Najmanovich R, and Schapira MStructural Chemistry of the Histone Methyltransferases Cofactor Binding Site. J. Chem. Inf.Model2011, 51, 612–623. [DOI] [PubMed] [Google Scholar]
• (103).Wilson JR, Jing C, Walker PA, Martin SR, Howell SA, Blackburn GM, Gamblin SJ, and Xiao BCrystal Structure and Functional Analysis of the Histone Methyltransferase SET7/9.Cell2002, 111, 105–115. [DOI] [PubMed] [Google Scholar]
• (104).Kwon T, Chang JH, Kwak E, Lee CW, Joachimiak A, Kim YC, Lee JW, and Cho YJMechanism of Histone Lysine Methyl Transfer Revealed by the Structure of SET7/9-AdoMet.EMBO J. 2003, 22,292–303. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (105).Xiao B, Jing C, Wilson JR, Walker PA, Vasisht N, Kelly G, Howell S, Taylor IA, Blackburn GM, and Gamblin SJStructure and Catalytic Mechanism of the Human Histone Methyltransferase SET7/9.Nature2003, 421, 652–656. [DOI] [PubMed] [Google Scholar]
• (106).Couture JF, Collazo E, Hauk G, and Trievel RCStructural Basis for the Methylation Site Specificity of SET7/9. Nat. Struct. Mol.Biol2006, 13, 140–146. [DOI] [PubMed] [Google Scholar]
• (107).Fang J, Feng Q, Ketel CS, Wang HB, Cao R, Xia L, Erdjument-Bromage H, Tempst P, Simon JA, and Zhang YPurification and Functional Characterization of SET8, a Nucleosomal Histone H4-Lysine 20-specificMethyltransferase. Curr. Biol2002, 12, 1086–1099. [DOI] [PubMed] [Google Scholar]
• (108).Couture JF, Collazo E, Brunzelle JS, and Trievel RCStructural and Functional Analysis of SET8, a Histone H4 Lys-20 Methyltransferase. GenesDev. 2005, 19, 1455–1465. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (109).Xiao B, Jing C, Kelly G, Walker PA, Muskett FW, Frenkiel TA, Martin SR, Sarma K, Reinberg D, Gamblin SJ, et al. Specificity and Mechanism of the Histone Methyltransferase Pr-Set7. Genes Dev.2005, 19, 1444–1454. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (110).Yao Y, Chen P, Diao J, Cheng G, Deng L, Anglin JL, Prasad BVV, and Song YSelective Inhibitors of Histone Methyltransferase DOT1L: Design, Synthesis, and CrystallographicStudies. J. Am. Chem. Soc2011, 133, 16746–16749. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (111).Poulin MB, Schneck JL, Matico RE, Mcdevitt PJ, Huddleston MJ, Hou W, Johnson NW, Thrall SH, Meek TD, and Schramm VLTransition State for the NSD2-catalyzed Methylation of Histone H3 Lysine 36. Proc. Natl.Acad. Sci. USA2016, 113, 1197–1201. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (112).Poulin MB, Schneck JL, Matico RE, Hou W, Mcdevitt PJ, Holbert M, and Schramm VLNucleosome Binding Alters the Substrate Bonding Environment of Histone H3 Lysine 36 MethyltransferaseNSD2. J. Am. Chem. Soc2016, 138, 6699–6702. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (113).Horowitz S, Yesselman JD, Al-Hashimi HM, and Trievel RCDirect Evidence for Methyl Group Coordination by Carbon-oxygen Hydrogen Bonds in the Lysine MethyltransferaseSET7/9. J. Biol. Chem2011, 286, 18658–18663. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (114).Horowitz S, and Trievel RCCarbon-oxygen Hydrogen Bonding in Biological Structure and Function. J. Biol.Chem2012, 287, 41576–41582. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (115).Horowitz S, Dirk LM, Yesselman JD, Nimtz JS, Adhikari U, Mehl RA, Scheiner S, Houtz RL, Al-Hashimi HM, and Trievel RCConservation and Functional Importance of Carbon-oxygen Hydrogen Bonding in AdoMet-dependentMethyltransferases. J. Am. Chem. Soc2013, 135, 15536–15548. [DOI] [PubMed] [Google Scholar]
• (116).Horowitz S, Adhikari U, Dirk LM, Del Rizzo PA, Mehl RA, Houtz RL, Al-Hashimi HM, Scheiner S, and Trievel RCManipulating Unconventional CH-based Hydrogen Bonding in a Methyltransferase via Noncanonical Amino AcidMutagenesis. ACS Chem. Biol2014, 9, 1692–1697. [DOI] [PubMed] [Google Scholar]
• (117).Fick RJ, Kroner GM, Nepal B, Magnani R, Horowitz S, Houtz RL, Scheiner S, and Trievel RCSulfur-oxygen Chalcogen Bonding Mediates AdoMet Recognition in the Lysine Methyltransferase SET7/9.ACS Chem. Biol2016, 11, 748–754. [DOI] [PubMed] [Google Scholar]
• (118).Mcginty RK, Kohn M, Chatterjee C, Chiang KP, Pratt MR, and Muir TWStructure-activity Analysis of Semisynthetic Nucleosomes: Mechanistic Insights into the Stimulation of Dot1L byUbiquitylated Histone H2B. ACS Chem. Biol2009, 4, 958–968. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (119).Huang J, Hsu YH, Mo C, Abreu E, Kiel DP, Bonewald LF, Brotto M, and Karasik DMETTL21C is a Potential Pleiotropic Gene for Osteoporosis and Sarcopenia Acting through the Modulation of theNFκb Signaling Pathway. J. Bone Miner Res2014, 29, 1531–1540. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (120).Kurash JK, Lei H, Shen Q, Marston WL, Granda BW, Fan H, Wall D, Li E, And Gaudet FMethylation of p53 by Set7/9 Mediates p53 Acetylation and Activity in vivo. Mol.Cell2008, 29, 392–400. [DOI] [PubMed] [Google Scholar]
• (121).Pradhan S, Chin HG, Esteve PO, and Jacobsen SESET7/9 Mediated Methylation of Non-histone Proteins in Mammalian Cells.Epigenetics2009, 4, 282–285. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (122).Kontaki H, and Talianidis ILysine Methylation Regulates E2F1-induced Cell Death. Mol Cell2010, 39, 152–160. [DOI] [PubMed] [Google Scholar]
• (123).Dhayalan A, Kudithipudi S, Rathert P, and Jeltsch ASpecificity Analysis-based Identification of New Methylation Targets of the SET7/9 Protein LysineMethyltransferase. Chem. Biol2011, 18, 111–120. [DOI] [PubMed] [Google Scholar]
• (124).Yang J, Huang J, Dasgupta M, Sears N, Miyagi M, Wang B, Chance MR, Chen X, Du Y, Wang Y, et al. Reversible Methylation of Promoter-bound STAT3 by Histone-modifying Enzymes. Proc. Natl.Acad. Sci. USA2010, 107, 21499–21504. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (125).Gaughan L, Stockley J, Wang N, Mccracken SR, Treumann A, Armstrong K, Shaheen F, Watt K, Mcewan IJ, Wang C, et al. Regulation of the Androgen Receptor by SET9-mediated Methylation. Nucleic AcidsRes. 2011, 39, 1266–1279. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (126).Ko S, Ahn J, Song CS, Kim S, Knapczyk-Stwora K, and Chatterjee BLysine Methylation and Functional Modulation of Androgen Receptor by Set9 Methyltransferase.Mol. Endocrinol2011, 25, 433–444. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (127).Barsyte-Lovejoy D, Li F, Oudhoff MJ, Tatlock JH, Dong A, Zeng H, Wu H, Freeman SA, Schapira M, Senisterra GA, et al. (R)-PFI-2 is a Potent and Selective Inhibitor of SETD7 Methyltransferase Activity inCells. Proc. Natl. Acad. Sci. USA2014, 111, 12853–12858. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (128).Liu X, Wang D, Zhao Y, Tu B, Zheng Z, Wang L, Wang H, Gu W, Roeder RG, and Zhu WGMethyltransferase Set7/9 Regulates p53 Activity by Interacting with Sirtuin 1 (SIRT1).Proc. Natl. Acad. Sci. USA2011, 108, 1925–1930. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (129).Xie Q, Hao Y, Tao L, Peng S, Rao C, Chen H, You H, Dong MQ, and Yuan ZLysine Methylation of FOXO3 Regulates Oxidative Stress-Induced Neuronal Cell Death. EMBORep. 2012, 13, 371–377. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (130).Calnan DR, Webb AE, White JL, Stowe TR, Goswami T, Shi X, Espejo A, Bedford MT, Gozani O, Gygi SP, et al. Methylation by Set9 Modulates Foxo3 Stability and Transcriptional Activity. Aging2012, 4, 462–479. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (131).Munro S, Khaire N, Inche A, Carr S, and La Thangue NBLysine Methylation Regulates the pRb Tumour Suppressor Protein. Oncogene2010, 29, 2357–2367. [DOI] [PubMed] [Google Scholar]
• (132).Yang XD, Huang B, Li MX, Lamb A, Kelleher NL, and Chen LFNegative Regulation of NF-Kappa B Action by Set9-mediated Lysine Methylation of the RelA Subunit.EMBO. J2009, 28, 1055–1066. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (133).Patnaik D, Chin HG, Esteve PO, Benner J, Jacobsen SE, and Pradhan SSubstrate Specificity and Kinetic Mechanism of Mammalian G9a Histone H3 Methyltransferase.J. Biol. Chem2004, 279, 53248–53258. [DOI] [PubMed] [Google Scholar]
• (134).Chin HG, Pradhan M, Esteve PO, Patnaik D, Evans TC Jr., and Pradhan SSequence Specificity and Role of Proximal Amino Acids of the Histone H3 Tail on Catalysis of Murine G9a Lysine 9Histone H3 Methyltransferase. Biochemistry2005, 44, 12998–13006. [DOI] [PubMed] [Google Scholar]
• (135).Chang Y, Sun L, Kokura K, Horton JR, Fukuda M, Espejo A, Izumi V, Koomen JM, Bedford MT, Zhang X, et al. MPP8 Mediates the Interactions Between DNA Methyltransferase Dnmt3a and H3K9 MethyltransferaseGLP/G9a. Nat. Commun2011, 2, 533. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (136).Huang J, Dorsey J, Chuikov S, Zhang XY, Jenuwein T, Reinberg D, and Berger SLG9a and Glp Methylate Lysine 373 in the Tumor Suppressor P53. J. Biol. Chem2010, 285, 9636–9641. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (137).Pless O, Kowenz-Leutz E, Knoblich M, Lausen J, Beyermann M, Walsh MJ, and Leutz AG9a-mediated Lysine Methylation Alters the Function of CCAAT/Enhancer-Binding Protein β.J. Biol. Chem2008, 283, 26357–26363. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (138).Lee JS, Kim Y, Kim IS, Kim B, Choi HJ, Lee JM, Shin HJR, Kim JH, Kim JY, Seo SB, et al. Negative Regulation of Hypoxic Responses via Induced Reptin Methylation. Mol. Cell2010, 39, 71–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (139).Ling BM, Bharathy N, Chung TK, Kok WK, Li S, Tan YH, Rao VK, Gopinadhan S, Sartorelli V, Walsh MJ, et al. Lysine Methyltransferase G9a Methylates the Transcription Factor MyoD and Regulates Skeletal MuscleDifferentiation. Proc. Natl. Acad. Sci. USA2012, 109, 841–846. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (140).Carlson SM, Moore KE, Green EM, Martin GM, and Gozani OProteome-wide Enrichment of Proteins Modified by Lysine Methylation. Nat. Protoc2014, 9, 37–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (141).Carlson SM, and Gozani OEmerging Technologies to Map the Protein Methylome. J. Mol. Biol2014, 426, 3350–3362. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (142).Moore KE, Carlson SM, Camp ND, Cheung P, James RG, Chua KF, Wolf-Yadlin A, and Gozani OA General Molecular Affinity Strategy for Global Detection and Proteomic Analysis of LysineMethylation. Mol Cell. 2013, 50,444–456. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (143).Islam K, Bothwell I, Chen Y, Sengelaub C, Wang R, Deng H, and Luo MBioorthogonal Profiling of Protein Methylation Using Azido Derivative of S-Adenosyl-L-methionine.J. Am. Chem. Soc2012, 134, 5909–5915. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (144).Rathert P, Dhayalan A, Murakami M, Zhang X, Tamas R, Jurkowska R, Komatsu Y, Shinkai Y, Cheng XD, and Jeltsch AProtein Lysine Methyltransferase G9a Acts on Non-histone Targets. Nat. Chem. Biol2008, 4, 344–346. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (145).Takawa M, Cho HS, Hayami S, Toyokawa G, Kogure M, Yamane Y, Iwai Y, Maejima K, Ueda K, Masuda A, et al. Histone Lysine Methyltransferase SETD8 Promotes Carcinogenesis by Deregulating PCNA Expression.Cancer Res. 2012, 72,3217–3227. [DOI] [PubMed] [Google Scholar]
• (146).Dhami GK, Liu H, Galka M, Voss C, Wei R, Muranko K, Kaneko T, Cregan SP, Li L, and Li SSDynamic Methylation of Numb by Set8 Regulates its Binding to p53 and Apoptosis. Mol.Cell2013, 50, 565–576. [DOI] [PubMed] [Google Scholar]
• (147).Knutson SK, Wigle TJ, Warholic NM, Sneeringer CJ, Allain CJ, Klaus CR, Sacks JD, Raimondi A, Majer CR, Song J, et al. A Selective Inhibitor of EZH2 Blocks H3K27 Methylation and Kills Mutant Lymphoma Cells.Nat. Chem. Biol2012, 8, 890–896. [DOI] [PubMed] [Google Scholar]
• (148).Lee JM, Lee JS, Kim H, Kim K, Park H, Kim JY, Lee SH, Kim IS, Kim J, Lee M, et al. EZH2 Generates a Methyl Degron that is Recognized by the DCAF1/DDB1/CUL4 E3 Ubiquitin LigaseComplex. Mol. Cell2012, 48, 572–786. [DOI] [PubMed] [Google Scholar]
• (149).Kim E, Kim M, Woo DH, Shin Y, Shin J, Chang N, Oh YT, Kim H, Rheey J, Nakano I, et al. Phosphorylation of EZH2 Activates STAT3 Signaling via STAT3 Methylation and Promotes Tumorigenicity of GlioblastomaStem-like Cells. Cancer Cell2013, 23, 839–852. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (150).Dasgupta M, Dermawan JK, Willard B, and Stark GRSTAT3-driven Transcription Depends upon the Dimethylation of K49 by EZH2. Proc. Natl. Acad.Sci. USA2015, 112, 3985–3990. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (151).Clancy KW, Russell AM, Subramanian V, Nguyen H, Qian Y, Campbell RM, and Thompson PRCitrullination/Methylation Crosstalk on Histone H3 Regulates ER-target Gene Transcription.ACS Chem. Biol2017, 12, 1691–1702. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (152).Del Rizzo PA, and Trievel RCSubstrate and Product Specificities of SET Domain Methyltransferases. Epigenetics2011, 6, 1059–1067. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (153).Zhang X, Huang Y, and Shi XEmerging Roles of Lysine Methylation on Non-histone Proteins. Cell Mol. Life Sci2015, 72, 4257–4272. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (154).Shi XB, Kachirskaia L, Yamaguchi H, West LE, Wen H, Wang EW, Dutta S, Appella E, and Gozani OModulation of p53 Function by SET8-mediated Methylation at Lysine 382. Mol. Cell2007, 27, 636–646. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (155).Fu W, Liu N, Qiao Q, Wang M, Min J, Zhu B, Xu RM, and Yang NStructural Basis for Substrate Preference of SMYD3, a SET Domain-containing Protein LysineMethyltransferase. J. Biol. Chem2016, 291, 9173–9180. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (156).Lanouette S, Davey JA, Elisma F, Ning Z, Figeys D, Chica RA, and Couture JFDiscovery of Substrates for a SET Domain Lysine Methyltransferase Predicted by Multistate Computational ProteinDesign. Structure2015, 23, 206–215. [DOI] [PubMed] [Google Scholar]
• (157).Boriack-Sjodin PA, Jin L, Jacques SL, Drew A, Sneeringer C, Scott MP, Moyer MP, Ribich S, Moradei O, and Copeland RAStructural Insights into Ternary Complex Formation of Human CARM1 with Various Substrates.ACS Chem. Biol2016, 11, 763–771. [DOI] [PubMed] [Google Scholar]
• (158).Chatterjee C, Mcginty RK, Fierz B, and Muir TWDisulfide-directed Histone Ubiquitylation Reveals Plasticity in hDOT1L Activation. Nat.Chem. Biol2010, 6, 267–269. [DOI] [PubMed] [Google Scholar]
• (159).Mcginty RK, Kim J, Chatterjee C, Roeder RG, and Muir TWChemically Ubiquitylated Histone H2B Stimulates hDOT1L-mediated Intranucleosomal Methylation.Nature2008, 453, 812–816. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (160).Wu H, Siarheyeva A, Zeng H, Lam R, Dong A, Wu XH, Li Y, Schapira M, Vedadi M, and Min JCrystal Structures of the Human Histone H4K20 Methyltransferases SUV420H1 and SUV420H2.FEBS Lett. 2013, 587,3859–3868. [DOI] [PubMed] [Google Scholar]
• (161).Sneeringer CJ, Scott MP, Kuntz KW, Knutson SK, Pollock RM, Richon VM, and Copeland RACoordinated Activities of Wild-type plus Mutant EZH2 Drive Tumor-associated Hypertrimethylation of Lysine 27 onHistone H3 (H3K27) in Human B-Cell Lymphomas. Proc. Natl. Acad. Sci. USA2010, 107, 20980–20985. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (162).Wigle TJ, Knutson SK, Jin L, Kuntz KW, Pollock RM, Richon VM, Copeland RA, and Scott MPThe Y641C Mutation of EZH2 Alters Substrate Specificity for Histone H3 Lysine 27 MethylationStates. FEBS Lett. 2011, 585,3011–3014. [DOI] [PubMed] [Google Scholar]
• (163).Majer CR, Jin L, Scott MP, Knutson SK, Kuntz KW, Keilhack H, Smith JJ, Moyer MP, Richon VM, Copeland RA, et al. A687V EZH2 is a Gain-of-function Mutation Found in Lymphoma Patients. FEBS Lett.2012, 586, 3448–3451. [DOI] [PubMed] [Google Scholar]
• (164).Mccabe MT, Graves AP, Ganji G, Diaz E, Halsey WS, Jiang Y, Smitheman KN, Ott HM, Pappalardi MB, Allen KE,et al. Mutation of A677 in Histone Methyltransferase EZH2 in Human B-Cell Lymphoma Promotes Hypertrimethylation of HistoneH3 on Lysine 27 (H3K27). Proc. Natl. Acad. Sci. USA2012, 109, 2989–2994. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (165).Ott HM, Graves AP, Pappalardi MB, Huddleston M, Halsey WS, Hughes AM, Groy A, Dul E, Jiang Y, Bai Y, et al. A687V EZH2 is a Driver of Histone H3 Lysine 27 (H3K27) Hypertrimethylation. Mol. CancerTher2014, 13, 3062–3073. [DOI] [PubMed] [Google Scholar]
• (166).Trievel RC, Beach BM, Dirk LMA, Houtz RL, and Hurley JHStructure and Catalytic Mechanism of a SET Domain Protein Methyltransferase. Cell2002, 111, 91–103. [DOI] [PubMed] [Google Scholar]
• (167).Zhang X, and Bruice TCHistone Lysine Methyltransferase SET7/9: Formation of a Water Channel Precedes Each MethylTransfer. Biochemistry2007, 46, 14838–14844. [DOI] [PubMed] [Google Scholar]
• (168).Zhang X, and Bruice TCProduct Specificity and Mechanism of Protein Lysine Methyltransferases: Insights from the Histone LysineMethyltransferase SET8. Biochemistry2008, 47, 6671–6677. [DOI] [PubMed] [Google Scholar]
• (169).Zhang X, and Bruice TCEnzymatic Mechanism and Product Specificity of SET-domain Protein Lysine Methyltransferases.Proc. Natl. Acad. Sci. USA2008, 105, 5728–5732. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (170).Zhang X, and Bruice TCMechanism of Product Specificity of Adomet Methylation Catalyzed by Lysine Methyltransferases: TranscriptionalFactor p53 Methylation by Histone Lysine Methyltransferase SET7/9. Biochemistry2008, 47, 2743–2748. [DOI] [PubMed] [Google Scholar]
• (171).Hu P, and Zhang YCatalytic Mechanism and Product Specificity of the Histone Lysine Methyltransferase SET7/9: an ab initio QM/MM-FEStudy with Multiple Initial Structures. J. Am. Chem. Soc2006, 128, 1272–1278. [DOI] [PubMed] [Google Scholar]
• (172).Guo HB, and Guo HMechanism of Histone Methylation Catalyzed by Protein Lysine Methyltransferase SET7/9 and Origin of ProductSpecificity. Proc. Natl. Acad. Sci. USA2007, 104, 8797–8802. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (173).Wang S, Hu P, and Zhang YAb initio Quantum Mechanical/Molecular Mechanical Molecular Dynamics Simulation of Enzyme Catalysis: the Case ofHistone Lysine Methyltransferase SET7/9. J. Phys. Chem. B2007, 111, 3758–3764. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (174).Hu P, Wang S, and Zhang YHow do SET-Domain Protein Lysine Methyltransferases Achieve the Methylation State Specificity? Revisited by abinitio QM/MM Molecular Dynamics Simulations. J. Am. Chem. Soc2008, 130, 3806–3813. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (175).Yao J, Chu Y, An R, and Guo HUnderstanding Product Specificity of Protein Lysine Methyltransferases from QM/MM Molecular Dynamics and FreeEnergy Simulations: the Effects of Mutation on SET7/9 beyond the Tyr/Phe Switch. J. Chem. Inf.Model2012, 52, 449–456. [DOI] [PubMed] [Google Scholar]
• (176).Schramm VLEnzymatic Transition State Theory and Transition State Analogue Design. J. Biol.Chem2007, 282, 28297–28300. [DOI] [PubMed] [Google Scholar]
• (177).Zhang J, and Klinman JPEnzymatic Methyl Transfer: Role of an Active Site Residue in Generating Active Site Compaction that Correlates withCatalytic Efficiency. J. Am. Chem. Soc2011, 133, 17134–17137. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (178).Zhang J, and Klinman JPConvergent Mechanistic Features Between the Structurally Diverse N- and O-Methyltransferases: GlycineN-Methyltransferase and Catechol O-Methyltransferase. J. Am. Chem. Soc2016, 138, 9158–9165. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (179).Wilson PB, and Williams IHInfluence of Equatorial CHO Interactions on Secondary Kinetic Isotope Effects for Methyl Transfer.Angew. Chem. Int. Ed2016, 55, 3192–3195. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (180).Schotta G, Sengupta R, Kubicek S, Malin S, Kauer M, Callen E, Celeste A, Pagani M, Opravil S, De La Rosa-Velazquez IA, et al. A Chromatin-wide Transition to H4K20 Monomethylation Impairs Genome Integrity and Programmed DNA Rearrangements inthe Mouse. Genes Dev. 2008, 22,2048–2061. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (181).Couture JF, Dirk LMA, Brunzelle JS, Houtz RL, and Trievel RCStructural Origins for the Product Specificity of SET Domain Protein Methyltransferases.Proc. Natl. Acad. Sci. USA2008, 105, 20659–20664. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (182).Collins RE, Tachibana M, Tamaru H, Smith KM, Jia D, Zhang X, Selker EU, Shinkai Y, and Cheng XDIn vitro and in vivo Analyses of a Phe/Tyr Switch Controlling Product Specificity of Histone LysineMethyltransferases. J. Biol. Chem2005, 280, 5563–5570. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (183).Del Rizzo PA, Couture JF, Dirk LMA, Strunk BS, Roiko MS, Brunzelle JS, Houtz RL, and Trievel RCSET7/9 Catalytic Mutants Reveal the Role of Active Site Water Molecules in Lysine MultipleMethylation. J. Biol. Chem2010, 285, 31849–31858. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (184).Bai Q, Shen Y, Yao X, Wang F, Du Y, Wang Q, Jin N, Hai J, Hu T, and Yang JModeling a New Water Channel that Allows SET9 to Dimethylate p53. Plos One2011, 6, E19856. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (185).Herz HM, Hu D, and Shilatifard AEnhancer Malfunction in Cancer. Mol Cell2014, 53, 859–666. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (186).Southall SM, Wong PS, Odho Z, Roe SM, and Wilson JRStructural Basis for the Requirement of Additional Factors for MLL1 SET Domain Activity and Recognition ofEpigenetic Marks. Mol. Cell2009, 33, 181–191. [DOI] [PubMed] [Google Scholar]
• (187).Thornton JL, Westfield GH, Takahashi YH, Cook M, Gao X, Woodfin AR, Lee JS, Morgan MA, Jackson J, Smith ER, et al. Context Dependency of Set1/COMPASS-mediated Histone H3 Lys4 Trimethylation. GenesDev. 2014, 28, 115–120. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (188).Takahashi YH, Lee JS, Swanson SK, Saraf A, Florens L, Washburn MP, Trievel RC, and Shilatifard ARegulation of H3K4 Trimethylation via Cps40 (Spp1) of COMPASS is Monoubiquitination Independent: Implication for aPhe/Tyr Switch by the Catalytic Domain of Set1. Mol. Cell Biol. 2009,29, 3478–3486. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (189).Schneider J, Wood A, Lee JS, Schuster R, Dueker J, Maguire C, Swanson SK, Florens L, Washburn MP, and Shilatifard AMolecular Regulation of Histone H3 Trimethylation by COMPASS and the Regulation of Gene Expression.Mol. Cell2005, 19, 849–856. [DOI] [PubMed] [Google Scholar]
• (190).Deshpande AJ, Deshpande A, Sinha AU, Chen L, Chang J, Cihan A, Fazio M, Chen CW, Zhu N, Koche R, et al. AF10 Regulates Progressive H3K79 Methylation and HOX Gene Expression in Diverse AML Subtypes.Cancer Cell2014, 26, 896–908. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (191).Wang R, Ibanez G, Islam K, Zheng W, Blum G, Sengelaub C, and Luo MFormulating a Fluorogenic Assay to Evaluate S-Adenosyl-L-Methionine Analogues as Protein MethyltransferaseCofactors. Mol. Biosyst2011, 7, 2970–2981. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (192).Collazo E, Couture JF, Bulfer S, and Trievel RCA Coupled Fluorescent Assay for Histone Methyltransferases. Anal. Biochem2005, 342, 86–92. [DOI] [PubMed] [Google Scholar]
• (193).Chen DH, Wu KT, Hung CJ, Hsieh M, and Li CEffects of Adenosine Dialdehyde Treatment on in vitro and in vivo Stable Protein Methylation in HelaCells. J. Biochem2004, 136, 371–376. [DOI] [PubMed] [Google Scholar]
• (194).Kraus D, Yang Q, Kong D, Banks AS, Zhang L, Rodgers JT, Pirinen E, Pulinilkunnil TC, Gong F, Wang YC, et al. Nicotinamide N-Methyltransferase Knockdown Protects against Diet-Induced Obesity.Nature2014, 508, 258–262. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (195).Ulanovskaya OA, Zuhl AM, and Cravatt BFNNMT Promotes Epigenetic Remodeling in Cancer by Creating a Metabolic Methylation Sink.Nat. Chem. Biol2013, 9, 300–306. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (196).Mentch SJ, Mehrmohamadi M, Huang L, Liu X, Gupta D, Mattocks D, Gomez Padilla P, Ables G, Bamman MM, Thalacker-Mercer AE, et al. Histone Methylation Dynamics and Gene Regulation Occur through the Sensing of One-carbonMetabolism. Cell Metab. 2015, 22,861–873. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (197).Ooi SK, Qiu C, Bernstein E, Li K, Jia D, Yang Z, Erdjument-Bromage H, Tempst P, Lin SP, Allis CD, et al. DNMT3L Connects Unmethylated Lysine 4 of Histone H3 to de novo Methylation of DNA.Nature2007, 448, 714–717. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (198).Collins RE, Northrop JP, Horton JR, Lee DY, Zhang X, Stallcup MR, and Cheng XThe Ankyrin Repeats of G9a and GLP Histone Methyltransferases are Mono- and Dimethyllysine BindingModules. Nat. Struct. Mol. Biol2008, 15, 245–250. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (199).Kuo AJ, Song J, Cheung P, Ishibe-Murakami S, Yamazoe S, Chen JK, Patel DJ, and Gozani OThe BAH Domain of ORC1 Links H4K20me2 to DNA Replication Licensing and Meier-Gorlin Syndrome.Nature2012, 484, 115–119. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (200).Kim D, Blus BJ, Chandra V, Huang P, Rastinejad F, and Khorasanizadeh SCorecognition of DNA and a Methylated Histone Tail by the MSL3 Chromodomain. Nat. Struct.Mol. Biol2010, 17, 1027–1029. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (201).Jacobs SA, and Khorasanizadeh SStructure of HP1 Chromodomain Bound to a Lysine 9-methylated Histone H3 Tail.Science2002, 295, 2080–2083. [DOI] [PubMed] [Google Scholar]
• (202).Flanagan JF, Mi LZ, Chruszcz M, Cymborowski M, Clines KL, Kim Y, Minor W, Rastinejad F, and Khorasanizadeh SDouble Chromodomains Cooperate to Recognize the Methylated Histone H3 Tail. Nature2005, 438, 1181–1185. [DOI] [PubMed] [Google Scholar]
• (203).Piazza I, Rutkowska A, Ori A, Walczak M, Metz J, Pelechano V, Beck M, and Haering CHAssociation of Condensin with Chromosomes Depends on DNA Binding by its HEAT-repeat Subunits.Nat. Struct. Mol. Biol2014, 21, 560–568. [DOI] [PubMed] [Google Scholar]
• (204).Min J, Allali-Hassani A, Nady N, Qi C, Ouyang H, Liu Y, Mackenzie F, Vedadi M, and Arrowsmith CHL3MBTL1 Recognition of Mono-and Dimethylated Histones. Nat. Struct. Mol. Biol2007, 14, 1229–1230. [DOI] [PubMed] [Google Scholar]
• (205).Roy S, Musselman CA, Kachirskaia I, Hayashi R, Glass KC, Nix JC, Gozani O, Appella E, and Kutateladze TGStructural Insight into p53 Recognition by the 53BP1 Tandem Tudor Domain. J. Mol.Biol2010, 398, 489–496. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (206).Kungulovski G, Mauser R, Reinhardt R, and Jeltsch AApplication of Recombinant TAF3 PHD Domain Instead of anti-H3K4me3 Antibody. EpigeneticsChromatin2016, 9, 11. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (207).Li H, Ilin S, Wang W, Duncan EM, Wysocka J, Allis CD, and Patel DJMolecular Basis for Site-specific Readout of Histone H3K4me3 by the BPTF PHD Finger of NURF.Nature2006, 442, 91–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (208).Pena PV, Davrazou F, Shi X, Walter KL, Verkhusha VV, Gozani O, Zhao R, and Kutateladze TGMolecular Mechanism of Histone H3K4me3 Recognition by Plant Homeodomain of ING2.Nature2006, 442, 100–103. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (209).Lan F, Collins RE, De Cegli R, Alpatov R, Horton JR, Shi X, Gozani O, Cheng X, and Shi YRecognition of Unmethylated Histone H3 Lysine 4 Links BHC80 to LSD1-mediated Gene Repression.Nature2007, 448, 718–722. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (210).Law JA, Du J, Hale CJ, Feng S, Krajewski K, Palanca AM, Strahl BD, Patel DJ, and Jacobsen SEPolymerase IV Occupancy at RNA-directed DNA Methylation Sites Requires SHH1.Nature2013, 498, 385–389. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (211).Musselman CA, Avvakumov N, Watanabe R, Abraham CG, Lalonde ME, Hong Z, Allen C, Roy S, Nunez JK, Nickoloff J, et al. Molecular Basis for H3K36me3 Recognition by the Tudor Domain of PHF1. Nat. Struct. Mol.Biol2012, 19, 1266–1272. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (212).Min J, Zhang Y, and Xu RMStructural Basis for Specific Binding of Polycomb Chromodomain to Histone H3 Methylated at Lys 27.Genes Dev. 2003, 17,1823–1828. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (213).Margueron R, Justin N, Ohno K, Sharpe ML, Son J, Drury WJ 3rd, Voigt P, Martin SR, Taylor WR, De Marco V, et al. Role of the Polycomb Protein EED in the Propagation of Repressive Histone Marks.Nature2009, 461, 762–767. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (214).He F, Umehara T, Saito K, Harada T, Watanabe S, Yabuki T, Kigawa T, Takahashi M, Kuwasako K, Tsuda K, et al. Structural Insight into the Zinc Finger CW Domain as a Histone Modification Reader.Structure2010, 18, 1127–11239. [DOI] [PubMed] [Google Scholar]
• (215).Musselman CA, Khorasanizadeh S, and Kutateladze TGTowards Understanding Methyllysine Readout. Biochim. Biophys. Acta2014, 1839, 686–693. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (216).Zhao Q, Rank G, Tan YT, Li H, Moritz RL, Simpson RJ, Cerruti L, Curtis DJ, Patel DJ, Allis CD, et al. PRMT5-mediated Methylation of Histone H4R3 Recruits DNMT3A, Coupling Histone and DNA Methylation in GeneSilencing. Nat. Struct. Mol. Biol2009, 16, 304–311. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (217).Cote J, and Richard STudor Domains Bind Symmetrical Dimethylated Arginines. J. Biol. Chem2005, 280, 28476–28483. [DOI] [PubMed] [Google Scholar]
• (218).Liu K, Guo Y, Liu H, Bian C, Lam R, Liu Y, Mackenzie F, Rojas LA, Reinberg D, Bedford MT, et al. Crystal Structure of TDRD3 and Methyl-Arginine Binding Characterization of TDRD3, SMN and SPF30.Plos One2012, 7, E30375. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (219).Sikorsky T, Hobor F, Krizanova E, Pasulka J, Kubicek K, and Stefl RRecognition of Asymmetrically Dimethylated Arginine by TDRD3. Nucleic Acids Res.2012, 40, 11748–11755. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (220).Yang Y, Lu Y, Espejo A, Wu J, Xu W, Liang S, and Bedford MTTDRD3 is an Effector Molecule for Arginine-methylated Histone Marks. Mol. Cell2010, 40, 1016–1123. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (221).Iberg AN, Espejo A, Cheng D, Kim D, Michaud-Levesque J, Richard S, and Bedford MTArginine Methylation of the Histone H3 Tail Impedes Effector Binding. J. Biol.Chem2008, 283, 3006–3610. [DOI] [PubMed] [Google Scholar]
• (222).Li H, Fischle W, Wang W, Duncan EM, Liang L, Murakami-Ishibe S, Allis CD, and Patel DJStructural Basis for Lower Lysine Methylation State-specific Readout by MBT Repeats of L3MBTL1 and an EngineeredPHD Finger. Mol. Cell2007, 28, 677–691. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (223).Trojer P, Li G, Sims RJ, Vaquero A, Kalakonda N, Boccuni P, Lee D, Erdjument-Bromage H, Tempst P, Nimer SD, et al. L3MBTL1, a Histone-methylation-dependent Chromatin Lock. Cell2007, 129, 915–928. [DOI] [PubMed] [Google Scholar]
• (224).Gao C, Herold JM, Kireev D, Wigle T, Norris JL, and Frye SBiophysical Probes Reveal a “Compromise” Nature of the Methyl-Lysine Binding Pocket inL3MBTL1. J. Am. Chem. Soc2011, 133, 5357–5362. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (225).Kamps JJ, Huang J, Poater J, Xu C, Pieters BJ, Dong A, Min J, Sherman W, Beuming T, Matthias Bickelhaupt F, et al. Chemical Basis for the Recognition of Trimethyllysine by Epigenetic Reader Proteins. Nat.Commun2015, 6, 8911. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (226).Wen H, Li Y, Xi Y, Jiang S, Stratton S, Peng D, Tanaka K, Ren Y, Xia Z, Wu J, et al. ZMYND11 Links Histone H3.3K36me3 to Transcription Elongation and Tumour Suppression.Nature2014, 508, 263–268. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (227).Iwase S, Xiang B, Ghosh S, Ren T, Lewis PW, Cochrane JC, Allis CD, Picketts DJ, Patel DJ, Li H, et al. ATRX ADD Domain Links an Atypical Histone Methylation Recognition Mechanism to Human Mental-retardationSyndrome. Nat. Struct. Mol. Biol2011, 18, 769–776. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (228).Bian C, Xu C, Ruan J, Lee KK, Burke TL, Tempel W, Barsyte D, Li J, Wu M, Zhou BO, et al. Sgf29 Binds Histone H3K4me2/3 and is Required for SAGA Complex Recruitment and Histone H3Acetylation. EMBO J. 2011, 30,2829–28242. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (229).Arita K, Isogai S, Oda T, Unoki M, Sugita K, Sekiyama N, Kuwata K, Hamamoto R, Tochio H, Sato M, et al. Recognition of Modification Status on a Histone H3 Tail by Linked Histone Reader Modules of the EpigeneticRegulator UHRF1. Proc. Natl. Acad. Sci. USA2012, 109, 12950–12955. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (230).Vezzoli A, Bonadies N, Allen MD, Freund SM, Santiveri CM, Kvinlaug BT, Huntly BJ, Gottgens B, and Bycroft MMolecular Basis of Histone H3K36me3 Recognition by the PWWP Domain of Brpf1. Nat. Struct.Mol. Biol2010, 17, 617–619. [DOI] [PubMed] [Google Scholar]
• (231).Patel DJ, and Wang ZReadout of Epigenetic Modifications. Annu. Rev. Biochem2013, 82, 81–118. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (232).Kilic S, Bachmann AL, Bryan LC, and Fierz BMultivalency Governs HP1α Association Dynamics with the Silent Chromatin State. Nat.Commun2015, 6, 7313. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (233).Ruthenburg AJ, Li H, Patel DJ, and Allis CDMultivalent Engagement of Chromatin Modifications by Linked Binding Modules. Nat. Rev. Mol.Cell Biol2007, 8, 983–994. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (234).Du J, Zhong X, Bernatavichute YV, Stroud H, Feng S, Caro E, Vashisht AA, Terragni J, Chin HG, Tu A, et al. Dual Binding of Chromomethylase Domains to H3K9me2-containing Nucleosomes Directs DNA Methylation inPlants. Cell2012, 151, 167–180. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (235).Fasting C, Schalley CA, Weber M, Seitz O, Hecht S, Koksch B, Dernedde J, Graf C, Knapp EW, and Haag RMultivalency as a Chemical Organization and Action Principle. Angew. Chem. Int.Ed2012, 51, 10472–10498. [DOI] [PubMed] [Google Scholar]
• (236).Rao RC, and Dou YHijacked in Cancer: the KMT2 (MLL) Family of Methyltransferases. Nat. Rev. Cancer2015, 15, 334–346. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (237).Stavropoulos P, Blobel G, and Hoelz ACrystal Structure and Mechanism of Human Lysine-specific Demethylase-1. Nat. Struct. Mol.Biol2006, 13, 626–632. [DOI] [PubMed] [Google Scholar]
• (238).Chen F, Yang H, Dong Z, Fang J, Wang P, Zhu T, Gong W, Fang R, Shi YG, Li Z, et al. Structural Insight into Substrate Recognition by Histone Demethylase LSD2/KDM1b. CellRes. 2013, 23, 306–309. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (239).Nowak RP, Tumber A, Johansson C, Che KH, Brennan P, Owen D, and Oppermann UAdvances and Challenges in Understanding Histone Demethylase Biology. Curr. Opin. Chem.Biol2016, 33, 151–159. [DOI] [PubMed] [Google Scholar]
• (240).Ng SS, Kavanagh KL, Mcdonough MA, Butler D, Pilka ES, Lienard BM, Bray JE, Savitsky P, Gileadi O, Von Delft F, et al. Crystal Structures of Histone Demethylase JMJD2A Reveal Basis for Substrate Specificity.Nature2007, 448, 87–91. [DOI] [PubMed] [Google Scholar]
• (241).Cole PAChemical Probes for Histone-modifying Enzymes. Nat. Chem. Biol2008, 4, 590–597. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (242).Zaragoza JPT, Yosca TH, Siegler MA, Moenne-Loccoz P, Green MT, and Goldberg DPDirect Observation of Oxygen Rebound with an Iron-hydroxide Complex. J. Am. Chem.Soc2017, 139, 13640–13643. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (243).Black JC, Van Rechem C, and Whetstine JRHistone Lysine Methylation Dynamics: Establishment, Regulation, and Biological Impact. Mol.Cell2012, 48, 491–507. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (244).Krishnan S, Horowitz S, and Trievel RCStructure and Function of Histone H3 Lysine 9 Methyltransferases and Demethylases.Chembiochem2011, 12, 254–263. [DOI] [PubMed] [Google Scholar]
• (245).Horton JR, Upadhyay AK, Hashimoto H, Zhang X, and Cheng XStructural Basis for Human PHF2 Jumonji Domain Interaction with Metal Ions. J. Mol.Biol2011, 406, 1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (246).Herranz N, Dave N, Millanes-Romero A, Pascual-Reguant L, Morey L, Diaz VM, Lorenz-Fonfria V, Gutierrez-Gallego R, Jeronimo C, Iturbide A, et al. Lysyl Oxidase-like 2 (LOXL2) Oxidizes Trimethylated Lysine 4 in Histone H3. FEBSJ. 2016, 283, 4263–4273. [DOI] [PubMed] [Google Scholar]
• (247).Accari SL, and Fisher PREmerging Roles of Jmjc Domain-containing Proteins. Int. Rev. Cell Mol. Biol2015, 319, 165–220. [DOI] [PubMed] [Google Scholar]
• (248).Nicholson TB, and Chen TLSD1 Demethylates Histone and Non-histone Proteins. Epigenetics2009, 4, 129–132. [DOI] [PubMed] [Google Scholar]
• (249).Zhang X, Tanaka K, Yan J, Li J, Peng D, Jiang Y, Yang Z, Barton MC, Wen H, and Shi XRegulation of Estrogen Receptor α by Histone Methyltransferase SMYD2-Mediated ProteinMethylation. Proc. Natl. Acad. Sci. USA2013, 110, 17284–17289. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (250).Abu-Farha M, Lanouette S, Elisma F, Tremblay V, Butson J, Figeys D, and Couture JFProteomic Analyses of the SMYD Family Interactomes Identify HSP90 as a Novel Target for SMYD2.J. Mol. Cell Biol2011, 3, 301–308. [DOI] [PubMed] [Google Scholar]
• (251).Wu CY, Hsieh CY, Huang KE, Chang C, and Kang HYCryptotanshinone Down-regulates Androgen Receptor Signaling by Modulating Lysine-specific Demethylase 1Function. Int. J. Cancer2012, 131, 1423–1434. [DOI] [PubMed] [Google Scholar]
• (252).Bremang M, Cuomo A, Agresta AM, Stugiewicz M, Spadotto V, and Bonaldi TMass Spectrometry-based Identification and Characterisation of Lysine and Arginine Methylation in the HumanProteome. Mol. Biosyst2013, 9, 2231–2247. [DOI] [PubMed] [Google Scholar]
• (253).Lott K, Li J, Fisk JC, Wang H, Aletta JM, Qu J, and Read LKGlobal Proteomic Analysis in Trypanosomes Reveals Unique Proteins and Conserved Cellular Processes Impacted byArginine Methylation. J. Proteomics2013, 91, 210–225. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (254).Guo A, Gu H, Zhou J, Mulhern D, Wang Y, Lee KA, Yang V, Aguiar M, Kornhauser J, Jia X, et al. Immunoaffinity Enrichment and Mass Spectrometry Analysis of Protein Methylation. Mol. CellProteomics2014, 13, 372–387. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (255).Liu H, Galka M, Mori E, Liu X, Lin YF, Wei R, Pittock P, Voss C, Dhami G, Li X, et al. A Method for Systematic Mapping of Protein Lysine Methylation Identifies Functions for HP1β in DNA DamageResponse. Mol. Cell2013, 50, 723–735. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (256).Afjehi-Sadat L, and Garcia BAComprehending Dynamic Protein Methylation with Mass Spectrometry. Curr. Opin. Chem.Biol2013, 17, 12–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (257).Ong SE, Mittler G, and Mann MIdentifying and Quantifying in vivo Methylation Sites by Heavy Methyl SILAC. Nat.Methods2004, 1, 119–126. [DOI] [PubMed] [Google Scholar]
• (258).Cheng D, Vemulapalli V, and Bedford MTMethods Applied to the Study of Protein Arginine Methylation. Methods Enzymol.2012, 512, 71–92. [DOI] [PubMed] [Google Scholar]
• (259).Bothwell IR, Islam K, Chen Y, Zheng W, Blum G, Deng H, and Luo MSe-Adenosyl-L-selenomethionine Cofactor Analogue as a Reporter of Protein Methylation. J.Am. Chem. Soc2012, 134, 14905–14912. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (260).Bothwell IR, and Luo MLarge-Scale, Protection-Free Synthesis of Se-Adenosyl-L-Selenomethionine Analogues and their Application asCofactor Surrogates of Methyltransferases. Org. Lett2014, 16, 3056–3059. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (261).Garcia BA, Mollah S, Ueberheide BM, Busby SA, Muratore TL, Shabanowitz J, and Hunt DFChemical Derivatization of Histones for Facilitated Analysis by Mass Spectrometry. Nat.Protoc2007, 2, 933–938. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (262).Plazas-Mayorca MD, Zee BM, Young NL, Fingerman IM, Leroy G, Briggs SD, and Garcia BAOne-pot Shotgun Quantitative Mass Spectrometry Characterization of Histones. J. ProteomeRes2009, 8, 5367–5374. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (263).Blair LP, Avaritt NL, Huang R, Cole PA, Taverna SD, and Tackett AJMasssquirm: An Assay for Quantitative Measurement of Lysine Demethylase Activity.Epigenetics2011, 6, 490–469. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (264).Peters W, Willnow S, Duisken M, Kleine H, Macherey T, Duncan KE, Litchfield DW, Luscher B, and Weinhold EEnzymatic Site-specific Functionalization of Protein Methyltransferase Substrates with Alkynes for ClickLabeling. Angew. Chem. Int. Ed2010, 49, 5170–5173. [DOI] [PubMed] [Google Scholar]
• (265).Wang R, Zheng W, Yu H, Deng H, and Luo MLabeling Substrates of Protein Arginine Methyltransferase with Engineered Enzymes and MatchedS-Adenosyl-L-Methionine Analogues. J. Am. Chem. Soc2011, 133, 7648–7651. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (266).Willnow S, Martin M, Luscher B, and Weinhold EA Selenium-based Click AdoMet Analogue for Versatile Substrate Labeling with Wild-type ProteinMethyltransferases. Chembiochem2012, 13, 1167–1173. [DOI] [PubMed] [Google Scholar]
• (267).Guo H, Wang R, Zheng W, Chen Y, Blum G, Deng H, and Luo MProfiling Substrates of Protein Arginine N-Methyltransferase 3 with S-Adenosyl-L-MethionineAnalogues. ACS Chem. Biol2014, 9, 476–484. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (268).Blum G, Islam K, and Luo MBioorthogonal Profiling of Protein Methylation (BPPM) Using an Azido Analog ofS-Adenosyl-L-Methionine. Curr. Protoc. Chem. Biol2013, 5, 45–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (269).Blum G, Bothwell IR, Islam K, and Luo MProfiling Protein Methylation with Cofactor Analog Containing Terminal Alkyne Functionality.Curr. Protoc. Chem. Biol2013, 5, 67–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (270).Wang R, Islam K, Liu Y, Zheng W, Tang H, Lailler N, Blum G, Deng H, and Luo MProfiling Genome-wide Chromatin Methylation with Engineered Posttranslation Apparatus within LivingCells. J. Am. Chem. Soc2013, 135, 1048–1056. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (271).Cao XJ, Arnaudo AM, and Garcia BALarge-scale Global Identification of Protein Lysine Methylation in vivo.Epigenetics2013, 8, 477–485. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (272).Nguyen H, Allali-Hassani A, Antonysamy S, Chang S, Chen LH, Curtis C, Emtage S, Fan L, Gheyi T, Li F, et al. LLY-507, a Cell-Active, Potent, and Selective Inhibitor of Protein Lysine Methyltransferase SMYD2.J. Biol. Chem2015, 290, 13641–13653. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (273).Kaniskan HU, and Jin JChemical Probes of Histone Lysine Methyltransferases. ACS Chem. Biol2014, 10, 40–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (274).Frye SVThe Art of the Chemical Probe. Nat. Chem. Biol2010, 6, 159–161. [DOI] [PubMed] [Google Scholar]
• (275).Daigle SR, Olhava EJ, Therkelsen CA, Majer CR, Sneeringer CJ, Song J, Johnston LD, Scott MP, Smith JJ, Xiao Y, et al. Selective Killing of Mixed Lineage Leukemia Cells by a Potent Small-Molecule DOT1L Inhibitor.Cancer Cell2011, 20, 53–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (276).Basavapathruni A, Jin L, Daigle SR, Majer CR, Therkelsen CA, Wigle TJ, Kuntz KW, Chesworth R, Pollock RM, Scott MP, et al. Conformational Adaptation Drives Potent, Selective and Durable Inhibition of the Human Protein MethyltransferaseDOT1L. Chem. Biol. Drug. Des2012, 80, 971–980. [DOI] [PubMed] [Google Scholar]
• (277).Van Aller GS, Pappalardi MB, Ott HM, Diaz E, Brandt M, Schwartz BJ, Miller WH, Dhanak D, Mccabe MT, Verma SK, et al. Long Residence Time Inhibition of EZH2 in Activated Polycomb Repressive Complex 2. ACSChem. Biol2013, 9, 622–629. [DOI] [PubMed] [Google Scholar]
• (278).Mccabe MT, Ott HM, Ganji G, Korenchuk S, Thompson C, Van Aller GS, Liu Y, Graves AP, Della Pietra A III, Diaz E, et al. EZH2 Inhibition as a Therapeutic Strategy for Lymphoma with EZH2-Activating Mutations.Nature2012, 492, 108–112. [DOI] [PubMed] [Google Scholar]
• (279).Konze KD, Ma A, Li F, Barsyte-Lovejoy D, Parton T, Macnevin CJ, Liu F, Gao C, Huang XP, Kuznetsova E, et al. An Orally Bioavailable Chemical Probe of the Lysine Methyltransferases EZH2 and EZH1. ACSChem. Biol2013, 8, 1324–1334. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (280).Chen C, Zhu H, Stauffer F, Caravatti G, Vollmer S, Machauer R, Holzer P, Mobitz H, Scheufler C, Klumpp M, Tiedt R, et al. Discovery of Novel DOT1L Inhibitors Through a Structure-based Fragmentation Approach. ACSMed. Chem. Lett2016, 7, 735–740. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (281).Scheufler C, Mobitz H, Gaul C, Ragot C, Be C, Fernandez C, Beyer KS, Tiedt R, and Stauffer FOptimization of a Fragment-based Screening Hit toward Potent DOT1L Inhibitors Interacting in an Induced BindingPocket. ACS Med. Chem. Lett2016, 7, 730–734. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (282).Mobitz H, Machauer R, Holzer P, Vaupel A, Stauffer F, Ragot C, Caravatti G, Scheufler C, Fernandez C, Hommel U, et al. Discovery of Potent, Selective, and Structurally Novel DOT1L Inhibitors by a Fragment LinkingApproach. ACS Med. Chem. Lett2017, 8, 338–343. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (283).Lee GM, and Craik CSTrapping Moving Targets with Small Molecules. Science2009, 324, 213–215. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (284).Qi W, Zhao K, Gu J, Huang Y, Wang Y, Zhang H, Zhang M, Zhang J, Yu Z, Li L, et al. An Allosteric PRC2 Inhibitor Targeting the H3K27me3 Binding Pocket of EED. Nat. Chem.Biol2017, 13, 381–388. [DOI] [PubMed] [Google Scholar]
• (285).Liu F, Barsyte-Lovejoy D, Li F, Xiong Y, Korboukh V, Huang XP, Allali-Hassani A, Janzen WP, Roth BL, Frye SV, et al. Discovery of an in vivo Chemical Probe of the Lysine Methyltransferases G9a and GLP. J.Med. Chem2013, 56, 8931–8942. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (286).Sweis RF, Wang Z, Algire M, Arrowsmith CH, Brown PJ, Chiang GG, Guo J, Jakob CG, Kennedy S, Li F, et al. Discovery of A-893, a New Cell-Active Benzoxazinone Inhibitor of Lysine Methyltransferase SMYD2.ACS Med. Chem. Lett2015, 6, 695–700. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (287).Bromberg KD, Mitchell TR, Upadhyay AK, Jakob CG, Jhala MA, Comess KM, Lasko LM, Li C, Tuzon CT, Dai Y, et al. The SUV4-20 Inhibitor A-196 Verifies a Role for Epigenetics in Genomic Integrity. Nat.Chem. Biol2017, 13, 317–324. [DOI] [PubMed] [Google Scholar]
• (288).Mitchell LH, Boriack-Sjodin PA, Smith S, Thomenius M, Rioux N, Munchhof M, Mills JE, Klaus C, Totman J, Riera TV, et al. Novel Oxindole Sulfonamides and Sulfamides: EPZ031686, the First Orally Bioavailable Small Molecule SMYD3Inhibitor. ACS Med. Chem. Lett2016, 7, 134–138. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (289).Van Aller GS, Graves AP, Elkins PA, Bonnette WG, Mcdevitt PJ, Zappacosta F, Annan RS, Dean TW, Su DS, Carpenter CL, et al. Structure-based Design of a Novel SMYD3 Inhibitor that Bridges the SAM- and MEKK2-binding Pockets.Structure2016, 24, 774–781. [DOI] [PubMed] [Google Scholar]
• (290).Grembecka J, He S, Shi A, Purohit T, Muntean AG, Sorenson RJ, Showalter HD, Murai MJ, Belcher AM, Hartley T, et al. Menin-MLL Inhibitors Reverse Oncogenic Activity of MLL Fusion Proteins in Leukemia. Nat.Chem. Biol2012, 8, 277–284. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (291).Ma A, Yu W, Xiong Y, Butler KV, Brown PJ, and Jin JStructure-activity relationship Studies of SETD8 Inhibitors. Medchemcomm2014, 5, 1892–1898. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (292).Butler KV, Ma A, Yu W, Li F, Tempel W, Babault N, Pittella-Silva F, Shao J, Wang J, Luo M, et al. Structure-based Design of a Covalent Inhibitor of the SET Domain-containing Protein Lysine MethyltransferaseSETD8. J. Med. Chem2016, 59, 9881–9889. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (293).Blum G, Ibanez G, Rao X, Shum D, Radu C, Djaballah H, Rice JC, and Luo MSmall-molecule Inhibitors of SETD8 with Cellular Activity. ACS Chem. Biol2014, 9, 2471–2478. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (294).Lafave LM, Beguelin W, Koche R, Teater M, Spitzer B, Chramiec A, Papalexi E, Keller MD, Hricik T, Konstantinoff K, et al. Loss of BAP1 Function Leads to EZH2-dependent Transformation. Nat. Med2015, 21, 1344–1349. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (295).Milite C, Feoli A, Viviano M, Rescigno D, Mai A, Castellano S, and Sbardella GProgress in the Development of Lysine Methyltransferase SETD8 Inhibitors.Chemmedchem2016, 11, 1680–1685. [DOI] [PubMed] [Google Scholar]
• (296).Osborne T, Roska RLW, Rajski SR, and Thompson PRIn situ Generation of a Bisubstrate Analogue for Protein Arginine Methyltransferase 1. J.Am. Chem. Soc2008, 130, 4574–4575. [DOI] [PubMed] [Google Scholar]
• (297).Jafari R, Almqvist H, Axelsson H, Ignatushchenko M, Lundback T, Nordlund P, and Martinez Molina DThe Cellular Thermal Shift Assay for Evaluating Drug Target Interactions in Cells. Nat.Protoc2014, 9, 2100–2122. [DOI] [PubMed] [Google Scholar]
• (298).Simon MD, Chu F, Racki LR, De La Cruz CC, Burlingame AL, Panning B, Narlikar GJ, and Shokat KMThe Site-specific Installation of Methyl-Lysine Analogs into Recombinant Histones.Cell2007, 128, 1003–1012. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (299).Jia G, Wang W, Li H, Mao Z, Cai G, Sun J, Wu H, Xu M, Yang P, Yuan W, et al. A Systematic Evaluation of the Compatibility of Histones Containing Methyl-Lysine Analogues with BiochemicalReactions. Cell Res. 2009, 19,1217–1220. [DOI] [PubMed] [Google Scholar]
• (300).Chen Z, Notti RQ, Ueberheide B, and Ruthenburg AJQuantitative and Structural Assessment of Histone Methyllysine Analogue Engagement by Cognate Binding ProteinsReveals Affinity Decrements Relative to those of Native Counterparts. Biochemistry2018, 57, 300–304. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (301).Wright TH, Bower BJ, Chalker JM, Bernardes GJ, Wiewiora R, Ng WL, Raj R, Faulkner S, Vallee MR, Phanumartwiwath A, et al. Posttranslational Mutagenesis: a Chemical Strategy for Exploring Protein Side-chain Diversity.Science2016, 354, Aag1465. [DOI] [PubMed] [Google Scholar]
• (302).Chalker JM, Lercher L, Rose NR, Schofield CJ, and Davis BGConversion of Cysteine into Dehydroalanine Enables Access to Synthetic Histones Bearing Diverse Post-TranslationalModifications. Angew. Chem. Int. Ed2012, 51, 1835–1839. [DOI] [PubMed] [Google Scholar]
• (303).Yang A, Ha S, Ahn J, Kim R, Kim S, Lee Y, Kim J, Soll D, Lee HY, and Park HSA Chemical Biology Route to Site-specific Authentic Protein Modifications. Science2016, 354, 623–626. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (304).Terasaka N, Iwane Y, Geiermann AS, Goto Y, and Suga HRecent Developments of Engineered Translational Machineries for the Incorporation of Non-canonical Amino Acids intoPolypeptides. Int. J. Mol. Sci2015, 16, 6513–6531. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (305).Chin JWExpanding and Reprogramming the Genetic Code. Nature2017, 550, 53–60. [DOI] [PubMed] [Google Scholar]
• (306).Guo J, Wang J, Lee JS, and Schultz PGSite-specific Incorporation of Methyl- and Acetyl-Lysine Analogues into Recombinant Proteins.Angew. Chem. Int. Ed2008, 47, 6399–6401. [DOI] [PubMed] [Google Scholar]
• (307).Wang ZA, Zeng Y, Kurra Y, Wang X, Tharp JM, Vatansever EC, Hsu WW, Dai S, Fang X, and Liu WRA Genetically Encoded Allysine for the Synthesis of Proteins with Site-Specific LysineDimethylation. Angew. Chem. Int. Ed2017, 56, 212–216. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (308).Ai HW, Lee JW, and Schultz PGA Method to Site-specifically Introduce Methyllysine into Proteins in E. coli. Chem.Commun2010, 46, 5506–5508. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (309).Wang Y-S, Wu B, Wang Z, Huang Y, Wan W, Russell WK, Pai P-J, Moe YN, Russell DH, and Liu WRA Genetically Encoded Photocaged Nε-Methyl-L-Lysine. Mol. Biosyst2010, 6, 1575–1578. [DOI] [PubMed] [Google Scholar]
• (310).Wang ZA, and Liu WRProteins with Site-Specific Lysine Methylation. Chemistry2017, 23, 11732–11737. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (311).Nguyen DP, Alai MMG, Kapadnis PB, Neumann H, and Chin JWGenetically Encoding Nε-Methyl-L-Lysine in Recombinant Histones. J. Am. Chem.Soc2009, 131, 14194–14195. [DOI] [PubMed] [Google Scholar]
• (312).Nguyen DP, Alai MMG, Virdee S, and Chin JWGenetically Directing ε-N, N-Dimethyl-L-Lysine in Recombinant Histones. Chem.Biol2010, 17, 1072–1076. [DOI] [PubMed] [Google Scholar]
• (313).Li J, Yu J, Zhao J, Wang J, Zheng S, Lin S, Chen L, Yang M, Jia S, Zhang X, et al. Palladium-triggered Deprotection Chemistry for Protein Activation in Living Cells. Nat.Chem2014, 6, 352–361. [DOI] [PubMed] [Google Scholar]
• (314).Klan P, Solomek T, Bochet CG, Blanc A, Givens R, Rubina M, Popik V, Kostikov A, and Wirz JPhotoremovable Protecting Groups in Chemistry and Biology: Reaction Mechanisms and Efficacy.Chem. Rev2013, 113, 119–191. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (315).Yanagisawa T, Ishii R, Fukunaga R, Kobayashi T, Sakamoto K, and Yokoyama SMultistep Engineering of Pyrrolysyl-tRNA Synthetase to Genetically Encode Nε-(O-Azidobenzyloxycarbonyl)Lysine for Site-Specific Protein Modification. Chem. Biol2008, 15, 1187–1197. [DOI] [PubMed] [Google Scholar]
• (316).Li J, Jia S, and Chen PRDiels-Alder Reaction-triggered Bioorthogonal Protein Decaging in Living Cells. Nat. Chem.Biol2014, 10, 1003–1005. [DOI] [PubMed] [Google Scholar]
• (317).Li J, and Chen PRDevelopment and Application of Bond Cleavage Reactions in Bioorthogonal Chemistry. Nat.Chem. Biol2016, 12, 129–137. [DOI] [PubMed] [Google Scholar]
• (318).Xie X, Li XM, Qin F, Lin J, Zhang G, Zhao J, Bao X, Zhu R, Song H, Li XD, et al. Genetically Encoded Photoaffinity Histone Marks. J. Am. Chem. Soc2017, 139, 6522–6525. [DOI] [PubMed] [Google Scholar]
• (319).Yang T, Li XM, Bao X, Fung YM, and Li XDPhoto-Lysine Captures Proteins that Bind Lysine Post-translational Modifications. Nat.Chem. Biol2016, 12, 70–72. [DOI] [PubMed] [Google Scholar]
• (320).Allis CD, and Muir TWSpreading Chromatin into Chemical Biology. Chembiochem2011, 12, 264–279. [DOI] [PubMed] [Google Scholar]
• (321).Dann GP, Liszczak GP, Bagert JD, Muller MM, Nguyen UTT, Wojcik F, Brown ZZ, Bos J, Panchenko T, Pihl R, et al. ISWI Chromatin Remodellers Sense Nucleosome Modifications to Determine Substrate Preference.Nature2017, 548, 607–611. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (322).Nguyen UT, Bittova L, Muller MM, Fierz B, David Y, Houck-Loomis B, Feng V, Dann GP, and Muir TWAccelerated Chromatin Biochemistry Using DNA-barcoded Nucleosome Libraries. Nat.Methods2014, 11, 834–840. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (323).Wang P, Dong S, Shieh JH, Peguero E, Hendrickson R, Moore MAS, and Danishefsky SJErythropoietin Derived by Chemical Synthesis. Science2013, 342, 1357–1360. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (324).Zhang Q, Johnston EV, Shieh JH, Moore MA, and Danishefsky SJSynthesis of Granulocyte-macrophage Colony-stimulating Factor as Homogeneous Glycoforms and Early Comparisons withYeast Cell-derived Material. Proc. Natl. Acad. Sci. USA2014, 111, 2885–2890. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (325).Shah NH, Dann GP, Vila-Perello M, Liu Z, and Muir TWUltrafast Protein Splicing is Common among Cyanobacterial Split Inteins: Implications for ProteinEngineering. J. Am. Chem. Soc2012, 134, 11338–11341. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (326).David Y, Vila-Perello M, Verma S, and Muir TWChemical Tagging and Customizing of Cellular Chromatin States Using Ultrafast Trans-splicingInteins. Nat. Chem2015, 7, 394–402. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (327).Stevens AJ, Sekar G, Shah NH, Mostafavi AZ, Cowburn D, and Muir TWA Promiscuous Split Intein with Expanded Protein Engineering Applications. Proc. Natl.Acad. Sci. USA2017, 114, 8538–8543. [DOI] [PMC free article] [PubMed] [Google Scholar]
• (328).Henager SH, Chu N, Chen Z, Bolduc D, Dempsey DR, Hwang Y, Wells J, and Cole PAEnzyme-catalyzed Expressed Protein Ligation. Nat. Methods2016, 13, 925–927. [DOI] [PMC free article] [PubMed] [Google Scholar]