Biochem. J. (2013) 453, 241–247 (Printed in Great Britain)
241
doi:10.1042/BJ20130439
Convergent evolution of chromatin modification by structurally distinct enzymes: comparative enzymology of histone H3 Lys27 methylation by human polycomb repressive complex 2 and vSET Brooke M. SWALM1,2 , Kenneth K. HALLENBECK1 , Christina R. MAJER2 , Lei JIN, Margaret Porter SCOTT2 , Mikel P. MOYER2 , Robert A. COPELAND2 and Tim J. WIGLE2,3 Epizyme Inc., 400 Technology Square, Cambridge, MA 02139, U.S.A.
H3K27 (histone H3 Lys27 ) methylation is an important epigenetic modification that regulates gene transcription. In humans, EZH (enhancer of zeste homologue) 1 and EZH2 are the only enzymes capable of catalysing methylation of H3K27. There is great interest in understanding structure–function relationships for EZH2, as genetic alterations in this enzyme are thought to play a causal role in a number of human cancers. EZH2 is challenging to study because it is only active in the context of the multi-subunit PRC2 (polycomb repressive complex 2). vSET is a viral lysine methyltransferase that represents the smallest protein unit capable of catalysing H3K27 methylation. The crystal structure of this minimal catalytic protein has been solved and researchers have suggested that vSET might prove useful as an EZH2 surrogate for the development of active site-directed inhibitors. To test this proposition, we conducted comparative enzymatic analysis of human EZH2 and vSET and report that, although both enzymes share
INTRODUCTION
Histone lysine methylation is a dynamic process that exerts control over gene expression by influencing the conformational state of chromatin and by also serving as a recognition locus for the recruitment of transcription factors, enzymes and other proteins to selective locations on the histone. Sequential methylation of lysine results in four distinct states of the lysine side chain (zero, one, two or three methyl groups on the ε-nitrogen), each associated with unique transcriptional consequences. All of these methylation reactions are catalysed by a family of enzymes known as the PKMTs (protein lysine methyltransferases). PKMTs universally use SAM (S-adenosylmethionine) as the methyl group donor and transfer the methyl group to the ε-nitrogen of a lysine side chain through an SN 2 reaction mechanism [1,2]. With the exception of the enzyme DOT1L, all known human PKMTs share a similar catalytic domain of approximately 130 amino acids referred to as the SET [Su(var)3–9/enhancer of zeste/trithorax domain] domain [3]. The SET domain contains recognition elements for SAM and lysine binding and confers substrate specificity through interactions between the histone protein surrounding the methylaccepting lysine and enzyme residues in close proximity to the mouth of the lysine-binding channel [4,5]. Methylation of H3K27 (histone H3 Lys27 ) is a transcriptionally repressive mark that plays a critical role during development and differentiation, and is implicated in several forms of human cancer
similar preferences for methylation of H3K27, they diverge in terms of their permissiveness for catalysing methylation of alternative histone lysine sites, their relative preferences for utilization of multimeric macromolecular substrates, their active site primary sequences and, most importantly, their sensitivity to inhibition by drug-like small molecules. The cumulative data led us to suggest that EZH2 and vSET have very distinct active site structures, despite the commonality of the reaction catalysed by the two enzymes. Hence, the EZH2 and vSET pair of enzymes represent an example of convergent evolution in which distinct structural solutions have developed to solve a common catalytic need. Key words: convergent evolution, epigenetics, enhancer of zeste homologue 2 (EZH2), histone H3 Lys27 (H3K27), methyltransferase, vSET.
[6–14]. In multicellular organisms, this reaction is exclusively catalysed by a multi-protein complex referred to as PRC2 (polycomb repressive complex 2) that contains either of the SET-domain PKMTs EZH (enhancer of zeste homologue) 1 (PRC2EZH1 ) or EZH2 (PRC2EZH2 ) [15]. This same reaction is also catalysed by a PKMT of Paramecium bursaria chlorella virus 1, known as vSET [16–19]. vSET represents the minimal protein structural unit for PKMT activity, consisting of 119 amino acids (∼ 13.5 kDa) that aligns well with the canonical SET domain of larger PKMTs [16]. The small size of this viral enzyme makes it ideal for structural studies aimed at understanding the critical elements of substrate recognition and catalysis by PKMTs. Indeed, high-resolution crystal structures of vSET have previously been reported (PDB codes 3KMT, 3KMJ and 3KMA) [17,19,20]. In contrast, the large multi-protein PRC2EZH2 complex has to date not been amenable to crystallographic methods. Given the keen interest in PRC2EZH2 as a potential target for cancer drug discovery [21,22], we speculated whether it is possible to use vSET as a structurally defined surrogate of PRC2EZH2 for the design of inhibitors and other active site-directed ligands. This approach rests on the reasonable assumption that the common enzymatic activity of PRC2 and vSET is conferred by common structural features of molecular recognition. In the present study we test this underlying assumption by comparing the enzymatic details of catalysis by these two
Abbreviations used: EZH, enhancer of zeste homologue; H3K9, histone H3 Lys9 ; H3K27, histone H3 Lys27 ; H3K27me3, trimethylated H3K27; Ni-NTA, Ni2 + -nitrilotriacetate; PKMT, protein lysine methyltransferase; PRC2, polycomb repressive complex 2; SAM, S -adenosylmethionine; SAH, S adenosylhomocysteine; SET domain, Su(var)3–9/enhancer of zeste/trithorax domain. 1 Brooke Swalm, Christina Majer, Margaret Scott, Mikel Moyer, Robert Copeland and Tim Wigle are all employees and shareholders of Epizyme. 2 These authors contributed equally to this work. 3 To whom correspondence should be addressed (email
[email protected]). c The Authors Journal compilation c 2013 Biochemical Society
242
B. M. Swalm and others
enzymes. We find that, although the overall mechanism of catalysis is similar for the two enzymes, active site interactions with ligands nevertheless differ in significant ways. Together, these two enzymes thus represent an example of convergent evolution of enzyme function [23,24]. MATERIALS AND METHODS Reagents and equipment
The vSET gene was synthesized based on a published sequence (GenBank® accession number AAC96946.1) and cloned into the pET28a vector (Novagen) between NdeI and HindIII sites with a N-terminal His tag and thrombin cleavage site. The protein was expressed in BL21-Gold(DE3) Escherichia coli (Agilent) in LB medium with 50 μg/ml kanamycin and induced by 0.3 mM IPTG at 16 ◦ C for 16 h. The harvested cell pellet was suspended in lysis buffer containing 25 mM Tris/HCl, pH 7.6, 300 mM NaCl, 5 % (v/v) glycerol and 5 mM 2-mercaptoethanol, and cells were lysed by sonication. Cell debris was cleared by centrifigation and the supernatant was added on to a Ni-NTA (Ni2 + -nitrilotriacetate; Qiagen) column equilibrated with lysis buffer. The Ni-NTA column was first washed with lysis buffer, followed by lysis buffer supplemented with 20 mM imidazole, then with buffer containing 25 mM Tris/HCl, pH 8, 300 mM NaCl, 5 % (v/v) glycerol, 5 mM 2-mercaptoethanol and 50 mM imidazole. The protein was eluted with buffer containing 25 mM Tris/HCl, pH 8, 200 mM NaCl, 5 % (v/v) glycerol, 5 mM 2-mercaptoethanol and 250 mM imidazole. Fractions containing the target protein were pooled, dialysed against lysis buffer to remove imidazole and then concentrated. The concentrated sample was loaded on to a S-75 column (GE Healthcare) and equilibrated with buffer containing 20 mM Tris/HCl, pH 8, 200 mM NaCl and 5 mM 2mercaptoethanol. Fractions containing vSET were pooled and concentrated for assay use, yielding a stock that was >90 % pure as judged by capillary electrophoresis. Four-component PRC2EZH2 (EZH2, Suz12, EED and RbAp48) was purified to >95 % purity and 1:1 stoichiometry (judged by capillary electrophoresis) as previously described using a FLAG tag on the EED subunit [11]. Chicken erythrocyte monoand oligo-nucleosomes were purified as described previously [25]. Flashplates (384-well) and Microscint 0 scintillation fluid were purchased from PerkinElmer. Multiscreen HTS glass fibre filter-binding plates (96-well) were obtained from Millipore. [3 H]SAM was obtained from American Radiolabeled Chemicals with a specific activity of 80 Ci/mmol. Unlabelled SAM, SAH (S-adenosylhomocysteine) and Sinefungin were obtained from Sigma–Aldrich. Recombinant histone H3 was purchased from New England Biolabs, and recombinant histone H3/H4 tetramer and recombinant histone H2A/H2B/H3/H4 octamer were produced by XTAL Biostructures at >95 % purity. All peptides were synthesized and HPLC purified to >95 % purity by 21st Century Biochemicals. EPZ005687 was synthesized by Epizyme as previously described [21] and GSK126 was purchased from Xcessbio. Flashplates and filter-binding plates were read on a TopCount NXT microplate reader (PerkinElmer), and flashplates were washed in a Biotek Elx-405 with 0.1 % Tween 20 before being read. All enzymatic assays were performed in 384-well and 96well V-bottom polypropylene microplates (Greiner). Scanning for activity against a protein substrate panel using filter-binding microplate assays
For histone substrates, 1× assay buffer containing 20 mM Bicine, pH 7.6, 0.002 % Tween 20, 0.005 % bovine skin c The Authors Journal compilation c 2013 Biochemical Society
gelatin and 0.5 mM DTT was used, and, when used with nucleosomes, the same assay buffer was supplemented with 100 mM KCl. Reactions (50 μl) were carried out at 25 ◦ C in 96-well polypropylene microplates and contained 10 nM enzyme, 200 nM [3 H]SAM and 200 nM protein substrate. Reactions were terminated by adding an excess of unlabelled SAM to outcompete the incorporation of [3 H]SAM. Quenched reactions were added to 96-well filter-binding plates and the membranes were washed three times with 200 μl of 10 % tricarboxylic acid followed by washing once with 200 μl of 95 % ethanol. The membranes were air dried and 30 μl of Microscint 0 was added before reading on a TopCount NXT instrument. Scanning for enzymatic activity against a histone peptide panel using flashplate format
A library of biotinylated histone peptides was solubilized in either water or DMSO, and 1 μl was spotted into 384-well polypropylene microplates. Reactions (50 μl) were carried out at 25 ◦ C in 384-well polypropylene microplates and contained 8 nM vSET or 8 nM PRC2EZH2 , 250 nM [3 H]SAM and 1 μM peptide. Reactions were terminated after 1 h by adding an excess of unlabelled SAM to outcompete the incorporation of [3 H]SAM. The reaction mixture was transferred to a flashplate, incubated for 1 h at room temperature (25 ◦ C), then washed with 0.l% Tween 20 and read on a TopCount NXT instrument. Determination of steady-state mechanism using flashplate format with peptide substrates
Peptide and SAM were titrated, and 8 nM vSET or 8 nM PRC2EZH2 were added to initiate the reaction. Reactions (50 μl) were carried out in assay buffer at 25 ◦ C in 384-well polypropylene microplates and samples at various time points were taken by adding an excess of unlabelled SAM to outcompete the incorporation of [3 H]SAM. The reaction mixture was transferred to a flashplate, incubated for 1 h at room temperature, then washed with 0.l% Tween 20 and read on a TopCount NXT instrument. Double substrate titrations were fitted to the following equation for a ternary complex mechanism to determine steady-state K m values: v=
kcat [E][A][B] α K A K B + α K B [A] + α K A [B] + [A][B]
(1)
where E is enzyme, A is substrate A (SAM) and B is substrate B (peptide). Self-assembled monolayer desorption/ionization MS analysis of peptide methylation
Reactions (50 μl) were carried out in assay buffer at 25 ◦ C and contained 4 nM vSET enzyme, 50 nM peptide substrate and 1 μM SAM. Reactions were terminated by the addition of 100 mM NaCl and a 2 μl sample of each reaction was analysed by SAMDI Tech using self-assembled monolayer desorption/ionization time-offlight MS [26]. Determination of enzyme inhibition K i values
Inhibitors were pre-incubated with vSET or PRC2EZH2 in assay buffer for 30 min at 25 ◦ C. Reactions (50 μl) were initiated by the addition of SAM and a peptide representing histone H3 residues 21–44 containing C-terminal biotin (appended to a C-terminal amide-capped lysine). The final concentrations
Convergent evolution of histone H3 Lys27 methylation by human PRC2 and vSET
Figure 1
Comparison of required complexes for vSET and EZH2 activiry
(A) Sequence alignment of the SET domain of vSET, human EZH2 and human EZH1. The sequence identity between the SET domain of human EZH2 and EZH1 is 93 %, whereas the sequence identity between human EZH2 and vSET is 23 %. Asterisks indicate identical residues. (B) Two 13.5 kDa vSET molecules form an anti-co-operative homodimer where one vSET is active, and the other is inactive. (C) PRC2EZH2 is minimally active as a three-component complex of EZH2 (86 kDa), Suz12 (83 kDa) and EED (50 kDa).
of reagents for the vSET reactions were 2 nM enzyme, 300 nM [3 H]SAM, 1000 nM unlabelled SAM and 25 nM peptide. Final concentrations of reagents for the PRC2EZH2 reactions were 4 nM enzyme, 200 nM [3 H]SAM, 1600 nM unlabelled SAM and 200 nM peptide. The 100 % inhibition control consisted of 1 mM final concentration of the product inhibitor SAH, whereas the 0 % inhibition control consisted of 2 % DMSO. vSET reaction mixtures were incubated for 60 min at 25 ◦ C and PRC2EZH2 reaction mixtures were incubated for 90 min at 25 ◦ C and quenched with an excess of unlabelled SAM, transferred on to a 384-well flashplate, washed after 30 min with 0.l% Tween 20 and read on a TopCount NXT instrument. Concentration–response curves of percentage inhibition were plotted as a function of inhibitor concentration and fitted in GraphPad Prism to determine the IC50 values. These were converted into K i values using Cheng–Prusoff relationships [27] and assuming knowledge of the inhibition modality (see below), competing substrate K m value and concentration. RESULTS AND DISCUSSION
A fundamental tenet of biochemistry is that function follows structure in such a manner that cogent structure–activity relationships exist throughout Nature. However, there are also many examples throughout Nature in which distinct structural (or mechanical) solutions have evolved to solve a common functional need, at the molecular through to the organismal level. This latter situation appears to be the case in terms of enzyme-catalysed methylation of H3K27. Figure 1(A) compares the amino acid sequences of the SET domains of human EZH2 and EZH1 with that of vSET. Human EZH2 and EZH1 demonstrate a high degree of overall sequence conservation (93 % identity in the SET domain), and previous studies suggest that these two enzymes, each in the context of the whole PRC2 complex, share similar
243
enzymatic characteristics ([28] and T. Wigle, unpublished work). Hence, for the rest of the present paper, we focus our attention on comparisons between PRC2EZH2 and vSET; unless otherwise stated, the characteristics reported for PRC2EZH2 are similar to those found for PRC2EZH1 . The degree of sequence identity between the human and viral proteins, on the other hand, is quite low (23 % identity relative to the SET domain of EZH2). Despite the significant differences between the enzyme active sites at the primary sequence level, the data summarized in Figure 1(A) do not necessarily imply significant differences in protein folding, hence three-dimensional architecture. However, a clear distinction in enzymatic activity between PRC2EZH2 and vSET is realized at the three-dimensional structure level in that vSET folds into a tertiary structure that supports H3K27 methylation as a homodimer [20], without the need for additional protein subunits (Figure 1B), whereas EZH2 is only active in the context of the PRC2 complex (Figure 1C), requiring a minimum of two additional protein subunits, EED and Suz12, for activity and is optimally active in a four- or five-component complex with RbAp48 and AEBP2 (adipocyte enhancer-binding protein 2) [30,31]. In the present study, we explore similarities and differences between vSET and PRC2EZH2 enzymatic activity under identical buffer and temperature conditions. Despite these structural differences, PRC2EZH2 and vSET display similar preferences for substrates containing the equivalent of the H3K27 site. To investigate this, we tested the ability of each enzyme to catalyse methylation of large physiologically relevant substrates, i.e. nucleosomes and histones, and a peptide library representing all lysine sites on human histones H3 and H4. Figure 2 reveals that oligonucleosome substrates stimulate higher activity from both vSET and PRC2EZH2 than do mononucleosome substrates. These data are consistent with the ability of these enzymes to rapidly propagate the methylation of H3K27 along contiguous stretches of chromatin. Both enzymes are postulated to use H3K27me3 (trimethylated H3K27) recognition on one nucleosome unit to anchor and physically place the enzyme in close proximity to a neighbouring unmethylated H3K27 residue to enhance the efficiency of methylation. Although the overall ability to use H3K27me3, the product of both enzymes’ enzymatic activity, to stimulate additional proximal H3K27 methylation is conceptually similar for PRC2EZH2 and vSET, this process differs at the structural and mechanical level. vSET accomplishes this nucleosome-walking using anti-co-operative homodimers, where one inactive vSET molecule bound to H3K27me3 places an activated vSET molecule in close proximity to an unmethylated H3K27 residue [20]. PRC2EZH2 , however, utilizes the WD40 repeat domain present in the EED subunit to recognize H3K27me3 and orientate the EZH2 subunit to a proximal unmethylated H3K27 residue [32,33]. Additionally, both enzymes show distinct preferences for recombinant histone H3 as substrate, with little or no activity detectable on recombinant histone H4. A clear difference between the enzymes is apparent when histone H3/H4 tetramers and histone H2A/H2B/H3/H4 octamers are used as substrates. PRC2EZH2 shows a significant enhancement of activity on these substrates, whereas the activity for vSET is actually reduced with these substrates relative to recombinant histone H3. This observation can be rationalized by previous studies indicating that the RbAp48 subunit of PRC2 binds to histone H4 [34,35], which is present in the recombinant histone tetramers and octamers. Combining this observation with the recognition of H3K27me3 by EED, PRC2EZH2 activity appears to be more dependent on recognition elements distal to the direct site of methyl transfer (e.g. H3K27) in the context of protein multimers as substrates. c The Authors Journal compilation c 2013 Biochemical Society
244
B. M. Swalm and others
Figure 2 The activity of vSET and PRC2EZH2 on a panel of nucleosome and histone substrates The activity of the enzymes was compared by performing a reaction with 8 nM enzyme, 200 nM [3 H]SAM and 200 nM of the indicated substrate. Mono- and oligo-nucleosomes were purified from chicken erythrocytes, whereas recombinant tetramer and octamer respectively refer to histone H3/H4 tetramers or histone H2A/H2B/H3/H4 octamers. The activity was measured by capturing the substrates on to a filter binding plate and scintillation counting to determine the incorporation of 3 H-labelled methyl groups. Results are means + − S.D. for three experiments.
To investigate further the specificity of each enzyme for particular amino acid sequences, the velocity of methylation catalysed by PRC2EZH2 and vSET was measured against a set of overlapping peptides walking the length of histone H3 and histone H4 in five-amino-acid increments. Inspection of Figure 3 reveals that there is a strong preference of both enzymes for histone H3 peptides. vSET showed modest activity on peptides that contained H3K9 (histone H3 Lys9 ), and much greater activity on those that contained H3K27. In contrast, PRC2EZH2 was more selective and only showed activity on peptides that contained H3K27. The ARKS amino acid sequence is common between both the H3K9 (QTARKSTGG) and H3K27 (KAARKSAPA) marks; it appears that PRC2EZH2 is more stringent in discriminating among the residues flanking this sequence than is the smaller viral enzyme. To compare further the catalysis of H3K27 methylation by each enzyme, steady-state kinetic analysis of vSET activity on a peptide substrate containing the H3K27 residue was performed. Figure 4 illustrates typical data from the steady-state analysis in which we have simultaneously varied the concentration of SAM and peptide and measured the resulting reaction velocity. The cumulative data shown in Figure 4(A) are best fitted globally to the velocity equation for a ternary complex mechanism [36], yielding the following values of steady-state parameters: K m SAM = 1377 nM; K m peptide = 26 nM; α = 11; and kcat = 0.037 s − 1 . A diagnostic signature of a ternary complex enzyme mechanism is a series of intersecting lines when these data are shown as a double reciprocal plot, as illustrated in Figure 4(B). Additionally, competition studies were performed using the SAH and H3K27me3 peptide products. For vSET, SAH is competitive with SAM when the unmethylated H3K27 substrate peptide is saturating, and non-competitive with the unmethylated H3K27 peptide when SAM is saturating. Conversely, the H3K27me3 peptide is competitive with the unmethylated H3K27 substrate when SAM is saturating and non-competitive with SAM when the unmethylated H3K27 substrate peptide is saturating (results not shown). This pattern of product inhibition is most consistent with a random ordered mechanism involving abortive complex formation, or one of the Theorell– Chance ordered mechanisms [37]. Similar studies using PRC2EZH2 demonstrated that this enzyme also utilizes a ternary complex enzyme mechanism of catalysis (Figures 4C and 4D), c The Authors Journal compilation c 2013 Biochemical Society
Figure 3 The activity of vSET and EZH2 on a panel of overlapping biotinylated histone peptides covering the entire length of histone H3 The activity of vSET and PRC2EZH2 was compared by incubating 8 nM enzyme with 200 nM [3 H]SAM and 1000 nM of the biotinylated peptides. The activity was measured by capturing the peptides on to a streptavidin-coated flashplate and performing scintillation counting to determine the incorporation of the 3 H-labelled methyl groups. There was no activity detectable on any histone H4 peptides for either enzyme. The average results from duplicate experiments are plotted.
however, enzyme activation observed when using the H3K27me3 peptide does not allow one to distinguish between a compulsory or random ordered mechanism of catalysis. Both PRC2EZH2 and vSET catalyse the mono-, di- and trimethylation of H3K27. Previous data have demonstrated that human PRC2EZH2 displays a clear pattern of substrate use with respect to the methylation state of H3K27, thus PRC2EZH2 is most efficient at catalysing the first methylation reaction (from zero to one methyl group), and progressively less efficient at catalysing di- and tri-methylation of H3K27. As shown in Figure 5(A), vSET shows this same pattern of substrate utilization. The kinetic parameters determined from the steady-state experiments illustrated in Figure 5(A) are summarized in Table 1, where they are compared with the corresponding values for human PRC2EZH2 . The sequential catalysis of mono-, di- and tri-methylation of H3K27 can be accomplished by either a processive or distributive mechanism of catalysis. Processive catalysis presumes that the enzyme remains bound to a particular substrate molecule until all three rounds of methylation have been completed. In contrast, a distributive mechanism implies dissociation and rebinding of enzyme and substrate after each round of catalysis. Recent studies of the combined activities of the wild-type and Tyr641 /Ala677 mutant PRC2EZH2 in subsets of non-Hodgkin’s lymphoma cell lines clearly demonstrate that the enzyme functions through a distributive mechanism of catalysis [11]. To determine whether vSET operates by a processive or distributive mechanism, we followed the accumulation of each reaction product (mono-, di- and tri-methylated peptide) as a function of time after initiating the reaction with unmethylated peptide substrate. The results of these studies are illustrated in Figure 5(B). For a distributive mechanism the various reaction products are expected to accumulate sequentially so that one would expect to see a monotonic diminution of the substrate concentration, transient and sequential accumulation then loss of the monomethyl and dimethyl intermediates, and a lag followed by saturable accumulation of the final trimethyl product. In contrast, for a processive mechanism one would expect to see essentially
Convergent evolution of histone H3 Lys27 methylation by human PRC2 and vSET
Figure 4
245
Steady-state K m values for peptide substrate and SAM in vSET and PRC2EZH2 enzymatic reactions
A double substrate titration of a biotinylated histone H3 21–44 peptide with unmodified Lys27 and SAM was performed and the corresponding velocities were globally fitted to the equation for + a random order ternary complex mechanism. (A) For vSET the peptide K m value was determined to be 26 + − 3 nM, whereas the SAM K m value was 1377 − 320 nM. (B) For vSET, a plot of 1/velocity against 1/[SAM] yields a series of lines converging, diagnostic of a ternary enzyme complex mechanism. (C) Similarly, a double titration was performed on PRC2EZH2 and (D) a plot of 1/velocity against 1/[SAM] yields a series of lines converging, diagnostic of a ternary enzyme complex mechanism for PRC2EZH2 . Results are means + − S.D. for three experiments.
Table 1 Comparing the interaction of vSET and PRC2EZH2 with a series of active site ligands Product inhibitors and substrates are examined using kinetic parameters determined from steady-state experiments. Organic small molecules known to inhibit PRC2EZH2 , GSK126 and EPZ005687 were also examined for their ability to inhibit vSET. Unless otherwise stated, data shown are the means + − S.D. of three replicates.
K m , K 1/2 ∗ or K i (μM)
Figure 5
vSET is a distributive lysine methyltransferase
Ligand
Type
vSET
PRC2EZH2
SAM SAH Sinefungin H3 21–44, K27me0 H3 21–44, K27me1 H3 21–44, K27me2 H3 21–44 K27me3 EPZ005687 GSK126
Substrate Product Inhibitor Substrate Substrate Substrate Product Inhibitor Inhibitor
1.300 + − 0.320 1.299 + − 0.260 >50 0.021 + − 0.003 0.012 + − 0.006 0.008 + − 0.004 0.810 + − 0.170 >100 >100
1.244 + − 0.195 6.907 + − 0.210 >50 0.157 + − 0.012 0.337 + − 0.026 0.144 + − 0.011 Enzyme activation observed 0.019 + − 0.006† 0.002 + − 0.001
(A) The velocity of vSET methylation on histone H3 peptide substrates spanning residues 21–44 decreases as the methylation state of Lys27 goes from unmethylated to dimethylated. Activity was measured in a flashplate by scintillation counting of the 3 H-labelled methyl group incorporation into peptides representing the following methylation states H3K27me0 (䊉), H3K27me1 (䊏) or H3K27me2 (䉱). The results are means + − S.D. for three experiments. (B) The methylation of a histone H3 peptide spanning residues 21–44 by vSET was observed over time using SAMDI (self-assembled monolayer desorption/ionization) time-of-flight MS. The observed methylation state is indicated by the following symbols: H3K27me0 (䊉), H3K27me1 (䊏) or H3K27me2 (䉱) and H3K27me3 (䉬). The broken line at the 4 nM mark represents the amount of enzyme present in the reaction. Since the amount of mono- and di-methylated products accumulate to greater than 4 nM, vSET must be catalysing methylation of the peptide in a distributive manner. Data plotted are averages from duplicate experiments.
∗ As previously described [11,38,39], the peptides displayed sigmoidal behaviour with PRC2EZH2 and the data were fitted using a K 1/2 calculation rather than classic Michaelis–Menten fits to determine the concentration of peptide resulting in half-maximal velocity. †This value is the mean + − S.D. for six replicate experiments.
monotonic diminution of substrate and accumulation of the final product, with no accumulation of any intermediate species beyond the concentration of the enzyme present in the reaction (4 nM in the present case). The data presented in Figure 5(B) unambiguously demonstrate that vSET methylates H3K27 by a distributive mechanism, in concordance with previously reported data for this enzyme and for human PRC2EZH2 . To investigate further the apparent similarities between the catalytic mechanisms of PRC2EZH2 and vSET, the interaction of each enzyme with a series of ligands directed at the SAMand lysine-binding pockets were compared. Table 1 contrasts the K m values of the methyl donor SAM and histone H3 substrate
peptides containing un-, mono- and di-methylated Lys27 for each enzyme. In addition, the K i of the SAM-mimetic sinefungin and the product-based inhibitors SAH or histone H3 peptide with trimethylated Lys27 were directly compared. Inspection of these data reveals that PRC2EZH2 and vSET appear to have a similar affinity for nucleoside-based ligands, but differ significantly in their interactions with histone H3-based ligands. The K m values for SAH and K i values for SAH are within 3-fold of one another, whereas sinefungin does not appreciably inhibit either enzyme up to 100 μM. However, the K m values for histone H3 peptides with zero, one or two methyl groups on Lys27 vary between the two c The Authors Journal compilation c 2013 Biochemical Society
246
B. M. Swalm and others
enzymes 8–28-fold. Even more profound is the effect seen when attempting to measure the inhibition of enzymatic activity using an H3 peptide bearing a trimethylated Lys27 . vSET is inhibited by this product peptide with a K i of 0.8 μM, whereas PRC2EZH2 activity is enhanced 2–3-fold over the same concentration range of product peptide. The activation of EZH2 by the H3K27me3 product is consistent with a previous study on this effect [32] and is more evidence that PRC2EZH2 is more tightly regulated by interactions with distal sites on the nucleosome substrate. Perhaps the most striking difference between vSET and PRC2EZH2 is their relative sensitivity to small organic SAMcompetitive inhibitors. Two potent EZH2 selective-inhibitors have recently been reported and studied as antiproliferative drugs against non-Hodgkin’s lymophoma cells bearing mutations within EZH2: EPZ005687 (K i = 24 nM) and GSK126 (K i = 0.5 nM) [21,22]. In contrast with the potent inhibition of PRC2EZH2 demonstated by both of these compounds, the enzymatic activity of vSET was totally insensitive to these compounds at all concentrations tested (up to 100 μM). These data clearly reflect significant structural differences between the SAM-binding pockets of the two enzymes with respect to recognition elements important in drug-like inhibitor binding. In summary, despite the commonality of reaction catalysed, there appear to be meaningful differences in active site structure and interactions with small-molecule inhibitors between PRC2EZH2 and vSET that suggest convergent evolution of H3K27 methylation function for these two evolutionarily distant enzymes. Despite similarities in the interaction and utilization of nucleoside-based ligands, there are several key differences that exist between these enzymes with respect to size, overall structure and substrate recognition. Furthermore, the lack of inhibition of vSET by EPZ005687 and GSK126 reflects significant divergence of active site structure between these two enzymes that fundamentally precludes the use of vSET as a meaningful surrogate for PRC2 in studies aimed at understanding the structure–function relationship for the latter enzyme and especially in the design of pharmacological agents based on inhibition of enzyme activity.
AUTHOR CONTRIBUTION Margaret Scott, Mikel Moyer, Robert Copeland and Tim Wigle designed the experiments. Brooke Swalm, Kenneth Hallenbeck, Christina Majer and Tim Wigle performed the experiments. Lei Jin and Margaret Scott supplied the enzymes. Robert Copeland and Tim Wigle wrote the paper.
ACKNOWLEDGMENTS We thank Mr Michael Scholle at SAMDI Tech for assistance with self-assembled monolayer desorption/ionization time-of-flight MS and Dr Kevin Kuntz for supplying the inhibitor EPZ005687 and for helpful discussion in the preparation of this paper.
FUNDING This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.
REFERENCES 1 Zhang, X. and Bruice, T. C. (2008) Enzymatic mechanism and product specificity of SET-domain protein lysine methyltransferases. Proc. Natl. Acad. Sci. U.S.A. 105, 5728–5732 2 Smith, B. C. and Denu, J. M. (2009) Chemical mechanisms of histone lysine and arginine modifications. Biochim. Biophys. Acta 1789, 45–57 c The Authors Journal compilation c 2013 Biochemical Society
3 Richon, V. M., Johnston, D., Sneeringer, C. J., Jin, L., Majer, C. R., Elliston, K., Jerva, L. F., Scott, M. P. and Copeland, R. A. (2011) Chemogenetic analysis of human protein methyltransferases. Chem. Biol. Drug Des. 78, 199–210 4 Wu, H., Min, J., Lunin, V. V., Antoshenko, T., Dombrovski, L., Zeng, H., Allali-Hassani, A., Campagna-Slater, V., Vedadi, M., Arrowsmith, C. H. et al. (2010) Structural biology of human H3K9 methyltransferases. PLoS ONE 5, e8570 5 Rathert, P., Dhayalan, A., Murakami, M., Zhang, X., Tamas, R., Jurkowska, R., Komatsu, Y., Shinkai, Y., Cheng, X. and Jeltsch, A. (2008) Protein lysine methyltransferase G9a acts on non-histone targets. Nat. Chem. Biol. 4, 344–346 6 Varambally, S., Dhanasekaran, S. M., Zhou, M., Barrette, T. R., Kumar-Sinha, C., Sanda, M. G., Ghosh, D., Pienta, K. J., Sewalt, R. G., Otte, A. P. et al. (2002) The polycomb group protein EZH2 is involved in progression of prostate cancer. Nature 419, 624–629 7 Kleer, C. G., Cao, Q., Varambally, S., Shen, R., Ota, I., Tomlins, S. A., Ghosh, D., Sewalt, R. G., Otte, A. P., Hayes, D. F. et al. (2003) EZH2 is a marker of aggressive breast cancer and promotes neoplastic transformation of breast epithelial cells. Proc. Natl. Acad. Sci. U.S.A. 100, 11606–11611 8 Bracken, A. P., Pasini, D., Capra, M., Prosperini, E., Colli, E. and Helin, K. (2003) EZH2 is downstream of the pRB-E2F pathway, essential for proliferation and amplified in cancer. EMBO J. 22, 5323–5335 9 Simon, J. A. and Lange, C. A. (2008) Roles of the EZH2 histone methyltransferase in cancer epigenetics. Mutat. Res. 647, 21–29 10 Morin, R. D., Johnson, N. A., Severson, T. M., Mungall, A. J., An, J., Goya, R., Paul, J. E., Boyle, M., Woolcock, B. W., Kuchenbauer, F. et al. (2010) Somatic mutations altering EZH2 (Tyr641) in follicular and diffuse large B-cell lymphomas of germinal-center origin. Nat. Genet. 42, 181–185 11 Sneeringer, C. J., Scott, M. P., Kuntz, K. W., Knutson, S. K., Pollock, R. M., Richon, V. M. and Copeland, R. A. (2010) Coordinated activities of wild-type plus mutant EZH2 drive tumor-associated hypertrimethylation of lysine 27 on histone H3 (H3K27) in human B-cell lymphomas. Proc. Natl. Acad. Sci. U.S.A. 107, 20980–20985 12 Yap, D. B., Chu, J., Berg, T., Schapira, M., Cheng, S. W., Moradian, A., Morin, R. D., Mungall, A. J., Meissner, B., Boyle, M. et al. (2011) Somatic mutations at EZH2 Y641 act dominantly through a mechanism of selectively altered PRC2 catalytic activity, to increase H3K27 trimethylation. Blood 117, 2451–2459 13 Morin, R. D., Mendez-Lago, M., Mungall, A. J., Goya, R., Mungall, K. L., Corbett, R. D., Johnson, N. A., Severson, T. M., Chiu, R., Field, M. et al. (2011) Frequent mutation of histone-modifying genes in non-Hodgkin lymphoma. Nature 476, 298–303 14 McCabe, M. T., Graves, A. P., Ganji, G., Diaz, E., Halsey, W. S., Jiang, Y., Smitheman, K. N., Ott, H. M., Pappalardi, M. B., Allen, K. E. et al. (2012) Mutation of A677 in histone methyltransferase EZH2 in human B-cell lymphoma promotes hypertrimethylation of histone H3 on lysine 27 (H3K27). Proc. Natl. Acad. Sci. U.S.A. 109, 2989–2994 15 Shen, X., Liu, Y., Hsu, Y. J., Fujiwara, Y., Kim, J., Mao, X., Yuan, G. C. and Orkin, S. H. (2008) EZH1 mediates methylation on histone H3 lysine 27 and complements EZH2 in maintaining stem cell identity and executing pluripotency. Mol. Cell 32, 491–502 16 Manzur, K. L., Farooq, A., Zeng, L., Plotnikova, O., Koch, A. W., Sachchidanand and Zhou, M. M. (2003) A dimeric viral SET domain methyltransferase specific to Lys27 of histone H3. Nat. Struct. Biol. 10, 187–196 17 Qian, C., Wang, X., Manzur, K., Sachchidanand, Farooq, A., Zeng, L., Wang, R. and Zhou, M. M. (2006) Structural insights of the specificity and catalysis of a viral histone H3 lysine 27 methyltransferase. J. Mol. Biol. 359, 86–96 18 Mujtaba, S., Manzur, K. L., Gurnon, J. R., Kang, M., Van Etten, J. L. and Zhou, M. M. (2008) Epigenetic transcriptional repression of cellular genes by a viral SET protein. Nat. Cell. Biol. 10, 1114–1122 19 Li, J., Wei, H. and Zhou, M. M. (2011) Structure-guided design of a methyl donor cofactor that controls a viral histone H3 lysine 27 methyltransferase activity. J. Med. Chem. 54, 7734–7738 20 Wei, H. and Zhou, M. M. (2010) Dimerization of a viral SET protein endows its function. Proc. Natl. Acad. Sci. U.S.A. 107, 18433–18438 21 Knutson, S. K., Wigle, T. J., Warholic, N. M., Sneeringer, C. J., Allain, C. J., Klaus, C. R., Sacks, J. D., Raimondi, A., Majer, C. R., Song, J. et al. (2012) A selective inhibitor of EZH2 blocks H3K27 methylation and kills mutant lymphoma cells. Nat. Chem. Biol. 8, 890–896 22 McCabe, M. T., Ott, H. M., Ganji, G., Korenchuk, S., Thompson, C., Van Aller, G. S., Liu, Y., Graves, A. P., Della Pietra, 3rd, A., Diaz, E. et al. (2012) EZH2 inhibition as a therapeutic strategy for lymphoma with EZH2-activating mutations. Nature 492, 108–112 23 Kuriyan, J., Krishna, T. S., Wong, L., Guenther, B., Pahler, A., Williams, Jr, C. H. and Model, P. (1991) Convergent evolution of similar function in two structurally divergent enzymes. Nature 352, 172–174 24 Gherardini, P. F., Wass, M. N., Helmer-Citterich, M. and Sternberg, M. J. (2007) Convergent evolution of enzyme active sites is not a rare phenomenon. J. Mol. Biol. 372, 817–845
Convergent evolution of histone H3 Lys27 methylation by human PRC2 and vSET 25 Fang, J., Wang, H. and Zhang, Y. (2004) Purification of histone methyltransferases from HeLa cells. Methods Enzymol. 377, 213–226 26 Min, D. H., Yeo, W. S. and Mrksich, M. (2004) A method for connecting solution-phase enzyme activity assays with immobilized format analysis by mass spectrometry. Anal. Chem. 76, 3923–3929 27 Cheng, Y. and Prusoff, W. H. (1973) Relationship between the inhibition constant (K I ) and the concentration of inhibitor which causes 50 per cent inhibition (I 50 ) of an enzymatic reaction. Biochem. Pharmacol. 22, 3099–3108 28 Margueron, R., Li, G., Sarma, K., Blais, A., Zavadil, J., Woodcock, C. L., Dynlacht, B. D. and Reinberg, D. (2008) Ezh1 and Ezh2 maintain repressive chromatin through different mechanisms. Mol. Cell 32, 503–518 29 Reference deleted 30 Kuzmichev, A., Nishioka, K., Erdjument-Bromage, H., Tempst, P. and Reinberg, D. (2002) Histone methyltransferase activity associated with a human multiprotein complex containing the enhancer of zeste protein. Genes Dev. 16, 2893–2905 31 Cao, R. and Zhang, Y. (2004) SUZ12 is required for both the histone methyltransferase activity and the silencing function of the EED-EZH2 complex. Mol. Cell 15, 57–67 32 Margueron, R., Justin, N., Ohno, K., Sharpe, M. L., Son, J., Drury, 3rd, W. J., Voigt, P., Martin, S. R., Taylor, W. R., De Marco, V. et al. (2009) Role of the polycomb protein EED in the propagation of repressive histone marks. Nature 461, 762–767
247
33 Xu, C., Bian, C., Yang, W., Galka, M., Ouyang, H., Chen, C., Qiu, W., Liu, H., Jones, A. E., MacKenzie, F. et al. (2010) Binding of different histone marks differentially regulates the activity and specificity of polycomb repressive complex 2 (PRC2). Proc. Natl. Acad. Sci. U.S.A. 107, 19266–19271 34 Verreault, A., Kaufman, P. D., Kobayashi, R. and Stillman, B. (1996) Nucleosome assembly by a complex of CAF-1 and acetylated histones H3/H4. Cell 87, 95–104 35 Murzina, N. V., Pei, X. Y., Zhang, W., Sparkes, M., Vicente-Garcia, J., Pratap, J. V., McLaughlin, S. H., Ben-Shahar, T. R., Verreault, A., Luisi, B. F. and Laue, E. D. (2008) Structural basis for the recognition of histone H4 by the histone-chaperone RbAp46. Structure 16, 1077–1085 36 Copeland, R. A. (2005) Evaluation of Enzyme Inhibitors in Drug Discovery. A Guide for Medicinal Chemists and Pharmacologists, John Wiley and Sons, Hoboken 37 Purich, D. L. (2010) Enzyme Kinetics Catalysis and Control: a Reference of Theory and Best-Practice Methods, Academic Press, Amsterdam 38 Wigle, T. J., Knutson, S. K., Jin, L., Kuntz, K. W., Pollock, R. M., Richon, V. M., Copeland, R. A. and Scott, M. P. (2011) The Y641C mutation of EZH2 alters substrate specificity for histone H3 lysine 27 methylation states. FEBS Lett. 585, 3011–3014 39 Majer, C. R., Jin, L., Scott, M. P., Knutson, S. K., Kuntz, K. W., Keilhack, H., Smith, J. J., Moyer, M. P., Richon, V. M., Copeland, R. A. and Wigle, T. J. (2012) A687V EZH2 is a gain-of-function mutation found in lymphoma patients. FEBS Lett. 586, 3448–3451
Received 25 March 2013/6 May 2013; accepted 17 May 2013 Published as BJ Immediate Publication 17 May 2013, doi:10.1042/BJ20130439
c The Authors Journal compilation c 2013 Biochemical Society
