Sumoylation of the human histone H4 tail inhibits p300-mediated transcription by RNA polymerase II in cellular extracts

The post-translational modification of histones by the small ubiquitin-like modifier (SUMO) protein has been associated with gene regulation, centromeric localization, and double-strand break repair in eukaryotes. Although sumoylation of histone H4 was specifically associated with gene repression, this could not be proven due to the challenge of site-specifically sumoylating H4 in cells. Biochemical crosstalk between SUMO and other histone modifications, such as H4 acetylation and H3 methylation, that are associated with active genes also remains unclear. We addressed these challenges in mechanistic studies using an H4 chemically modified at Lys12 by SUMO-3 (H4K12su) and incorporated into mononucleosomes and chromatinized plasmids for functional studies. Mononucleosome-based assays revealed that H4K12su inhibits transcription-activating H4 tail acetylation by the histone acetyltransferase p300, as well as transcription-associated H3K4 methylation by the extended catalytic module of the Set1/COMPASS (complex of proteins associated with Set1) histone methyltransferase complex. Activator- and p300-dependent in vitro transcription assays with chromatinized plasmids revealed that H4K12su inhibits both H4 tail acetylation and RNA polymerase II-mediated transcription. Finally, cell-based assays with a SUMO-H4 fusion that mimics H4 tail sumoylation confirmed the negative crosstalk between histone sumoylation and acetylation/methylation. Thus, our studies establish the key role for histone sumoylation in gene silencing and its negative biochemical crosstalk with active transcription-associated marks in human cells.


Introduction
Chromatin is Nature's elegant architectural solution to the challenge of packing approximately 3 billion base-pairs of human genomic DNA in an average nuclear volume of only about 500 cubic microns. Histones constitute the main protein component of chromatin and their reversible post-translational modifications (PTMs), or marks, regulate chromatin structure and function by a range of direct and indirect mechanisms. 1 Based upon their association with either transcriptionally active or silenced regions of chromatin, histone marks were proposed to constitute an epigenetic code for gene function. 2 As a consequence of their early discovery and the development of modification-specific chemical and molecular biological tools, marks such as methylation, 3 acetylation 4 and ubiquitylation 5 have been extensively investigated in vitro and in cell culture. In contrast, histone modification by the small ubiquitin-like modifier (SUMO) protein is a poorly understood mark due both to its very low abundance in cells, which prevents the isolation of sumoylated histones in quantities required for biochemical analysis, and to a lack of sumoylated histone-specific antibodies for cellular studies. First reported in human HEK293T and P493-6 B cells by Shiio and Eisenman, 6 histone sumoylation also occurs in yeast, 7 parasitic protozoans, 8 and plants. 9 Similar to histone ubiquitylation, sumoylation occurs on all core histones, the linker histone H1, the histone variants H2A.Z and H2A.X and the centromeric histone variant Cse4 in yeast. 10,11 Pioneering efforts to identify specific lysine sites of sumoylation identified K12 in histone H4 as a major recurring site of sumoylation by SUMO-2/3 (H4K12su), 12,13 although multiple proximal lysines in the H4 N-terminal tail may also be enzymatically sumoylated in vitro. 10 Genetic studies in yeast and human cells have typically associated H4 sumoylation with the repression of gene transcription, although mechanistic studies of the direct roles for histone sumoylation in human cells have remained intractable due to the dynamic nature and low abundance of sumoylation. 6,14 In an effort to understand the direct effects of H4K12su in chromatin, we previously applied a disulfide-directed chemical sumoylation strategy to generate uniformly and sitespecifically sumoylated nucleosome arrays. 15 Biophysical studies of chromatin-array compaction remarkably showed that H4K12su is incompatible with the compact chromatin structures seen in transcriptionally silent heterochromatin. Subsequent biochemical studies revealed that H4K12su stimulates intranucleosomal activity of the H3K4me2-specific histone demethylase LSD1. 16 These studies suggested that sumoylated H4 does not directly enable heterochromatin formation and may instead act by recruiting LSD1 to genes. However, a potentially direct effect of histone H4 sumoylation on promoter-driven transcription by RNA polymerase II (RNAPII) and associated initiation factors that are key for efficient eukaryotic gene transcription has remained unknown.
Pioneering studies of the reconstitution of class II promoter-driven accurate eukaryotic transcription in both nuclear extracts and purified systems has led to insights into roles for histone modifications in gene function. 17,18 The ability to reconstitute chromatinized plasmid templates using chemically modified histones enables studies of the roles of specific histone modifications in transcription and investigations of their crosstalk with key enzymes associated with transcription initiation and elongation. 19 Multiple proteins involved in gene transcription bind to and modify histone tails, which enables the remodeling of chromatin prior to and during transcription. One such modification, acetylation of lysine side-chains on H3 and H4 by the acetyltransferase p300, is necessary for efficient activator-driven transcription of both 11-nm chromatin 20 and 30-nm linker histone H1-containing heterochromatin, likely through mechanisms that include direct decompaction of chromatin upon H4K16 acetylation and octamer eviction by the chromatin remodeler NAP1. 21,22 Due to their proposed opposing roles in gene transcription, we investigated the precise nature of biochemical crosstalk between histone sumoylation and histone acetylation by p300.
Histone H4 site-specifically sumoylated at Lys12 (H4K12su) was synthesized with the aid of a traceless ligation auxiliary, 2-aminooxyethanethiol, 23 and then incorporated into histone octamers for subsequent reconstitution of cognate mononucleosomes and chromatinized plasmids. Each sumoylated substrate was subjected to acetyltransferase assays with the full-length p300 enzyme, which revealed a consistent inhibition of acetylation in the H4K12su tail. Consistent with this observation and requirements for both H3 and H4 acetylation for in vitro transcription of chromatin, 20 replacing wild-type (wt) H4 with H4K12su in chromatinized plasmid templates dramatically inhibited p300-dependent, RNAPII-mediated transcription in vitro. Bottom-up mass spectrometry on chromatinized histones, following a novel in-gel desumoylation protocol, revealed decreased acetylation in H4K12su by p300 when compared to wt H4. Consistent with a role in gene repression, H4K12su also inhibited H3K4 methylation by the extended catalytic module of the Set1/COMPASS methyltransferase complex. 24 Collectively, our observations provide the first unambiguous biochemical demonstration that sumoylated histone H4 directly inhibits RNAPII-mediated transcription from chromatin templates, and reveal its direct negative crosstalk with histone acetylation by p300 and methylation by Set1/COMPASS that are strongly associated with active gene transcription.
Histone octamer acetylation by p300. We previously showed that H4K12su stimulates activity of the H3K4me1/2 demethylase, LSD1, in the context of a LSD1-CoREST sub-complex. 16 The stimulation of histone deacetylase (HDAC) activity of the Set3c complex in yeast was also recently proposed for sumoylated histone H2B. 25 Although the erasure of specific methyl and acetyl marks in the H3 and H4 tails may facilitate the transcriptionally repressed state of chromatin, there remains no information regarding the re-installation of these marks by the corresponding writer enzymes in the presence of H4K12su. Key among the histone acetyltransferases is the enzyme p300 that is recruited to chromatin by transcriptional activators for histone tail acetylation prior to transcription initiation. 20,26,27 Given its essential role in transcription, we investigated the effect of H4K12su on histone acetylation by p300 prior to and during transcription by RNAPII.
To investigate the direct biochemical crosstalk between H4K12su and acetylation, a Western-blot-based histone acetyltransferase (HAT) assay was developed using a sequenceindependent pan-acetyllysine antibody to detect lysine acetylation in all four histones ( Figure   S1A). In order to effectively compare acetylation of wt H4 with H4K12su and to strictly exclude any acetylation of the surface-exposed lysines in SUMO-3 attached to H4 (Figure S1B), it was proteolyzed from H4K12su prior to analysis. To this end, after acetylation by p300, the assay product was heat inactivated at 65 o C for 10 min, followed by addition of the purified catalytic domain of human SENP2 containing residues 365-590 ( Figure S1C). 28 Heat inactivation precluded p300 activity during desumoylation, and enabled the direct comparison of acetylation status in H4 and H4K12su.
Histone octamers containing wt H4 were first acetylated with full-length p300 immunoaffinity purified from HEK293T cells with an N-terminal FLAG epitope-tag ( Figure S1D). 29 Western blot analysis showed the robust acetylation of all four histones ( Figure 1E top panel and S1E). This is consistent with previous in vitro assays that revealed acetylation of all four histones by p300. 30 Strikingly, H4 from octamers containing H4K12su was devoid of acetylation, including H4K16ac ( Figure 1E bottom panel), which is strongly associated with chromatin decompaction and active gene transcription. 21,31,32 Importantly, the inhibition of H4 tail acetylation does not arise from allosteric inactivation of p300, based on the observation that the other histones in H4K12su octamers were acetylated to the same extent as in wt H4 octamers. Additionally, Western blots confirmed that SUMO-3 did not inhibit p300 autoacetylation, which is associated with robust acetyltransferase activity ( Figure S1F). Hence the inhibition of H4 acetylation in H4K12su is likely due to lysine acetylation site-occlusion by proximal SUMO-3 in the H4 tail.
Mononucleosome acetylation by p300. The histone acetylation assay was next undertaken with mononucleosomes containing either wt H4 or H4K12su. We failed to see significant nucleosome acetylation with pan-acetyllysine antibodies with or without pre-incubation of p300 with acetyl-CoA prior to the addition of nucleosomes ( Figure S1G). Due to the significantly decreased activity of p300 with nucleosomal substrates, a [ 3 H]-acetyl-CoA co-factor was employed and the transfer of acetyl groups to histones observed by fluorography. H4 acetylation was also suppressed in mononucleosomes containing H4K12su, but not in unmodified H4 mononucleosomes ( Figure 1F).
Our results unequivocally suggested that SUMO-3 in the H4 tail is inhibitory toward p300mediated acetylation of chromatin, a process that is necessary for active gene transcription.
Based on the lower acetyltransferase activity observed with mononucleosomes than with octamers, we wondered if dsDNA may inhibit p300 activity. To test this, an equimolar amount of free 147 bp Widom 601 dsDNA was included in the octamer acetylation assay. The presence of DNA was sufficient to inhibit p300 activity to a similar extent as observed with mononucleosomes ( Figure S1H). This unexpected inhibition of p300 activity by free DNA suggests that additional factors, such as transcription factor and RNAPII binding, enable robust p300 activity on histones during transcription initiation and elongation.
The effect of H4K12su on cell-free transcription from chromatinized templates. Based on our observation that H4K12su inhibits the acetylation of key H4 tail residues that are associated with p300-dependent active transcription, including H4K16 that is acetylated in euchromatin, we sought to investigate the direct effect of H4K12su on transcription in our reconstituted cell-free system.
The plasmid DNA template consisted of five gal4 binding sites and a ~400 bp G-less cassette. 27 Due to the absence of any engineered strong nucleosome positioning sequences in the template, chromatinization was undertaken with the histone chaperone NAP1 and the chromatin remodelers Acf1 and ISWI ( Figure S2). Limited micrococcal nuclease digestion of the transcription templates revealed the periodic spacing of nucleosomes in chromatin assembled with either wt H4 or H4K12su octamers, clearly indicating that H4K12su does not inhibit the formation of recombinant chromatin ( Figure 2B). In this background, addition of the transcription activator Gal4-VP16, p300, acetyl-CoA, [a-32 P]-CTP, rNTPs, and transcriptional machinery from a HeLa nuclear extract resulted in the transcription of a 365 base RNA from the chromatin template assembled with wt H4 histones ( Figure 2C). Surprisingly, and in contrast to the direct structural decompaction of chromatin by H4K12su, transcription from templates assembled with H4K12su was drastically inhibited when compared with templates assembled with wt H4. The addition of Trichostatin A (TSA), a nanomolar inhibitor of class I and II HDACs, did not lead to significant changes in transcription, indicating that the repressive effect of H4 K12su is not significantly mediated through HDAC1 in chromatinized templates assembled with non-acetylated histones. 33 Importantly, our results unambiguously demonstrated transcriptional repression when sitespecifically sumoylated H4 was present in chromatin.

H4 acetylation is inhibited prior to gene transcription in chromatin containing
H4K12su. Based on our observations with sumoylated octamer and nucleosome substrates, we wondered if the inhibition of transcription by H4K12su also correlated with diminished H4 tail acetylation by p300. Previous in vitro transcription studies with chromatinized plasmids containing either K-to-R mutations in the H4 tail or truncated H4 missing tail residues 1-19 revealed an ~80 % reduced transcriptional output relative to transcription from chromatin containing wt H4. 20 Chromatinized plasmids containing either wt H4 or H4K12su were incubated with Gal4-VP16, p300 and acetyl-CoA for 30 min to enable steps preceding transcription; and the histones were subsequently resolved by SDS-PAGE and analyzed by tandem mass spectrometry after chemical propionylation, trypsination, and separation by capillary-liquid chromatography ( Figure S3). 34 A critical innovation in the bottom-up analysis workflow was our use of the SENP2 catalytic domain to desumoylate H4K12su within the polyacrylamide gel matrix after SDS-PAGE. This procedural step was important to generate the same H4(4-17) tryptic peptide from wt H4 and H4K12su after p300-mediated acetylation. The H4(4-17) peptide contains K5,8,12 and 16 that are known to be acetylated by p300 in vitro and in vivo. 30 Analysis of the H4(4-17) tryptic peptides arising from wt H4 revealed a remarkable degree of hyperacetylation within 30 min. The most abundant peptide corresponded to the K5, 8,12,16 tetra-acetylated form with some tri-acetylated species also present ( Figure 3A, Table 1, and Figures   S4-S5). This is consistent with the fact that p300 acetylates histones to facilitate transcription. [20][21][22]27 No significant degree of monoacetylation was observed, and a low abundance of diacetylated peptide was detected after manually searching the MS-MS spectra over the expected elution time ( Figure S6). In comparison, chromatin assembled with H4K12su generated significantly fewer hyperacetylated peptides, with approximately equal amounts of tri-and di-acetylated H4 (4-17) peptides (Figures 3B and Figures S7-S8). Small amounts of unmodified H4(4-17) peptides were also observed ( Figure S9). This clearly indicates that H4K12su directly inhibits p300-mediated H4 tail acetylation in the steps prior to transcription. Given the importance of H4 tail acetylation for efficient transcription, H4K12su likely inhibits transcription, in-part, by directly inhibiting p300 activity on H4.
Since all four acetylated states of H4(4-17) were observed during tandem-MS-based analysis of p300 assay products with octameric and chromatin substrates, we interrogated the sitespecificity of p300 in the H4 tail. Consistent with previous reports, we observed that K5 and K8 are preferred sites in the double acetylated H4 tail, over acetylation at K12 and K16 (Table 2). 35 Additionally, K12 was preferentially acetylated over K16 in the triply acetylated H4 tail peptide (Table 3). These observations were consistent between H4 or H4K12su containing substrates, indicating that the intrinsic substrate preference of p300 is unchanged in the presence of SUMO-3 (Table 4).

H3K4 methylation by COMPASS is inhibited in nucleosomes containing H4K12su. Along with
H4K16ac, trimethylation at Lys 4 in histone H3 (H3K4me3) is a key mark associated with active gene transcription. 36 In humans, H3K4me3 is installed by the SETD1A/B and MLL-1/2 enzyme complexes, while the corresponding yeast enzyme complex is COMPASS (Complex of proteins associated with Set1). 37 Importantly, the catalytic module that imparts enzymatic activity and product specificity is evolutionarily conserved in animals and yeast, and consists of Set1 and the subunits Swd1, Bre2, Swd3 and Sdc1 in COMPASS. In an effort to understand the mechanism of auto-regulation in SET1/MLL enzymatic complexes, we recently reported the reconstitution and structural characterization of an extended catalytic module (eCM) of COMPASS that contains both the nSET domain of Set1 and the Spp1 subunit ( Figure 4A). 24 Although ubiquitylation at H2BK120 stimulates the methyltransferase activity of the eCM, it is not absolutely critical for nucleosome methylation by SET1/COMPASS complexes in vitro. 38 Based on our previous observation that H4K12su biochemically opposes the presence of H3K4me2 in nucleosomes by stimulating the activity of the H3K4me1/2 demethylase LSD1, we asked if H4K12su also directly opposes the installation of H3K4me3 in nucleosomes. Recent cryo-EM structures of the COMPASS eCM bound to the nucleosome core particle show significant spatial separation between the disordered H4 tail and the eCM ( Figure 4A), thereby making it hard to predict any biochemical crosstalk between sumoylation and methylation. 24,39 In order to shed light on this problem, methylation assays were undertaken with mononucleosome substrates containing either wt H4 or H4K12su and the six-subunit eCM ( Figure S10). The degree of H3K4me1/2/3 was measured by Western blot with antibodies specific for the different H3 methylation states ( Figure   4B). These experiments clearly showed that H4K12su inhibits the installation of H3K4me1/2/3 on nucleosomes. Moreover, the negative biochemical crosstalk arises from the presence of the Spp1 subunit in the eCM, because the core 5 protein catalytic module (Set1, Swd1, Bre2, Swd3, Sdc1) remained active on nucleosomes with or without the presence of H4K12su ( Figure 4C). Thus, we conclude that H4K12su in nucleosomes engages in negative biochemical crosstalk with p300mediated acetylation within the same H4 tail, in cis, and may engage in negative biochemical crosstalk with the COMPASS-mediated methylation in the H3 tail, in trans.

Discussion
Histone marks in eukaryotic chromatin represent a range of biological pathways that modulate chromatin structure and function. 40 Marks may directly alter chromatin structure either through their steric bulk or by changing the charge of amino acid side-chains. Additionally, marks may recruit chromatin-modifying enzymes that change the modification state of other histones within a nucleosome. The biochemical crosstalk between marks is considered positive when one mark directs installation of another, and negative when one mark opposes the installation of another. The specific sumoylation of histones in chromatin was associated with the repression of gene transcription through early studies in yeast and human cells. 6,14 We previously discovered the negative biochemical crosstalk between H4 sumoylation and H3K4me2 in nucleosomes mediated by the CoREST-LSD1 sub-complex, which suggested that sumoylation of actively transcribed regions enriched in H3K4me2 may lead to histone demethylation and silencing. 16 However, the direct effect of H4K12su in chromatin on p300-mediated gene transcription by RNAPII remained entirely unknown. Our semisynthesis of H4K12su using the ligation auxiliary, 2-aminooxyethanethiol, enabled the first interrogation of biochemical crosstalk between histone sumoylation and p300-catalyzed histone acetylation in reconstituted octamers, nucleosomes and chromatinized plasmids. Interestingly, we found that although H4K12su has no significant impact on the acetylation of other histones, in either histone octamer or nucleosomal contexts, it significantly impedes acetylation in the unstructured H4 N-terminal tail. While the simplest explanation is that steric bulk of the 93-amino acid SUMO-3 at K12 prevents acetylation at proximal lysines, this should not be taken for granted as ubiquitin-family modifications at lysines do not always occlude enzymatic activity at proximal sites. For example, the ubiquitylation at K119 in the H2A C-terminus by the polycomb repressive complex 1 (PRC1) E3 ligases Ring1B/Bmi1 does not inhibit ubiquitylation at K124/K127/K129 in the H2A tail by the BRCA1-BARD1 heterodimeric E3 ligase. 41 Therefore, reduced acetylation may equally arise from a specific spatial orientation of the H4 tail upon its sumoylation that limits access to the p300 active site.
Given the essential role of acetylated lysines in the H4 tail on chromatin structure and gene transcription in vitro and in cells, 20 we surmised that diminished H4 acetylation may adversely influence gene transcription. Surprisingly, replacing wt H4 with H4K12su had no significant effect on the efficiency of plasmid chromatinization by the histone chaperone Nap1 and remodelers ACF1 and ISWI, resulting in regularly positioned nucleosomes for both wt H4 and H4K12su.
Notably, however, an analysis of these chromatinized plasmid substrates in activator-and p300dependent transcription assays with nuclear extracts revealed a strong repression of transcription by H4K12su. Toward a further understanding of the H4 acetylation events whose loss leads to this repression, bottom-up mass spectral analysis with data-independent acquisition of tryptic H4 peptides after SDS-PAGE resolution of the transcription assay components demonstrated reduced acetylation in the H4(4-17) peptide when chromatin was reconstituted with H4K12su. To the best of our knowledge, this is the first mass-spectral analysis of acetylation in the tail of chromatinassociated H4 following in vitro transcription, and also the first demonstration that the catalytic domain of the SUMO-specific protease SENP2 can desumoylate histones in SDS-PAGE gels.
Given the inherent challenges of detecting SUMO target sites in substrates, due to the lack of a convenient trypsin cleavage site at the C-terminus of SUMO, the ability to selectively remove SUMO using SENP2 may be particularly useful for analyzing sumoylated proteins in complex mixtures that require some degree of separation by SDS-PAGE. The fact that chromatinized plasmids containing H2B ubiquitylated at K120 show similar levels of transcription to unmodified chromatin suggests that the inhibition of transcription we observed may not strictly be due to the steric bulk of SUMO in chromatin. 19 In addition to histone acetylation, another key histone mark associated with active transcription and enriched at promoter regions is H3K4me3. Installed by the Set1 containing COMPASS complex in yeast and the SET1/MLL1-4 family of methyltransferases in humans, H3K4me3 activates transcription in p53-and p300-dependent transcription from chromatinized plasmids 38 . And although H2BK120ub stimulates the methylation of H3K4, it is not absolutely critical for SET1 complex activity. 38 Structures of the 5-protein core catalytic module of COMPASS 42 and the 6-protein extended catalytic module were recently reported. 24 While the CM complex does not change methyltransferase activity in the presence of H2BK120ub, the eCM demonstrates some activity on nucleosomes that is further enhanced by the presence of H2BK120ub. 24 Consistent with this observation, the human SET1 complex also retains some in vitro activity on chromatinized templates lacking H2BK120ub. 38 From the cryo-EM structure of the eCM complex bound to nucleosomes, and the relative position of the unstructured H4 tail ( Figure 4A), we wondered if H4K12su would have an effect on H3K4me3 methylation by COMPASS. We discovered that although the CM is not hindered by the presence of H4K12su in nucleosomes, the eCM is significantly hindered by SUMO. From the differences in composition of the two subcomplexes, we propose that SUMO may sterically interact with the Spp1 subunit in the eCM and may reduce nucleosome binding and/or productive catalysis by the Set1 protein.
Interestingly, our results from methyltransferase assays, in conjunction with previous observations that H4 tail sumoylation inhibits chromatin compaction, appear to indicate that SUMO attached to the H4 tail does not extend away from the nucleosome, but instead may occupy a fixed space that prevents the close apposition of both adjacent nucleosomes and histone-modifying enzymes in chromatin. Future structural studies will aim to identify the precise placement of SUMO in sumoylated nucleosomes. As histone acetylation by p300 also stimulates SETD1 activity on chromatin, the direct effect of H4K12su on p300-stimulated SETD1 methylation at H3K4 remains an interesting question. 38 Collectively, the disruption of H4 tail acetylation and H3 tail methylation by the presence of H4K12su along with the inhibition of p300-mediated transcription from chromatinized templates have revealed multiple biochemical pathways by which histone sumoylation may inhibit gene transcription ( Figure 5). These results have shed light on important aspects of chromatin regulation by histone H4 sumoylation and provide a strong mechanistic basis for the proposed roles for SUMO from studies in yeast and cultured human cells. Table   Resource

Electrospray ionization mass spectrometry
Routine peptide/protein mass spectrometry was performed by direct infusion on a Bruker (Billerica, MA) Esquire ion-trap mass spectrometer operating in positive mode.
coli BL21(DE3) cells. 15 The insoluble histones were extracted from exclusion bodies with 6 M Gn-HCl, 10 mM Tris, pH 7.5. Histones were precipitated by dialysis against Millipore water in Spectra/Por 6 3.5 kDa molecular weight cut-off dialysis tubing and lyophilized to dryness. Crude histones were dissolved in 6 M Gn-HCl and purified by preparative C4 RP-HPLC.

Semisynthesis of H4K12su
The

Fluorography
After SDS-PAGE separation of methylation assay components, gels were soaked in Amplify solution (GE Healthcare, Chicago, IL) for 30 min before drying on a vacuum air dryer. Dried gels were exposed to X-ray film for 1 week at -80 °C. Images were developed and fixed using Kodak (Rochester, NY) GBX solutions.

Chromatin assembly and MNase digestion
Chromatin assembly and micrococcal nuclease analysis proceeded essentially as described previously. 22

In-gel desumoylation and propionylation of lysine residues
Histone modification analysis by mass spectrometry was conducted as previously described with one key modification.

In-gel tryptic digestion and peptide extraction
Histones were digested by rehydrating gel pieces with 12.5 ng/µL trypsin (Pierce) in 50 mM NH4HCO3 on ice for 30 min, and then incubation at 30 °C overnight. The bicarbonate solution after overnight digestion contained histone peptides that were transferred into a new empty tube.
Peptides were further extracted from the trypsinized gel pieces by the sequential addition of 20 µL     a Peptides were chemically propionylated before and after trypsinization to cap unmodified lysine side-chains and newly generated N-termini. b MS-MS spectra observed contained major fragments for the shown modification pattern over other potential patterns, however, no singly acetylated peptides were observed for wt H4. c Acetylation at K12 is not possible for H4K12su. d The triply acetylated peptide from H4K12su is blocked from acetylation at K12, but is propionylated after in-gel desumoylation. PSM= peptide spectral match. n.d. = not detected.