Abstract

Posttranslational modifications (PTMs) of core histones, such as histone methylation, play critical roles in a variety of biological processes including transcription regulation, chromatin condensation and DNA repair. In T. brucei, no domain recognizing methylated histone has been identified so far. TbTFIIS2-2, as a potential transcription elongation factors in T. brucei, contains a PWWP domain in the N-terminus which shares low sequence similarity compared with other PWWP domains and is absent from other TFIIS factors. In the present study, the solution structure of TbTFIIS2-2 PWWP domain was determined by NMR spectroscopy. TbTFIIS2-2 PWWP domain adopts a global fold containing a five-strand β-barrel and two C-terminal α-helices similar to other PWWP domains. Moreover, through systematic screening, we revealed that TbTFIIS2-2 PWWP domain is able to bind H4K17me3 and H3K32me3. Meanwhile, we identified the critical residues responsible for the binding ability of TbTFIIS2-2 PWWP domain. The conserved cage formed by the aromatic amino acids in TbTFIIS2-2 PWWP domain is essential for its binding to methylated histones.

Introduction

Posttranslational modifications (PTMs) of core histones, such as phosphorylation, acetylation, methylation and ubiquitination, affect chromatin structure which has been shown to have consequences for these downstream activities including transcription activation, chromatin condensation and signaling for DNA repair [1]. As one type of PTMs, histone methylation is generally associated with transcriptional activity of chromatin, and these effects are dependent upon the particular lysine that is methylated. For example, H3K9 and H4K20 methylation in human are associated with silent chromatin or heterochromatin, and H3K4 and H3K79 methylation are associated with euchromatin [2]. To execute these effects, histone methylation should be recognized by the readers of the histone code. Readers of histone methylation use different domains including Chromodomains, MBT (malignant brain tumor) fingers, PHD (plant homeo domain), Tudor and PWWP domains to recognize this modification [3].

The PWWP domain, named after the conserved motif Pro(P)–-Trp(W)–-Trp(W)–-Pro(P), displays affinities for DNA and methylated histones. The PWWP-containing proteins are usually involved in the processes of transcriptional regulation, DNA repair and DNA methylation [4]. Many PWWP domain-containing proteins which are mostly associated with chromatin have been identified in eukaryotes. In human, there are more than 22 PWWP domain-containing proteins, such as WHSC1 (Wolf–Hirschhorn syndrome candidate 1), HDGF2 (Hepatoma-derived growth factor 2), BRPF1 (bromodomain and PHD finger-containing protein 1), BRPF2, BRPF3, DNMT3A (DNA methyltransferases) and DNMT3B. Many of them were reported to possess methyl-lysine recognition activity: DNMT3A PWWP domain binds H3K36me3 [5], HDGF2 PWWP domain binds H3K79me3 and H4K20me3 [6] and ZMYND11 PWWP domain binds H3.3K36me3 [7].

Trypanosoma brucei, causing sleeping sickness in humans and nagana in cattle in sub-Saharan Africa, is an early branching unicellular eukaryotic parasite. As a result, the nuclear transcription of T. brucei displays unusual features, such as pol II transcription in T. brucei seems to be essentially constitutive and regulation is mainly post­transcriptional [8]. Consistent with this regulation, a smaller set of histone PTMs was identified in T. brucei [9,10]. The tails of the four core histones of T. brucei are highly diverged from those of mammals and yeasts, and many well-conserved modifications [9,10], such as the conserved H3K9, are absent from T. brucei. On the other hand, some PTMs are conserved between T. brucei and humans and might represent an ancient basal repertoire of histone modifications. For example, the homolog of human H4K20me3 is H4K18me3 and the corresponding homolog of human H3K36me3 is H3K32me3 in T. brucei [9,10].

The PWWP domain and PWWP-containing proteins in T. brucei are still little studied. TbTFIIS2-2, as a paralog of transcription elongation factor TbTFIIS2-1, contains a PWWP domain in the N-terminus which is absent from other TFIIS factors. In the present study, we determined the solution structure of TbTFIIS2-2 PWWP domain by NMR and investigated its interaction with methylated histones.

Experimental procedures

Cloning, expression and protein purification

The DNA encoding TbTFIIS2-2 PWWP domain (residues 1–110) was amplified from T. brucei genomic DNA by PCR and cloned into the vector pET-22b(+) (Novagen). The recombinant vectors were transformed into expression host BL21 (DE3). The recombinant TbTFIIS2-2 PWWP domain was expressed and purified as described previously [11]. 15N, 13C-labeled TbTFIIS2-2 PWWP domain was prepared in the same way except for that super broth was replaced by M9 medium containing 2.5 g/l 99% 13C-glucose and 0.5 g/l 99% 15NH4Cl as the sole carbon and nitrogen source, respectively.

NMR spectroscopy, data processing and structure calculation

All NMR spectra were acquired at 293K on a Bruker DMX 600 spectrometer. 1H-15N HSQC, CBCANH, CBCA(CO)NH, HBHA(CO)NH, HC(CO)NH, H(CC)ONH, 15N-edited NOESY and 13C-edited NOESY spectra were recorded for structure analysis. NMR data were processed with NMRpipe and NMRDraw software [12] and then evaluated using SPARKY3 [13]. All softwares were run on a Linux system. Firstly, we got medium- and long-range distance restraints from 15N-edited NOESY for structure calculations. Then, chemical shifts of five types of nuclei: 13Cα, 13Cβ, 13CO, 1Hα and 15NH were analyzed to predict backbone torsion angle restraints using TALOS program [14]. Dihedral angel restraints were introduced in consecutive steps. Structure was calculated using the program CYANA [15]. Dihedral angle restraints, hydrogen bonds and all other NOEs were introduced in consecutive steps. Five hundred conformers were totally calculated independently, of which 20 lowest-energy structures were selected and analyzed by MOLMOL [16] and PROCHECK [17] online (http://nihserver.mbi.ucla.edu/SAVES_3/).

NMR chemical shift perturbation

Chemical shift perturbation was performed to investigate the ability of the TbTFIIS2-2 PWWP domain recognizing its ligands. 15N-labeled TbTFIIS2-2 PWWP domain (0.2 mM) was titrated with different DNA fragments to a molar ratio (PWWP/DNA: 1 : 10) and different methylated peptides to a molar ratio (PWWP/peptide: 1 : 5). Peptides and DNA stock solutions in identical buffer were titrated with a sample dilution of less than 10% of total volume. 1H-15N HSQC spectra were acquired for every titration for analysis. Kd values were calculated by a nonlinear least-squares analysis in Kaleidagraph using the equation: 
formula
where CL is the concentration of the peptide, CP is the concentration of the protein, Δδ is the observed normalized chemical shift change and Δδmax is the normalized chemical shift change at saturation, calculated as 
formula
where δ is the chemical shift in ppm. α = 0.14 for most residues, α = 0.2 for glycine [18]. The threshold values used as a criterion for significance were 60% for intensity reduction and 0.1 ppm for chemical shift perturbation. The sequences of DNA fragments and peptides are shown in Supplementary Tables S1 and S2.

Site-directed mutagenesis

A series of point mutations (F18A, W21L and F18A/W21A) were introduced into the recombinant pET-22b (+)-TbTFIIS2-2 PWWP vector. The recombinant plasmid pET-22b (+)-TbTFIIS2-2 PWWP was used as a template, amplified by PCR using PrimeSTARTM HS DNA polymerase (TaKaRa, Dalian, China) and two complementary (partially overlapping) primers containing the desired mutation. 50 µl PCR was carried out with 50–100 ng templates, 10 µM primer pair, 200 µM dNTPs and 2 U of DNA polymerase and started at 98°C for 2 min, followed by 25 cycles of 98°C for 40 s, 60°C for 40 s, and a final extension at 72°C for 6 min and 30 s. After PCR, the PCR product was digested with Dpn I (TaKaRa, Dalian, China) overnight to remove methylated parental non-mutated plasmid, and then transformed into Escherichia coli BL21 (DE3).

Fluorescence polarization assays

Fluorescence polarization assays (FPAs) were performed at 293 K using a SpectraMax M5 Microplate Reader system (Molecular Devices). The wavelengths of fluorescence excitation and emission were 485 and 528 nm, respectively. Each well of a 96-well plate contained 1 µM 5′-FITC-labeled H3K32me3 or H4K17me3 peptides and different amounts of TbTFIIS2-2 PWWP domain or TbTFIIS2-2 PWWP domain mutants (concentrations from 0 to ∼4 mM) with a final volume of 200 µl. For each assay, peptide-free controls (TbTFIIS2-2 PWWP domain or TbTFIIS2-2 PWWP domain mutants) were included. The fluorescence polarization data were analyzed essentially as described previously [19]. The sequences of 5′-FITC-labeled peptides (H3K32me3 and H4K17me3) are shown in Supplementary Table S2.

Results

TbTFIIS2-2 PWWP domain shares low sequence similarity with other PWWP domains

TbTFIIS2-2, distinct from other TFIIS factors, can be divided into two parts: an additional PWWP domain at the N-terminus and an unconserved region at the C-terminus (Figure 1A). Sequence alignments of the PWWP domains were performed using ClustalW2 and the output files were processed using ESPript (http://espript.ibcp.fr/ESPript/ESPript/) [20,21]. TbTFIIS2-2 PWWP domain shares less than 10% sequence identity and similarity compared with other PWWP domains (Figure 1B). Although TbTFIIS2-2 PWWP domain only contains the conserved motif (P–W–W–P) and the several similar residues preceding this motif in comparison with other PWWP domains, it exhibits common secondary structure including five β-sheets and two C-terminal α-helices similar to other PWWP domains (Figure 1B).

Domain architecture and sequence alignments of TbTFIIS2-2 PWWP.

Figure 1.
Domain architecture and sequence alignments of TbTFIIS2-2 PWWP.

(A) Diagram of the TbTFIIS2-2 PWWP domain architecture. (B) Multiple sequence alignments of TbTFIIS2-2 PWWP domain with other PWWP domains by ClustalW and ESPript. Identical residues were enclosed in red box, and conserved residues are colored in red. The secondary structure of TbTFIIS2-2 PWWP domain determined is also included at the top of the sequences.

Figure 1.
Domain architecture and sequence alignments of TbTFIIS2-2 PWWP.

(A) Diagram of the TbTFIIS2-2 PWWP domain architecture. (B) Multiple sequence alignments of TbTFIIS2-2 PWWP domain with other PWWP domains by ClustalW and ESPript. Identical residues were enclosed in red box, and conserved residues are colored in red. The secondary structure of TbTFIIS2-2 PWWP domain determined is also included at the top of the sequences.

Solution structure of the TbTFIIS2-2 PWWP domain

The solution structure of TbTFIIS2-2 PWWP domain was determined by a set of NMR spectra. The atomic co-ordinates for 20 lowest-energy structures of TbTFIIS2-2 PWWP domain have been deposited in the Protein Data Bank with the PDB ID code 2NAS. The assembly of the 20 structures and a best representative are shown in Figure 2A,B. Statistics of the solution structure of TbTFIIS2-2 PWWP domain are summarized in Table 1. In total, 2134 NOE distance restraints and 126 dihedral angle restraints were included in the structure calculation (Table 1). The residues in the secondary structure are well defined with the RMSD of 0.38 Å for the backbone. The statistical parameters in Table 1 indicate the high-quality NMR structures of TbTFIIS2-2 PWWP domain. The Ramachandran plot was analyzed by PROCHECK [17] to check the quality of the structure. The data showed that 80.4% residues are in the most favored region, 17.4% and 2.2% residues are within the additional and generously allowed regions, respectively.

NMR structure of TbTFIIS2-2 PWWP domain.

Figure 2.
NMR structure of TbTFIIS2-2 PWWP domain.

(A) The ensemble of 20 lowest-energy structures. The figures were produced with MOLMOL and Pymol. (B) The ribbon representation of the representative structure of TbTFIIS2-2 PWWP domain with the secondary structure elements highlighted. (C) Electrostatic surface diagram of the lowest-energy conformation of TbTFIIS2-2 PWWP domain is shown from two different orientations 180° apart (red, negative; blue, positive; white, neutral).

Figure 2.
NMR structure of TbTFIIS2-2 PWWP domain.

(A) The ensemble of 20 lowest-energy structures. The figures were produced with MOLMOL and Pymol. (B) The ribbon representation of the representative structure of TbTFIIS2-2 PWWP domain with the secondary structure elements highlighted. (C) Electrostatic surface diagram of the lowest-energy conformation of TbTFIIS2-2 PWWP domain is shown from two different orientations 180° apart (red, negative; blue, positive; white, neutral).

Table 1
NMR structural statistics
NMR restraints in the structure calculation 
 Intraresidue 236 
 Sequential (|i − j| = 1) 740 
 Medium-range (1<|i − j| < 5) 514 
 Long-range (|i − j| ≥5) 620 
 Hydrogen bonds 24 
 Total distance restraints 2134 
 Dihedral angle restraints 126 
Residual violations 
 CYANA target functions, Å 1.10 
 NOE upper distance constrain violation  
 Maximum, Å 0.12 ± 0.04 
 Number > 0.2 Å 0 ± 0 
 Dihedral angle constrain violations  
 Maximum, ° 4.51 ± 0.36 
 Number > 5° 0 ± 0 
 Van der Waals violations  
 Maximum, Å 0.28 ± 0.01 
Number > 0.2 Å 2 ± 1 
Average structural RMSD to the mean co-ordinates (Å) 
 Secondary structure backbone1 0.32 
 All backbone atoms2 0.38 
 All heavy atoms2 0.98 
Ramachandran statistics, % of all residues 
 Most favored regions 80.4 
 Additional allowed regions 17.4 
 Generously allowed regions 2.2 
 Disallowed regions 0.0 
NMR restraints in the structure calculation 
 Intraresidue 236 
 Sequential (|i − j| = 1) 740 
 Medium-range (1<|i − j| < 5) 514 
 Long-range (|i − j| ≥5) 620 
 Hydrogen bonds 24 
 Total distance restraints 2134 
 Dihedral angle restraints 126 
Residual violations 
 CYANA target functions, Å 1.10 
 NOE upper distance constrain violation  
 Maximum, Å 0.12 ± 0.04 
 Number > 0.2 Å 0 ± 0 
 Dihedral angle constrain violations  
 Maximum, ° 4.51 ± 0.36 
 Number > 5° 0 ± 0 
 Van der Waals violations  
 Maximum, Å 0.28 ± 0.01 
Number > 0.2 Å 2 ± 1 
Average structural RMSD to the mean co-ordinates (Å) 
 Secondary structure backbone1 0.32 
 All backbone atoms2 0.38 
 All heavy atoms2 0.98 
Ramachandran statistics, % of all residues 
 Most favored regions 80.4 
 Additional allowed regions 17.4 
 Generously allowed regions 2.2 
 Disallowed regions 0.0 
1

Includes residues in secondary structure: 10–14, 20–26, 42–46, 53–57, 63–67, 78–82, 84–103.

2

Obtained for residues L10–Q103 since no long NOEs were identified for amino acids 1–9 and 104–110.

The solution structure of TbTFIIS2-2 PWWP domain adopts a classical PWWP fold, consisting of three motifs: a canonical β-barrel core, an insertion motif between the second and third β-strands and a C-terminal α-helix bundle (Figure 2B). The five-strand antiparallel β-barrel core is followed by two α-helices at the C-terminus. The central β-barrel is composed of strand 1 (residues 10–14), strand 2 (residues 20–26), strand 3 (residues 42–46), strand 4 (residues 53–57) and strand 5 (residues 63–67). All of the β-strands are linked by tight turns, except that β2 and β3 are joined by a flexible loop and the N- and C-termini of the structure are on the same side of the molecule, similar to other PWWP domains (Figure 2B). The C-terminal region consists of a short α-helix (residues 78–82) and a long α-helix (residues 84–103) which is common in PWWP domains. The long α-helix is fixed in a groove between β2 and the β3/β4 loop through extensive contact with residues on one side of the β-barrel (Figure 2B). The whole TbTFIIS2-2 PWWP protein is highly negatively charged except for the surface of β3 and β4 and the C-terminal tail (Figure 2C). This is also consistent with its primary sequence containing lots of acidic amino acids. This feature implies that the TbTFIIS2-2 PWWP domain might not interact with negatively charged molecules such as DNA.

Structural comparison between TbTFIIS2-2 PWWP domain and other PWWP domains

The structure of TbTFIIS2-2 PWWP domain was submitted to the program DALI to search for its similar structures [22]. Indeed, the result showed that the structures most similar to that of TbTFIIS2-2 PWWP domain in PDB entries are those of PWWP domains of human. The Cα RMSD between TbTFIIS2-2 PWWP domain (PDB ID: 2NAS) and human BRPF1 PWWP domain (PDB ID: 2X4Y), WHSC1L1 PWWP domain from Homo sapiens (PDB ID: 2DAQ) and Pdp1 PWWP domain from Schizosaccharomyces pombe are 3.8, 3.2 and 4.6 Å, with corresponding Z-scores of 5.9, 5.6 and 5.0, respectively. Structural and topological comparisons of TbTFIIS2-2 PWWP domain with the PWWP domain of human BRPF1 PWWP (PDB ID: 2X4Y) [23] and the Pdp1 PWWP domain from S. pombe (PDB ID: 2L89) [24] showed that the TbTFIIS2-2 PWWP domain is similar to the canonical PWWP fold (Figure 3).

Structural (upper) and corresponding topological (below) comparisons between TbTFIIS2-2 PWWP domain from Trypanosoma brucei (PDB ID: 2NAS), BRPF1 PWWP domain from Homo sapiens (PDB ID: 2X4Y) and Pdp1 PWWP domain from Schizosaccharomyces pombe (PDB ID: 2L89).

Figure 3.
Structural (upper) and corresponding topological (below) comparisons between TbTFIIS2-2 PWWP domain from Trypanosoma brucei (PDB ID: 2NAS), BRPF1 PWWP domain from Homo sapiens (PDB ID: 2X4Y) and Pdp1 PWWP domain from Schizosaccharomyces pombe (PDB ID: 2L89).

These PWWP domains all share the classical fold except for the insertion motif between β2 and β3 and the C-terminal helix bundle. The red box encloses the variant insertion motifs.

Figure 3.
Structural (upper) and corresponding topological (below) comparisons between TbTFIIS2-2 PWWP domain from Trypanosoma brucei (PDB ID: 2NAS), BRPF1 PWWP domain from Homo sapiens (PDB ID: 2X4Y) and Pdp1 PWWP domain from Schizosaccharomyces pombe (PDB ID: 2L89).

These PWWP domains all share the classical fold except for the insertion motif between β2 and β3 and the C-terminal helix bundle. The red box encloses the variant insertion motifs.

Some differences were also observed such as the insertion region between β2 and β3 and the C-terminal α-helix bundle. β2 and β3 formed an extended antiparallel β-sheet are connected by a flexible loop in TbTFIIS2-2 PWWP domain and S. pombe Pdp1 PWWP domain, whereas the corresponding β-sheet in human BRPF1 PWWP domain is joined by an α-helix and a short β-sheet (Figure 3). The C-terminal regions of the three PWWP domains are very variable because the number of α-helix in this region is different (Figure 3). However, the three PWWP domains contain the common α-helix which is placed in a groove between β2 and the β3/β4 loop in a fixed direction (Figure 3). Sequence analysis indicated that the structural variations of PWWP domains are consistent with their different primary sequences in these regions (Figure 3).

TbTFIIS2-2 PWWP domain binds to H4K17me3 and H3K32me3 through its conserved aromatic cage

The PWWP domain has been shown to bind either DNA or methyl-lysine histones. Both TbTFIIS2-1 and TbTFIIS2-2 contain a PWWP domain, which may contribute to transcription regulation by recognizing methylated histones or DNA. However, our previous work on TbTFIIS2-1 PWWP domain indicated that TbTFIIS2-1 PWWP domain is not able to bind to methylated histones or DNA [25]. As described previously, the surface of TbTFIIS2-2 PWWP domain is highly negatively charged (Figure 2C), we therefore speculated that this PWWP domain might not bind to DNA either. To verify it, we used NMR titration to measure the DNA-binding ability of the PWWP domain. AT- and GC-rich ssDNA and dsDNA sequences (Supplementary Table S1) were used to titrate TbTFIIS2-2 PWWP domain to a molar ratio of 1:10 (PWWP:DNA). No peak in HSQC spectra was observed to shift obviously (Supplementary Figure S1), which suggested that TbTFIIS2-2 PWWP domain has no DNA-binding affinity, similar to TbTFIIS2-1 PWWP domain [25].

According to the identified methyl-lysine histones in T. brucei [3,9,10], six peptides: H3K32me3, H3K76me3, H4K2me2, H4K17me3, H4K18me2 and H4K18me3 (Supplementary Table S2) were synthesized. By NMR chemical shift perturbation, we investigated the interactions between TbTFIIS2-2 PWWP domain and six synthetic peptides. The results showed that the resonances of TbTFIIS2-2 PWWP domain in HSQC spectra underwent obvious shift only upon addition of H4K17me3 and H3K32me3 peptides (H4K17me3 and H3K32me3 in T. brucei are the counterparts of human H4K20me3 and H3K36me3, respectively) [3], suggesting that TbTFIIS2-2 PWWP domain is able to recognize trimethylated H4K17 and H3K32 (Figure 4 and Supplementary Figure S2). Analysis of the chemical shift perturbations afforded the dissociation constants (Kd) of 0.68 ± 0.16 and 0.64 ± 0.15 mM for the binding of TbTFIIS2-2 PWWP domain to trimethylated H4K17 and H3K32 (Figure 4C), respectively, which are comparable with the Kd of 2.7 ± 0.2 mM for the binding of Brpf1 PWWP domain to H3K36me3 and Kd of 6.0 ± 1.7 mM for the interactions between Pdp1 PWWP domain and H4K20me3 [23,24]. The similar binding affinity of these PWWP domains implied that H4K17me3 and H3K32me3 might be the endogenous interaction partners of TbTFIIS2-2 PWWP domain. By comparison, TbTFIIS2-1 PWWP domain did not have the ability [25]. To further confirm this interaction, H4K17 and H3K32 with different degrees of methylation were also investigated. TbTFIIS2-2 PWWP domain only recognized H4K17me3 and H3K32me3 (Supplementary Figures S3 and S4). Meanwhile, the residues with obvious perturbation before and after addition of H4K17me3 peptides were almost the same as those of H3K32me3 (Figure 4C,D). The residues showing a substantial shift of resonance or obvious reduction in peak intensity included L13, K14, Q15, D16, F18, W21, F24, W43, C45, C46, S51, T53 and L54. Interestingly, these residues are located at one end of the β-barrel where a conserved aromatic cage is formed by F18, W21 and C46 (Figure 4D–F).

The TbTFIIS2-2 PWWP domain binds to H3K32me3 and H4K17me3.

Figure 4.
The TbTFIIS2-2 PWWP domain binds to H3K32me3 and H4K17me3.

1H-15N HSQC NMR spectra of the 15N-labelled TbTFIIS2-2 PWWP domain in the absence (red) and presence (green) of H4K17me3 (A) and H3K32me3 (B) peptides, showing chemical shift perturbations of many residues. (C) Plots of the chemical shift changes of seven well-resolved amide resonances against H4K17me3 and H3K32me3 peptide concentrations. Each dissociation constant is determined by fitting the data to a single-site ligand binding model, and the overall Kd are 0.68 ± 0.16 mM and 0.64 ± 0.15 mM for H3K32me3 and H4K17me3 peptides by averaging the individual ones, respectively. (D) The ratio of peak intensity before and after the addition of H4K17me3 or H3K32me3. (E) The histogram displays 1H-15N chemical shift changes observed in the corresponding spectra of the TbTFIIS2-2 PWWP domain shown in (A and B). The residues (Pro) that could not be assigned were labeled with the black pentagram. The residues that had significant changes and were untraceable in 1H-15N HSQC spectra after the additions of H4K17me3 or H3K32me3 were labeled with the red pentagram. The lines represent the significance level of intensity reductions >60% (intensity ratios <40%) and chemical shift changes >0.1 ppm, respectively. All the comparisons were performed at TbTFIIS2-2 PWWP/peptide molar ratio of 1 : 5. (F) Ribbon representation of TbTFIIS2-2 PWWP domain displaying residues colored in red (Δδ>0.1 ppm or intensity ratios <40%) which were involved in its interactions with H4K17me3 and H3K32me3.

Figure 4.
The TbTFIIS2-2 PWWP domain binds to H3K32me3 and H4K17me3.

1H-15N HSQC NMR spectra of the 15N-labelled TbTFIIS2-2 PWWP domain in the absence (red) and presence (green) of H4K17me3 (A) and H3K32me3 (B) peptides, showing chemical shift perturbations of many residues. (C) Plots of the chemical shift changes of seven well-resolved amide resonances against H4K17me3 and H3K32me3 peptide concentrations. Each dissociation constant is determined by fitting the data to a single-site ligand binding model, and the overall Kd are 0.68 ± 0.16 mM and 0.64 ± 0.15 mM for H3K32me3 and H4K17me3 peptides by averaging the individual ones, respectively. (D) The ratio of peak intensity before and after the addition of H4K17me3 or H3K32me3. (E) The histogram displays 1H-15N chemical shift changes observed in the corresponding spectra of the TbTFIIS2-2 PWWP domain shown in (A and B). The residues (Pro) that could not be assigned were labeled with the black pentagram. The residues that had significant changes and were untraceable in 1H-15N HSQC spectra after the additions of H4K17me3 or H3K32me3 were labeled with the red pentagram. The lines represent the significance level of intensity reductions >60% (intensity ratios <40%) and chemical shift changes >0.1 ppm, respectively. All the comparisons were performed at TbTFIIS2-2 PWWP/peptide molar ratio of 1 : 5. (F) Ribbon representation of TbTFIIS2-2 PWWP domain displaying residues colored in red (Δδ>0.1 ppm or intensity ratios <40%) which were involved in its interactions with H4K17me3 and H3K32me3.

The structures of the PWWP domain of human BRPF1 in complex with H3K36me3 [23] and human ZMYND11 PWWP domain in complex with H3.3K36me3 [7] have shed light on the molecular mechanism of methylated histone binding by PWWP domains. In the BRPF1–H3K36me3 complex structure, the H3K36me3 peptide is accommodated in an aromatic cage at one end of the β-barrel formed by three aromatic residues. The aromatic cage within ZMYND11 PWWP domain is also essential for the trimethyl-lysine binding [7]. The structural comparison between the TbTFIIS2-2 PWWP domain and other PWWP domains in complex with methylated peptides revealed that all of these PWWP domains share high structural similarities in the β-barrel core and sequence analysis showed that the aromatic residues (F18, W21) of TbTFIIS2-2 PWWP domain for methylated lysine recognition are conserved (Figure 5A,B). Except F18 and W21, the third aromatic residue forming the conserved cage in PWWP domain is substituted by C46 in TbTFIIS2-2 PWWP domain (Figure 5A). However, this difference has no effect on its recognitions of H4K17me3 and H3K32me3, suggesting the first two residues (F18, W21) are more important for binding to the methylated histones.

TbTFIIS2-2 PWWP domain interacts with H4K17me3 and H3K32me3 by the two aromatic residues (F18, W21).

Figure 5.
TbTFIIS2-2 PWWP domain interacts with H4K17me3 and H3K32me3 by the two aromatic residues (F18, W21).

(A) Sequence alignments of TbTFIIS2-2 PWWP domain with other PWWP domains with methyl-lysine binding ability. All of these PWWP domains with methyl-lysine binding ability contain the three aromatic residues (labeled by green triangles) formed an aromatic cage. The alignment was generated by using ClustalW and ESPript. (B) Comparison of TbTFIIS2-2 PWWP domain with the K36me3-binding PWWP domain of ZMYND11. The structures of TbTFIIS2-2 PWWP domain and ZMYND11 PWWP domain (PDB: 4N4H) are aligned and shown in cyan and gray, respectively. F291, W294 and F310 of ZMYND11 PWWP domain that is critical for H3.3K36me3 binding are shown as sticks in blue. The corresponding residues, F18, W21 and C46 of TbTFIIS2-2 PWWP domain, are shown as sticks in red. H3.3K36me3 peptide is shown in yellow with trimethylated lysine 36 shown as sticks. The aromatic rage of TbTFIIS2-2 PWWP domain is critical for H3K32me3 and H4K17me3 binding. FPAs of the wild-type TbTFIIS2-2 PWWP domains and mutants (F18A, W21L and F18A/W21A), with 5′-FITC-labelled H4K17me3 (C) and H3K32me3 (D). (E) Coomassie blue-stained SDS–PAGE gel of proteins used in FPAs.

Figure 5.
TbTFIIS2-2 PWWP domain interacts with H4K17me3 and H3K32me3 by the two aromatic residues (F18, W21).

(A) Sequence alignments of TbTFIIS2-2 PWWP domain with other PWWP domains with methyl-lysine binding ability. All of these PWWP domains with methyl-lysine binding ability contain the three aromatic residues (labeled by green triangles) formed an aromatic cage. The alignment was generated by using ClustalW and ESPript. (B) Comparison of TbTFIIS2-2 PWWP domain with the K36me3-binding PWWP domain of ZMYND11. The structures of TbTFIIS2-2 PWWP domain and ZMYND11 PWWP domain (PDB: 4N4H) are aligned and shown in cyan and gray, respectively. F291, W294 and F310 of ZMYND11 PWWP domain that is critical for H3.3K36me3 binding are shown as sticks in blue. The corresponding residues, F18, W21 and C46 of TbTFIIS2-2 PWWP domain, are shown as sticks in red. H3.3K36me3 peptide is shown in yellow with trimethylated lysine 36 shown as sticks. The aromatic rage of TbTFIIS2-2 PWWP domain is critical for H3K32me3 and H4K17me3 binding. FPAs of the wild-type TbTFIIS2-2 PWWP domains and mutants (F18A, W21L and F18A/W21A), with 5′-FITC-labelled H4K17me3 (C) and H3K32me3 (D). (E) Coomassie blue-stained SDS–PAGE gel of proteins used in FPAs.

To further confirm that the aromatic cage is important for TbTFIIS2-2 PWWP domain to recognize methylated peptides, we engineered three mutations including F18A, W21L and F18/W21A (the protein with C46A mutation precipitated significantly at the concentration suitable for study) and tested the binding abilities of these mutants by FPAs. The results showed that the mutations of F18A and W21L reduced the histone-binding activity of TbTFIIS2-2 PWWP domain significantly and double mutation lost the binding ability (Figure 5C–E). This result confirmed that the two residues play an important role in histone binding. The result of FPAs also indicated that the ability of TbTFIIS2-2 PWWP domain binding to H4K17me3 is similar to that of TbTFIIS2-2 PWWP domain binding to H3K32me3.

Although the two PWWP domains of TbTFIIS2-1 and TbTFIIS2-2 both adopt conserved PWWP fold, the conserved aromatic residues necessary for binding methylated histone are absent from TbTFIIS2-1 PWWP domain according to the sequence alignments, which might explain the reason why TbTFIIS2-1 PWWP domain has no ability of recognizing methylated histone [25]. This difference further confirms the two aromatic residues are necessary for PWWP domains to bind methylated histone and also implies that TbTFIIS2-1 and TbTFIIS2-2 might play a different role in histone recognition.

Discussion

In T. brucei, three TFIIS orthologs named as TbTFIIS1, TbTFIIS2-1 and TbTFIIS2-2 have been identified [26]. A significant difference of TFIIS between trypanosoma and other eukaryotes is reflected by a unique PWWP domain in TbTFIIS2-1 and TbTFIIS2-2, which has not been found in any other TFIIS factors so far. As a histone code reader, the PWWP domain may contribute to transcription regulation by recognizing methylated histones or DNA. In the present study, we determined the solution structure of TbTFIIS2-2 PWWP domain. Although PWWP domains share low sequence similarity, TbTFIIS2-2 PWWP domain adopts a global fold containing a five-strand β-barrel and two C-terminal α-helices similar to other PWWP domains, albeit some differences exist.

Although core histones are the most evolutionarily conserved proteins, there are substantial differences between trypanosome histones and those of higher eukaryotes, especially in the highly modified amino­terminal tails [3,9,10]. The absence of many well-conserved post-transcriptional modifications (PTMs) and the discovery of some unusual and apparently trypanosome-specific PTMs make it very difficult to determine which kind of methyl-lysine histones the PWWP domains interact with. A survey of T. brucei PTMs using Edman degradation and mass spectrometry has revealed some PTMs of histones [9,10]. In our previous study, we indicated that TbTFIIS2-1 PWWP domain shows no binding ability to DNA or methylated histones [25]. Here, through systematic screening, we revealed that TbTFIIS2-2 PWWP domain interacts with H3K32me3 and H4K17me3, which is the first domain identified to bind methyl-lysine histones in T. brucei.

Similar to other PWWP domains, TbTFIIS2-2 PWWP domain binds to methylated histone through a conserved aromatic cage. In most PWWP domains bearing the binding ability of methylated histone, the conserved aromatic cage is formed by three aromatic residues: the third residue (W/Y) of the P–W–W–P motif, the residue (F/Y/W) immediately preceding this motif and the residue (F/Y/W) located at the end of the β3-strand. TbTFIIS2-2 PWWP domain has two aromatic residues (F18 and W21) whose mutation abolished its ability of binding to methylated histones, suggesting the importance of these two residues in histone binding. However, the third aromatic residue is absent from TbTFIIS2-2 PWWP domain and substituted by Cys46 which has no effect on its recognition of H4K17me3 and H3K32me3, implying that the third residue forming the conserved cage might not be essential for the binding affinity. In total, the pattern of TbTFIIS2-2 PWWP domain binding to methylated histones is similar to those of other PWWP domains.

Histone methylation is always implicated in the regulation of gene transcription. The H4K17me3 and H3K32me3 in T. brucei are corresponding to human H4K20me3 and H3K36me3, respectively. In human, H4K20me3 is often associated with silent chromatin or heterochromatin and required for DNA-damage checkpoint activation [2730]. As the counterpart of H4K20me3 in trypanosoma, H4K17me3 might also be involved in these similar processes. Besides, human H3K36me3 is not restricted to actively transcribed regions only and may contribute to the composition of heterochromatin, in combination with other histone modifications [31]. In other eukaryotes, methylation of H3K36 by lysine methyltransferase Set2 is linked to transcriptional elongation by RNA polymerase II [32]. Therefore, the recognition of H3K32me3 by TbTFIIS2-2 PWWP domain might imply its role in the transcription regulation.

Abbreviations

     
  • FPAs

    fluorescence polarization assays

  •  
  • MBT

    malignant brain tumor

  •  
  • PHD

    plant homeo domain

  •  
  • PTMs

    posttranslational modifications

Author Contribution

R.W., J.G., S.L. and X.T. designed the research. R.W., J.G., S.L. and J.Z. performed the experiments and data analysis, and R.W., J.G., X.Z., C.X. and X.T. wrote the paper.

Funding

This work was supported by the National Natural Science Foundation of China [grant number U1332137 to X.T. and 31500601 to S.L. and 31570737 to C.X. and 21573205 to Z.Z.].

Acknowledgements

We thank F. Delaglio and A. Bax for providing NMRPipe and NMRDraw, T. D. Goddard and D. Kneller for Sparky, T. Herrmann, K. Wüthrich and Peter Güntert for CYANA, R. Koradi and K. Wüthrich for MOLMOL.

Competing Interests

The Authors declare that there are no competing interests associated with the manuscript.

References

References
1
Strahl
,
B.D.
and
Allis
,
C.D.
(
2000
)
The language of covalent histone modifications
.
Nature
403
,
41
45
2
Sims
,
R.J.
,
Nishioka
,
K.
and
Reinberg
,
D.
(
2003
)
Histone lysine methylation: a signature for chromatin function
.
Trends Genet.
19
,
629
639
3
Figueiredo
,
L.M.
,
Cross
,
G.A.
and
Janzen
,
C.J.
(
2009
)
Epigenetic regulation in African trypanosomes: a new kid on the block
.
Nat. Rev. Microbiol.
7
,
504
513
4
Stec
,
I.
,
Nagl
,
S.B.
,
van Ommen
,
G.J.B.
and
den Dunnen
,
J.T.
(
2000
)
The PWWP domain: a potential protein–protein interaction domain in nuclear proteins influencing differentiation?
FEBS Lett.
473
,
1
5
5
Dhayalan
,
A.
,
Rajavelu
,
A.
,
Rathert
,
P.
,
Tamas
,
R.
,
Jurkowska
,
R.Z.
,
Ragozin
,
S.
et al.  (
2010
)
The Dnmt3a PWWP domain reads histone 3 lysine 36 trimethylation and guides DNA methylation
.
J. Biol. Chem.
285
,
26114
26120
6
Wu
,
H.
,
Zeng
,
H.
,
Lam
,
R.
,
Tempel
,
W.
,
Amaya
,
M.F.
,
Xu
,
C.
et al.  (
2011
)
Structural and histone binding ability characterizations of human PWWP domains
.
PLoS ONE
6
,
e18919
7
Wen
,
H.
,
Li
,
Y.
,
Xi
,
Y.
,
Jiang
,
S.
,
Stratton
,
S.
,
Peng
,
D.
et al.  (
2014
)
ZMYND11 links histone H3.3K36me3 to transcription elongation and tumour suppression
.
Nature
508
,
263
268
8
Clayton
,
C.E.
(
2002
)
Life without transcriptional control? From fly to man and back again
.
EMBO J.
21
,
1881
1888
9
Janzen
,
C.J.
,
Fernandez
,
J.P.
,
Deng
,
H.
,
Diaz
,
R.
,
Hake
,
S.B.
and
Cross
,
G.A.
(
2006
)
Unusual histone modifications in Trypanosoma brucei
.
FEBS Lett.
580
,
2306
2310
10
Mandava
,
V.
,
Fernandez
,
J.P.
,
Deng
,
H.
,
Janzen
,
C.J.
,
Hake
,
S.B.
and
Cross
,
G.A.
(
2007
)
Histone modifications in Trypanosoma brucei
.
Mol. Biochem. Parasitol.
156
,
41
50
11
Zhang
,
J.
,
Dai
,
K.
,
Liao
,
S.
and
Tu
,
X.
(
2013
)
1H, ¹³C and 15N resonance assignments of TbTFIIS2-2 PWWP domain from trypanosoma brucei
.
Biomol. NMR Assign.
7
,
207
209
12
Delaglio
,
F.
,
Grzesiek
,
S.
,
Vuister
,
G.W.
,
Zhu
,
G.
,
Pfeifer
,
J.
and
Bax
,
A.D.
(
1995
)
NMRPipe: a multidimensional spectral processing system based on UNIX pipes
.
J. Biomol. NMR
6
,
277
293
13
Goddard
,
T.D.
and
Kneller
,
D.G.
(
1993
)
SPARKY3
,
University of California
,
San Francisco
14
Shen
,
Y.
,
Delaglio
,
F.
,
Cornilescu
,
G.
and
Bax
,
A.
(
2009
)
TALOS+: a hybrid method for predicting protein backbone torsion angles from NMR chemical shifts
.
J. Biomol. NMR
44
,
213
223
15
Güntert
,
P.
,
Mumenthaler
,
C.
and
Wüthrich
,
K.
(
1997
)
Torsion angle dynamics for NMR structure calculation with the new program DYANA
.
J. Mol. Biol.
273
,
283
298
16
Koradi
,
R.
,
Billeter
,
M.
and
Wüthrich
,
K.
(
1996
)
MOLMOL: a program for display and analysis of macromolecular structures
.
J. Mol. Graph.
14
,
51
55
17
Laskowski
,
R.A.
,
MacArthur
,
M.W.
,
Moss
,
D.S.
and
Thornton
,
J.M.
(
1993
)
PROCHECK: a program to check the stereochemical quality of protein structures
.
J. Appl. Crystallogr.
26
,
283
291
18
Jurkowska
,
R.Z.
,
Qin
,
S.
,
Kungulovski
,
G.
,
Tempel
,
W.
,
Liu
,
Y.
,
Bashtrykov
,
P.
et al.  (
2017
)
H3k14ac is linked to methylation of H3K9 by the triple Tudor domain of SETDB1
.
Nat. Commun.
8
,
2057
19
Liu
,
Y.
,
Huang
,
H.
,
Zhou
,
B.O.
,
Wang
,
S.S.
,
Hu
,
Y.
,
Li
,
X.
et al.  (
2010
)
Structural analysis of Rtt106p reveals a DNA binding role required for heterochromatin silencing
.
J. Biol. Chem.
285
,
4251
4262
20
Gouet
,
P.
,
Courcelle
,
E.
and
Stuart
,
D.I.
(
1999
)
ESPript: analysis of multiple sequence alignments in PostScript
.
Bioinformatics
15
,
305
308
21
Higgins
,
D.G.
and
Sharp
,
P.M.
(
1988
)
CLUSTAL: a package for performing multiple sequence alignment on a microcomputer
.
Gene
73
,
237
244
22
Holm
,
L.
and
Sander
,
C.
(
1993
)
Protein structure comparison by alignment of distance matrices
.
J. Mol. Biol.
233
,
123
138
23
Vezzoli
,
A.
,
Bonadies
,
N.
,
Allen
,
M.D.
,
Freund
,
S.M.
,
Santiveri
,
C.M.
,
Kvinlaug
,
B.T.
et al.  (
2010
)
Molecular basis of histone H3K36me3 recognition by the PWWP domain of Brpf1
.
Nat. Struct. Mol. Biol.
17
,
617
619
24
Qiu
,
Y.
,
Zhang
,
W.
,
Zhao
,
C.
,
Wang
,
Y.
,
Wang
,
W.
,
Zhang
,
J.
et al.  (
2012
)
Solution structure of the Pdp1 PWWP domain reveals its unique binding sites for methylated H4K20 and DNA
.
Biochem. J.
442
,
527
538
25
Wang
,
R.
,
Zhang
,
J.
,
Liao
,
S.
and
Tu
,
X.
(
2016
)
Solution structure of TbTFIIS2-1 PWWP domain from Trypanosoma brucei
.
Proteins, Struct. Funct. Bioinformatics
84
,
912
919
26
Uzureau
,
P.
,
Daniels
,
J.P.
,
Walgraffe
,
D.
,
Wickstead
,
B.
,
Pays
,
E.
,
Gull
,
K.
et al.  (
2008
)
Identification and characterization of two trypanosome TFIIS proteins exhibiting particular domain architectures and differential nuclear localizations
.
Mol. Microbiol.
69
,
1121
1136
27
Botuyan
,
M.V.
,
Lee
,
J.
,
Ward
,
I.M.
,
Kim
,
J.E.
,
Thompson
,
J.R.
,
Chen
,
J.
et al.  (
2006
)
Structural basis for the methylation state-specific recognition of histone H4-K20 by 53BP1 and Crb2 in DNA repair
.
Cell
127
,
1361
1373
28
Sanders
,
S.L.
,
Portoso
,
M.
,
Mata
,
J.
,
Bähler
,
J.
,
Allshire
,
R.C.
and
Kouzarides
,
T.
(
2004
)
Methylation of histone H4 lysine 20 controls recruitment of Crb2 to sites of DNA damage
.
Cell
119
,
603
614
29
Schotta
,
G.
,
Lachner
,
M.
,
Sarma
,
K.
,
Ebert
,
A.
,
Sengupta
,
R.
,
Reuter
,
G.
et al.  (
2004
)
A silencing pathway to induce H3-K9 and H4-K20 trimethylation at constitutive heterochromatin
.
Genes Dev.
18
,
1251
1262
30
Couture
,
J.F.
,
Collazo
,
E.
,
Brunzelle
,
J.S.
and
Trievel
,
R.C.
(
2005
)
Structural and functional analysis of SET8, a histone H4 Lys-20 methyltransferase
.
Genes Dev.
19
,
1455
1465
31
Chantalat
,
S.
,
Depaux
,
A.
,
Héry
,
P.
,
Barral
,
S.
,
Thuret
,
J.Y.
,
Dimitrov
,
S.
et al.  (
2011
)
Histone H3 trimethylation at lysine 36 is associated with constitutive and facultative heterochromatin
.
Genome Res.
21
,
1426
1437
32
Krogan
,
N.J.
,
Kim
,
M.
,
Tong
,
A.
,
Golshani
,
A.
,
Cagney
,
G.
,
Canadien
,
V.
et al.  (
2003
)
Methylation of histone H3 by Set2 in Saccharomyces cerevisiae is linked to transcriptional elongation by RNA polymerase II
.
Mol. Cell. Biol.
23
,
4207
4218

Author notes

*

These authors contributed equally to this work.