Fusion protein based on double-stranded DNA deaminase and epigenetic modification detection method

Abstract

The invention discloses a fusion protein based on double-stranded DNA deaminase and an epigenetic modification detection method. The fusion protein comprises a molecule having a specific affinity for an antibody, a linker peptide, and a double-stranded DNA deaminase. Double-stranded DNA deaminase is guided to a specific antibody through a molecule with specific affinity to the antibody, so that a deamination reaction is performed on a DNA sequence of a target site of the antibody, and a chromatin-associated protein targeted by the antibody or the position of the chromatin-associated protein modified on the double-stranded DNA sequence is detected according to the determined position of the DNA deamination site.

Description

Technical Field

[0001] The present invention relates to a fusion protein based on double-stranded DNA deaminase and an epigenetic modification detection method. Background Art

[0002] Transcription factors, chromatin-associated proteins, and histone modifications play a key role in gene regulation and epigenetic research. They work together to regulate the on or off state of genes. Transcription factors bind to specific DNA sequences to promote or inhibit the transcription of related genes. At the same time, the openness of chromatin and the state of histone modifications affect the openness of specific genomic regions, thereby controlling the functional state of the genome. For example, different histone modifications such as acetylation, methylation, and ubiquitination can regulate the compactness of chromatin and the level of gene transcription, affecting the level of gene expression. In recent years, with the development of high-throughput technology, a variety of technologies have emerged to detect the localization of transcription factors, chromatin-associated proteins and histone modifications on the genome.

[0003] Chromatin immunoprecipitation (ChIP) is a classic method for studying protein-DNA interactions, which can reveal transcription factor binding sites and histone modification characteristics. In 2007, ChIP technology was combined with high-throughput second-generation sequencing to form chromatin immunoprecipitation sequencing technology (ChIP-Seq), which significantly promoted the development of epigenetic and gene regulation research. . ChIP-Seq uses formaldehyde to fix cells or tissues, randomly interrupt chromatin after lysis, precipitate the target protein-DNA complex through specific antibodies, and then extract DNA by reverse cross-linking and sequence. Sequencing data can be used to identify protein binding sites throughout the genome, providing key information for studying gene regulation. However, the application of ChIP-Seq is limited by high sample requirements, limited signal-to-noise ratio, and high dependence on antibody quality. .

[0004] In recent years, researchers have developed CUT&RUN (Cleavage Under Targets and Release Using Nuclease) (10,11) and CUT&Tag (Cleavage Under Targets and Tagmentation) technologies. . Both methods utilize Protein A that can bind to antibodies, and achieve antibody-guided targeted DNA cutting or transposition by expressing microsphere nuclease (CUT&RUN) or Tn5 transposase (CUT&Tag) fused to Protein A and binding to target protein-specific antibodies. CUT&RUN uses MNase to shear DNA near the target protein, while CUT&Tag uses Tn5 to insert linkers at both ends of the DNA fragment to mark the binding area. These two methods significantly simplify the experimental process, improve the signal-to-noise ratio and sensitivity, and greatly reduce the amount of sample required, and can even be applied to single-cell analysis. However, since both CUT&RUN and CUT&Tag involve DNA fragmentation, they are only applicable to second-generation sequencing technology, cannot integrate single-molecule long-read sequencing, and have problems with compatibility with multiple genomics technologies.

[0005] In addition, some new technologies developed in recent years have further expanded the research possibilities. In 2022, the targeted methylation and long-read sequencing technology (DiMeLo-Seq) will fuse N6 adenine methyltransferase with Protein A to perform 6mA methylation near the target protein binding site, and combine nanopore sequencing or PacBio HiFi sequencing to detect protein-DNA interactions on DNA molecules. Although DiMeLo-Seq's amplification-free strategy has the advantages of high accuracy and long read length, in order to retain methylation information, its DNA samples cannot be amplified and the demand for starting materials is high (1-2 million cells / experiment), which limits its application in rare samples.

[0006] CUT&Tag, CUT&RUN and DiMeLo-Seq all use protein A to bind to the Fc segment of the antibody to achieve the combination of enzyme and antibody. However, protein A has different affinities for antibodies from different species. For example, it has a strong affinity for rabbit antibodies, but a weak affinity for mouse antibodies. With the continuous development of related technologies, two different technical routes have emerged. One route is nano-CT, which further replaces the protein A in Cut&Tag with a more specific nano-antibody to recognize the Fc segment of antibodies from specific species. The other method is to fuse protein A and protein G to achieve a more universal affinity for antibodies of different species. .

[0007] The discovery of double-stranded DNA deaminases provides a new tool for genomic research. Unlike single-stranded DNA deaminases, double-stranded DNA deaminases such as DddA can catalyze cytosine deamination reactions in double-stranded DNA, converting cytosine into uracil. In 2020, Joseph Mougous and David Liu's research group first reported the development of DdCBE by fusing double-stranded deaminase DddA with TALEN, achieving single-base precise editing of the mitochondrial genome, providing an important tool for the study of mitochondrial genetic diseases. Joseph Mougous's research group subsequently established 3D-seq technology, which overexpressed DddA and a specific transcription factor in uracil-tolerant (ung gene knockout) living bacteria, converting cytosine near the transcription factor's genomic binding site into uracil. The uracil was then detected as thymine during PCR amplification and next-generation sequencing to determine the localization of the transcription factor. However, DddA is highly toxic in living cells and can only be used in uracil-tolerant bacteria (ung gene knockout). In addition, each assay requires fusion expression with a specific transcription factor, which limits its throughput and application in eukaryotic cells. Summary of the invention

[0008] In order to overcome the above-mentioned defects of the prior art, the inventors have developed a method for detecting the interaction between chromatin-associated proteins and DNA based on targeted double-stranded DNA deaminase. The inventors expressed protein A and protein G with specific affinity for antibodies by fusion with double-stranded DNA deaminase, and the antibodies had specific affinity for chromatin-associated proteins or chromatin-associated protein modifications. The double-stranded DNA deaminase was guided to the position where specific chromatin-associated proteins or chromatin-associated protein modifications bind to DNA by utilizing the characteristics of protein A-protein G protein binding antibodies, and the double-stranded DNA deaminase deaminated the DNA at that position, thereby achieving the localization of chromatin-associated proteins or chromatin-associated protein modifications on double-stranded DNA sequences. This process not only provides a more accurate modification detection method, but also avoids the DNA fragmentation problem that may exist in common modification detection methods, making it an ideal DNA binding protein detection technology. This detection method of the present invention is named ANDIE (Antibody-associated DeamInasE).

[0009] To achieve the above-mentioned object, the present invention provides a fusion protein in the first aspect, which comprises, from the N-terminus to the C-terminus: a molecule having a specific affinity for an antibody, a connecting peptide and a double-stranded DNA deaminase, wherein the molecule having a specific affinity for the antibody is Protein A and / or Protein G, and the antibody is an antibody having a specific affinity for a chromatin-associated protein or a chromatin-associated protein modification.

[0010] In one embodiment of the first aspect of the present invention, the double-stranded DNA deaminase is the active domain of double-stranded DNA deaminase DddAtox.

[0011] In one embodiment of the first aspect of the present invention, the Protein A sequence is Seq ID No. 1; the Protein G protein sequence is Seq ID No. 2; the connecting peptide sequence is Seq ID No. 3; and / or the double-stranded DNA deaminase sequence is SeqID No. 4.

[0012] In one embodiment of the first aspect of the present invention, the chromatin-associated protein is at least one protein selected from DNA binding proteins, transcription factors and histones, or a protein bound to at least one protein among DNA binding proteins, transcription factors and histones.

[0013] The present invention also provides a polynucleotide encoding the above-mentioned fusion protein of the present invention.

[0014] The present invention also provides a nucleic acid vector, which comprises the polynucleotide of the present invention as described above.

[0015] The invention also provides a host cell, which comprises the above-mentioned polynucleotide of the invention or the above-mentioned nucleic acid vector of the invention.

[0016] In a second aspect, the present invention provides a method for detecting the position of a chromatin-associated protein or a chromatin-associated protein modification on a double-stranded DNA sequence, the method comprising:

[0017] Step a. connecting a molecule having a specific affinity for an antibody to a double-stranded DNA deaminase, wherein the antibody has a specific affinity for a chromatin-associated protein or a chromatin-associated protein modification, and through the binding of the antibody to the chromatin-associated protein or the chromatin-associated protein modification, the double-stranded DNA deaminase performs a targeted deamination reaction on a DNA sequence near the chromatin-associated protein or the chromatin-associated protein modification bound by the antibody,

[0018] Step b. detecting the position of the chromatin-associated protein or chromatin-associated protein modification bound by the antibody on the double-stranded DNA sequence by determining the position of the DNA deamination site,

[0019] Wherein, the molecule having specific affinity for the antibody is Protein A and / or Protein G having specific affinity for the Fc domain of the antibody.

[0020] In one embodiment of the second aspect of the present invention, the double-stranded DNA deaminase refers to an enzyme that can deaminize cytosine and / or adenine on double-stranded DNA.

[0021] In one embodiment of the second aspect of the present invention, the molecule having specific affinity for the antibody is linked to a double-stranded DNA deaminase as a protein fusion and expressed.

[0022] In one embodiment of the second aspect of the present invention, the linkage between the molecule having specific affinity for the antibody and the double-stranded DNA deaminase is affinity linkage.

[0023] In one embodiment of the second aspect of the present invention, the linkage between the molecule having specific affinity for the antibody and the double-stranded DNA deaminase is covalent linkage.

[0024] In one embodiment of the second aspect of the present invention, the chromatin-associated protein is at least one protein selected from DNA binding proteins, transcription factors and histones, or a protein bound to at least one protein among DNA binding proteins, transcription factors and histones.

[0025] In one embodiment of the second aspect of the present invention, the DNA deamination site is determined by detecting the bases that undergo sequence conversion after PCR amplification and sequencing.

[0026] In one embodiment of the second aspect of the present invention, the binding sequence of the antibody and DNA is determined by analyzing the position of the deamination site.

[0027] In one embodiment of the second aspect of the present invention, the chromatin-associated protein modification comprises trimethylation of lysine 4 on histone H3, or acetylation of lysine 27 on histone H3.

[0028] Beneficial Effects

[0029] Unlike DNA cutting-based methods (such as CUT&Tag), the ANDIE of the present invention keeps the genomic DNA from being fragmented after the deamination reaction. Unlike current non-cutting methods (such as DiMeLo-seq), the ANDIE of the present invention can retain cytosine conversion information during PCR amplification. This feature makes the ANDIE of the present invention compatible with other genomic analysis methods and supports subsequent joint analysis. In order to verify this advantage, by seamlessly combining the ANDIE of the present invention with single-cell Hi-C, the inventors developed ANDIE combined with Hi-C technology to achieve simultaneous mapping of histone modifications and three-dimensional genome structures in single cells.

[0030] Furthermore, the present invention can be more flexibly applied to detection objects targeting various antibodies or antibody fragments by fusing double-stranded DNA deaminase with Protein A and / or Protein G. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 The schematic diagram shows the map of ProteinAG-DddA protein expression nucleic acid vector pETDuet-1::pagdddA + dddAI. The inserted pagdddA gene and dddAI gene are driven by two independent lac promoters respectively.

[0032] Figure 2 The purified pAG-DddA protein was stained with Coomassie blue on SDS-PAGE gel.

[0033] Figure 3 A sequence marker diagram showing the nucleotide frequencies surrounding the positions of cytosine converted by pAG-DddA in genomic DNA.

[0034] Figure 4 Schematic diagram of lysed nuclei incubated with specific antibodies conjugated to pAG-DddA, which induces targeted deamination of cytosine near the antibody binding region. Subsequently, through library construction and high-throughput sequencing, the conversion sites can be accurately determined, thereby achieving precise localization and map construction of histone modifications.

[0035] Figure 5 The conversion ratio (y-axis) around the center of the H3K4me3 (left) and H3K27ac (right) modification peaks is shown. The x-axis represents the distance from the center of the ChIP-seq peak. The gray line is the IgG control group. From these data, the conversion enrichment of H3K4me3 and H3K27ac can be observed.

[0036] Figure 6The conversion ratio of H3K4me3 (black) and H3K27ac (grey) modifications within the transcription start site (TSS) region is shown (y-axis). The x-axis indicates the distance from the center of the transcription start site.

[0037] Figure 7 The conversion ratio of H3K4me3 and H3K27ac modifications is shown in the whole genome. The heat map reflects the deamination efficiency in different DNA sequence environments, where rows and columns represent the adjacent bases respectively.

[0038] Figure 8 Shown: Distribution plots showing TC conversion ratios for H3K4me3 (left) and H3K27ac (right). Dark curves represent positive reads and light curves represent negative reads. The average TC conversion ratio for H3K4me3 positive reads was 0.218, while that for negative reads was 0.046 (P < 0.0001, two-sided Mann-Whitney U rank test); the average ratio for H3K27ac positive reads was 0.229, while that for negative reads was 0.051 (P < 0.0001, two-sided Mann-Whitney U rank test).

[0039] Fig. 9 It is an aggregate enrichment map and heat map showing the distribution of H3K4me3 signals at transcription start sites (TSSs), comparing the results of ANDIE, CUT&Tag and ChIP-seq. All three methods showed significant enrichment and consistent distribution characteristics at TSSs.

[0040] Fig.10 It is a Pearson correlation heat map showing the correlation between H3K4me3 (left) and H3K27ac (right). The results show that ANDIE showed high correlation in repeated experiments and had high correlation with CUT&Tag and ChIP-seq, while the correlation of the IgG control group was very low.

[0041] Fig.11 The principal component analysis (PCA) plot shows the results of clustering samples based on the genome-wide histone modification profile, with data generated by ChIP-seq, CUT&Tag, and ANDIE. Each point represents a biological replicate, with H3K4me3 in dark (upper left) and H3K27ac in light (upper right). The PCA plot clearly shows the separation of H3K4me3 and H3K27ac marks mapped by the three methods on the principal components.

[0042] Fig.12The genome browser visualization shows H3K4me3 (above) and H3K27ac (below) signals at representative genomic sites, which is a screenshot of the software. ANDIE accurately reproduces the peaks observed by CUT&Tag and ChIP-seq at different sites, showing high consistency.

[0043] Fig.13 This is a schematic diagram of the ANDIE combined with Hi-C workflow. Antibody-directed deamination occurs near the targeted histone modification. This is followed by restriction endonuclease digestion and proximity pair ligation in the Hi-C protocol to capture chromatin interactions, followed by library construction and sequencing.

[0044] Fig.14 This is a H3K4me3 spectrum in the Chr8 (37-40 Mb) region, which is a screenshot of the software, showing genomic visualization from ChIP-seq, merged ANDIE combined with Hi-C data, and single-cell ANDIE combined with Hi-C and single-cell sortChIC data.

[0045] Fig.15 ANDIE combined with Hi-C and ChIP-seq signal enrichment analysis showing H3K4me3 around transcription start sites (TSSs).

[0046] Fig.16 Shown is a comparison of the number of H3K4me3 detections in single cells between single-cell ANDIE combined with Hi-C, sortChIC, and Paired-Tag.

[0047] Fig.17 The correlation of A / B compartment scores between ANDIE combined Hi-C and scMicro-C (25) is shown. DETAILED DESCRIPTION

[0048] As used herein, the term "fusion protein" refers to a hybrid polypeptide comprising protein domains from at least two different proteins. The "fusion protein" of the present invention comprises: a molecule having a specific affinity for an antibody, and a double-stranded DNA deaminase, wherein, for example, the molecule having a specific affinity for an antibody or the double-stranded DNA deaminase can be located at the amino terminal (N-terminal) portion of the fusion protein; another protein, such as a double-stranded DNA deaminase or a molecule having a specific affinity for an antibody can be located at the carboxyl terminal (C-terminal) of the fusion protein.

[0049] A linker peptide may be provided between the molecule having a specific affinity for the antibody or the double-stranded DNA deaminase.

[0050] The connecting peptide only needs to ensure that there is enough space between the N-terminal protein and the C-terminal protein to exert activity. For example, sequences commonly used in the art can be used, such as GSGGSSGGSSGSGDP, SGGSSGGSSGSETPGTSESATPESSGGSSGGS, GGGGSGGGGS, GGSGGSDDDKEF, GGSDDDKEF, DDDKEF, EAAAKEAAAK, AAAKEAAAKA, EAAAK, SGGSGGSGGS, GGGGS, GGGGGG and SGGS.

[0051] The molecules having specific affinity for antibodies are selected from Protein A and Protein G. Both Protein A and Protein G are proteins that bind to the Fc segment of antibodies, but the types of antibodies bound by the two are different. Those skilled in the art can flexibly select Protein A and Protein G as needed.

[0052] The present invention preferably uses both Protein A and Protein G to cover more types of antibodies, so that the fusion protein of the present invention and its application method can be combined with a wider range of antibodies, thereby having a wider range of application scenarios.

[0053] In the present invention, the double-stranded DNA deaminase can catalyze the deamination reaction of cytosine in double-stranded DNA to convert cytosine into uracil. After the conversion, uracil is detected as thymine during PCR amplification and next-generation sequencing to determine the action position of the double-stranded DNA deaminase.

[0054] In the prior art, DddA has high toxicity in living cells and can only be used for uracil-tolerant bacteria (ung gene knockout), and each detection requires fusion expression with a specific transcription factor, which limits its flux and application in eukaryotic cells. In this regard, the present invention overcomes its toxicity by expressing double-stranded DNA deaminase and double-stranded DNA deaminase inhibitory protein in living cells.

[0055] In the embodiment of the present invention, DddAtox, i.e., the active domain of double-stranded DNA deaminase, is used to facilitate expression and operation. In the following embodiments of the present invention, DddAtox, the active domain of double-stranded DNA deaminase, is actually used. It is understood by those skilled in the art that DddAtox has the same activity as the wild-type double-stranded DNA deaminase, and can be used interchangeably in the technical scheme of the present invention.

[0056] In one embodiment of the first aspect of the present invention, the double-stranded DNA deaminase is the active domain of double-stranded DNA deaminase DddAtox, the sequence of which is shown in SEQ ID No. 4.

[0057] In one embodiment of the present invention, the sequence of Protein A can be a sequence of its wild type or known variant, such as a sequence from databases such as GenBank, EMBL, UniProt and PIR, or a sequence having 80%, 85%, 90%, 95%, 98% or 99% sequence homology thereto, as long as the ability to bind to the Fc segment of an antibody is retained. For example, the antibody binding domain of Protein A is used in the examples of the present application, and the sequence is SEQ ID No. 1.

[0058] The protein G protein sequence can be a sequence of its wild type or known variant, such as a sequence from databases such as GenBank, EMBL, UniProt and PIR, or a sequence having 80%, 85%, 90%, 95%, 98% or 99% sequence homology thereto, as long as it retains its ability to bind to the Fc segment of an antibody. For example, the antibody binding domain B1 of Protein G is used in the examples of the present application, and the sequence is SEQ ID No. 2.

[0059] For information on the antibody-binding domains of Protein A and Protein G, see, for example, Chimeric IgG-binding receptors engineered from staphylococcal protein A and streptococcal protein G, which is incorporated herein by reference. .

[0060] The connecting peptide sequence is SEQ ID No. 3, or a sequence having 80%, 85%, 90%, 95%, 98% or 99% sequence homology thereto, as long as it does not affect the activity of Protein A and / or Protein G, and the activity of double-stranded DNA deaminase.

[0061] The double-stranded DNA deaminase sequence is SEQ ID No. 4 or a sequence having 80%, 85%, 90%, 95%, 98% or 99% sequence homology thereto, as long as it retains the ability to catalyze cytosine deamination reaction in double-stranded DNA.

[0062] In one embodiment of the present invention, the chromatin-associated protein is at least one protein selected from DNA binding proteins, transcription factors and histones, or a protein that binds to at least one protein among DNA binding proteins, transcription factors and histones.

[0063] The protein that binds to at least one of DNA binding proteins, transcription factors and histones refers to a protein that can specifically bind to at least one of DNA binding proteins, transcription factors and histones. In the case where the chromatin-associated protein is a protein that binds to at least one of DNA binding proteins, transcription factors and histones, the antibody of the present invention indirectly binds to at least one of DNA binding proteins, transcription factors and histones.

[0064] In one embodiment of the present invention, a polynucleotide is provided, which encodes the fusion protein of the present invention. Due to the degeneracy of codons, those skilled in the art can know that there are multiple nucleic acid sequences that can encode the fusion protein provided in the present application, which are not limited here. Those skilled in the art can perform appropriate codon optimization and select appropriate nucleic acid sequences for expressing the fusion protein provided in the present application according to the different expression purposes or hosts.

[0065] In one embodiment of the present invention, a nucleic acid vector is provided, which comprises the polynucleotide of the present invention.

[0066] In some embodiments, the nucleic acid vector includes a cloning vector and an expression vector. The expression nucleic acid vector can be used for prokaryotic cell expression systems, eukaryotic cell (mammalian cells and insect cells) expression systems, and cell-free expression systems. Both viral and non-viral expression vectors can be used to produce deaminase fusion proteins in eukaryotic host cells. Non-viral vectors and systems include plasmids, additional vectors (typically with expression cassettes for expressing proteins or RNA) and human artificial chromosomes. Virus-derived vectors include, but are not limited to, lentiviral vectors, retroviral vectors, adenoviral vectors, adeno-associated viral vectors, poxvirus vectors, and herpesvirus vectors.

[0067] In one embodiment of the present invention, a host cell is provided, which comprises the polynucleotide of the present invention or the nucleic acid vector of the present invention.

[0068] The host cell can be selected according to the type of the expression vector. The polynucleotide and / or the expression vector can be delivered into the host cell using any suitable method known in the art, which is not limited here.

[0069] [Method for detecting the position of chromatin-associated protein or chromatin-associated protein modification on double-stranded DNA sequence]

[0070] In one embodiment of the present invention, a method for detecting the position of a chromatin-associated protein or a chromatin-associated protein modification on a double-stranded DNA sequence is provided, the method comprising:

[0071] Step a. connecting a molecule having a specific affinity for an antibody to a double-stranded DNA deaminase, wherein the antibody has a specific affinity for a chromatin-associated protein or a chromatin-associated protein modification, and through the binding of the antibody to the chromatin-associated protein or the chromatin-associated protein modification, the double-stranded DNA deaminase performs a targeted deamination reaction on a DNA sequence near the chromatin-associated protein or the chromatin-associated protein modification bound by the antibody,

[0072] Step b. detecting the position of the chromatin-associated protein or chromatin-associated protein modification bound by the antibody on the double-stranded DNA sequence by determining the position of the DNA deamination site,

[0073] Wherein, the molecule having specific affinity for the antibody is Protein A and / or Protein G, or a fragment thereof having specific affinity for the antibody.

[0074] The chromatin-associated protein is at least one protein selected from the group consisting of DNA binding proteins, transcription factors and histones, or a protein that binds to at least one protein among DNA binding proteins, transcription factors and histones.

[0075] In step a, due to the binding of the molecule having a specific affinity to the antibody to the antibody and the binding of the antibody to the chromatin-associated protein or the chromatin-associated protein modification, the double-stranded DNA deaminase connected to the molecule having a specific affinity to the antibody is guided by the antibody to the chromatin-associated protein or the chromatin-associated protein modification, and performs a deamination reaction on the DNA sequence near the chromatin-associated protein or the chromatin-associated protein modification.

[0076] In one embodiment of the present invention, a molecule having a specific affinity for an antibody is linked to a double-stranded DNA deaminase. Since the antibody has a specific affinity for chromatin-associated proteins or chromatin-associated protein modifications, the double-stranded DNA deaminase is targeted to the DNA sequence bound by the chromatin-associated protein or chromatin-associated protein modification through the binding of the antibody to the chromatin-associated protein or chromatin-associated protein modification and the binding of the molecule having a specific affinity for the antibody to the antibody, and a deamination reaction is performed on the DNA sequence bound by the chromatin-associated protein or chromatin-associated protein modification.

[0077] In one embodiment of the present invention, the deamination reaction occurs on the DNA sequence bound to the chromatin-associated protein or the chromatin-associated protein modification, as well as the DNA sequence within 300bp, 200bp, 150bp, 100bp upstream, and 300bp, 200bp, 150bp, 100bp downstream.

[0078] In one embodiment of the present invention, the position of the chromatin-associated protein bound by the antibody or the position of the chromatin-associated protein modification on the double-stranded DNA sequence is detected by determining the position of the DNA deamination site.

[0079] In one embodiment of the present invention, the DNA deamination sites are determined by detecting bases that undergo sequence conversion after PCR amplification and sequencing.

[0080] In one embodiment of the present invention, the binding sequence of the antibody to DNA is determined by analyzing the position of the deamination site.

[0081] In one embodiment of the present invention, the deamination reaction is a reaction that catalyzes the deamination of cytosine in double-stranded DNA to convert cytosine into uracil.

[0082] In one embodiment of the present invention, a method for detecting the interaction between chromatin-associated proteins and DNA by targeting double-stranded DNA deaminase. For example, Protein A and Protein G are fused with double-stranded DNA deaminase and expressed. When an antibody having a specific affinity for chromatin-associated proteins or chromatin-associated protein modifications binds to the chromatin-associated proteins or chromatin-associated protein modifications, the double-stranded DNA deaminase is guided to the DNA where the specific chromatin-associated protein or chromatin-associated protein modification is located by binding of the Protein A-Protein G protein to the antibody, thereby activating the double-stranded DNA deaminase to deaminate the DNA at that location, and the positioning of the chromatin-associated protein or chromatin-associated protein modification on the double-stranded DNA sequence is achieved by recognizing the deamination sequence.

[0083] This process not only provides a more accurate modification detection method, but also avoids the DNA fragmentation problem that may exist in common modification detection methods, making it an ideal DNA binding protein detection technology. This detection method of the present invention is named ANDIE.

[0084] Unlike DNA cutting-based methods (such as CUT&Tag), the ANDIE of the present invention keeps the genomic DNA from being fragmented after the deamination reaction. Unlike current non-cutting methods (such as DiMeLo-seq), the ANDIE of the present invention can retain cytosine conversion information during PCR amplification. This feature makes the ANDIE of the present invention compatible with other genomic analysis methods and supports subsequent joint analysis. In order to verify this advantage, by seamlessly combining the ANDIE of the present invention with single-cell Hi-C, the inventors developed ANDIE combined with Hi-C technology to achieve simultaneous mapping of histone modifications and three-dimensional genome structures in single cells.

[0085] Furthermore, the present invention can be more flexibly applied to detection objects targeting various antibodies or antibody fragments by fusing double-stranded DNA deaminase with Protein A and / or Protein G.

[0086] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0087] Example:

[0088] Materials and Methods

[0089] 1. Construction of expression nucleic acid vector

[0090] Expression vector construction: ProteinAG-DddA (PAG-DddA) and DddAI expression constructs were obtained from Genescript via gene synthesis service.

[0091] Construct a ProteinAG-DddA expression vector based on pETDuet-1, in which:

[0092]

[0093]

[0094] The coding sequence of ProteinAG-DddA was synthesized and cloned into MCS-1 (NcoI and NotI sites, retaining the N-terminal hexahistidine tag).

[0095] In PAG-DddA, the protein sequence HHHHHH is a hexahistidine tag.

[0096]

[0097] The coding sequence of the immunity protein DddAI was synthesized and cloned into MCS-2 (NdeI and XhoI sites, removing the C-terminal S tag). The pETDuet-1::pagdddAtox + dddAI vector was transformed into E. coli DH5α and BL21 strains and stored at -20°C.

[0098] 2. Bacterial strain culture conditions

[0099] Escherichia coli (E. coli) strains transformed with nucleic acid vectors were cultured in LB medium at 37°C or on LB solid medium supplemented with agar (LBA, 1.5% w / v). If necessary, ampicillin (100 μg / mL) or IPTG (0.5 mM) was added to the medium. E. coli DH5α and BL21 were used for nucleic acid vector storage and protein expression, respectively.

[0100] 3. Purification of PAG-DddAtox

[0101] Briefly, to purify His-tagged PAG-DddAtox complexed with DddAI, 2 L of LB culture was inoculated with E. coli BL21 (pETDuet-1::pagdddA + dddAI) at a ratio of 1:100 and grown overnight at 37°C. When the OD600 of the culture reached approximately 0.6, 0.5 mM IPTG was added and the culture was shaken at 18°C ​​for 16 h. The bacterial cell pellet was collected by centrifugation at 4000 g for 30 min and resuspended in 50 ml of lysis buffer (50 mM Tris-HCl pH 8.0, 500 mM NaCl, 10 mM imidazole, 1 mg / mL lysozyme and protease inhibitor cocktail). The bacterial cell pellet was lysed by ultrasonication (five pulses of 10 s each), and the supernatant after ultrasonication was centrifuged at 25,000 g for 30 min to remove cell debris. The His-tagged PAG-DddAtox-DddAI complex was purified from the supernatant using a nickel column. The DddAtox-DddAI complex was eluted using elution buffer (50 mM Tris-HCl pH 7.5, 300 mM imidazole, 500 mM NaCl, 30 mM imidazole, 1 mM DTT). The eluted PAG-DddA-DddAI complex was denatured by adding 50 ml 8 M urea denaturation buffer (50 mM Tris-HCl pH 7.5, 300 mM imidazole, 500 mM NaCl, 1 mM DTT) at 4°C for 16 hours to completely separate PAG-DddA from DddAI. The denatured proteins were separated again by a nickel column in 8 M urea denaturation buffer, and PAG-DddA could bind to the column due to the His tag. The column was washed with 50 ml 8 M urea denaturation buffer to remove residual DddAI. Subsequently, the column was washed with denaturing buffer with gradually decreasing urea concentration (6 M, 4 M, 2 M, 1 M) and finally washed with washing buffer without urea. PAG-DddAtox bound to the column was eluted by 5 ml elution buffer. The eluted PAG-DddA was separated by size exclusion chromatography by fast protein liquid chromatography (FPLC) and gel filtration on a Superdex200 column (GE Healthcare), using a size buffer (20 mM Tris-HCl pH7.5, 200 mM NaCl, 1 mM DTT, 5% (w / v) glycerol). The purity of the eluted PAG-DddA was evaluated by SDS-PAGE gel and Coomassie Brilliant Blue staining. The purified protein was stored at -80°C.

[0102] 4. Eukaryotic Cell Culture and Treatment Conditions

[0103] Drosophila Schneider 2 (S2) was cultured in Schneider's Drosophila Medium (Gibco, Product No. 21720024) supplemented with 10% fetal bovine serum (FBS) (Cellmax, Product No. SA301.02) and 1% penicillin-streptomycin (Pen-Strep) (Gibco, Thermo Fisher Scientific, Product No. 15140122) at 25°C without CO2. When used, S2 cells were collected by centrifugation at 300g for 5 minutes and resuspended in 1× PBS. When fixation (paraformaldehyde, PFA) was required, 3 million S2 cells were resuspended in 500 μL 0.5% PFA and fixed for 10 minutes. The fixed cells were collected by centrifugation at 600 g for 5 min, resuspended in 300 µL DPBS (Product No.: 14040117, Thermo Fisher Scientific)-1% BSA (Product No.: A1933, Sigma), washed once, left to stand for 5 min to eliminate the effect of PFA, and finally washed three times with 1% BSA-PBS.

[0104] Human cell line GM12878 cells were cultured in RPMI 1640 medium (Cat. No. 11875093, Gibco) supplemented with 10% fetal bovine serum (FBS) (Cat. No. 10099141C, Gibco) at 37°C in a 5% CO2 incubator. If necessary, 1 million cells were crosslinked with 1 ml of freshly prepared 0.5% PFA fixation buffer for 10 min at room temperature (RT). To terminate the fixation reaction, cells were resuspended in 1 ml of 1% BSA / PBS, left to stand for 5 min to eliminate the effect of PFA, and finally washed three times with 1% BSA-PBS.

[0105] 5. Solution Preparation

[0106] Table 1. Deaminase Incubation Buffer Composition

[0107]

[0108] Table 2. Wash buffer composition

[0109]

[0110] Table 3. Antibody buffer composition

[0111]

[0112] Table 4. High Salt Buffer Composition

[0113]

[0114] Table 5. 1% BSA / PBS composition

[0115]

[0116] Table 6. 0.5% PFA composition

[0117]

[0118] 6. Detection of specific histone modifications in Drosophila cells using pAG-DddA (ANDIE)

[0119] Rinse S2 cells with PBS and resuspend in PBS and count using a cell counter. Next, centrifuge 3×10^6 cells from each sample at 300×g for 5 minutes. To lightly fix the cells, resuspend approximately 3×10^6 S2 cells in 500μL of 0.5% PFA and rotate for 5 minutes or shake by hand to allow the fixative to fully contact the cells. Then, add 300μL of 1% BSA / PBS and mix, centrifuge at 500×g for 4 minutes at 4°C and remove the supernatant. Next, resuspend the cells in 300μL of 1% BSA / PBS, incubate for 5 minutes while rotating, and resuspend in 200μL PBS and count the cells. Take 5×10^5 S2 cells into a PCR tube, centrifuge at 800×g for 5 minutes at 4°C and remove the supernatant. Then, 1 μL of 20% Np-40 was added to 200 μL of incubation buffer (to a final concentration of 0.1%), and S2 cells were placed in a PCR tube, and 100 μL of the resulting liquid was added to resuspend the cells, and ice bathed for 5 minutes. Immediately, the cells were centrifuged at 800 × g for 5 minutes at 4°C, and the supernatant was removed. Next, the cells were rinsed once with 100 μL of wash buffer, centrifuged at 800 × g for 5 minutes at 4°C, and the supernatant was removed. Subsequently, 1 μL of antibody (one IgG control group, one H3K4me3 experimental group, and one H3K27ac experimental group) was added to 50 μL of antibody buffer and mixed, and the precipitate was resuspended and incubated at room temperature for 2 hours. After that, the S2 cells were centrifuged at 300 × g for 5 minutes, and then washed once with 100 μL of cooled antibody buffer, and gently pipetted about 10 times. After resuspension, the cells were rotated on a 4°C rotator for 5 minutes, and then centrifuged at 300 × g to remove the supernatant. Gently dissolve the pellet with 50 μL high salt buffer containing 2.5 μL PA / G-DddA and place on a rotator for 1 hour, rotating and mixing in a cold room at 4°C. Centrifuge at 300×g for 5 minutes and remove the supernatant. Then, wash twice with 150 μL of cold high salt buffer, gently pipetting about 10 times. After resuspending, place on a rotator at 4°C for 5 minutes, then centrifuge at 300×g and remove the supernatant. For each sample, gently dissolve the pellet with 100 μL incubation buffer and incubate at 37°C for 1 hour. Finally, centrifuge S2 cells at 800×g for 5 minutes, remove the supernatant, and add 100 μL of room temperature 0.2% SDS-PBS (i.e., add 2 μL 10% SDS to 98 μL PBS) and 10 μL proteinase K, and let stand at 55°C overnight. After overnight, genomic DNA was extracted and purified using Zymo Research D4010 Genomic DNA Clean & Concentrator®-10 (or 0.5× Vazyme beads), and the resulting DNA concentration was determined using Qubit.The sequencing library was prepared using Vazyme TD502 (TD503) TruePrep DNA Library Preparation Kit.

[0120] 7. Co-detection of histone modifications and chromatin structure in single cells using pAG-DddA (ANDIE combined with Hi-C)

[0121] The inventors performed the ANDIE experiment as described above using 5×10^4 GM12878 cells fixed with 0.5% PFA. After the antibody-directed deamination reaction, the cells were fixed again with 1.5% PFA and the Dip-C cell fixation process was performed according to the previously reported method. .

[0122] [Construction, expression and deamination detection of fusion vector]

[0123] According to the above “1. Construction of expression nucleic acid vector”, construct: ProteinAG-DddAtox (PAG-DddAtox) expression vector and DddAI expression construct.

[0124] The active region DddAtox of the wild-type DddA protein was used. The inventors put the His tag, Protein A, Protein G, and a linker sequence GSGGSSGGSSGSGDP and the coding gene of DddAtox into the same open reading frame (ORF) through gene synthesis, and put the coding gene of DddAI (active neutralizing protein of DddA) into another open reading frame, which were driven by two independent promoters ( Figure 1 After protein expression was induced, DddDddAtoxA was separated from DddAI by denaturation treatment to obtain purified protein ( Figure 2 ). DddAtox (pAG-DddAtox) fused with Protein A and Protein G still has deamination activity on double-stranded DNA, and its deamination sites are mainly TC ( Figure 3 ), which is consistent with previous reports.

[0125] [Genome-wide histone modification detection - ANDIE]

[0126] The detection process of whole-genome histone modification using PAG-DddA is as follows: first, the histone modification sites are identified by incubating the permeabilized cell nuclei with specific antibodies; then PAG-DddA is added and incubated under low temperature conditions to allow PAG-DddA to bind to the antibody and reduce nonspecific deamination reactions. Subsequently, the unbound antibodies and PAG-DddA in the cell nuclei are removed by washing under low temperature conditions. PAG-DddA bound to the antibody then catalyzes the deamination reaction of DNA at 37 degrees to achieve antibody-guided targeted DNA deamination. The DNA then undergoes PCR library amplification. During the amplification process, cytosine deaminated to uracil can be converted to thymine, so that the deamination conversion site can be detected after sequencing. DNA regions with a higher conversion ratio can be determined as binding regions on the DNA occupied by histone modifications ( Figure 4 ).

[0127] The inventors named this detection method ANDIE (Antibody-directed deamination) and applied it to the detection of lysine acetylation at position 27 (H3K27ac) and lysine trimethylation at position 4 (H3K4me3) of histone H3. Among them, H3K27ac is usually used as a marker of active enhancers, while H3K4me3 is a common marker of the promoter region of actively transcribed genes. The inventors analyzed the data of ANDIE performed with H3K4me3 or H3K27ac antibodies. At the center of the ChIP-seq peak of H3K4me3 or H3K27ac, ANDIE of H3K4me3 or H3K27ac showed a significant increase in conversion ( Figure 5 ). Notably, ANDIE performed using IgG controls showed almost no background conversion, further validating the specificity of the method ( Figure 5 In addition, at transcription start sites (TSSs), ANDIE of H3K4me3 and H3K27ac also showed significant signal enrichment, with a signal periodicity of about 150 bp, which is consistent with the known distribution pattern of nucleosomes at TSS positions ( Figure 6 ). These experimental results fully demonstrate the high specificity and sensitivity of ANDIE in histone modification detection.

[0128] From the whole genome level, ANDIE's global TC conversion rate in histone modification detection is about 2% ( Figure 7 ). In order to accurately define the positive signal of histone modification, the inventors set a standard: within a specific histone modification region, a positive result requires at least three converted TC sites in the paired-end 150 bp NGS sequencing read. Specifically, the minimum conversion rate for H3K4me3 is 7%, while that for H3K27ac is 6% ( Figure 8 In histone modification-positive regions, the average TC conversion rate of positive reads in target regions (average of approximately 20%) was approximately five times higher than that of negative reads in non-target regions (average of approximately 4% after excluding sequencing reads with zero conversion), further demonstrating the reliability of ANDIE in histone modification detection.

[0129] To verify the performance of ANDIE, the inventors compared it with the currently widely used detection methods - CUT&Tag and ChIP-seq. In the detection of the H3K4me3 mark, the three methods all showed consistent enrichment patterns when centered on active TSS ( Fig. 9 ). To further evaluate the correlation between these methods, the inventors calculated the correlation coefficients of ANDIE, CUT&Tag, and ChIP-seq. The results showed that the correlation coefficients of ANDIE repeated experiments for H3K4me3 were all over 0.95, and the correlation coefficients with the other two methods were all over 0.8 ( Fig.10 ). This result shows that ANDIE is highly consistent with other mature methods in terms of repeatability and reliability. Similarly, ANDIE for H3K27ac also showed high experimental repeatability and was highly consistent with the results between CUT&Tag and ChIP-seq ( Fig.10 ).

[0130] Although H3K4me3 and H3K27ac are usually located in different regions of the genome, they often co-exist in highly active regulatory elements. In order to further compare the differences in H3K4me3 and H3K27ac detection obtained by ANDIE, CUT&Tag, and ChIP-seq, the inventors performed principal component analysis (PCA). The analysis results showed that the repeated experiments of ANDIE, CUT&Tag, and ChIP-seq were clustered together, and the distribution of H3K4me3 and H3K27ac marks along the principal components was clearly separated ( Fig.11 To further validate these results, the inventors selected two genomic regions and used ANDIE to perform a visual comparison of histone marks ( Fig.12). In the Chr2R (6.55-6.8 Mb) region, CUT&Tag and ChIP-seq produced almost identical H3K4me3 and H3K27ac peaks, and ANDIE also reproduced this pattern with specific signals. In contrast, near Chr2R (5.56-5.86 Mb), CUT&Tag and ChIP-seq produced sharp H3K4me3 peaks, while H3K27ac showed a wider distribution, and ANDIE was also able to reproduce this phenomenon. These experimental results show that ANDIE can provide comparable sensitivity and resolution to CUT&Tag and ChIP-seq in different genomic regions, with excellent signal quality and experimental repeatability.

[0131] [ANDIE combines Hi-C and single-cell detection]

[0132] The deamination-based detection method can maintain the integrity of genomic DNA and retain cytosine conversion information through the PCR amplification process. This feature makes ANDIE compatible with other genomic analysis methods (such as chromatin conformation capture technology Hi-C). The inventors call this combined method ANDIE combined with Hi-C. After antibody-guided deamination reactions in specific histone modification regions, chromatin interactions are subsequently captured by restriction endonuclease digestion and proximity pairing ligation technology ( Fig.13 ).

[0133] The inventors conducted a validation experiment of single-cell ANDIE combined with Hi-C to simultaneously detect H3K4me3 patterns and genome-wide three-dimensional chromatin contacts in the same cell. The ANDIE combined with Hi-C signals at the same location were highly consistent with the ChIP-seq results of H3K4me3 detection ( Fig.14 Compared to the sparse and low-resolution signals observed in single cells by the sortChIC method, single-cell ANDIE combined with Hi-C showed stronger signal intensity and lower background noise in H3K4me3-enriched regions ( Fig.14 ). Consistent with ChIP-seq, H3K4me3-positive reads from ANDIE combined with Hi-C showed significant enrichment in the transcription start site (TSS) region ( Fig.15 ). It is worth noting that compared with other single-cell methods such as sortChIC or Paired-Tag, single-cell ANDIE combined with Hi-C increased the number of H3K4me3 detection reads by about 10 times, demonstrating the sensitivity and accuracy of this method in histone modification profiling ( Fig.16A strong A / B compartment correlation (r = 0.967) was observed between ANDIE combined with Hi-C and high-resolution scMicro-C 3D genomic data, indicating that ANDIE combined with Hi-C can accurately capture chromatin structure at the single-cell level ( Fig.17 ). These results strongly demonstrate that ANDIE combined with Hi-C technology can associate histone modification profiles with chromatin structure.

[0134] References: 1.Cramer P (2019) Organization and regulation of genetranscription. Nature 573(7772):45-54. 2.Bannister AJ&Kouzarides T (2011) Regulation of chromatin by histonemodifications. Cell Res 21(3):381-395. 3.Orlando V (2000) Mapping chromosomal proteins in vivo byformaldehyde-crosslinked-chromatin immunoprecipitation. Trends in Biochemical Sciences 25(3):99-104. 4.Solomon MJ, Larsen PL,&Varshavsky A (1988) Mapping protein–DNA interactions in vivo with formaldehyde: evidence that histone H4 is retained on a highly transcribed gene. Cell 53(6):937-947. 5.Johnson DS, Mortazavi A, Myers RM,&Wold B (2007) Genome-widemapping of in vivo protein-DNA interactions. Science 316(5830):1497-1502. 6. Barski A , et al.(2007) High-resolution profiling of histonemethylations in the human genome. Cell 129(4):823-837. 7.Robertson G , et al. (2007) Genome-wide profiles of STAT1 DNAassociation using chromatin immunoprecipitation and massively parallelsequencing. Nature Methods 4(8):651-657. 8.Mikkelsen TS , et al. (2007) Genome-wide maps of chromatin state inpluripotent and lineage-committed cells. Nature 448(7153):553-U552. 9.Park PJ (2009) ChIP-seq: advantages and challenges of a maturingtechnology. Nature Reviews Genetics 10(10):669-680. 10.Meers MP, Bryson TD, Henikoff JG,&Henikoff S (2019) Improved CUT&RUN chromatin profiling tools. Elife 8. 11.Skene PJ&Henikoff S (2017) An efficient targeted nuclease strategyfor high-resolution mapping of DNA binding sites. Elife 6. 12.Kaya-Okur HS , et al. (2019) CUT&Tag for efficient epigenomicprofiling of small samples and single cells. bioRxiv . 13.Skene PJ, Henikoff JG,&Henikoff S (2018) Targeted in situ genome-wide profiling with high efficiency for low cell numbers. Nature Protocols 13(5):1006-1019. 14.Altemose N , et al. (2022) DiMeLo-seq: a long-read, single-moleculemethod for mapping protein-DNA interactions genome wide. Nature Methods 19(6):711-+. 15.Bartosovic M&Castelo-Branco G (2023) Multimodal chromatinprofiling using nanobody-based single-cell CUT&Tag. Nat Biotechnol 41(6):794-805. 16.Mok BY , et al. (2020) A bacterial cytidine deaminase toxin enablesCRISPR-free mitochondrial base editing. Nature 583(7817):631-+. 17.Gallagher LA , et al. (2022) Genome-wide protein-DNA interactionsite mapping in bacteria using a double-stranded DNA-specific cytosinedeaminase. Nat Microbiol 7(6):844-855. 18.Eliasson M , et al. (1988) Chimeric IgG-binding receptors engineeredfrom staphylococcal protein A and streptococcal protein G. J Biol Chem 263(9):4323-4327. 19.Tan L (2021) Determining the 3D genome structure of a singlemammalian cell with Dip-C. STAR Protoc 2(3):100622. 20.Tan L, Xing D, Chang CH, Li H,&Xie XS (2018) Three-dimensionalgenome structures of single diploid human cells. Science 361(6405):924-928.。

Claims

1. A fusion protein comprising, from the N-terminus to the C-terminus: a molecule having a specific affinity for an antibody, a connecting peptide and a double-stranded DNA deaminase, wherein: The molecule having specific affinity for the antibody is Protein A and / or Protein G, and the antibody is an antibody having specific affinity for chromatin-associated proteins or chromatin-associated protein modifications.

2. The fusion protein according to claim 1, wherein The protein A sequence is SEQ ID No. 1; The protein G protein sequence is SEQ ID No. 2; The connecting peptide sequence is SEQ ID No. 3; and / or The double-stranded DNA deaminase is a double-stranded DNA deaminase domain DddAtox with a sequence of SEQ ID No.

4.

3. The fusion protein according to claim 1, wherein The chromatin-associated protein is at least one protein selected from the group consisting of DNA binding proteins, transcription factors and histones, or a protein that binds to at least one protein selected from the group consisting of DNA binding proteins, transcription factors and histones. A polynucleotide encoding the fusion protein according to any one of claims 1 to 3. A nucleic acid vector comprising the polynucleotide according to claim 4 . A host cell comprising the polynucleotide according to claim 4 or the nucleic acid vector according to claim 5.

7. A method for detecting the position of a chromatin-associated protein or a chromatin-associated protein modification on a double-stranded DNA sequence, the method comprising: Step a. connecting a molecule having a specific affinity for an antibody to a double-stranded DNA deaminase, wherein the antibody has a specific affinity for a chromatin-associated protein or a chromatin-associated protein modification, and through the binding of the antibody to the chromatin-associated protein or the chromatin-associated protein modification, the double-stranded DNA deaminase performs a targeted deamination reaction on a DNA sequence near the chromatin-associated protein or the chromatin-associated protein modification bound by the antibody, Step b. detecting the position of the chromatin-associated protein or chromatin-associated protein modification bound by the antibody on the double-stranded DNA sequence by determining the position of the DNA deamination site, Wherein, the molecule having specific affinity for the antibody is Protein A and / or Protein G having specific affinity for the Fc domain of the antibody.

8. The method according to claim 7, wherein the double-stranded DNA deaminase refers to an enzyme or a deaminase domain thereof that can deaminize cytosine and / or adenine on double-stranded DNA.

9. The method according to claim 7, wherein: A molecule having a specific affinity for the antibody is linked to a double-stranded DNA deaminase as a protein fusion and expressed, or The molecule having a specific affinity for the antibody is linked to the double-stranded DNA deaminase by affinity, or The link between the molecule having a specific affinity for the antibody and the double-stranded DNA deaminase is covalent.

10. The method according to claim 7, wherein the chromatin-associated protein is at least one protein selected from the group consisting of DNA binding proteins, transcription factors and histones, or a protein that binds to at least one protein among DNA binding proteins, transcription factors and histones.

11. The method according to claim 7, wherein the DNA deamination site is determined by detecting the bases that undergo sequence conversion after PCR amplification and sequencing.

12. The method according to claim 7, wherein the binding sequence of the antibody to DNA is determined by analyzing the position of the deamination site.

13. The method according to claim 7, wherein the chromatin-associated protein modification comprises trimethylation of lysine 4 on histone H3, or acetylation of lysine 27 on histone H3.

Images