Reverse-ChIP Techniques for Identifying Locus-Specific Proteomes: A Key Tool in Unlocking the Cancer Regulome

Abstract

A phenotypic hallmark of cancer is aberrant transcriptional regulation. Transcriptional regulation is controlled by a complicated array of molecular factors, including the presence of transcription factors, the deposition of histone post-translational modifications, and long-range DNA interactions. Determining the molecular identity and function of these various factors is necessary to understand specific aspects of cancer biology and reveal potential therapeutic targets. Regulation of the genome by specific factors is typically studied using chromatin immunoprecipitation followed by sequencing (ChIP-Seq) that identifies genome-wide binding interactions through the use of factor-specific antibodies. A long-standing goal in many laboratories has been the development of a 'reverse-ChIP' approach to identify unknown binding partners at loci of interest. A variety of strategies have been employed to enable the selective biochemical purification of sequence-defined chromatin regions, including single-copy loci, and the subsequent analytical detection of associated proteins. This review covers mass spectrometry techniques that enable quantitative proteomics before providing a survey of approaches toward the development of strategies for the purification of sequence-specific chromatin as a 'reverse-ChIP' technique. A fully realized reverse-ChIP technique holds great potential for identifying cancer-specific targets and the development of personalized therapeutic regimens.

Keywords: cancer regulome; chromatin architecture; genome regulation; locus-specific chromatin isolation; mass spectrometry proteomics; promoter pulldown proteomics; proximity labeling; quantitative mass spectrometry; telomeres; transcriptional regulation.

Figure 1

Comparison of ChIP-Seq with reverse-ChIP. ChIP-Seq (a) and reverse- ChIP (b) follow a similar experimental workflow. Cells are typically (though not always) cross-linked (i) so proteins and nucleic acids that comprise chromatin fragments are covalently bonded together. Subsequently, chromatin is fragmented (often through sonication) (ii) to solubilize chromatin complexes. The next step is epitope binding (iii). In ChIP-Seq, antibodies bind to the factor of interest. A variety of strategies have been employed to generate epitopes that bind to and isolate a specific genomic locus (see Section 3 of the review). The epitope-tagged region is purified via magnetic beads (iv) and molecules are de-cross-linked (v) for downstream analysis (vi). In ChIP-Seq, nucleic acids are isolated and identified through PCR (targeted) or sequencing (genome-wide). In Reverse-ChIP, the protein fraction is collected and identified via Western blot (targeted) or mass spectrometry (proteome-wide).

Figure 2

Approaches for quantitative mass spectrometry. (a) Workflow for a bottom-up mass spec proteomics experiment. Cells are lysed (i) and proteins are digested (ii). Peptide fragments are separated by physical properties using liquid chromatography coupled to mass spectrometry (iii), and bioinformatic analysis of produced spectra (iv) identifies peptides present in the sample. (b) Metabolic labeling for quantitative mass spectrometry. Cells in different conditions are grown on media that incorporate labeled amino acids of differing mass, which are incorporated into cells (e.g., Arg⁰, Arg⁶, or Arg¹⁰ [isotopically abundant Arg or Arg incorporating ¹³C₆ and ¹³C₆¹⁵N₄, respectively]). Combined cell lysate is subjected to the conventional bottom-up proteomics workflow. Peptides present in both samples are doubled in m/z, while peptides uniquely present in one sample show a single signal. (c) Chemical labeling for quantitative mass spectrometry. Samples are processed according to the bottom-up proteomics workflow in parallel. Peptides are labeled with isobaric chemical tags before samples are combined for LC-MS. The mass of the tag is chosen for secondary fragmentation in MS/MS mode for ratiometric comparison of peptide quantity between multiplexed samples. To avoid ratio compression due to the co-isolation of contaminating peptides during MS2, an additional fragmentation step can be added for increased specificity (MS3, see Section 2.2.5). (d) Label-free approaches for quantitative mass spectrometry. Peptides are aligned across runs based on elution time on the LC, m/z, and spectra in MS and MS2 modes. Concentration is determined by the size of the peak in the chromatogram for peak area measurements as determined through integration (i). The number of times a particular m/z spectrum from a peptide is observed from a sample is used in spectral counting (ii).

Figure 3

Nucleic acid-based approaches for locus-specific chromatin isolation. (a) Synthetic oligo baits can help identify proteins that recognize specific DNA sequences. Cells are lysed and split into control and sample pools. Biotinylated synthetic oligonucleotides with the bait sequence of interest or a scrambled control are added to the cell lysate to allow for proteins to bind. Sequence-specific interactors (orange with red stars) only interact with the bait sequence, while non-sequence-specific interactors bind to both bait and control. (b) Proteomics of isolated chromatin (PICh) and related protocols (HyCCAPP [yeast], RChIP [plants]) represent a conceptual shift in nucleic acid-based identification of sequence-specific binders by directly isolating the chromatin complexes from cells. After cross-linking and sonication to solubilize chromatin, a biotinylated oligonucelotide probe is added by the experimenter. DNA in chromatin complexes is denatured to allow for the oligo probe to form stable hybrids with the chromatin region of interest. Streptavidin-based purification isolates the chromatin region from the rest of the cellular milieu for downstream analysis. (c) Global exonuclease-based enrichment of chromatin-associated proteins for proteomics (GENECAPP) directly isolates chromatin complexes of interest similar to PICh via an alternative capture method. The addition of an exonuclease by the experimenter creates defined DNA overhangs that are targeted by magnetic bead-attached complementary oligonucleotides. This is in contrast to the denaturation-hybridization approach of PICh and related techniques. The chromatin complexes with the complementary oligo are purified via magnetic beads for downstream analysis.

Figure 4

Protein-based purification strategies. (a) Recombinase sites are inserted to flank a region of interest (ROI) in recombinase excision. The cells are then engineered to express the recombinase that recognizes the introduced sequences to excise the region of interest as a small circular plasmid separate from bulk chromatin. Density centrifugation separates the ROI into the supernatant, and the protein fraction is collected for analysis. (b) For insertional chromatin immunoprecipitation (iChIP), an experimenter engineers cells to include an exogenous binding element (EBE) not natively present near the ROI (LexA-binding element in the initial report). The cells containing the EBE are then further engineered to express the protein that recognizes the EBE (LexA protein in the initial report). Since neither the EBE nor the protein that recognizes it is present in wild-type cells, antibody purification against the introduced protein can specifically isolate the ROI. (c) Targeted chromatin purification (TChP) follows a similar strategy to iChIP. The exogenous binding element introduced in TChP is the tet-responsive element (TRE). The cells are then engineered to express an affinity tag-modified TetOn protein, which only binds TRE in the presence of doxycycline. The inclusion or exclusion of doxycycline by the experimenter determines whether the specific locus or background nonspecific binders are isolated. (d) In TAL-based strategies, a transcription activator-like (TAL) protein is designed to specifically recognize and bind near the ROI. Cells are then engineered to express the TAL to enable purification. This protocol differs from iChIP and TChP in that the engineering of the recognition element is performed at the protein level and does not require the insertion of a recognition element near the ROI. (e) Epi-Decoder is a sequencing-based strategy in yeast. A library of yeast in which the ROI has been ‘barcoded’ with a unique molecular identifier up- and downstream is crossed with a library of yeast in which each individual protein has been tagged with a tandem affinity purification (TAP) tag. The result is a new library of yeast in which each protein is associated with a unique barcode. Conventional ChIP-Seq with an antibody directed at the common TAP tag allows for the identification of specific binding by any protein at the ROI based on sequencing counts of the unique barcode.

Figure 5

CRISPR-based purification approaches. (a) For CRISPR-enChIP and analogous strategies, an appropriate cell model system is engineered to incorporate an affinity-tagged dCas9 protein. This engineered cell line is then modified to express sgRNAs targeting the region of interest (ROI), allowing the dCas9-sgRNA complex to bind to chromatin. The ROI can then be purified with an antigen to the affinity tag on dCas9. The modular nature of sgRNA allows for relatively rapid targeting of new sites compared with more labor-intensive iChIP and TAL-based strategies. (b) The Cas9-locus-associated proteome (CLASP) technique is an in vitro version of CRISPR-enChIP strategies utilizing recombinant dCas9. Using recombinant dCas9 allows for the isolation of the ROI without needing to engineer the cell line to express the dCas9 protein. This stands in contrast to TAL-based strategies where recombinant TAL expression is not sufficient for locus isolation. However, purification efficiency is lower. Empirically, about an order of magnitude more cells are needed for in vitro strategies using recombinant protein. (c) For the CRISPR affinity purification in situ of regulatory elements (CAPTURE) technique, cells are engineered to express a dCas9 protein that has been modified to incorporate a site for in vivo biotinylation by a biotin ligase. The first-generation CAPTURE required engineering cells to express the biotin ligase as well, though the second-generation version takes advantage of endogenous biotin ligases. The biotinylated-dCas9-chromatin complex can then be purified using streptavidin beads for analysis of locus-specific binders. (d) Proximity labeling approaches rely on engineering cells to express dCas9 fused to proximity labeling enzymes, such as BirA or APEX2. The region of interest is targeted with designed sgRNAs, so the proximity labeling enzyme can covalently link biotin to proteins in the ROI. The spatial and temporal resolution of tagging depends on the specific proximity labeling enzyme used. Since proteins present at the locus are directly tagged with biotin in live cells, there is no need for a cross-linking step to covalently bind chromatin complexes together.