iChIP‐SILAC analysis identifies epigenetic regulators of CpG methylation of the p16INK4A gene

Allele‐specific epigenetic events regulate the expression of specific genes such as tumor suppressor genes. Methods to biochemically identify epigenetic regulators remain limited. Here, we used insertional chromatin immunoprecipitation (iChIP) to address this issue. iChIP combined with quantitative mass spectrometry identified DNA methyltransferase 1 (DNMT1) and epigenetic regulators as proteins that potentially interact with a region of the p16INK4A gene that is CpG‐methylated in one allele in HCT116 cells. Some of the identified proteins are involved in the CpG methylation of this region, and of these, DEAD‐box helicase 24 (DDX24) contributes to CpG methylation by regulating the protein levels of DNMT1. Thus, iChIP is a useful method to identify proteins which bind to a target locus of interest.

Gene expression is finely regulated by epigenetic mechanisms including DNA and/or histone modifications [e.g., transcriptional silencing by DNA (CpG) methylation and transcriptional activation by H3K4me1 and H3K27ac] [1].Aberrant epigenetic regulation leads to the development of intractable diseases such as cancer [2].For example, tumor suppressor genes are frequently downregulated in cancer cells by methylation of CpG islands in promoter regions (epigenetic abnormality in cancer) [3].DNA demethylating agents such as 5-aza-2 0 -deoxycytidine are used as anticancer drugs despite their genome-wide effects [4].Identification of molecules that specifically bind to genes with abnormal CpG methylation may help elucidate the mechanisms underlying aberrant epigenetic regulation and may contribute to the development of target gene-specific epigenetic drugs.
To identify molecules that bind to a target genomic region and thus examine the mechanisms of aberrant epigenetic regulation, we developed a locus-specific chromatin immunoprecipitation (locus-specific ChIP) method, which can be used to isolate specific genomic regions [5].Molecules (DNA, RNA, and proteins) bound to the isolated genomic region can be comprehensively identified by downstream analysis using mass spectrometry (MS) and next-generation sequencing.Locus-specific ChIP consists of insertional ChIP (iChIP) [5][6][7] and engineered DNA-binding moleculemediated ChIP (enChIP) [5,8,9].The iChIP method consists of three steps: (I) locus targeting by insertion of the binding elements of LexA, a bacterial DNAbinding protein (LexA BE), in/around a target genomic region in a cell; (II) expression of the 3xFLAG-tagged DNA-binding domain of LexA (3xFNLDD) in the same cell; and (III) isolation of the target genomic region by affinity purification of 3xFNLDD.We previously demonstrated that iChIP/enChIP combined with stable isotope labeling using amino acids in cell culture Abbreviations Ab, antibody; ANOVA, analysis of variance; BE, binding element; ChIP, chromatin immunoprecipitation; DDX24, DEAD-box helicase 24; DNMT1, DNA methyltransferase 1; enChIP, engineered DNA-binding molecule-mediated ChIP; FBL, fibrillarin; GFP, green fluorescent protein; iChIP, insertional ChIP; MS, mass spectrometry; RT, reverse transcription; SILAC, stable isotope labeling using amino acids in cell culture; USP7, ubiquitin specific peptidase 7; ZFN, zinc finger nuclease.
(SILAC), a quantitative MS analysis (iChIP/enChIP-SILAC), are useful tools to identify proteins that interact with target genomic regions [10,11].Unlike enChIP, iChIP isolates only the LexA BE-inserted allele, which is ideal for biochemical analysis of a genomic region in an allele of interest.
The Knudson two-hit theory of tumor suppressor genes suggests that genetic mutations and/or epigenetic silencing of the two alleles are necessary for tumor development.In some cases, a mutation is introduced in one allele and epigenetic events silence gene transcription on the other allele [3].For example, in gastrointestinal cancers, the promoter region of the p16 INK4A gene encoding a cell cycle regulator is highly CpG-methylated and gene transcription is silenced [12].Identifying proteins that interact with the epigenetically silenced allele will provide insight into epigenetic abnormalities involved in cancer development and may lead to the development of novel molecular targeted agents to restore the abnormal epigenetic status and restart gene transcription.
In this study, iChIP-SILAC was used for analysis of proteins interacting with the epigenetically silenced (CpG-methylated) p16 INK4A gene in one allele in the human colon cancer cell line HCT116 (Fig. 1).Proteins including DNA methyltransferase 1 (DNMT1) and epigenetic regulators were identified.The findings of this study suggest that iChIP is a useful screening tool for identification of proteins potentially binding to a target locus of interest.

Construction of the targeting plasmid pTNL-LexA-p16
The nucleotide sequences 6.8 kbp upstream and 4.7 kbp downstream of the insertion site of LexA BE were amplified by PCR using HCT116 genomic DNA.Briefly, the 6.8 kbp fragment was amplified and inserted between the Xho I and BamH I sites of the plasmid pTNLrSP4 [13] to construct pTNL-p16XhoBam.The 4.7 kbp fragment was separately amplified as 0.3 and 4.4 kbp fragments.LexA BE from 8 9 LexA-binding elements/pMD20 (#48807; Addgene) [14] and the 4.7 kbp fragment (0.3 and 4.4 kbp fragments) were inserted sequentially between the Xba I and Not I sites of pTNLrSP4 to construct pTNL-LexA-p16XN.The BamH I-Not I fragment (Neo r ) of pTNL-p16XhoBam was replaced with the BamH I-Not I fragment (Neo r -LexA-p16XN) of pTNL-LexA-p16XN to construct pTNL-LexA-p16.The primers used in this study are shown in Table S1.

Gene targeting
CompoZr Custom Zinc Finger Nucleases targeting the p16 INK4A gene in HCT116 cells were purchased from Sigma-Aldrich.The Zinc Finger Nuclease (ZFN)-binding sites (upper cases) and cutting site (lower cases) are as follows: CCAAACACCCCGATTCaatttgGCAGTTAGGAA GGTTGTA.The targeting plasmid pTNL-LexA-p16 is not cleavable by this ZFN because the right ZFN-binding site is separated by the Neo r gene and LexA BE (GCAGTTA GGA-Neo r -LexA BE-AGGTTGTA).For gene targeting, HCT116 cells (8 9 10 5 cells) were transfected with linearized pTNL-LexA-p16 (Not I cut, 36 lg) and pooled ZFN mRNA (2 lg) using the TransIT-mRNA Transfection Kit (#MIR2225; Takara Bio, Shiga, Japan) according to the manufacturer's instructions.The transfected cells were cultured in the presence of G418 (0.4 mgÁmL À1 ) in a 10 cm dish.The surviving cell colonies were individually picked up and expanded as targeted cell lines (#102, #109, #113, and #170).To eliminate the floxed Neo r , the targeted cell lines were transfected with the green fluorescent protein (GFP)-fused Cre expression vector pCAG-Cre:GFP (#13776; Addgene) [15] using Lipofectamine 2000 transfection reagent (#11668027; Thermo Fisher Scientific) according to the manufacturer's instructions.After sorting of GFP-positive cells, the pooled cells were cultured and expanded.

Genotyping
The insertion of LexA BE into the target site was confirmed by genotyping PCR using extracted genomic DNA and KOD FX (Toyobo, Osaka, Japan) according to the manufacturer's instructions.The primers used in this study are shown in Table S1.The insertion was confirmed by Southern blot analysis with a probe targeting the Neo r gene as described previously [11].

Monoallelic locus tagging
Step 2 Expression of a capture protein Negative control cell Target cell

Step 3 Evaluation of yields of the target allele
Step 4 Identification of allele-specific interacting proteins Proteins to be analyzed Group, Rosemont, IL, USA), and anti-a-Tubulin Ab (#017-25031; Fujifilm Wako, Osaka, Japan), as described previously [11].In some cases, the immunoblotted membranes were cut before addition of the Abs.Band intensities on immunoblots were analyzed using IMAGEJ (https://imagej.net/ij/index.html).

Statistical analysis
GRAPHPAD PRISM 6 (GraphPad, San Diego, CA, USA) was used for statistical analysis using one-way analysis of variance (ANOVA).

iChIP-SILAC
iChIP-SILAC was performed as described previously [11].Briefly, two cell lines were cultured in SILAC Heavy and Light medium.After formaldehyde crosslinking, the cells (ca. 1 9 10 8 each) were collected in one tube and subjected to iChIP procedures to specifically isolate the target genomic regions.After SDS/PAGE followed by CBB staining, the stained parts were excised into six pieces and subjected to MS analysis at DNA-chip Development Center for Infectious Diseases (RIMD, Osaka University).The results including the identified peptide information are shown in Table S2.

Reverse transcription (RT)-PCR
Total RNA was subjected to RT with ReverTra Ace qPCR RT Master Mix with gDNA Remover (#FSQ-301; Toyobo) according to the manufacturer's instructions.The synthesized cDNA was used for PCR with KOD FX (#KFX-101; Toyobo).The primers used in this study are shown in Table S1.If necessary, each amplicon was subjected to DNA sequence analysis (Eurofins Genomics, Tokyo, Japan).DNA sequence data were analyzed using SNAPGENE VIEWER (https://www.snapgene.com/snapgene-viewer/).

Bisulfite-PCR and bisulfite-sequencing
Bisulfite treatment, PCR amplification, and DNA sequencing analysis were performed as described previously [16].

Knockdown experiments
ON-TARGETplus SMARTpool siRNA, which contains four kinds of siRNA sequences for each target RNA, was purchased from GE Healthcare (Buckinghamshire, UK) for use in siRNA-mediated knockdown experiments.ON-TARGETplus Non-targeting pool (#D-001810-10-05), which contains four kinds of siRNA sequences, was also purchased from GE Healthcare for negative control experiments.The detailed sequence information is provided in Table S3.HCT116 cells (3 9 10 4 cells) were transfected with 7.5 nM (final concentration) of siRNA using Lipofectamine RNAi MAX (#13778030; Thermo Fisher Scientific).After 2-4 days, DNA or RNA was extracted from the transfected cells.Alternatively, siRNA transfection was repeated each 3 or 4 days, cells were cultured for an additional 3-8 days (7-12 day culture in total), and then DNA was extracted.

Gene targeting for iChIP analysis
In HCT116 cells, a CpG island in the p16 INK4A gene promoter region is methylated on one allele [16], and its transcription is silenced (Fig. 1, Fig. S1).By contrast, the gene on the other allele contains a single guanine (G) insertion in the first exon, which leads to a frameshift mutation; therefore, the full-length protein is not synthesized from the transcribed mRNA (Fig. 2A).We first designed an iChIP procedure to   identify proteins that interact with the methylated CpG island in one allele (Fig. 1).In the first step, eight copies of LexA BE (0.16 kbp in total) were inserted 0.18 kbp upstream of the transcription start site of the p16 INK4A gene (the 5 0 edge of the CpG island) in HCT116 cells by homologous recombination (Figs 1 and 2A, Fig. S1).In this context, the insertion would stochastically occur for each or both allele(s).After single cell cloning, we confirmed the insertion of LexA BE by Southern blotting and genotyping PCR (PCR1 and PCR2) (Fig. 2B).In addition, we amplified the p16 INK4A gene in the non-target allele (PCR3) and confirmed the DNA sequences (G 9 4 or G 9 5) by DNA sequencing analysis.LexA BE was inserted in the target site of one allele in the established cell lines #102, #109, #113, and #170; among them, #102 and #109 possessed LexA BE in the CpG-methylated allele, whereas in #113 and #170, LexA BE was present in the unmethylated allele (Fig. 2B).Although the targeting construct did not include the G-insertion mutation (Fig. 2A), cell lines #113 and #170 retained this mutation on the targeted G-inserted allele.This result suggests that a partial sequence of the targeting construct would be utilized for homologous recombination in the cells.Next, we eliminated the neomycin-resistance gene, which was inserted concomitantly with the gene targeting construct, from the representative cell lines #109 and #113 by transient expression of the Cre recombinase.The cell lines #109-2 and #109-7 were established as targeted cell lines possessing LexA BE in the CpG-methylated allele, whereas #113-5 and #113-12 had LexA BE in the unmethylated allele (Fig. 2C).

Examination of the methylation status of the p16 INK4A gene
Next, we examined whether the LexA BE insertion affected the allele-specific properties of the p16 INK4A gene.Similar to the parental HCT116 line, the p16 INK4A gene was transcribed from the G-inserted allele in the established cell lines (Fig. 3A,B).In addition, bisulfite-PCR showed that the cell lines retained the CpG-methylated and unmethylated alleles (Fig. 3C).We cloned the bisulfite-PCR amplicons in plasmids and confirmed the G 9 5 (G-insertion) or G 9 4 (non-G-insertion) sequence of the clones by DNA sequencing analysis.The DNA clones for the methylation-specific primer set included only the G 9 4 sequence, whereas those for the unmethylation-specific primer set contained only the G 9 5 sequence (Fig. 3D).These results demonstrate that the LexA BE insertion does not affect the allele-specific status of CpG methylation or gene transcription (Fig. 3E).

Isolation of the p16 INK4A gene locus by iChIP
The 3xFNLDD construct was expressed in #109-2 and #113-5 cells for affinity isolation of the targeted p16 INK4A region by iChIP.3xFNLDD expression was confirmed by immunoblotting with an anti-FLAG antibody (Ab) (Fig. 4A) and its expression did not change CpG methylation status at or transcription from the target locus (Fig. S2a-d).Next, iChIP with an anti-FLAG Ab was performed to isolate the target region from #109-2-1 and #113-5-1 (Fig. 4B).The yield of the target region was approximately 8% of the input for both cell lines, whereas that of an irrelevant locus (SOX2) was negligible.These results demonstrate locus-specific, efficient, and comparable target isolation from both cell lines by iChIP.

Identification of proteins interacting with the methylated p16 INK4A CpG island by iChIP-SILAC
To identify proteins that bind to the methylated p16 INK4A CpG island, we performed iChIP-SILAC analysis (Fig. 4C).Cellular cultivation in the SILAC medium (#109-2-1 in Heavy medium and #113-5-1 in Light medium) did not affect the locus-specific, efficient, and comparable target isolation by iChIP (Fig. S2e).iChIP-SILAC identified 518 candidate proteins that interacted with the p16 INK4A CpG island (Fig. 4C, Fig. S2f, Table S2).A Heavy/Light value > 1.0 indicates that the identified proteins were more abundant in #109-2-1 than in #113-5-1 (Fig. 4C, Table S2).DNMT1, which is considered to be present on methylated genomic regions, was detected as a protein as expected, and its Heavy/Light value was 1.203 (Table 1, Table S2).As shown in Table 1 and Fig. 4C, 68 proteins showed a Heavy/Light value > 1.3 (arbitrary threshold).The 68 proteins were categorized according to their molecular functions and biological processes annotated in the UniProt database (https://www.uniprot.org/)(Fig. 4D,E, Table S4).The identified proteins included DNA-binding proteins, chromatin regulators, methyltransferases, and transcriptional regulators.Table 1 lists the proteins related to chromatin condensation and histone modification.For example, DNA topoisomerase 2-alpha (TOP2A) localizes to heterochromatin regions and functions in chromatin condensation [17,18].Histones H1x and H1.0 are linker histones and are classified as strong condensers of  chromatin structure [19].Nucleolar complex protein 2 homolog (NOC2L) is an inhibitor of histone acetyltransferase [20].UHRF1 is a key regulator of CpG methylation [21][22][23].COOH-terminal binding protein (CTBP1) interacts with Pc2, the polycomb group protein, and suppresses the transcription of the p16 INK4A gene in human cells [24][25][26].The more abundant detection of these proteins at the methylated locus indicates that iChIP-SILAC is a useful method for the identification of proteins related to epigenetic gene silencing events in a target locus.

Proteins related to methylation of the p16 INK4A
CpG island To examine the involvement of the identified proteins in the CpG methylation event of the p16 INK4A gene, we performed siRNA-mediated knockdown experiments and evaluated the CpG methylation status in HCT116 cells.We selected 28 proteins including histones and enzymes such as UHRF1 as siRNA targets (Fig. S3a).CpG is demethylated passively through DNA replication and actively by enzymes such as those belonging to the ten-eleven translocation (TET) family [27].If CpG methylation of the target gene is inhibited by knockdown of the identified proteins, the amounts of CpG methylation would be expected to decrease gradually or actively by the TET family.We therefore examined for CpG demethylation at later time points (e.g., 12 days).In some cases, the siRNA treatment was continued for 12 days (Fig. S3a).However, cells in which siRNAs suppressed cell proliferation were harvested at earlier time points [2 or 7 (8) days after siRNA treatment].After genomic DNA extraction from the harvested cells, we performed bisulfite-PCR, DNA cloning, and DNA sequencing analysis.Some siRNAs caused demethylation of the CpG island of the methylated allele (CpG demethylation: ≥ 4 sites, Fig. S3).The same experiments were repeated with the selected siRNAs to confirm the induction of demethylation (Fig. 5).In addition, we confirmed the downregulation of the siRNA targets at the mRNA level (Fig. S4).Compared with the negative control, siRNAs against FBL, EIF2S3, DDX24, and UHRF1 showed remarkable demethylation of the CpG island (CpG demethylation: ≥ 4 sites, Fig. 5C).
In this regard, compared with the negative control, CpG demethylation at 0-2 site(s) was dominant,  whereas CpG demethylation at three sites was rare (≤ 20% of analyzed amplicons, Table 2).By contrast, in two independent experiments, more than 30% of the analyzed amplicons reproducibly showed CpG demethylation at ≥ 4 sites when siRNAs against FBL, EIF2S3, DDX24, and UHRF1 were used, whereas no amplicons showed the same levels of CpG demethylation when the negative control siRNA was used (Table 2).In addition, siRNAs against DDX24 and UHRF1 resulted in statistically higher levels of demethylation than that of the negative control siRNA in at least one experiment (Fig. 5C, Fig. S3c).These results suggest the potential involvement of FBL, EIF2S3, and DDX24 in the maintenance of methylation of the p16 INK4A CpG island in addition to UHRF1.The siRNAs against FBL and EIF2S3 strongly suppressed cell proliferation.However, the suppression effect may not be directly related to CpG demethylation because the siRNA against VCP, which also showed suppression of cell proliferation, did not induce CpG demethylation (Fig. S3).We speculate that FBL and EIF2S3 possess various functions, one of which may be related to cell proliferation.

DDX24 regulates DNMT1 protein levels
Next, we examined the interaction of FBL, EIF2S3, and DDX24 with DNMT1 to evaluate their involvement in CpG methylation events.We expressed 3xFLAG-tagged human proteins (3xFLAG-DDX24, 3xFLAG-FBL, and 3xFLAG-EIF2S3) in HCT116 and performed immunoprecipitation followed by immunoblot analysis.Figure 6A shows that the 3xFLAG-tagged proteins were effectively immunoprecipitated by the anti-FLAG Ab at the same levels although their nuclear expression levels are different.We found that DNMT1 co-immunoprecipitated with DDX24 but not with FBL or EIF2S3 (Fig. 6A, Fig. S5).These results suggest that DDX24 controls CpG methylation of the p16 INK4A gene via its interaction with DNMT1.DDX24 interacts with USP7 [28], which binds to DNMT1 for the maintenance of CpG methylation [29][30][31].We also confirmed coimmunoprecipitation of DDX24 with USP7 in HCT116 (Fig. 6A).
To determine the mechanism of DDX24-mediated CpG methylation, we examined the effects of knockdown of DDX24 on DNMT1 protein levels using plasmids expressing shRNA that target human DDX24 mRNA.As shown in Fig. 6B, shRNA-mediated knockdown decreased the expression of DDX24 by approximately 30-50% in HCT116.Interestingly, DNMT1 protein levels appeared slightly decreased by DDX24 knockdown (Fig. 6B).Therefore, we analyzed DNMT1 band intensities in more detail.As shown in Fig. 6C, knockdown of DDX24 moderately but significantly decreased DNMT1 protein levels.Therefore, by forming a complex with USP7 and DNMT1, DDX24 regulates DNMT1 protein levels to maintain the CpG methylation status of the p16 INK4A gene in HCT116.USP7 is a deubiquitinating enzyme that protects target proteins from degradation by the ubiquitin-proteasome  system [32,33].Because DDX24 knockdown did not affect USP7 protein levels (Fig. 6B), it seems likely that it disturbed the association between USP7 and DNMT1.In this context, we propose that DNMT1 unbound to USP7 is degraded, resulting in CpG demethylation (Fig. 6D).

Discussion
In this study, we used iChIP-SILAC for the identification of proteins that interact with the CpG-methylated p16 INK4A gene in HCT116 cells.The identified proteins included epigenetic regulators such as UHRF1 and DNMT1, confirming the reliability of iChIP analysis.However, a limitation of iChIP is that it requires genetic manipulation of a target locus, which may change the physiological states of the locus (e.g., alteration of epigenetic modifications and interacting proteins).Therefore, it is necessary to confirm that the intrinsic properties of the target locus (e.g., CpG methylation and transcription status) are retained after insertion of LexA BE and expression of 3xFNLDD, although the possibility that the manipulations have undetected effects cannot be excluded.In this regard, the results presented in Fig. 3 and Fig. S2a-d suggest that the intrinsic properties of the target locus were retained, at least partially.The identification of UHRF1 and DNMT1 as proteins that are more abundantly associated with the CpG-methylated allele also supports this notion.It is noteworthy, however, that although the list of identified proteins (Table S2) includes bona fide proteins interacting with CpG-methylated genomic regions such as UHRF1 and DNMT1, it is necessary to validate whether the identified proteins regulate methylation of the target locus under physiological conditions.In this regard, we demonstrated that several identified proteins were involved in the CpG methylation events in non-manipulated cells, showing that at least some of the proteins identified by iChIP function in physiological conditions.In HCT116 cells, the p16 INK4A gene on two alleles can be discriminated by differences in the DNA sequence (G 9 4 and G 9 5).However, in general, target genes may not retain such convenient sequence differences.Even in such cases, evaluation of CpG methylation status (e.g., by bisulfite-PCR) by considering the LexA BE insertion enables discrimination of the targeted allele.If the inserted LexA BE is detected only in the CpG-unmethylated allele, there are two possible explanations: (a) CpG methylation-derived heterochromatinization disturbs the insertion of LexA BE, and (b) CpG demethylation is induced by the insertion of LexA BE.In this case, the LexA BE insertion site should be changed.Knockdown of some of the identified proteins affected the CpG methylation status of the p16 INK4A gene (Fig. 5).DDX24 is a DEAD-box helicase that binds to single-and double-stranded RNA [34].We found that DDX24 regulated DNMT1 protein levels to maintain the CpG methylation status of the p16 INK4A gene in HCT116.EIF2S3 is a subunit of eukaryotic translation initiation factor 2. Although gene mutations in the EIF2S3 gene cause MEHMO (Mental retardation, Epileptic seizures, Hypogenitalism, Microcephaly, Obesity) syndrome [35], the biochemical roles of EIF2S3 in CpG methylation remain unknown.FBL catalyzes 2 0 -O methylation (2 0 -O-Me) of ribosomal RNA by binding to the enhancer of zeste homolog 2 (EZH2), a transcription repressor, and histone lysine methyltransferase [36].FBL also methylates histone H2A at glutamine 104 (H2AQ104me) [37].These epigenetic functions of FBL may contribute to the allele-specific CpG methylation of the p16 INK4A gene.It will be interesting to examine the molecular mechanisms in more detail in future studies.
Various epigenetic events including DNA/histone modification collaborate for the refined regulation of genome functions.In this study, we identified DDX24 as a novel regulator of CpG methylation using iChIP analysis.However, it is still unclear how the epigenetic machinery is regulated by the identified proteins.In this regard, we believe that sharing the list of identified proteins with the scientific community is important and informative for elucidating the underlying mechanisms, as well as for identifying novel targets for epigenetic drugs against cancer.In addition, we believe that iChIP is a useful tool for the biochemical analysis of the molecular mechanisms involved in genomic functions such as genome imprinting and epigenetic gene silencing/activation.The proposed step-by-step strategy for iChIP analysis could be of value for researchers to target other epigenetically regulated loci.

Fig. 1 .
Fig. 1.Procedures of iChIP analysis.iChIP-SILAC experiments were performed according to the steps shown on the left hand side of the figure.Experimental details are shown on the right hand side.

Fig. 2 .
Fig. 2. Insertion of LexA BE in the CpG island around the p16 INK4A promoter region.(A) Targeting strategy.A targeting construct including LexA BE and floxed neomycin-resistance (Neo r ) gene was used for gene targeting.The p16 INK4A gene contains a single guanine (G) insertion in the first exon of one allele.(B) Southern blot analysis and genotyping PCR to confirm insertion of LexA BE and the knock-in allele.The Southern blot bands are shown with an arrow.For PCR genotyping, the extension time of PCR3 was designed to amplify the non-targeted but not the targeted allele.The amplicons of PCR3 were subjected to DNA sequencing analysis to verify the non-targeted allele (G 9 4 or G 9 5).The target allele was chosen based on the results of the sequencing analysis.(C) Genotyping PCR to confirm Cre-mediated deletion of the floxed Neo r gene.Original images are shown in the Supporting Information (Supporting Image file).

Fig. 3 .
Fig. 3. Examination of the allele-specific properties of the p16 INK4A gene.(A) Reverse transcription (RT)-PCR to examine the expression of p16 INK4A mRNA in the cell lines retaining LexA BE in the p16 INK4A promoter region.GAPDH mRNA was detected as an internal control.WT: parental HCT116.(B) Results of DNA sequence analysis of the amplicons (p16) shown in (A).Signals around the G 9 4/G 9 5 positions are shown.(C) Bisulfite-PCR analysis using methylation-specific and unmethylation-specific primer sets.(D) Results of DNA sequence analysis of the amplicons shown in (C).Amplicons were cloned, and five clones (each amplicon) were analyzed by DNA sequence analysis to confirm the G 9 4/G 9 5 sequence.(E) Summary of the properties of the p16 INK4A gene in the cell lines.Original images are shown in the Supporting Information (Supporting Image file).

Fig. 4 .
Fig. 4. iChIP-SILAC analysis.(A) Immunoblot analysis to confirm the expression of 3xFNLDD in the cell lines retaining LexA BE in the p16 INK4A promoter region.The result of CBB staining is shown as a protein loading control.Original images are shown in the Supporting Information (Supporting Image file).(B) Isolation of the p16 INK4A promoter region by iChIP.The primer positions for real-time PCR and the yields of iChIP are shown in the upper and lower panels, respectively.(C) Schematic and results of iChIP-SILAC.For iChIP-SILAC analysis, the cultured cell lines were fixed, mixed in one tube, and then subjected to iChIP procedures followed by liquid chromatograph-mass spectrometry (LC-MS/MS).The result of CBB staining is shown in Fig.S2f.(D, E) Molecular functions (D) and biological processes (E) of 68 proteins (Heavy/Light ratio: > 1.3) annotated in the UniProt database.More than two molecular functions or biological processes are annotated to some proteins.Details are shown in TableS4.

FEBS
Letters 598 (2024) 1094-1109 ª 2024 The Authors.FEBS Letters published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies.

1101FEBS
Letters 598 (2024) 1094-1109 ª 2024 The Authors.FEBS Letters published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies.

Fig. 5 .-
Fig. 5. Roles of the identified proteins in methylation of the p16 INK4A CpG island.(A) Schematic of siRNA experiments in HCT116 cells.(B) Positions of the primer sets used for bisulfite-sequencing analysis.The graphic view of the CpG island was drawn by MethPrimer [39].(C) Demethylation of the CpG island (35 CpG sites) by siRNA-mediated knockdown.After bisulfite-PCR, DNA cloning, and DNA sequencing analysis of each clone, the methylation status of the CpG-methylated G 9 4 allele was evaluated.Positions of CpG demethylation at three sites are shown with dotted lines.At least 10 independent DNA clones were analyzed for each siRNA.P-values (***< 0.001, one-way ANOVA with Dunnett's test) are shown when the number of demethylated sites was statistically higher than that of the negative control siRNA.The experiments were performed twice independently.See also Fig. S3.

Fig. 6 .
Fig. 6.DDX24 regulates DNMT1 protein levels.(A) Interaction of DDX24 with DNMT1.The 3xFLAG-tagged human proteins were expressed in HCT116.After nuclear extraction, immunoprecipitation was performed with an anti-FLAG Ab.Untransfected cells were used as a negative control.Nuclear extracts (1% Input) and the immunoprecipitation (IP) fractions were analyzed by immunoblot (IB) analysis.Identical results were obtained from independent experiments, and a representative result is shown.(B) Knockdown of DDX24 decreases protein levels of DNMT1.Two kinds of shRNA were designed to target human DDX24 sequences (#5 and #6).shRNA targeting GFP was used as a negative control.Equal amounts of whole cell lysates were subjected to IB analysis.a-Tubulin was used as an internal control.Relative band intensity of DDX24 is indicated under each band image.Identical results were obtained from independent experiments, and a representative result is shown.(C) Semi-quantitative analysis of immunoblot band intensities.DNMT1 band intensity was measured and normalized to that of a-Tubulin.Results of three independent knockdown experiments are shown as mean AE standard deviation.P-values (*< 0.05, **< 0.01, ***< 0.001, one-way ANOVA with Tukey's test) are shown when statistical significance was achieved.(D) Schema showing the potential mechanism of DDX24-mediated maintenance of CpG methylation.DDX24 knockdown reduces DNMT1 protein levels, which results in CpG demethylation.Yellow circle: CpG methylation.Original images are shown in the Supporting Information (Supporting Image file).
FEBS Letters 598 (2024) 1094-1109 ª 2024 The Authors.FEBS Letters published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies.

Table 1 .
(Continued).from the latest database probably due to separated registration of HSPA1A and HSPA1B.; b DNMT1 is shown as a reference although Heavy/Light ratio is < 1.3.

Table 2 .
Summary of siRNA experiments for selected targets.
FEBS Letters 598 (2024) 1094-1109 ª 2024 The Authors.FEBS Letters published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies.