Identification of Novel MeCP2 Cancer-Associated Target Genes and Post-Translational Modifications

Abnormal regulation of DNA methylation and its readers has been associated with a wide range of cellular dysfunction. Disruption of the normal function of DNA methylation readers contributes to cancer progression, neurodevelopmental disorders, autoimmune disease and other pathologies. One reader of DNA methylation known to be especially important is MeCP2. It acts a bridge and connects DNA methylation with histone modifications and regulates many gene targets contributing to various diseases; however, much remains unknown about how it contributes to cancer malignancy. We and others previously described novel MeCP2 post-translational regulation. We set out to test the hypothesis that MeCP2 would regulate novel genes linked with tumorigenesis and that MeCP2 is subject to additional post-translational regulation not previously identified. Herein we report novel genes bound and regulated by MeCP2 through MeCP2 ChIP-seq and RNA-seq analyses in two breast cancer cell lines representing different breast cancer subtypes. Through genomics analyses, we localize MeCP2 to novel gene targets and further define the full range of gene targets within breast cancer cell lines. We also further examine the scope of clinical and pre-clinical lysine deacetylase inhibitors (KDACi) that regulate MeCP2 post-translationally. Through proteomics analyses, we identify many additional novel acetylation sites, nine of which are mutated in Rett Syndrome. Our study provides important new insight into downstream targets of MeCP2 and provide the first comprehensive map of novel sites of acetylation associated with both pre-clinical and FDA-approved KDACi used in the clinic. This report examines a critical reader of DNA methylation and has important implications for understanding MeCP2 regulation in cancer models and identifying novel molecular targets associated with epigenetic therapies.


INTRODUCTION
Epigenetic dysregulation involving mutations or abnormal expression of DNA methylation readers has been associated with a broad spectrum of disorders that range from Rett Syndrome to human cancers (1)(2)(3)(4)(5)(6)(7), and alterations in both the writing and reading of epigenetic marks have been linked with tumor progression at every stage (8)(9)(10)(11)(12). Aberrant DNA methylation not only promotes disease progression but is targeted via therapeutics applied in the clinic (13)(14)(15). Because of the prevalence of abnormal epigenetic changes in tumor progression (16)(17)(18)(19)(20)(21)(22), exploitation of this property led to FDA approved "epigenetic" therapies (23,24). Interestingly, DNA methylation readers, such as methyl-CpG-binding protein 2 (MeCP2), bridge DNA methylation and histone modifications by binding to methylated DNA and recruiting co-repressor proteins (25)(26)(27)(28). While both normal and abnormal DNA methylation is read by MeCP2, much remains unknown about its role and regulation in cancer-associated pathologies. MeCP2 was shown early on to have an affinity for 5-methylcytosine in the context of methylated CG dinucleotides (mCG) (29,30) and methylated CH (mCH), where H = A/C/T. MeCP2 binds methylated cytosine (31)(32)(33) and shows selectivity for mCG sequences with adjacent A/T sequences (34). However, it also binds to hydoxymethylated cytosine (31,32,(35)(36)(37). While more investigation is needed, MeCP2 binding to mCH has been primarily noted on mCA (31-33, 35, 36, 38). Studies have also shown that MeCP2 binding in mouse brain is proportional to mCAC + mCG density wherein transcription is sensitive to MeCP2 occupancy (38). Additionally, MeCP2 regulates tumor suppressor genes (TSG) silencing, and serves as a critical bridge for histone methyltransferases (HMTs) (25), histone deacetylases (HDACs) (26,28,39), and other proteins that bind modified histones or that mediate nucleosome remodeling (27,40,41). Moreover, MeCP2 has been reported to be amplified in diverse cancer including human triple-negative breast cancers (TNBC), and it activates growth factor pathways targeted by activated Ras, MAPK and PI3K pathways (42). Novel interacting protein partners and gene targets in brain tissue have also been identified (43). These are the types of enigmatic and versatile properties of MeCP2 that have contributed to long-standing knowledge gaps. We previously reported that inhibition of SIRT1 triggers acetylation of endogenous MeCP2 at lysine (K171), a site that regulates MeCP2 interaction with HDAC1 and ATRX (44). These findings demonstrated that MeCP2 post-translational modifications (PTMs) can critically impact its function, yet few PTMs have been mapped despite the potential that they might affect substrate specificity (35,45). This knowledge gap is especially important given reports demonstrating unique characteristics of MeCP2 domains in determining binding specificity (46,47) and the impact of MeCP2 on chromatin-dependent regulation of epigenetic writers (48). In the present study, we have identified additional novel PTMs across the length of MeCP2 and target genes in cancer models. Our findings provide new insight on the versatile role of MeCP2 which is known to be critical in regulating gene imprinting (49), transcriptional activation and repression (50) in disparate conditions that range from autism to cancer (4,51,52).  26) cell lines used in this manuscript were purchased from ATCC which utilizes STR technology for cell authentication. Cells were used at a low passage (<20) within 6 months or less after receipt or resuscitation. MDA-MB-468, T47D, and BT549 cells were cultured in RPMI 1640 (Gibco). MCF10A and MCF12F were cultured in HuMEC medium supplemented with HuMEC supplement kit (Gibco). PC3 cells were cultured in ATCC formulated F-12K media (ATCC). MCF7 cells were propagated in MEM while MDA-MB-231 cells were cultured in DMEM (Gibco). T47D and MCF7 cells were cultured in media supplemented with 0.1% insulin (Sigma). All cells were grown in culture media supplemented with 1% pen-strep and 10% fetal bovine serum from GIBCO at 37°in 5% CO 2 .

Bioluminescent MDA-MB-468 Cells
The pGL4.50[luc2/CMV/Hygro] plasmid (E1310) which encodes the luciferase reporter gene luc2 (Photinus pyralis) was purchased from Promega. MDA-MB-468 cells were plated in a 6-well plate (Genesee) at the seeding density of 2 × 10 5 cells in order to reach 60% confluency at the time of transfection. Cells were transfected with 1 µg of the pGL4. 50 lentiviral transduction  particles purchased from Sigma for MeCP2 (TRCN0000330971,  TRCN0000330972) and Non-Targeting shRN control transduction particles (SHC002V). The transduction was enhanced with 5 mg/ml polybrene (Sigma Millipore) and 2× multiplicity of infection (MOI) viral particles was added to the media. After 24 h, culture media was replaced with fresh media for 2 days. Stable clones were selected with 6 mg/ml puromycincontaining media which was replaced every 3-4 days until selection was achieved and knockdown confirmed by Western blots and qPCR.

RT-PCR and qPCR
Total RNA was isolated using the Aurum ™ Total RNA Mini Kit (Bio Rad) and 2 µg of RNA was used to produce cDNA via the SuperScript ® III First-Strand Synthesis System for RT-PCR (Invitrogen). Intron-spanning primers designed for gene expression analysis are summarized in Table 1. All primers were validated by end-point PCR (RT-PCR), a minus reverse transcription control (−RT control) was included in all RT-PCR experiments. Equal amount of synthesized cDNA was used for qPCR using the Power UP SYBR Green (Thermofisher Scientific, A25778) and the CFX96 Real-Time System C1000 Touch Thermal Cycler (Bio Rad). b-actin gene expression was used as endogenous control for mRNA quantification, as is not a MeCp2 target gene in both cell lines studied and its expression didn't change after the depletion of MeCP2 in RNA-seq analysis.

Western Blots
Protein extracts were generated using RIPA lysis buffer supplemented with protease inhibitor cocktail (Thermo Fisher Scientific). The protein concentration was measured with the BCA method. Approximately 50 mg of protein from each sample was loaded on NuPAGE ™ 4-12% Bis-Tris Protein Gels (Thermo Fisher Scientific) and run at 175 V constant voltage. A constant voltage of 30 V was used for protein transfer onto polyvinylidene fluoride (PVDF) membranes (Millipore-Sigma). Blots were probed with rabbit anti-MeCP2 antibodies (1:1,000; Cell Signaling) and mouse anti-HA antibodies (1:2,000; Santa Cruz) overnight at 4°C. After three washes with tris-buffered saline and polysorbate 20 (TBST; Fisher scientific), blots were then incubated with anti-rabbit HRP conjugate secondary antibody (1:5,000) and anti-mouse HRP conjugate at room temperature for 1 h. After washing three times, chemiluminescence (Pierce ECL Western blotting substrate: Thermofisher Scientific, A25778) was then used to visualize protein bands. b-actin antibody (1:10,000; Santa Cruz) was used as control.
Immunofluorescence About 1 × 10 5 cells were plated on coverslips 48 h prior and they were washed with PBS and then fixed with 4% paraformaldehyde for 15 min. After washing with PBS they were then permeabilized with 0.2% Triton X-100 for 20 min and blocked with 5% BSA for 30 min. Following that, cells were incubated with primary antibodies MeCP2 (Cell Signaling, 3456S), or HA (Cell    Table 2 were used for analysis.

ChIP Sequencing
For ChIP-Seq experiments, ChIP DNA was prepared as described above library preparation was followed by high throughput sequencing with Illumina Hi-seq 2000 at GENEWIZ Corporation.

RNA Sequencing
RNA was prepared as described above, and library preparation and sequencing were performed at Center for Biotechnology & Genomics of Texas Tech University. RNA quality was determined using RNA Screen Tpe (Agilent). Ribosomal RNA depletion was achieved using NEB Next rRNA Depletion Kit (Human/Mouse/Rat) (NEB # E6310X). RNA fragmentation, double stranded cDNA and adaptor ligation was generated

ChIP Sequencing and RNA Sequencing Data Analysis
For ChIP-Seq analysis, the FASTQ files were analyzed using DNASTAR's Laser Gene software. MEME-ChIP was used to analyze MeCP2 binding motifs and TOMTOM to identify if those motifs were similar to known consensus sequences using the MEME Suite Programs http://meme-suite.org/index.html (53). We downloaded the FASTQ data sets of RRBS for MCF7 cells from the ENCODE portal (54) (https://www.encodeproject.org/) with the following identifiers: ENCSR943EFS, and ENCSR939RXT; then avisualized with Integrative Genomics Viewer (IGV). Venn diagrams to identify the overlapping genes were generated using the Venny tool https://bioinfogp.cnb.csic.es/tools/venny/index. html. For RNA-Seq analysis, the RNA-Seq reads were normalized by RPKM and assembled by mapping reads directly to the annotated human reference genome using the DNASTAR SeqMan software (DNASTAR, Inc., Madison, WI). Differential gene expression levels were quantified using Fisher's Exact Test Signal Search in the DNASTAR ArrayStar software package (DNASTAR, Inc., Madison, WI). Differentially expressed genes were filtered if they met the criterion for a two-fold change, a p-value that was less than.05 at a 95% confidence interval. For each comparison, genes were sorted based on fold change, from low to high. The results were ported into Excel spreadsheets where the log2 of the fold change for each gene was calculated.

RNA Analysis In Silico
Relative RNA expression of 20 selected genes in breast cancer and normal adjacent tissue was downloaded from UCSC Xena platform on 11th of April 2020 (1,092 breast cancer primary tumors and 114 normal tissues).

Liquid Chromatography/Mass Spectrometry (LC-MS/MS)
PC3 and MDA-MB-468 cells were cultured and seeded in p150 mm dishes at 37°C under atmospheric oxygen conditions. Once 70% confluent, cells were treated with DMSO, 2µM panobinostat, 10 µM Inhibitor-IV, 10 µM Inhibitor-VII, and 10 µM pracinostat for 45 min to 1.5 h and harvested in RIPA buffer (with complete protease inhibitor cocktail, 1 µM Trichostatin A and 1 mM nicotinamide). Protein concentration was quantified by the BCA method. Immunoprecipitation was performed using 4 mg of anti-MeCP2 antibody (Cell Signaling) and incubated for 2 h at 4°C. Protein A dynabeads (Invitrogen) were added to the immune-complex and incubated for 2 h at 4°C. IP protocol was followed as mentioned above. Beads were washed with RIPA buffer (four times) and autoclaved water (two times). Dry beads were shipped to Applied Biomics Inc. (Hayward, CA) for acetylation site identification by LC-MS/MS mass spectrometry on a fee-based service. The specific lysine residues that were acetylated, exhibited ion peaks at mass/charge (m/z) ratio of~126 as summarized in Figure 4A.  ACOT2  GGT TTT GCT GTG ATG GCT CT  AAG GTA GTG TGT GTT GGG GTA G  AP1M2  ACA GAG ATG TGC GGT GCA A  TTC CAC CCT CAG CCA TTG AT  AQP1  ACT TCA GCA ACC ACT GGG TAG  ATT TCC TGT CCT CTG GCT GTC  ATG4D  TGT ACC GTG GGC TTC TAT GC  CAC CTT TTC AGG GGT GGA CA  CCT6A  TTG GTT CAT TGC GAC TAC CA  TCA CTT GAG GCC AGG AGT TT  CHCHD2  AGT AAT GGC GTG ACC CAA TGT  TGG TTG GAA TTG GGA ACT TGA TG  DKK1  ATT GGC AGG AAC AGG ATG TGT  GAG TGG AAT GAG GAA GGA TTT GT  DLX1  GGT CTT CAT TTG TTG CTC GC  AAT CCT TCC TGC GCC CTA AT  DNMT1  CAA AAG GGG AAC CTT GTT CA  CCT GGG AGG AAG AAA TAG GG  DUX4  TTG CCA TTT GTT CAC TCT GC  TGA TCC TGG GGT TGA AAG TC  EGFR  CCT TGG TTG TCC TCC TCC TAA  JARID2  ACG AGT GTA TGG GTG AGT GC  CCA TTG CAG CCA TTT GTC CC  KDM1A  AAG CCA ACG GAC AAG CTG TA  ACA TCA CAT CAT CTC TAC CCT CA  KDM1B  AGT TTG GAA AAC CTG CAA CAC T  AGA GTA GGT GAT TTC GCT GGG  KDM2A  TGC TTC TCA ATG TGC TCT CCA  GCC AGG CTG AAA ACA CTT ACT T  KDM3B  AAC TCC TTT GCT CTC AGC GT  TCC AAA TCT TAC CTC CCC GTC  KDM4A  GGG TCA AAG CAC TTG GGG AT  GCT TCA CAG AGC AAC AAG GC  KMT2A  ACC ACC ATG TGA CTA TTG GAC TT  ACA GCT CTT ACA GCG AAC ACA  KMT2B  AAC CCC ACC CAT TTC CCT GTT  TGG GAG GCC AGG AAG TTG AA  KRI1  TGA TAA ATG CGG GGG TCC TT  TCC ATC CTA ATC CCT ACG CTG A  LANCL2  CCA TTA ACT TGG GAG GCT GA  GGA CTG CAA TGT CAC CAA TG  METTL7A  GCT CTG TGG ATG TGG TGG TC  CTC ACA CCC TTT CAC TCA CCG  MRPS17  TAG GTG CCA AGG ATG GTT TC  CTC CCA AAG GTC AAG GAT CA  NUPR1L  AAA GCC TGC GGA ACT TCA TA  GGA TGG TTT CGA TCT CCT GA  OXCT2  TTG ATG TCG TCC ACC GTC AG  GAC CTG GCG AAC TGG ATG AT  P2RY11  CTG GTG GTT GAG TTC CTG

Statistical Analysis
Statistical analysis was performed using unpaired student's t tests (Graph Pad Prism software) to assess whether differences observed in the various experiments were significant. All results are expressed as mean ± SEM and considered significant at *p <0.05, **p <0.01 and ***p <0.001.

MeCP2 Binds Novel Genes in Breast Cancer Cells Associated With Diverse Biological Functions
Since the discovery that MeCP2 regulates transcription and mutations in the gene cause Rett Syndrome, there has been considerable interest in what regulates its function and what downstream genes are targeted (55,56). DNA methylation and its readers influence transcription activation and repression in a context-dependent manner depending on the genomic location of binding (57,58). While this process is known to be frequently altered in cancers (59-61), many unknowns remain regarding the role of MeCP2 in regulating gene expression. Given abnormal DNA methylation in breast cancers (10,12,62,63) and MeCP2 amplification in cancers (41) Figure 1D). We further analyzed the methylation status for genes in MCF7 cells for which publically available Reduced Representation Bisulfite Sequencing (RRBS) data was available ( Figure S1). We found that MeCP2 binds to genes in MCF7 cells in regions where CpG methylation had been mapped such as SDK1, a cell adhesion molecule; Jagged 2 (JAG2), a Notch ligand; glycogenin 2 (GYG2), an enzyme involved in glycogen synthesis ( Figure 1C, Figure S1C). These novel MeCP2 targets as well as others in Figure S1 had not previously been linked with MeCP2, but have been linked with pathobiology associated with cancer (76)(77)(78)(79)(80)(81)(82)(83)(84) or genetic disorders such as Leigh syndrome (85) and Raine syndrome (86,87). We also found that MeCP2 binds to genomic regions devoid of CpG methylation such as for USP34, MGAM, GMDS, SLC45A4, SSU72, CAPN2 and PLXN2 ( Figures  S1B-C). Similarly, while these are novel targets of MeCP2, many have been implicated in diverse cancers (67,71,(88)(89)(90)(91)(92). This further shows the complexity of MeCP2 binding across the genome. To identify the DNA motifs associated with MeCP2 genomic binding, we analyzed the genomic fragments sequenced in our MeCP2 ChIP-Seq analyses performed in triplicate. The MEME-ChIP analysis revealed a motif consistent across three independent experiments for both MDA-MB-468 and MCF7 cells ( Figure 1E and Table 3).

MeCP2 Localizes to Novel Genes and Regulates Their Expression
We further determined the global occupancy of MeCP2 with respect to cellular functions and performed pathway analysis to identify the core pathways associated with the newly identified target genes in MDA-MB-468 and MCF7. We observed an enrichment of the gene expression, immune system, metabolism, metabolism of proteins, and signal transduction pathways (Figure 2A). We further randomly chose more than 100 genes identified in the triplicate analysis of MeCP2 ChIPseq in MDA-MB-468 cells and validated MeCP2 binding via MeCP2 ChIP-PCR, some of which are shown in Figure 2B. Consistent with our MeCP2 ChIPseq analyses, we found via MeCP2 ChIP-PCR that MeCP2 localizes to various gene promoters involved in diverse biological processes such as immune system regulation (IL6, ICAM3, and ICAM5), signal transduction (EGFR, WNT3A, and DKK1), transcription (KMT2A, SIRT1, HDAC1, DNMT1), developmental biology (DUX4) and lncRNAs with two different shRNA (sh1 and sh3) in MDA-MB-468 cells ( Figure S2A). We also observed by quantitative RT-qPCR a change in mRNA expression of novel targets in which were validated for knockdown ( Figure S2A). A minimum of three independent experiments showed that depletion of MeCP2 caused a change in the expression of several of the genes whose promoter it bound. We found that knockdown of MeCP2 in MDA-MB-468 cells caused an increase in some genes such as IL6, KDM3B, HIPK3, KDM3A, EGFR, and KMT2B and a reduction others such NUPR1L (also known as NUPR2), METTL7A, PSPH, LANCL2, MRPS17 and HDAC1.
( Figure 2C and Figure S2B). Together these results show the complexity of MeCP2-mediated regulation of gene expression.

MeCP2 Targets Genes With Differential Expression Between Breast Cancer and Normal Samples
To evaluate the global effect of MeCP2 on RNA expression we performed RNA-Seq in MDA-MB-468 cells. We analyzed three independent experiments of non-targeting control (NTC) versus sh1 MeCP2 and NTC versus sh3 MeCP2 and found changes in 899 genes and 875 genes, respectively ( Figure 3A). Overlap of ChIP-Seq hits and RNA-Seq hits showed 175 potential transcriptional targets of MeCP2 ( Figure 3B). A pathway enrichment analysis of these potential targets showed their participation in the immune system, metabolism, metabolism of proteins, and signal transduction, among other pathways (Table 4). Moreover, these genes were differentially expressed in normal vs. breast cancer tissue ( Figure  3C), and several of these target genes have been previously reported to be tumor suppressors (93)(94)(95)(96)(97)(98)(99) while others were reported to be oncogenes (100-103) ( Figure 3D).

Endogenous MeCP2 Is Acetylated at Key Lysine Residues and KDI Further Influence Its Acetylation Patterns
We previously reported that MeCP2 undergoes acetylation on Lys-171 in both MCF7 and RKO cells. We further demonstrated that a K171 acetylation mimetic did not perturb binding to select gene targets, but it diminished interaction of MeCP2 with binding partners such as ATRX and HDAC1 in colorectal cancer cells (44). In vivo and in vitro studies have demonstrated the importance of MeCP2 post-translational regulation (45,(104)(105)(106)(107), yet little has been done to comprehensively map novel MeCP2 PTMs. In the current study we wanted to extend our analyses and provide a comprehensive map of post-translational acetylation in other cancer cell line models. In order to further understand how MeCP2 is posttranslationally regulated in TNBC breast and prostate cancer cell lines, we systematically identified the specific lysines on endogenous MeCP2 where acetylation was induced upon lysine deacetylase inhibition (KDACi). We inhibited SIRT1, a class III lysine deacetylase, using 10 µM Inhibitor-IV or 10 µM Inhibitor-VII, as well as the class I/II/IV lysine deacetylases using 2 µM panobinostat and 10 µM pracinostat. Given the links between DNA methylation and/or aberrant expression of DNA methylation readers in prostate cancer (4,8) and TNBC (12,42), we focused on two model lines representing each cancer, PC3 and MDA-MB-468, respectively. Next, we performed immunoprecipitation of endogenous MeCP2 and analyzed the samples using LC-MS/MS. Figure 4A summarizes the specific lysine residues that were acetylated and exhibited ion peaks at mass/charge (m/z) ratio of~126 under basal (vehicle control) and KDI-induced conditions (i.e., cells treated with panobinostat, Inhibitor-IV, Inhibitor-VII, and pracinostat) (also see Figure S3). The mass spectrometry analyses showed that We found changes in acetylation patterns induced by exposure to both pre-clinical KDIs such as SIRT1 inhibitors and pracinostat as well as an FDA-approved inhibitor, panobinostat, which is used in the clinic to treat leukemias and lymphomas (23,108). Interestingly, some of the lysine residues detected as acetylation sites (K22 and K135) were also sites mutated in Rett Syndrome. Moreover, some of the lysine residues detected as acetylation sites (such as K135), have been previously reported as sites linked with ubiquitination (4). We found acetylated lysine residues across the length of the protein, including at the N-terminus, in the methyl-binding domain (MBD), in the intermediate domain (ID) and the transcriptional repression domain (TRD) as well at the C-terminus region ( Figure 4B). Together, these results indicate that MeCP2 is acetylated under basal and KDI-induced conditions in multiple cancer cell lines. Next, we wanted to determine the impact of K135 acetylation on MeCP2 subcellular localization. We chose to study this site since it is situated in a highly conserved MBD domain and is a  Figure S4B). Using immunofluorescence assays, we detected that HA-tagged wild-type MeCP2, deacetylation mutants (K135R), and acetylation mutants (K135Q) were mostly in the nucleus of stably expressing MDA-MB-468 cells ( Figure 5). These data demonstrate that post-translational acetylation on K135 lysine residue does not alter MeCP2 sub-cellular localization and calls for future studies to examine the role of acetylation at this residue as well as others identified in this report.

DISCUSSION
The present study provides valuable insight on two important fronts. First, we identify novel genes that are subject to MeCP2mediated regulation. Second, we provide a comprehensive identification of novel sites of post-translational acetylation associated with different cancer types and in response to multiple classes of deacetylase inhibitors. Concerning genomic analyses, these findings are important because we identify novel MeCP2 target genes linked with tumor progression which were not previously linked with MeCP2. While global DNA hypomethylation frequently occurs during tumorigenesis (60,61), the promoters of TSGs may undergo hypermethylation (109)(110)(111) and these aberrant changes in both the marks and the enzymes that modify them are being intensively examined for novel therapies (112)(113)(114)(115)(116). These epigenomic changes may instigate genomic instability or generate a heritable molecular signature, which enables tumor progression, so identification of novel genomic targets of MeCP2 is very important (117)(118)(119). Previous reports linked MeCP2 expression with ER status (3) and with BRCA1 promoter silencing (120), which provided further rationale for assessing genome-wide MeCP2 profiling in both MCF7 and MDA-MB-468 cells, which represent two subtypes of breast cancer. We found that MeCP2 binds to multiple regions of genes, including promoters, exons, and introns. These novel targets have been associated with a wide range of regulatory and signaling pathways. We found that there was an overlap of around 800 genes between the two cell lines, and there were distinct MeCP2 binding motif enrichments between both cell lines. We observed that not only did MeCP2 bind many novel gene targets, but its depletion also led to both increases and decreases in their corresponding RNA transcripts. This is especially important given that studies demonstrate MeCP2 binds to methylated cytosines and hydoxymethylated cytosines in mCH dinucleotides, a property wherein many unknowns remain (31-33, 35, 36, 38, 121). We discovered that MeCP2 localizes at various gene promoters involved in diverse processes such as autophagy (ATG4D), immune cell regulation (IL6), chromatin organization (KDM3B, KDM2A, KMT2B, KMT2A, KDM1A, HDAC1, HIST1H4F), circadian clock (SIRT1), developmental biology (EGFR, DKK1, SUMF2), extracellular matrix organization (ICAM5, ICAM3, ICAM1) and metabolism (EIF3G, SLC44A2, SUMF2, OXCT2, ACOT2, PSPH).
Recently, MeCP2 was shown to be amplified in human tumors and can mimic the function of activated Ras in cancer models (42), and also acts as a critical bridge linking information encoded in methylated DNA to epigenetic regulators (40,122). Although MeCP2 binds methyltransferases (25), co-repressors (123) histone deacetylases (26), chromatin modulators (41), and long noncoding RNAs (lncRNAs) (124,125) and other epigenetic regulators (4), much remains unknown about what regulates these interactions and what regulates binding to mCG vs. mCA dinucleotides as well as methylation-independent binding (126). However, our previous report provided some of the first insight into the role of post-translational regulation of MeCP2 binding to co-repressor proteins. We found that K171 acetylation regulates MeCP2 interaction with HDAC1 and ATRX (44). It is worth noting that the severity of Rett syndrome is influenced by the location of the mutation. For example, the site of the mutation can strongly impact cognitive and psychomotor skills as well as neonatal encephalopathy and death. Some mutations have been shown to disrupt conserved AT-hook regions and cause differential localization of a-Thalassemia/Mental Retardation Syndrome X-Linked (ATRX) (23). ATRX is a critical SWI/SNF-like chromatin remodeler (127) and our previous report demonstrates that a novel PTM on MeCP2 regulates MeCP2-ATRX binding, which is a critical aspect of MeCP2 function (43). Our present study identified an additional 17 novel sites of post-translational  (43). Based on previous findings one may reason that one or more of these novel PTMs may be influencing MeCP2 function in cancer progression. Another example of the impact of post-translational regulation comes from transgenic models involving single MeCP2 serine residues that undergo post-translational regulation which show distinct neurological defects (106,128), and phosphorylation of specific serine residues is enriched at specific gene promoters (104). However, much less is known about the role of MeCP2 acetylation as a regulatory switch in any context. Our more thorough mapping of novel MeCP2 acetylation PTMs performed here is a first step in defining their functional significance which is beyond the scope of the present study. Based on MeCP2 acetylation patterns induced by the various pre-clinical or clinical lysine deacetylase inhibitors, it is likely that KDACi's that target class I/II vs. class III HDACs will influence MeCP2 function in both common and distinct ways. Based on acetylation mapping one can also reason that MeCP2 interaction with different KDACs may lead to important role in cell-typespecific biology driven by unique acetylation patterns. We previously demonstrated that lysine acetylation serves as a regulatory switch in Wnt pathway signaling (129,130) and cancer-associated steroidogenesis (131,132). The current study provides yet another example of the scope of post-translational acetylation and may help explain how SIRT1 preferentially targets active (133,134) vs. repressed genes (135) depending on its deacetylation of specific non-histone partners (136,137). Future work may identify more factors involved in this SIRT1-MeCP2 regulatory network, and through such work, our understanding of the key molecular relationships in cancer may lead to deeper understanding of the mechanism of action of epigenetic therapies and KDAC inhibitors.

DATA AVAILABILITY STATEMENT
Sequences and processed ChIP-Seq and RNA-Seq data files were deposited in the NCBI Gene Expression Omnibus (GEO) database under accession number GSE160150 and the BioProject: PRJNA667107.

ACKNOWLEDGMENTS
Immunofluorescence images were generated in the Image Analysis Core Facility, supported in part by TTUHSC.

SUPPLEMENTARY MATERIAL
The