Abstract

   Aberrant methylation of the EphA7 promoter has been observed in
   cervical cancer (CC); however, its precise function and role in CC
   remain largely unknown. In this study, we investigated the role and
   molecular mechanisms of EphA7 promoter methylation in cervical
   carcinogenesis. First, our results indicated that the reactivation of
   EphA7 expression via a CRISPR demethylation tool (dCas9-Tet1) had
   antitumor effects. It restrained tumor proliferation and invasion while
   promoting apoptosis via the PI3K/AKT signaling pathway in both CaSki
   and SiHa cells. The upstream interacting factors were subsequently
   captured by CRISPR-mediated pull-down in situ, and the result revealed
   that SP1 and MAZ interacted with the promoter of EphA7. However, the
   perturbation results revealed that EphA7 expression was associated with
   SP1/DNMT1 but not MAZ. Furthermore, 17-β-estradiol (E2) can upregulate
   EphA7 expression through demethylation via the SP1/DNMT1 axis. A rescue
   experiment revealed that interference with SP1 expression could restore
   the effect of E2 on increasing the expression of EphA7 by upregulating
   estrogen receptor expression. In addition, EphA7 demethylation reduced
   the half-maximal inhibitory concentration (IC[50]) of cisplatin and
   paclitaxel. Pooled analysis revealed that EphA7 promoter
   hypermethylation was positively correlated with tumor purity but
   negatively correlated with immune cell infiltration, cytotoxic T
   lymphocyte (CTL) and immune checkpoint (IC) activity, and the
   expression of EphA7 was significantly positively correlated with tumor
   mutational burden (TMB), microsatellite instability (MSI) and the
   presence of single nucleotide variant (SNV) neoantigens, suggesting a
   better prognosis for patients with EphA7 promoter hypomethylation and
   high expression. Collectively, these findings indicate that targeted
   demethylation of the EphA7 promoter and restoration of endogenous EphA7
   expression by dCas9-Tet1 are promising therapeutic approaches and are
   favorable for the prognosis of CC patients.

   Subject terms: Targeted therapies, Prognostic markers, Molecular
   biology

Introduction

   Cervical cancer (CC) ranks fourth in the incidence and mortality of
   cancer in women, with an estimated 660,000 new cases and 350,000 deaths
   worldwide in 2022 [[32]1]. Recent statistics have shown that CC has
   moved up to become the third most common cause of cancer-related death
   among young women since 2019 in the United States [[33]2]. A
   significant upward trend in mortality rates for CC patients from
   2000–2018 in China was also observed [[34]3]. Therefore, CC is
   considered a great threat to female health.

   Despite significant advances in surgery, chemoradiotherapy and
   immunotherapies, drug resistance, recurrence, and metastasis of CC
   continue to significantly impact patient survival, with a 5-year
   survival rate of ~17.0% [[35]4]. Currently, since PI3K signaling is one
   of the most frequently aberrantly activated pathways in cancer, the use
   of PI3K inhibitors to treat CC patients has been tentative. However,
   the response to PI3K inhibitors varies widely among different patients,
   in particular, infectious and immune-mediated toxicities limit their
   use at this stage [[36]5]. Notably, further research into the molecular
   mechanisms of cervical carcinogenesis is essential for improving
   adjuvants and developing molecular-targeted compounds to accelerate
   precision medicine for CC.

   EphA7 (Eph receptor A7) is a hypermethylation biomarker identified in
   CC in our previous study [[37]6]. It belongs to the family of
   erythropoietin-producing hepatocellular carcinoma (Eph) receptors,
   which constitute the largest family of receptor tyrosine kinases (RTKs)
   [[38]7]. Studies have shown that EphA7 is considered an indirect
   upstream regulator of the PI3K/AKT signaling pathway [[39]8],
   contributing greatly to carcinogenesis, but its role is altered in
   different cancers [[40]9].

   In addition, since DNA methylation not only regulates gene expression
   but is also generally reversible, it holds enormous potential for
   directing clinical care [[41]10]. Furthermore, the development of the
   clustered regularly interspaced short palindromic repeats (CRISPR)
   system of sequence-targeted technologies provides a bright future for
   manipulating DNA methylation and the interaction of transcription
   factors (TFs) [[42]11], which provides additional insights into precise
   therapeutic approaches for CC. A previous study revealed that DNA
   methyltransferase 1 (DNMT1) is recruited by SP1 [[43]12], and SP1 has
   been reported to be involved in cervical carcinogenesis [[44]13,
   [45]14], which suggests that SP1 is a crucial TF that contributes to
   EphA7 methylation in CC.

   Therefore, this study aimed to systematically investigate the function
   and mechanism of the epigenetic modulation of EphA7 in CC via CRISPR
   techniques, and we revealed that targeted demethylation of the EphA7
   promoter inhibited CC progression via the PI3K/AKT pathway. In
   addition, we explored whether EphA7 promoter demethylation serves as an
   emerging prognostic and therapeutic target by examining its
   associations with estrogen, chemosensitivity, and immune checkpoints
   (ICs). Our findings provide new insights into biomarkers that integrate
   diagnosis with treatment for the targeted therapy of CC.

Materials and methods

Data collection and pooled analysis of EphA7

   EphA7 promoter methylation and expression data were extracted from The
   Cancer Genome Atlas (TCGA; [46]https://cancergenome.nih.gov/) and Gene
   Expression Omnibus (GEO; [47]http://www.ncbi.nlm.nih.gov/geo/). The CC
   specimens collected from TCGA-CESC (Cervical squamous cell carcinoma
   and endocervical adenocarcinoma) cohort (n = 306) were included in the
   study, and according to the median of EphA7 promoter methylation level,
   the hypermethylation and hypomethylation groups were separated. GEO2R
   was used to identify differentially expressed genes (DEGs) related to
   EphA7 in [48]GSE9750. Genes with adjusted P < 0.05 and |log2 FC | >2
   were considered DEGs. The DAVID bioinformatics tool enables Kyoto
   Encyclopedia of Genes and Genomes (KEGG) enrichment analyses to predict
   the function of EphA7. TCGA plot was used to analyze the associations
   between EphA7 expression and the tumor mutational burden (TMB),
   microsatellite instability (MSI), and single-nucleotide variant (SNV)
   neoantigens in TCGA cancers.

Cell lines and culture

   The CC cell lines epidermoid carcinoma CaSki (RRID: CVCL_1100) and
   squamous cell carcinoma SiHa (RRID: CVCL_0032; Shanghai
   Zhongqiaoxinzhou Biotech, China), human embryonic kidney cells
   (HEK293T: RRID: CVCL_0063; Genetic Testing Biotechnology Corporation,
   Suzhou, China) and the normal cervical immortalized epithelial cell
   line H8 (BeiNa Culture Collection, China) were cultured in the
   appropriate media (DMEM, MEM or RPMI 1640). CaSki and SiHa cell lines
   were selected for this study because they are hypermethylation of EphA7
   promoter and covered the important histological subtypes of CC. All
   cell lines were authenticated using short tandem repeat profiling
   within the last three years. All experiments were performed with
   mycoplasma-free cells. The cells were maintained in an environment
   supplemented with 10% FBS (BI, USA) and 1% antibiotics
   (penicillin/streptomycin) (Gibco, USA) at 37 °C with 5% CO[2]. For
   experimental treatments, cells were seeded at 70% confluence and
   exposed to 0.2 or 2 ng/ml 17-β estradiol (E2) (Solarbio) diluted in
   dimethyl sulfoxide (DMSO) for 48 h.

Plasmid design and construction

   As described in our previous study [[49]6], single guide RNAs (sgRNAs)
   for targeting EphA7 promoter demethylation (referred to as
   “act-sgRNAs”) were designed via CRISPR-ERA (crispr-era.stanford.edu).
   In accordance with our previous optimization results [[50]6], the
   combination of act-sgRNA1 + 3 was used to activate EphA7 in CaSki
   cells, and act-sgRNA1 + 2 was used in SiHa cells (Supplementary Table
   [51]1). Moreover, the top 5 potential off-target loci predicted by the
   COSMID web tool ([52]https://crispr.bme.gatech.edu/) were selected. The
   primers used are listed in Supplementary Table [53]2. SgRNAs for
   CRISPR-based pull-down in situ (Cap-sgRNAs) were also designed to
   minimize off-target cleavage using the public tool CRISPR-ERA
   (Supplementary Table [54]1). The SP1 knockout sgRNA (KO-sgRNA) was
   designed as the reference described [[55]15]. The annealed oligos were
   inserted into a modified pgRNA plasmid to create the sgRNA expression
   plasmids (Addgene #44248) with an AarI (Thermo, USA) site. For the
   knockdown of MAZ, target shRNA sequences (GATGCTGAGCTCGGCTTATAT) were
   subcloned and inserted into pLKO.1-shRNA lentivectors with EcoRI and
   AgeI (NEB, USA). The recombinant SP1 ([56]NM_138473.3) overexpression
   plasmid was constructed by inserting the full-length cDNA into the
   pLVX-mCherry vector with XhoI and EcoRI (NEB, USA). The sequences of
   the sgRNAs and primers used for plasmid construction are detailed in
   Supplementary Table [57]1.

Lentivirus production and stable cell line generation

   Lentiviruses expressing Fuw-dCas9-Tet1CD (Addgene #84475),
   lenti-Cas9-Blast (Addgene #52962), sgRNAs, pLVX-SP1, and shMAZ were
   produced by transfecting HEK293T cells with those plasmids along with
   standard packaging vectors (psPAX2 and pMD2.G). The filtered
   Fuw-dCas9-Tet1CD lentivirus was concentrated via ultracentrifugation
   (System Biosciences, USA), as previously described [[58]16].

   Stable cell lines with CRISPR-captured FB-dCas9 and BirA were generated
   by transfecting pEF1a-FB-dCas9-puro (Addgene #100548) and
   pEF1a-BirA-V5-neo (Addgene #100547) into cells using polyethylenimine
   transfection reagent (Serochem). Stably transfected cells were treated
   with drug selection for 2 weeks or with a BD FACS Aria cell sorter on
   the basis of the selectable markers of the plasmids. The verification
   results for these stable cell lines are provided in Fig. [59]S1A–C.

Western blotting analysis

   Western blotting analysis was performed as previously described
   [[60]17]. HRP-labeled secondary antibodies were applied for 1 h at room
   temperature. The signals were detected using an enhanced
   chemiluminescence kit (Millipore, USA). GAPDH (Bioss) served as a
   loading control. The detailed information for the antibodies used for
   Western blotting is provided in Supplementary Table [61]3.

Quantitative real-time PCR (qRT‒PCR)

   The harvested cells were dissolved in TRIzol reagent (Invitrogen, USA),
   and the total mRNA was extracted according to the manufacturer’s
   protocol. qRT‒PCR assays were performed with SYBR Green (Vazyme) and a
   Stratagene Mx3005p instrument (Agilent, USA). The amplification
   parameters were as follows: 95 °C for 30 s, followed by 40 cycles of
   95 °C for 10 s, and 60 °C for 30 s. The data were normalized to GAPDH
   expression, and the relative expression of target genes was calculated
   via the 2^−ΔΔCt method. The primers used for qRT‒PCR are available in
   Supplementary Table [62]2.

Methylation-specific PCR (MSP)

   The primers were designed via Methyl Primer Express v1.0 (Applied
   Biosystems, USA), and the primer sequences are listed in Supplementary
   Table [63]2. MSP was performed to assess the promoter methylation of
   EphA7, as previously described [[64]6].

Cell proliferation assay

   For the detection of cell growth, 5,000 cells were seeded into each
   well of a 96-well plate and cultured for 0, 24, 48, or 72 h. Afterward,
   50 μl of 1× MTT solution (KeyGEN BioTECH) was added and incubated for
   4 h. Then, 150 μl of DMSO was added to the wells to dissolve the
   contents, and the optical density was determined by measuring the
   absorbance (490 nm) at the indicated time points.

Transwell migration assays

   For Transwell migration assays, 10,000 cells were suspended in 200 μl
   of serum-free medium and seeded into the upper chamber of each insert.
   Subsequently, 600 μl of culture medium containing 15% FBS was added to
   the lower chamber of a 24-well plate. After incubation at 37 °C, the
   cells that migrated were fixed, stained with 4% paraformaldehyde and
   0.1% crystal violet, and rinsed with PBS. The cells attached to the
   chamber were observed under a microscope. The number of cells invading
   the Matrigel was considered to reflect the invasive capacity of the
   cells.

Colony formation assays

   For the clonogenicity analysis, 500 cells were seeded in six-well
   plates and cultured for 15 days. Afterward, the colonies were fixed
   with 4% paraformaldehyde for 30 min and stained with 0.05% crystal
   violet for 20 min.

Measurement of E2 concentration

   The supernatant was collected from SiHa-Tet1 and SiHa-Tet1 cells with
   sgRNAs, and the concentrations of E2 were measured via the Abbott
   Architect i2000SR system using chemiluminescence assays.

Cytotoxicity assay

   The cytotoxicity of cisplatin was assessed via MTT assays, as described
   previously [[65]17]. CaSki and SiHa cells were treated with increasing
   concentrations of cisplatin or paclitaxel for 48 h. The half-maximal
   inhibitory concentration (IC[50]) was obtained from the optimized
   standard curve of the percentage of viable cells plotted against the
   extract concentration on the ordinate.

In situ CRISPR-based pull-down assay

   The in situ CRISPR-based pull-down assay for the EphA7 promoter was
   performed as previously described [[66]11]. Affinity-purified chromatin
   was treated with RNaseA (Tiangen) and Protease K (G-Clone) and analyzed
   by qRT–PCR (Supplementary Table [67]2) to evaluate the sensitivity and
   specificity of cap-sgRNAs. TF detection was conducted after chemical
   cross-linking by formaldehyde and chromatin sonication. The proteins
   were separated via centrifugation and analyzed by western blotting.

Statistical analysis

   All image collection and data analyses were performed as double-blind
   experiments. Statistical analysis was performed using GraphPad Prism
   8.0 (GraphPad Software, USA). Student’s t test was used to evaluate the
   significance of differences between two groups conforming to a normal
   distribution, whereas a Mann–Whitney U test was used for data not
   conforming to a normal distribution. One-way ANOVA was employed for
   multiple groups comparison. The associations between tumor purity,
   immune cell infiltration, chemosensitivity, immune-related genes, and
   methylation or expression levels were investigated by calculating
   Spearman’s rank correlation coefficients. The Wilcoxon test was
   performed to analyze the differences in expression between the complete
   response group and the group that did not achieve a complete response.
   Overall survival was compared between groups using Kaplan‒Meier curves
   and log-rank tests. At least three independent replicates were assessed
   for each of the in vitro experiments, and the pooled data are presented
   as the means ± standard errors. A P value < 0.05 was considered to
   indicate statistical significance (*P < 0.05, **P < 0.01, ***P < 0.001
   and ns, not significant).

Results

Demethylation of the EphA7 promoter by the dCas9-Tet1 system inhibits CC
progression and induces apoptosis

   First, the methylation level of the EphA7 promoter was efficiently
   decreased by the dCas9-Tet1 tool (Fig. [68]S2A) in our previous study
   [[69]6]. The expression of EphA7 was restored (P < 0.05) at both the
   mRNA (Fig. [70]S2B) and protein (Fig. [71]1A) levels in CaSki and SiHa
   cells. In addition, consistent with findings from other reports, the
   off-target effect of dCas9-Tet1 on DNA methylation was minimal
   [[72]18]. Figure [73]S2C, D shows no significant variation in the mRNA
   expression levels of the top five potential off-target genes predicted
   by the COSMID web tool in the demethylation groups of both CaSki and
   SiHa cells.

Fig. 1. EphA7 demethylation via the dCas9-Tet1 system inhibited CC
progression and induced apoptosis.

   [74]Fig. 1
   [75]Open in a new tab

   A Western blotting was performed and revealed that EphA7 expression was
   upregulated in the dCas9-Tet1 with act-sgRNAs group (demethylation)
   compared with the dCas9-Tet1 group (control) in both CaSki and SiHa
   cells (n = 3). B Compared with the control, targeted demethylation of
   EphA7 effectively inhibited the proliferation of CaSki and SiHa cells,
   as determined via the MTT assay (P < 0.05) (n = 6). C Transwell assays
   revealed that EphA7 demethylation (n = 6) strongly reduced the number
   of migrated CC cells compared with that in the control group
   (P < 0.05). D The epithelial marker CDH1 (E-cadherin) was upregulated
   due to the demethylation of EphA7 (n = 3), whereas the mesenchymal
   markers CDH2 (N-cadherin) and Snail1 were downregulated in CaSki and
   SiHa cells. E Western blotting was used to verify that the
   demethylation of EphA7 increased the expression of Bax, downregulated
   the expression of Bcl-2 and promoted caspase-3 activity (P < 0.05) in
   CaSki and SiHa cells (n = 3). Error bars represent the means ± SDs, P
   values were calculated using two-tailed unpaired Student’s t tests
   (A–E). *P < 0.05, **P < 0.01, ***P < 0.001.

   The enriched DEGs between the EphA7 high- and low-methylation groups
   were subsequently extracted from TCGA-CESCs, and the results indicated
   that EphA7 promoter methylation disrupted biological functions related
   to proliferation, metastasis, and apoptosis (Fig. [76]S3A, B).

   Our MTT (Fig. [77]1B) and colony formation (Fig. [78]S3C) assays
   revealed that the demethylation of EphA7 profoundly blocked the
   proliferative activities of CaSki and SiHa cells, indicating that the
   demethylation of the EphA7 promoter inhibited the proliferation of CC
   cells. Fewer migrated CC cells were observed in the EphA7 promoter
   demethylation group than in the control groups (Fig. [79]1C).

   The epithelial–mesenchymal transition (EMT) is a key step in tumor
   invasion and metastasis [[80]19], and we further detected the
   expression levels of EMT-related genes in CC cells. As shown in Fig.
   [81]1D, the epithelial marker CDH1 was upregulated, whereas the
   mesenchymal markers CDH2 and Snail were downregulated in the
   EphA7-demethylation groups of both CaSki and SiHa cells. Overall, these
   findings imply that EphA7 demethylation might serve as an inhibitor of
   CC progression.

   Subsequently, we found that the demethylation of EphA7 promoted
   cleaved-caspase-3 activity (P < 0.05), increased the expression of Bax,
   while downregulated the expression of Bcl-2 (P < 0.05) (Figs. [82]1E
   and [83]S3D). The Bax/Bcl-2 ratio in CC cells was obviously increased
   (P < 0.05) in the EphA7-demethylation group (Fig. [84]1E).

Demethylation of EphA7 inhibits CC progression via the PI3K/AKT pathway

   We extracted RNA-seq data from [85]GSE9750 and subsequently performed a
   KEGG pathway enrichment analysis on the DEGs (Fig. [86]S4A). The
   results revealed that EphA7 significantly altered the PI3K/AKT
   signaling pathway (Fig. [87]2A, B).

Fig. 2. EphA7 demethylation inhibited CC progression via the PI3K/AKT
signaling pathway.

   [88]Fig. 2
   [89]Open in a new tab

   A These DEGs were obtained by comparing normal cervical samples with
   high EphA7 expression (n = 5) to CC samples with low EphA7 expression
   (n = 5) on the basis of EphA7 RNA-seq data ([90]GSE9750). KEGG pathway
   enrichment analysis revealed that EphA7 significantly altered the
   PI3K/AKT signaling pathway. B Signaling pathways regulated by EphA7
   according to the DAVID web tool. C The levels of the phosphorylated
   forms of PI3K and AKT were reduced, and the pPI3K/PI3K and pAKT/AKT
   ratios were decreased (P < 0.05) in the EphA7 demethylation group
   (n = 3). D The expression of EFNA5 and EphA2 was increased in the EphA7
   demethylation group (n = 3). Error bars represent the means ± SDs, P
   values were calculated using two-tailed unpaired Student’s t tests (C,
   D). *P < 0.05,  **P < 0.01, ***P < 0.001 and ns, not significant.

   As depicted in Fig. [91]2C, the pPI3K/PI3K and pAKT/AKT ratios were
   decreased in the EphA7 demethylation group (P < 0.05). Furthermore, the
   EphA7 protein was found to directly bind to EFNA5 and EphA2 according
   to the STRING analysis (Fig. [92]S4B). In addition, single-cell RNA
   sequencing revealed a positive association between EphA7 expression and
   EFNA5/EphA2 expression in human cells using Cellxgene tools (Fig.
   [93]S4C). Consistent with these findings, qRT‒PCR confirmed that the
   expression of EFNA5 and EphA2 was upregulated in the EphA7
   demethylation group (P < 0.05) (Fig. [94]2D).

SP1 regulates the methylation of EphA7 by interfering with DNMT1

   Eight predicted binding sites were identified in the EphA7 promoter
   region through an overlapping TF database, including JASPAR, hTFtarget,
   and ConTra v3 (Fig. [95]3A). SP1 and MAZ were found to directly
   interact with EphA7 via the Pathway Commons integrated tool (Fig.
   [96]3A). The predicted binding sites are listed in Supplementary Table
   [97]4, and the sites of capture-sgRNAs are shown in Fig. [98]S5A.
   Although the EphA7 promoter was hypermethylated in SiHa and CaSki cells
   (Fig. [99]S5B), as shown in Figs. [100]3B and [101]S5C, total SP1 and
   MAZ were significantly more highly expressed in SiHa cells than in
   CaSki cells. Furthermore, consistent with the experimental results, we
   also found that SP1 was indeed expressed at lower levels in
   adenocarcinoma (ADC) (Fig. [102]S5D) than in squamous cell carcinoma
   (SCC) from the TCGA. Next, in situ CRISPR pull-down techniques were
   used to elucidate the upstream regulatory mechanism of EphA7. The
   results verified that Cap-sgRNA1-4 effectively enriched fragments of
   the EphA7 promoter in SiHa cells (P < 0.05) (Fig. [103]S5E) and
   confirmed that SP1 and MAZ were indeed located in the EphA7 promoter
   region in SiHa cells (Fig. [104]3C). As controls of in situ pull-down
   technics, we performed parallel experiments in H8 cells of and the
   results showed that SP1 and MAZ were enriched as well (Fig. [105]S5E,
   [106]F).

Fig. 3. SP1 regulated the methylation of EphA7 by interfering with DNMT1.

   [107]Fig. 3
   [108]Open in a new tab

   A The JASPAR ([109]http://jaspar.genereg.net/), CONTA V3
   ([110]http://bioit2.irc.ugent. be/contra/v3), and hTFtatget
   ([111]http://bioinfo.life.hust.edu.cn/ hTFtarget) tools were integrated
   to predict transcription factor binding in the promoter of EphA7. SP1
   and MAZ were found to directly interact with EphA7 on the basis of the
   Pathway Commons ([112]https://www.pathwaycommons.org). B The mRNA and
   protein expression analysis showed that total SP1 and MAZ were
   significantly expressed in SiHa cells. C The in situ pull-down results
   verified that SP1 and MAZ were located in the promoter region of EphA7
   in SiHa cells. D Pathway commons tools indicating that SP1 and MAZ
   directly interact with EphA7 and DNMT1. Promoter methylation of EphA7
   was associated with DNMT1 (P < 0.05) in TCGA-CESCs (n = 306). E Western
   blotting was performed to demonstrate that SP1 was silenced in SiHa-SP1
   knockout cells (SiHa-SP1-KO) and that the protein expression levels of
   EphA7 were increased in SP1-KO cells (n = 3). F The mRNA expression
   levels of EphA7 were increased, and the expression of DNMT1 was
   decreased in SP1-KO cells (n = 3). G Knockout of SP1 markedly reduced
   the methylation level of EphA7 in SiHa cells. H, I The results of
   rescue assays showed that SP1 overexpression reversed the changes in
   the mRNA (P < 0.05) and protein expression of EphA7 in the SP1-KO
   group. Error bars represent the means ± SDs, P values were calculated
   using two-tailed unpaired Student’s t tests (E–G) and Spearman’s
   analysis (D). *P < 0.05, **P < 0.01,  ***P < 0.001.

   By applying the Pathway Commons, SP1 and MAZ were discovered to
   directly interact with EphA7 and DNMT1 (Fig. [113]3D). In addition,
   EphA7 promoter methylation was positively associated with DNMT1
   expression in TCGA-CESCs (Fig. [114]3D). Knockout of SP1 significantly
   increased the expression of EphA7 (Fig. [115]3E) and downregulated the
   expression of DNMT1 (P < 0.05) (Fig. [116]3F). We also found that the
   level of EphA7 methylation in SiHa cells was reduced in the SP1
   knockout group (Fig. [117]3G). However, MAZ knockout did not affect the
   expression of DNMT1(Fig. [118]S5G).

   Notably, both the protein and mRNA expression levels of EphA7 increased
   in SiHa-SP1-KO cells (Fig. [119]3E, F). Conversely, the overexpression
   of SP1 decreased EphA7 expression while upregulating DNMT1 (Fig.
   [120]S5H) in SiHa cells.

   We performed a rescue experiment using an SP1 overexpression plasmid in
   SiHa-SP1-KO cells to further confirm the direct association between SP1
   and EphA7. SP1 overexpression reversed the changes in EphA7 mRNA and
   protein expression in the SP1-KO group (Fig. [121]3H, I), indicating
   that EphA7 expression is directly regulated by SP1 in SiHa cells.

SP1 interacts with the ER to regulate EphA7 methylation

   As depicted in Fig. [122]2B, KEGG pathway enrichment analysis revealed
   that EphA7 obviously altered the estrogen signaling pathway (P < 0.05).
   According to Pathway Commons, SP1 interacted with ESR1 and ESR2 (Fig.
   [123]4A). Furthermore, the Integrative Interactome Database revealed
   that SP1 interacts with ESR1 in human uterine tissues (Fig. [124]4B).

Fig. 4. SP1 interacts with the ER to regulate EphA7 methylation.

   [125]Fig. 4
   [126]Open in a new tab

   A The Pathway Commons web tool demonstrated that SP1 interacted with
   ESR1 and ESR2. B According to Integrative Interactome Database
   ([127]http://ophid.utoronto.ca/iid), SP1 was found to interact with
   ESR1 in human uterine tissues. C The protein level of SP1 was obviously
   reduced (P < 0.05), and the protein level of EphA7 gradually increased
   in a dose-dependent manner with increasing concentration (P < 0.05)
   (n = 3). D The methylation level of EphA7 was obviously reduced in
   E2-treated SiHa cells. E The increase in EphA7 levels due to E2
   treatment was partly reversed by SP1 overexpression, as determined by a
   rescue assay (n = 3). Error bars represent the means ± SDs, P values
   were calculated using two-tailed unpaired Student’s t tests (E), and
   one-way ANOVA with Tukey’s multiple comparisons test (C). *P < 0.05,
   **P < 0.01, ***P < 0.001.

   Following the treatment of SiHa cells with E2, the mRNA expression of
   ESR1 and ESR2 generally increased (Fig. [128]S6). Following the
   treatment of SiHa cells with E2, the expression of SP1 (Fig. [129]4C)
   and the level of EphA7 methylation were obviously reduced (P < 0.05)
   (Fig. [130]4D). Furthermore, both the mRNA and protein levels of EphA7
   gradually increased in a dose-dependent manner after treatment with
   increasing E2 concentrations (Fig. [131]4C, E). In addition, as shown
   in Fig. [132]4E, the increased levels of EphA7 resulting from E2
   treatment were partially reversed by SP1 overexpression. Moreover, we
   detected that much less E2 was produced by SiHa-Tet1 or SiHa-Tet1 cells
   with sgRNAs, which demonstrates that the regulation of EphA7 is not
   affected by endogenous E2(Supplementary Table [133]5).

Demethylation of EphA7 is associated with a favorable response to
chemotherapy in CC

   Figure [134]5A shows greater EphA7 expression in the complete response
   (CR) group than in the noncomplete response (NCR) group (P < 0.05) in
   the [135]GSE56363 dataset. In addition, the association between EphA7
   expression and sensitivity to therapeutic drugs revealed a negative
   correlation with IC[50] values of cisplatin (P < 0.05) (Fig. [136]5B).

Fig. 5. The demethylation of EphA7 is associated with a favorable response to
chemotherapy in CC patients.

   [137]Fig. 5
   [138]Open in a new tab

   A The results showed that the expression of EphA7 was greater in the CR
   group than in the NCR group (P < 0.05) through the BEST web tool in the
   [139]GSE56363 dataset. B Analysis of the association between EphA7
   expression and sensitivity to chemotherapeutic drugs via Guangreshengwu
   ([140]https://grswsci.top/analyze). C EphA7 demethylation appreciably
   reduced the half-maximal IC[50] of cisplatin in CaSki and SiHa cells
   compared with that in the control group (n = 6). D The IC[50] of
   paclitaxel was lower in the experimental group than in the
   demethylation group (n = 6). E CRISPR-targeted demethylation of EphA7
   worked synergistically with cisplatin/paclitaxel, significantly
   increasing cisplatin/paclitaxel susceptibility. Error bars represent
   the mean ± SD, P values were calculated using two-tailed unpaired
   Student’s t tests (E). *P < 0.05. *P < 0.05, **P < 0.01, ***P < 0.001.

   Consistent with these results, our findings shown in Fig. [141]5C, D
   indicate that EphA7 demethylation significantly reduced the IC[50]
   values of cisplatin and paclitaxel in CaSki and SiHa cells compared
   with those in the control group. As shown in Fig. [142]5E,
   CRISPR-targeted demethylation of EphA7 worked synergistically with
   cisplatin/paclitaxel, significantly increasing cisplatin/paclitaxel
   susceptibility (P < 0.05). This finding suggested that EphA7
   demethylation effectively improves the chemosensitivity of CC cells and
   that the demethylation of EphA7 appears to be associated with a
   favorable response of CC to chemotherapy.

EphA7 methylation is a potential immunotherapy biomarker involving the tumor
microenvironment (TME) in CC

   Tumor purity is one way to determine the efficacy of immunotherapy, and
   CC patients with low tumor purity tend to have a better prognosis
   [[143]20]. As shown in Fig. [144]6A, the level of EphA7
   hypermethylation was positively correlated with the tumor purity of
   CESCs (cor = 0.115, P < 0.05).

Fig. 6. EphA7 is a potential immune therapy biomarker involving the TME in
CC.

   [145]Fig. 6
   [146]Open in a new tab

   A Tumor purity was positively correlated with the extent of EphA7
   methylation (cor = 0.115, P < 0.05) in CESCs according to the Linkomics
   tool ([147]http://www.linkedomics.org/login.php) in CESCs (n = 306). B
   The results showed that the infiltration score of CESCs (n = 306)
   negatively correlated with EphA7 methylation (cor = −0.16, P < 0.05)
   via GSCA ([148]https://guolab.wchscu.cn/ GSCA/#/). C Spearman
   correlation analyses showing the association between the methylation
   levels of EphA7 and immune cell infiltration via GSCA. D, E Promoter
   methylation of EphA7 was negatively correlated with CTLs
   (cor = −0.214), and the population with low EphA7 methylation had a
   better prognosis in the CTL top group on the basis of the TIDE
   algorithm ([149]http://tide.dfci.harvard.edu/). F Comparative analyses
   of IC expression profiles across EphA7 hypo/hypermethylation cohorts in
   TCGA-CESCs (n = 306). The upper triangle represents P value. The lower
   triangle in the heatmap represents the correlation coefficient. Red
   indicates a positive correlation, and blue indicates a negative
   correlation. G Patients with anti-PD-L1 antibodies had a better
   prognosis in the group with high EphA7 expression (P < 0.05) by the web
   tools BEST ([150]https://rookieutopia.com/app_direct/BEST/). H EphA7
   expression was significantly positively associated with SNV
   neoantigens, MSI status, and TMB (P < 0.05). The abbreviations and full
   names are listed in Supplementary Table [151]6. P values were
   calculated using two-tailed unpaired Student’s t tests (F) and
   Spearman’s analysis (A, B, D, H). *P < 0.05,**P < 0.01, ***P < 0.001.

   The immune infiltration score can predict the prognosis and efficacy of
   immunotherapy for multiple cancers [[152]21–[153]23]. GSCA revealed
   that EphA7 hypermethylation was negatively correlated with the
   infiltration score (cor = -0.16, P < 0.05) and infiltration of 8 immune
   cell types, including type 1 regulatory T (Tr1) cells, CD4 + T cells,
   central memory cells, cytotoxic cells, NK cells, and CD8 + T cells
   (Fig. [154]6B, D). Conversely, a positive association between EphA7
   hypermethylation and the infiltration of effector memory and neutrophil
   cells was observed (P < 0.05) (Fig. [155]6C). Notably, Kaplan‒Meier
   curves showed that high cytotoxic T lymphocyte (CTL) levels were
   associated with prolonged survival in the EphA7 hypomethylation group
   (Fig. [156]6E). The results revealed that patients with CC in the EphA7
   hypomethylation group had a favorable response to immunotherapy.

   Studies have shown that patients with high expression of ICs are more
   likely to respond to immunotherapy [[157]24]. We found that EphA7
   promoter methylation was negatively correlated with the expression of
   IC genes, and the EphA7 hypomethylation group presented higher mRNA
   expression levels of IC genes, including CTLA4, HAVCR2, LAG3, PDCD1,
   PDCD1LG2, TIGT, BTLA, SIRPA, and CD200R1 (P < 0.05) (Fig. [158]6F).
   Moreover, EphA7 promoter methylation was negatively correlated with the
   expression of chemokines, chemokine receptors, and major
   histocompatibility complex (MHC) genes (Fig. [159]S7), indicating that
   the EphA7 hypomethylation group may be more sensitive to immune
   checkpoint blockade (ICB) therapy.

   Finally, we observed that patients in the high EphA7 expression group
   who received anti-PD-L1 treatment had a better prognosis (P < 0.05)
   (Fig. [160]6G). EphA7 expression was positively correlated with TMB,
   MSI, and the presence of SNV neoantigens in CESC patients (P < 0.05)
   according to the TCGAplot and Guangreshengwu web tool (Fig. [161]6H).
   These findings may significantly contribute to the development of
   targeted immunotherapy and improved treatment strategies for CC.

Discussion

   In this study, we first demonstrated that EphA7 functions as a tumor
   suppressor in CC via the PI3K/AKT signaling pathway by activating
   endogenous EphA7 expression via demethylation. We also revealed that
   the demethylation of EphA7 is associated with a favorable response to
   antitumor therapy, which indicates that EphA7 methylation may be an
   emerging prognostic marker and therapeutic target.

   Previous research revealed that DNA methylation occurs at different
   steps of tumorigenesis, and given that epigenetic dysregulation is
   reversible, epigenetic modifications have been confirmed as potential
   targets for cancer therapy [[162]25]. Owing to the nonspecific nature
   and wide range of side effects of epi-drugs [[163]26], the CRISPR
   system for sequence-targeted demethylation has a broad future for
   manipulating DNA methylation and regulating endogenous expression in a
   targeted manner [[164]27], as evidenced by EphA7 demethylation in our
   previous and current studies.

   Numerous studies have shown that the PI3K/AKT pathway is abnormally
   activated in CC [[165]28–[166]30]. As an upstream factor, EphA7
   interferes with PTEN/PI3K/AKT signaling to regulate cell apoptosis in
   prostate tumors [[167]31] and laryngeal squamous cell carcinoma
   [[168]32]; however, little information on the role of EphA7 in CC has
   been obtained to date. Therefore, both the KEGG pathway enrichment
   analysis and the experimental findings confirmed that the demethylation
   of EphA7 in CC impairs the PI3K/AKT signaling pathway (Fig. [169]2C).

   In addition, Dr Xu and colleagues have harnessed CRISPR-dCas9
   technology by coupling it with FB/BirA to construct an in situ
   pull-down tool [[170]18], which allows us to further explore the
   interacting factors upstream of the target gene. Previous studies have
   shown that SP1 positively regulates the transcription of DNMT1,
   increasing the risk of hypermethylation of gene promoters [[171]33]. By
   combining bioinformatics information and the CRISPR-based in situ
   pull-down tool, we also successfully identified SP1 as a crucial factor
   that regulates the methylation of EphA7 by interfering with DNMT1 (Fig.
   [172]3G).

   Estrogen, a key hormone, plays a multifaceted role in CC. It has been
   found to decrease mitochondrial membrane permeability, promote the
   Warburg effect, and increase antiapoptotic protein levels, which favor
   the metabolic adaptation of CC cell lines and cervical squamous
   carcinoma survival [[173]34, [174]35]. Riera-Leal A et al. reported
   that cotreatment with E2 enhances chemotherapeutic agent sensitivity in
   SiHa cells (HPV positive) but reduces apoptosis in C33A cells (HPV
   negative) [[175]36]. Interestingly, Young MJ et al. further revealed
   that SP1 could interact with the estrogen receptor ER to coregulate
   gene expression [[176]37]. Our study discovered that despite the
   regulation of EphA7, substantially less E2 is secreted from SiHa cells,
   but exogenous E2 impairs SP1 function, leading to reduced EphA7
   methylation and the reversal of EphA7 expression. The present study
   provides strong evidence for the efficacy of estrogen replacement
   therapy in patients with cervical squamous cell carcinoma.

   Cisplatin monotherapy and combination regimens are standard
   chemotherapeutic approaches for treating CC [[177]38]. Unfortunately,
   drug resistance compromises the efficacy of chemotherapy. Therefore,
   researchers are actively exploring strategies to increase cancer cell
   sensitivity to drugs and improve the survival rate of CC patients. A
   previous study tested therapies combining cisplatin with other cancer
   drugs in cancer cell lines, but few anticancer drugs have been found to
   exert synergistic effects [[178]39]. Sun Hee Lee et al. discovered that
   DNA methylation inhibitors such as 5-aza-2-deoxycytidine can work
   synergistically with cisplatin to increase the efficacy of
   cisplatin-based chemotherapy [[179]40]. Consistent with our results,
   CRISPR-targeted demethylation of EphA7 works synergistically with
   cisplatin/paclitaxel, significantly increasing cisplatin/paclitaxel
   susceptibility (Fig. [180]5E).

   Emerging data indicated that tumor DNA methylation profiling may serve
   as a prognostic or predictive biomarker of the response to targeted
   therapy and immunotherapy [[181]41]. In addition, DNA methylation of
   both tumor cells and immune cells affects the TME and has been shown to
   play critical roles in influencing immune cell function and tumor
   immune evasion [[182]42]. As a major component of the TME, immune cell
   infiltration and tumor purity have been proven to contribute to tumor
   progression and the immunotherapy response [[183]43]. CTLs constitute a
   critical subset involved in adaptive immune responses [[184]44], and
   immunotherapy based on CTL-mediated tumor recognition and elimination
   has shown remarkable anticancer efficacy [[185]45]. Gray SM et al.
   reported that DNA methylation is associated with the acquisition of
   differential gene expression profiles unique to CTLs and that DNA
   methylation plays an important role in CTL differentiation and function
   [[186]46]. Our results indicated that EphA7 methylation was negatively
   related to the infiltration score and CTL population, highlighting its
   prognostic role in CC (Fig. [187]6D).

   As significant strides in immunotherapy, ICBs for recurrent or
   metastatic cancer have been reported [[188]47]. However, high
   expression of IC genes is widely adopted as a predictor of the
   immunotherapeutic response rate [[189]24]. Correlations between DNA
   methylation status and the expression of IC molecules and their
   prognostic significance have been demonstrated in multiple cancers
   [[190]48, [191]49]. We also found that EphA7 promoter methylation was
   negatively correlated with the expression of IC genes (Fig. [192]6F)
   and that patients who received anti-PD-L1 treatment and presented high
   EphA7 expression had a better prognosis (Fig. [193]6G). These results
   indicate that patients with EphA7 demethylation are more likely to
   benefit from immunotherapy. These findings provide valuable information
   for realizing more precise and personalized immunotherapy in the
   future. In addition, the TMB and MSI status serve as promising
   pan-cancer predictive biomarkers, guiding immunotherapy in the era of
   precision medicine. A higher TMB leads to more neoantigens, increasing
   the likelihood of T-cell recognition, and correlates with better ICBs
   outcomes [[194]50]. In CC, high-frequency MSI independently predicts
   clinical characteristics and prognosis [[195]51]. Therefore, these
   insights may lead to the proposal of new targets for the development of
   immunosuppressants for CC treatment.

   In conclusion, our results comprehensively revealed that EphA7 may
   function as a tumor suppressor in CC by restraining tumor proliferation
   and invasion, promoting apoptosis, and modulating the PI3K/AKT
   signaling pathway in CC cells. EphA7 represents a promising therapeutic
   target for tumor chemo-/immunotherapy in CC. These insights (Fig.
   [196]7) provide a strong foundation for identifying novel therapeutic
   targets for CC management, although future translational research
   should consider delivery systems.

Fig. 7. Targeted demethylation of EphA7 inhibits tumorigenesis via the
SP1/DNMT1 & PI3K/AKT axis and improves the response to multiple therapies in
cervical cancer.

   [197]Fig. 7
   [198]Open in a new tab

   As in our previous research, EphA7 was significantly hypermethylated in
   CC tissues, and the CRISPR-dCas9-Tet1 system was used to reduce the
   promoter methylation of EphA7 by reactivating its expression, which
   provided strong evidence that promoter methylation of EphA7 is
   inversely correlated with gene expression in CC. In this study, we
   found that the promoter demethylation of EphA7 via the
   CRISPR-dCas9-Tet1 system inhibited proliferation, migration and EMT,
   and induced apoptosis through the PI3K/AKT signaling pathway. Moreover,
   we found that E2 could impair SP1, reduce the methylation of EphA7, and
   reverse the expression of EphA7 by reducing the expression of DNMT1 via
   the SP1/ER axis. Furthermore, the chemosensitivity of CaSki and SiHa
   cells to cisplatin and paclitaxel was effectively improved with the
   demethylation of EphA7, and the demethylation of EphA7 was associated
   with a favorable response to therapy and a favorable prognosis for CC
   patients. In addition, pooled analysis indicated that EphA7
   hypermethylation was positively correlated with tumor purity but
   negatively correlated with immune infiltration, CTLs and IC activity,
   and the expression of EphA7 was significantly positively correlated
   with TMB, MSI, and SNV neoantigens, suggesting a better prognosis with
   EphA7 demethylation/high expression. Overall, specific demethylation of
   the EphA7 promoter and restoration of endogenous EphA7 expression by
   dCas9-Tet1 hold promising therapeutic applications and offer a
   favorable prognosis for CC.

Supplementary information

   [199]Supplemental file^ (3.9MB, docx)
   [200]Supplementary Table^ (19.1KB, xlsx)
   [201]Original data^ (2.5MB, pdf)

Author contributions

   Rong Wang: conceptualization, resources, investigation, methodology.
   Wenfan Zhang and Jing Zhao: conceptualization, methodology,
   writing–original draft, and writing–review. Xueting Fan: data curation,
   formal analysis, and validation. Shuang Chen: data curation, formal
   analysis, and visualization.

Funding

   This study is supported by the Natural Science Foundation of Tianjin
   Municipality, China (23JCYBJC00230).

Data availability

   All the data generated or analyzed during this study can be found in
   this published article and its supplementary information files. Only
   publicly available data were used for this study. The sources of these
   data and how we handled them are described in the Materials and Methods
   section. Additional data that support the findings of our study can be
   obtained from the corresponding author upon request.

Competing interests

   The authors declare no competing interests.

Ethics

   Ethical approval was not required for this study as it utilized data
   from publicly available databases, including The Cancer Genome Atlas
   (TCGA) and the Gene Expression Omnibus (GEO), and human-derived cell
   lines (HEK293T, CaSki, SiHa, H8) were commercialized. All data and
   samples were anonymized and ethically sourced, complying with
   international ethical standards.

Footnotes

   Edited by Domagoj Vucic

   Publisher’s note Springer Nature remains neutral with regard to
   jurisdictional claims in published maps and institutional affiliations.

   These authors contributed equally: Wenfan Zhang, Jing Zhao.

Supplementary information

   The online version contains supplementary material available at
   10.1038/s41419-025-07512-4.

References