Epigenetic silencing L3MBTL4 sensitizes esophageal cancer to DNA-PKcs inhibitor

ABSTRACT

Background

Esophageal squamous cell carcinoma (ESCC) remains a deadly disease, with no effective therapeutics available for advanced stages. The application of the “synthetic lethality” principle to cancers with abnormal epigenetic changes provides more opportunities for developing novel therapeutic strategies. It is necessary to identify more molecules that are involved in the DNA damage repair response or cell fate determination to reach this end. Malignant brain tumor (MBT) domain proteins are important for development and cell fate. L3MBTL4 is a new member of this family, but its function remains to be clarified.

Methods

Lentiviral infection was used to re-express L3MBTL4. Immunoprecipitation, immunofluorescence, comet, homologous recombination (HR), and non-homologous end joining (NHEJ) reporter assays were performed to explore the mechanism involved.

Results

The expression of L3MBTL4 was regulated by methylation of the promoter region. L3MBTL4 inhibited cell proliferation and colony formation, and induced G1/S arrest and apoptosis in ESCC cells. L3MBTL4 promoted ATM signaling and inhibited NHEJ signaling by interacting with KU70. Epigenetic silencing of L3MBTL4 sensitized ESCC cells to NU7441, a DNA-PKcs inhibitor, both in vitro and in vivo.

Conclusion

L3MBTL4 is a potential tumor suppressor and methylation of L3MBTL4 is a sensitive marker of DNA-PKcs inhibitors.

1. Introduction

Esophageal cancer is a deadly disease that includes two major tissue types with distinct biological characteristics.^1,2 Esophageal squamous cell carcinoma (ESCC) accounts for approximately 90% of cases, and the incidence of esophageal adenocarcinoma (EAC) is increasing markedly.³ ESCC has been linked to alcohol, tobacco, environmental pollution, and human papillomavirus. Squamous cell dysplasia closely related to ESCC is polyclonal and heterogeneous.⁴ Common mutations include TP53, CDKN2A, PIK3CA, PTEN, MLL2, KDM6A, ZNF750, and NOTCH1/3. Mutations in TP53 and CDKN2A are similar in squamous dysplasia and ESCC.⁴ The increased frequency of NOTCH1 mutations is also associated with age. Somatic mutations occur frequently in healthy cells during life, most of which accumulate passively without affecting cell behavior.⁵ Overall, most mutations are likely to occur via intrinsic mutational processes linked to age or transcription, and there is no clear evidence related to external mutagenic processes. Occasionally, some key gene mutations have been linked to diseases, and targeted therapy has mainly focused on gain-of-function mutations in cancer.^5-7 Despite the importance of somatic mutations in cell behavior and diseases, the underlying genetic regulatory mechanism has been challenged in a recent study.⁸ The color of butterfly wings (red, yellow, white, and black stripes) has been recognized to be controlled by a protein-coding gene, Cortex. It was upended by the discovery that visible traits were controlled by long noncoding RNA.⁸ Abnormal epigenetic changes have been applied to cancer diagnosis and therapeutics. However, most trials have been unsuccessful in solid tumors, as the “epidrugs” unselected targeting cancer and normal cells can induce extensive toxicity.^9-11 This approach is encouraging to target DNA damage response (DDR) pathways in cancer, with the successful application of PARP inhibitors in BRCA1/2 defective tumors.¹² Although various mutations in DDR-related genes have been detected in different cancers, many of these events are probably passenger mutations without functional roles in DDR.¹³ Experimental studies have shown promising results for epigenetic-based synthetic lethal therapeutic strategies for cancer.^14,15

The malignant brain tumor (MBT) domain harbors approximately 100 amino acid residues, which are highly conserved between Drosophila melanogaster and humans.^16,17 The genomes of multicellular eukaryotes encode several proteins containing conserved MBT repeats.^16,17 The temperature-sensitive mutant lethal (3) malignant brain tumor (L3MBT) was obtained from a developing fly through genetic screening for malignant transformations. Malignant overgrowth of the larval brain has resulted in L3MBT mutants accounting for 100% of individuals and die at the end of the third instar larva phase.^17,18 It was discovered that four homologs of MBT proteins are encoded by the human genome, and some of them have been found to be related to important signaling pathways, including human L3MBT-like pathways (L3MBTL)1, 2, 3, and 4.¹⁷ L3MBTL1 and L3MBTL2 have been identified to be involved in DDR.^19-21 L3MBTL3 is associated with cancer development through negative regulation of HIF-1α.²² L3MBTL4 is a novel member of the MBT gene family and is located on chromosome 18p11. Reduced L3MBTL4 expression and genomic region deletions have been reported in breast cancer.²³ However, the role of L3MBTL4 in cancer needs to be clarified.

2. Results

2.1. L3MBTL4 expression is regulated by promoter region methylation in ESCC

The Cancer Genome Atlas (TCGA) database (http://xena.ucsc.edu/) was used to assess the epigenetic regulation of L3MBTL4 in ESCC. Compared with normal esophageal tissue, the levels of L3MBTL4 expression were decreased in cancer samples (Figure S1A). A reverse association was observed between L3MBTL4 expression and CpG site methylation in the promoter region around the transcription start site (TSS) (Figure S1B and S1C, P < 0.0001). To validate that L3MBTL4 expression is regulated by promoter region CpG methylation, the levels of L3MBTL4 expression and promoter region methylation status were detected by semi-quantitative reverse transcription polymerase chain reaction (RT-PCR) and methylation-specific PCR (MSP) in esophageal cancer cells. L3MBTL4 was not expressed in KYSE30, KYSE150, KYSE410, and KYSE510 cells; low levels of expression were detected in KYSE180, KYSE450, KYSE520, and COLO680N cells (Figure 1A). Complete methylation was detected in KYSE30, KYSE150, KYSE410, and KYSE510 cells, and partial methylation was observed in KYSE180, KYSE450, KYSE520, and COLO680N cells (Figure 1B). The results imply that the reduction of L3MBTL4 expression is associated with promoter region methylation. To further verify that the expression of L3MBTL4 is regulated by DNA methylation, 5-aza-2’-deoxycytidine (5-aza), a DNA demethylation reagent, was applied. Treatment with 5-aza restored the expression of L3MBTL4 in completely methylated cells, and elevated the expression levels of L3MBTL4 in partially methylated cells, suggesting that DNA methylation regulates L3MBTL4 expression (Figure 1A). Thereafter, the methylation status of L3MBTL4 was examined in tissue samples. L3MBTL4 was methylated in 13.85% (9/65) of esophageal dysplasia samples and 24.06% (249/1035) of primary esophageal cancer samples (Figure 1C and D). Methylation of L3MBTL4 was significantly associated with gender (P < 0.05) and tumor differentiation (P < 0.01, Table 1).

Figure 1.

The expression and methylation status of L3MBTL4 in ESCC cells and tissue samples. (A) The expression of L3MBTL4 in ESCC cells. 5-aza: 5-aza-2'-deoxycytidine; GAPDH: internal control; (–): absence of 5-aza; (+): presence of 5-aza. (B) MSP results in ESCC cells. U: unmethylation alleles; M: methylation alleles; IVD: in vitro methylated DNA, which serves as a methylation control; NL: normal peripheral lymphocytes DNA, which serves as an unmethylation control. (C) Representative MSP results of esophageal intraepithelial neoplasia (EIN). (D) Representative MSP results for esophageal carcinoma (EC) samples. (E) Representative immunohistochemistry staining of L3MBTL4 in esophageal tumor samples and adjacent noncancerous tissue samples. Scale bar: 100 μm (top); 50 μm (bottom). (F) Immunohistochemistry staining score. ****P < 0.0001. (G) Bar diagram indicates an inverse association between the levels of L3MBTL4 and the DNA methylation status. *P < 0.05.

Table 1.

The association of L3MBTL4 methylation and clinical factors in ESCC.

Variables	Number	L3MBTL4 methylation status		P
		Unmethylation n = 786 (75.94%)	Methylation n = 249 (24.06%)
Age (year)				0.725
≤50	63	49	14
>50	972	737	235
Gender				0.015*
Female	334	238	96
Male	701	548	153
Smoking				0.647
No	570	436	134
Yes	465	350	115
Alcohol abuse				0.810
No	742	562	180
Yes	293	224	69
Cancer family history				0.210
Negative	807	620	187
Positive	228	166	62
Tumor size (cm)				0.672
≤4	540	413	127
>4	495	373	122
Differentiation				0.006**
Well or moderate	721	565	156
Poor	314	221	93
TNM stage				0.887
I–II	657	498	159
III–IV	378	288	90
Lymph node metastasis				0.962
No	539	409	130
Yes	496	377	119

P-values are obtained from the chi-squared test.

^* P < 0.05.

^** P < 0.01.

The levels of L3MBTL4 expression were evaluated in 40 cases of available matched cancer and adjacent tissue samples using immunohistochemistry (IHC). The level of L3MBTL4 was higher in adjacent tissues than in cancer tissue samples (Figure 1E and F, P < 0.0001). L3MBTL4 was expressed in both the nucleus and cytoplasm. Among the 24 cases of L3MBTL4 expression was reduced, 14 cases were methylated (58.33%), while only 3 cases were methylated (18.75%) in 16 cases of normal level expressed cancer samples. The ratio of L3MBTL4 methylation was significantly higher in cancer samples with reduced expression than in normal levels (Figure 1G, P < 0.05), indicating epigenetic regulation of L3MBTL4 expression.

2.2. L3MBTL4 suppressed cell proliferation, induced G1/S arrest and apoptosis in ESCC cells

The MTT assay was utilized to evaluate cell viability. Before and after re-expression of L3MBTL4, the OD values were 1.104 ± 0.047 vs. 0.826 ± 0.039 (P < 0.0001) and 1.189 ± 0.019 vs. 0.863 ± 0.021 (P < 0.0001) in KYSE30 and KYSE150 cells, respectively (Figure 2A). These results indicate an inhibitory role of L3MBTL4 in cell proliferation. The clone numbers were 456.7 ± 41.9 vs. 300.7 ± 17.0 and 457.3 ± 48.4 vs. 232.3 ± 3.8 in L3MBTL4 silenced and re-expressing KYSE30 and KYSE150 cells, respectively (both P < 0.01, Figure 2B). These results demonstrated that L3MBTL4 reduced the colony number.

Figure 2.

Effect of L3MBTL4 on ESCC cell proliferation, colony formation, the cell cycle, and apoptosis. (A) MTT assay showing the effect of L3MBTL4 on ESCC cell proliferation. (B) Effect of L3MBTL4 on colony formation. The scale bar represents the average clone numbers. (C) Bar diagram showing the distribution of the cell phases according to the flow cytometry results. (D) Western blot results of G1/S phase regulators normalized to that of β-actin. n = 3 in each group. (E) Apoptotic results for L3MBTL4 unexpressed and re-expressed cells. (F) Western blot results of apoptotic-related proteins, normalized to β-actin. n = 3 in each group. Vector: empty vector control; L3MBTL4: L3MBTL4 expressing vector. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

In L3MBTL4 unexpressed and overexpressed KYSE30 cells, the cell phase distribution was 37.2 ± 1.9% vs. 49.1 ± 0.9% in the G0/G1 phase (P < 0.001), 46.4 ± 3.7% vs. 32.8 ± 0.6% in the S phase (P < 0.01), and 16.5 ± 2.1% vs. 18.1 ± 0.7% in the G2/M phase (Figure 2C). In L3MBTL4 -silenced and re-expressed KYSE150 cells, the G0/G1 phase was 41.4 ± 3.4% vs. 56.7 ± 0.7% (P < 0.01), the S phase was 39.1 ± 2.1% vs. 28.8 ± 0.4% (P < 0.01), and the G2/M phase was 19.6 ± 2.0% vs. 14.5 ± 0.3% (P < 0.05) (Figure 2C). These results indicated that L3MBTL4 induced G1/S arrest in ESCC cells. The levels of G1/S phases key regulators were detected by Western blotting. The levels of cyclin D1, cyclin A2, cyclin E1, and CDK2 were reduced by the re-expression of L3MBTL4 in KYSE30 and KYSE150 cells (Figure 2D), supporting the above results.

The influence of L3MBTL4 on apoptosis was evaluated using cell cytometry. The ratio of apoptotic cells was 6.9 ± 1.7% vs. 14.6 ± 0.6% and 6.2 ± 0.5% vs. 13.3 ± 0.6% in L3MBTL4 unexpressed and re-expressed KYSE30 and KYSE150 cells, respectively (P < 0.01, P < 0.0001, Figure 2E), suggesting that ESCC cell apoptosis was induced by L3MBTL4. Apoptosis-related proteins were detected by Western blotting. Increased levels of cleaved-caspase3 and Bax and reduced Bcl-2 were observed by re-expressing L3MBTL4 in KYSE30 and KYSE150 cells (Figure 2F), indicating the role of L3MBTL4 in inducing cell apoptosis.

2.3. L3MBTL4 involves in DNA damage repair in ESCC

To explore the mechanism of L3MBTL4 in ESCC, an immunoprecipitation (IP) assay was performed. KU70 had the highest score after analyzing the distinct extra band in the L3MBTL4 overexpressed KYSE150 cell lysate by mass spectrometry (Figure 3A). The combination of L3MBTL4 and KU70 was verified by Co-IP and reciprocal Co-IP (Figure 3B), which was validated by colocalizing of L3MBTL4 and KU70 in the nucleus with immunofluorescence assay (Figure 3C). KU70 is a key component of non-homologous end joining (NHEJ) signaling, suggesting the possible role of L3MBTL4 in the DDR. Two members of the MBT gene family serve as DDR molecules.^19,24

Figure 3.

L3MBTL4 promotes DNA damage repair in ESCC. (A) IP assay and silver staining. Red arrow: differential band. (B) Co-IP and reciprocal Co-IP results. (C) Immunofluorescence assay showing the colocalization of L3MBTL4 and KU70. (D) IC50 values of cisplatin in L3MBTL4 unexpressed and re-expressed cells. (E) Comet assay in low-dose cisplatin-treated ESCC cells. The data are presented as the mean ± SD of three independent experiments. **P < 0.01, ****P < 0.0001.

A DNA damage cell model induced by low-dose cisplatin was used to evaluate the effect of L3MBTL4 in DDR. First, the cytotoxicity of cisplatin was evaluated in ESCC cells. The IC50 value was 5.2 ± 0.5  μM vs. 11.9 ± 1.8  μM in KYSE30 cells and 6.3 ± 1.0  μM vs. 16.1 ± 2.3  μM in KYSE150 cells before and after re-expressing L3MBTL4, respectively (both P < 0.01, Figure 3D). These results demonstrate that L3MBTL4 reduces cisplatin toxicity in these cells.

The impact of L3MBTL4 on DNA double-strand break (DSB) repair was further validated using the comet assay. Compared to epigenetically silenced KYSE30 and KYSE150 cells, overexpression of L3MBTL4 reduced the comet tail moment in both cell lines (Figure 3E). These results suggest that L3MBTL4 promotes DDR in ESCC cells.

2.4. L3MBTL4 activates ATM/CHK2 signaling and inhibit NHEJ signaling

DSB repair plays a crucial role in maintaining genomic integrity, including homologous recombination (HR) and NHEJ, which are the major signaling pathways involved. HR and NHEJ efficiency assays were utilized to detect the impact of L3MBTL4 on the DDR. The efficiency of HR was significantly decreased by L3MBTL4 knockdown in U2OS cells, a classical DDR cell model, suggesting a role for L3MBTL4 in promoting HR (Figure 4A). The efficiency of NHEJ signaling was increased by L3MBTL4 knockdown (Figure 4A), suggesting that L3MBTL4 inhibits NHEJ signaling. Thereafter, the role of L3MBTL4 in the DDR was detected in KYSE30 and KYSE150 cells (Figure 4B). To explore the impact of L3MBTL4 on the DDR, the key components of these pathways were detected using western blotting. The levels of p-ATM and p-CHK2, two molecules representing ATM activation, were increased by re-expressing L3MBTL4 in KYSE30 and KYSE150 cells, demonstrating the promoting role of L3MBTL4 in ATM signaling (Figure 4C). P-ATR and p-CHK1 are the two major activating components of ATR signaling, and no obvious changes were observed by re-expressing L3MBTL4, suggesting that L3MBTL4 has no effect on ATR signaling in ESCC cells (Figure 4D). The P-DNA-dependent protein kinase catalytic subunit (P-DNA-PKcs) represents an active molecule for NHEJ signaling. In KYSE30 and KYSE150 cells, the level of p-DNA-PKcs was reduced after restoration of L3MBTL4 expression, suggesting that L3MBTL4 inhibits NHEJ signaling in ESCC cells (Figure 4E). The role of L3MBTL4 in NHEJ was further validated by detecting p-DNA-PKcs before and after siRNA-mediated knockdown of KU70 (Figure 4F).

Figure 4.

L3MBTL4 activates ATM/CHK2 signaling and inhibits NHEJ signaling. (A) HR and NHEJ efficiencies were evaluated in U2OS-DR-GFP and U2OS-EJ5 cells before and after L3MBTL4 knockdown. (B) HR and NHEJ efficiencies were evaluated in KYSE30 and KYSE150 cells with or without L3MBTL4 expression. (C–E) Levels of ATM/CHK2, ATR/CHK1, and NHEJ signaling in L3MBTL4 unexpressed and re-expressed ESCC cells after treatment with 1  μM cisplatin for 48  h, as detected by immunoblotting, normalized to that of β-actin. n = 3 in each group. (F) The effects of KU70 knockdown on the NHEJ pathways for L3MBTL4 unexpressed and re-expressed ESCC cells normalized to that of β-actin. n = 3 in each group. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

2.5. Loss of L3MBTL4 sensitizes ESCC cells to DNA-PKcs inhibitor

As L3MBTL4 inhibits NHEJ signaling, the sensitivity of ESCC cells to the DNA-PKcs inhibitor NU7441 was tested. The IC50 value of NU7441 was evaluated in distinct cells. Under treatment with 0.5  μM cisplatin, the IC50 value was 6.2 ± 2.2  μM vs. 21.6 ± 2.6  μM and 6.2 ± 1.0  μM vs. 25.2 ± 4.2  μM in L3MBTL4 unexpressed and re-expressed KYSE30 and KYSE150 cells, respectively (both P < 0.01, Figure 5A). The IC50 value of NU7441 was increased by L3MBTL4. These results suggest that the lack of L3MBTL4 expression increases the sensitivity of ESCC cells to NU7441.

Figure 5.

Loss of L3MBTL4 expression sensitized ESCC cells to NU7441. (A) IC50 assay showing the sensitivity of ESCC cells to NU7441 under the treatment of 0.5  μM cisplatin. (B) Representative colony formation results showing the synthetic lethal effect of the L3MBTL4 defect and NU7441. (C) Western blot results for NHEJ signaling and γ-H2AX in L3MBTL4 unexpressed and re-expressed cells under the treatment of cisplatin and NU7441 for 48 h, normalized to that of β-actin. n = 3 in each group. Presenting the mean ± SD for three independent experiments. **P < 0.01, ***P < 0.001, ****P < 0.0001.

To explore the potential role of L3MBTL4 defects and DNA-PK inhibitors in synthetic lethality, we performed low-dose cisplatin treatment and colony formation assays. Before and after re-expression of L3MBTL4 in KYSE30 cells and KYSE150 cells, the colony efficiency was normalized to 21.9 ± 4.8% vs. 38.0 ± 5.6% and 13.3 ± 1.4% vs. 30.9 ± 7.3% under combined cisplatin and NU7441 treatment, respectively (both P < 0.01, Figure 5B). The above results suggest that the loss of L3MBTL4 expression reduced the relative colony formation efficiency, indicating the synthetic lethal role of L3MBTL4 depletion and NU7441. This effect was further validated by measuring the levels of the DNA damage marker γ-H2AX (Figure 5C).

2.6. Silencing L3MBTL4 expression sensitized ESCC cell xenografts to NU7441

To further validate the role of L3MBTL4 in the DDR, a KYSE150 cell xenograft model was used. Upon treatment with cisplatin, levels of p-ATM and p-CHK2 were increased, while p-DNA-PKcs was decreased in L3MBTL4 re-expressed KYSE150 cell xenografts, validating the promoting role of L3MBTL4 in ATM signaling and its inhibitory role in the NHEJ pathway in vivo (Figure 6A).

Figure 6.

Synthetic lethality of L3MBTL4 methylation and DNA-PK inhibitor in vivo. (A) Immunohistochemistry results in L3MBTL4 unexpressed and re-expressed KYSE150 cell xenografts under treatment with cisplatin. (B) L3MBTL4 unexpressed and re-expressed KYSE150 cell xenografts under the treatment with 2 mg/kg cisplatin, 10 mg/kg NU7441 and combination of 2 mg/kg cisplatin with 10 mg/kg NU7441. (C) Growth curves of xenograft tumors. (D) and (E) Normalized tumor volume and weight in L3MBTL4 unexpressed and re-expressed KYSE150 cell xenografts. ****P < 0.0001. (F) A working model for synthetic lethality of L3MBTL4 methylation and DNA-PK inhibitor in ESCC cells.

The response of L3MBTL4 -deficient ESCC cells to a DNA-PKcs inhibitor was then investigated. The normalized tumor volumes were 82.8 ± 5.9% vs. 88.5 ± 10.6% (P > 0.05), 59.2 ± 3.4% vs. 83.6 ± 6.1% (P < 0.0001), and 25.6 ± 2.5% vs. 45.8 ± 5.4% (P < 0.0001) in L3MBTL4 unexpressed and overexpressed xenografts in the cisplatin, NU7441, or combined cisplatin and NU7441 treatment groups, respectively (Figure 6B–D). The normalized tumor weight was 77.2 ± 7.0% vs. 82.3 ± 9.2% (P > 0.05), 41.5 ± 4.2% vs. 66.6 ± 8.9% (P < 0.0001), and 10.9 ± 2.4% vs. 31.9 ± 4.3% (P < 0.0001) in L3MBTL4 unexpressed and overexpressed xenografts in the cisplatin, NU7441, or combined cisplatin and NU7441 treatment groups, respectively (Figure 6E). Tumor volume and weight were reduced in the combined cisplatin and NU7441 treatment group in L3MBTL4-silenced cell xenografts, suggesting a synthetic lethality efficiency in vivo.

3. Discussion

Current precision medicine mainly focuses on searching for actionable driver gene mutations in cancers.^25,26 Although benefits have been obtained, suitable cancer patients remains limited, and genomic resources are almost exhausted.^25,27 Classical epigenomic-based cancer therapy is mainly focused on targeting epigenetic regulators.^11,28 Advancing in synthetic lethal therapeutic strategies is promoting researchers to develop a variety of approaches for cancer patients with undruggable genetic or epigenetic abnormal changes, which related to DDR or cell fate-determining signaling pathways.²⁹ The approach of classical targeting therapy is to search for cancer-specific driver mutations and directly inhibit activated signaling pathways.^30-32 Synthetic lethality indirectly targets the partnering pathway of cell fate determination or DDR signaling, which is loss of function by mutations or aberrant epigenetic changes in cancers, to precisely induce cell death.^33-35 Current experimental techniques are insufficient for investigating synthetic lethality gene pairs at the genomic scale.^26,36 To identify novel cancer markers, it is necessary to understand the mechanisms of each gene and discover the compensatory pathways in cell fate decisions. L3MBTL4 was found to be frequently methylated in esophageal dysplasia and ESCC. Further study demonstrated that L3MBTL4 serves as a tumor suppressor in ESCC. Promoter region methylation was validated to regulate L3MBTL4 expression.

To understand the mechanism of L3MBTL4 in ESCC, IP was performed using L3MBTL4 antibodies. KU70 was found to bind to L3MBTL4, which was further validated by reciprocal IP and western blotting. KU70 is a key component of NHEJ. KU70, KU80, and DNA-PKcs form a complex to initiate NHEJ.³⁷ Members of the MBT gene family, L3MBTL1 and L3MBTL2, play important roles in the DSB repair response.^19,20 Then it was found that depletion depletion of L3MBTL4 expression sensitized ESCC cells to cisplatin. The role of L3MBTL4 in DSB repair was validated using a comet assay. Further studies demonstrated that L3MBTL4 promotes ATM signaling and inhibits NHEJ signaling. Epigenetic silencing of L3MBTL4 inhibits ATM signaling and activates NHEJ signaling. These findings offer the opportunity to pursue therapeutic targets.‌ Targeting DNA-PKcs has been used for cancer therapy in preclinical models.³⁷ To harness this epigenetic marker for cancer therapy, NU7441 was tested in L3MBTL4 epigenetically silenced ESCC cells. Both in vitro and in vivo studies demonstrated that the loss of L3MBTL4 expression sensitized ESCC cells to NU7441, suggesting the synthetic lethality role of the epigenetic silencing of L3MBTL4 and the DNA-PKcs inhibitor (Figure 6F). Beyond DNA repair, DNA-PKcs has been shown to be involved in tumorigenesis through transcription regulation, telomere maintenance, metabolic regulation and immune escape.³⁸ NU7441, a conceptional proofed DNA-PKcs inhibitor, has been shown to induce the G2/M checkpoint and mitotic catastrophe in cancer cells.³⁹ Therefore, in addition to enhanced synthetic lethality efficiency after cisplatin treatment, the baseline synthetic lethality interaction of L3MBTL4 deficiency and NU7441 has also been observed in xenograft mice models (Figure 6B–E). The choice of pathways for NU7441 under different circumstances, such as the complex roles of DNA-PKcs, remain to be elucidated. The synthetic lethality of L3MBTL4 deficiency and NU7441 may be extended to other DNA-PKcs inhibitors, which are utilized in clinical trials in combination with radiotherapy and chemotherapy.⁴⁰ Our findings support the proof-of-concept of synthetic lethality in L3MBTL4-silenced ESCC cells and suggest the predictive value of L3MBTL4 methylation for combination with DNA-PKcs inhibitors and radiotherapy or platinum-based chemotherapy in ESCC patients. The function and regulatory network of L3MBTL4 in cancer remain to be fully understood. Extensive studies are necessary by using patient-derived xenograft models, pre-clinical, and clinical trials for translational relevance.

The synthetic lethality paradigm provides a rational basis for combined cancer therapy by targeting partner signaling pathways for patients with genetic or epigenetic abnormal changes. However, most studies have focused on classical DDR gene mutations. The regulatory networks for the DDR and cell fate decision-making are extremely complex. In addition to protein-coding genes, noncoding RNAs play important roles in the cell phenotype in a dynamic fashion, including DDR and cell fate regulation.^41,42 A deeper understanding of these mechanisms will broaden the application of combined cancer therapies and reduce drug resistance. There are still some challenges for impeding cancer therapy, including genetic and epigenetic heterogeneity and the switch of signaling transductions.

4. Conclusion

L3MBTL4 is frequently methylated in ESCC. L3MBTL4 is a novel tumor suppressor and is involved in the DDR. Epigenetic silencing L3MBTL4 is a synthetic lethal marker for DNA-PKcs inhibitor treatment of ESCC cells.

5. Materials and methods

5.1. Cell lines and esophageal tissue samples

The human ESCC cell lines KYSE30 (#CL0189), KYSE150 (#CL0493), KYSE180 (#CL0605), KYSE410 (#CL0190), KYSE450 (#CL0606), KYSE510 (#CL0191), and KYSE520 (#CL0913) were acquired from the cell bank of Hunan Fenghui Biotechnology Co., Ltd. COLO680N (#CL0928) was acquired from Procell Company. The cell lines were authenticated via short tandem repeat (STR) profiling. All the cell lines were confirmed to be mycoplasma negative prior to the experiments with a CellCare Quick mycoplasma detection kit (CYTOCH, #F250003BD, China). ESCC cell lines were cultured in RPMI 1640 medium (Gibco, #31800089, USA) supplemented with 10% fetal bovine serum (Gemini, #900-108, USA) and 1% penicillin/streptomycin (Beyotime, #C0222, China). Tissue samples, including 65 cases of esophageal dysplasia and 1035 cases of ESCC, were obtained from the Chinese PLA General Hospital. Before surgery, chemoradiotherapy was not performed in these patients, and TNM staging was classified according to the guidelines (AJCC 2019). The procedure was approved by the Institutional Review Board of the Chinese PLA General Hospital (IRB no. 20090701-015) in accordance with the principles of the Helsinki Declaration (1964). Informed written consent was obtained from all the patients for the purpose of this study.

5.2. 5-aza treatment, PCR, MSP, and IHC detection

The procedures of 5-aza treatment, RT-PCR, and methylation-specific PCR (MSP) were performed as previously described.⁴³ The primer sequences and antibodies used for IHC are listed in Tables S1 and S2, respectively.

5.3. L3MBTL4 expressing constructs and monoclonal cell screening

The coding sequence (CDS) of human L3MBTL4 (NM_173464) was inserted into the pCDH-CMV-MCS-puro vector and validated by the Sanger sequence. L3MBTL4-expressing or empty vectors were transfected into HEK293 cells using Lipofectamine 3000 (Invitrogen, #L3000008, USA). The lentiviral supernatant was added to the culture medium. L3MBTL4-expressing cells were obtained by selecting puromycin (Beyotime, #ST551, China)-treated cells for 3 d, with 2.5 μg/ml for KYSE30 cells and 1.5 μg/ml for KYSE150 cells. L3MBTL4-expressing single-cell clones were obtained by limited dilution screening and validated by western blotting.

5.4. L3MBTL4 siRNA knockdown techniques

RNAiMax was used for siRNA knockdown (JTS Scientific, Beijing, China) following the manufacturer’s instructions (Invitrogen, #13778150, USA). The sequences are listed in Table S1.

5.5. MTT, colony formation, and cell flow cytometry assays

The MTT assay, colony formation assay, and cell flow cytometry were performed as previously described.⁴³ The MTT assay was used to evaluate cell proliferation and IC50 analysis (KeyGEN Biotech, # KGT5251, China). KYSE30 and KYSE150 cells (2 × 10³ per well) were seeded in 96-well plates to evaluate cell proliferation, and the sensitivity to cisplatin (Selleck, #S1166, USA) and NU7441 (MCE, #HY-11006, USA) was analyzed by seeding cells at 2 × 10³ cells per well. For the colony formation assay, cells (1 × 10³ cells/well) were seeded in 6-well plates and grown for 12 d. The efficacy of cisplatin and NU7441 was evaluated as previously described.¹⁴ The cell cycle and apoptosis were analyzed using a cell cycle detection kit (KeyGEN Biotech, #KGA9101, China) and an Annexin V-FITC/PI Apoptosis Detection Kit (KeyGEN Biotech, #KGA1101, China), according to the manufacturer’s instructions, using a FACS Caliber flow cytometer (BD, FACSAria™ Fusion, USA). Each experiment was repeated three times.

5.6. Western blot, immunoprecipitation, and immunofluorescence colocalization assays

The antibodies used are listed in Table S2, and co-immunoprecipitation (Co-IP) and western blotting were performed as previously described.⁴⁴ For the immunofluorescence assay, cells were seeded on coverslips for 24 h and washed with PBS. After fixation with 4% paraformaldehyde, the cells were treated with 0.3% Triton X-100. After blocking with BSA and goat serum, the coverslips were incubated with L3MBTL4 and KU70 antibodies overnight and then incubated with CoraLite488 (Proteintech, #RGAM002, China) and CoraLite594-conjugated antibodies (Proteintech, #RGAR004, China). Coverslips were stained with DAPI (ZSGB-BIO, #ZLI-9600, China), and fluorescent images of L3MBTL4 and KU70 were captured using confocal microscopy (Olympus, OlyVIA, Japan).

5.7. HR and NHEJ assay

HR and NHEJ assays were employed to determine the HR and NHEJ repair efficiency, as described previously.¹⁴ Briefly, U2OS cells expressing direct repeat GFP (DR-GFP) or end joining 5 GFP (EJ5-GFP) were transfected with siL3MBTL4 and scrambled control for 24 h, and I-Scel expression and empty lentivirus vectors were transfected. The efficiency of HR and NHEJ was evaluated using FACS after 24 h (BD Biosciences, USA). L3MBTL4 silenced or re-expressed KYSE30 and KYSE150 cells were grown in 6-well plates and transfected with 1.25  μg of pCVL Traffic Light Reporter 1.1 (Sce target) Ef1a Puro plasmids and 1.25  μg of I-SceI plasmids. The cells were harvested and analyzed by FASC for growing 48 h. The data were analyzed using FlowJo V10 (BD Biosciences, USA). The experiments were performed in triplicate.

5.8. Comet assay

Comet assay was conducted in accordance with previously described methods.^14,45 Cisplatin (0.5  μM) was used to induce DNA damage in ESCC cells. Briefly, 150 μl of 0.5% normal melting agarose was added to a comet slide. The cells were collected and mixed with 0.5% low-melting agarose. Subsequently, 80 μl of the mixture was carefully covered onto the coated slides. A pre-chilled lysing solution was applied to immerse the slides at 4 °C. The slides were then placed in an alkali unwinding solution for 15 min. Other procedures were performed as previously described.¹⁴ The experiments were performed in triplicate.

5.9. Evaluating the synthetic lethal role of L3MBTL4 epigenetic silencing and DNA-PKcs inhibitor in ESCC cell xenograft mouse model

To evaluate the in vivo efficiency of L3MBTL4 defects, four-week-old female BALB/c nude mice (16–18 g/mouse, n = 48) were used (SPF Biotechnology Company, Beijing, China), maintaining pathogen-free conditions. The mice were randomly divided into two groups via a random number table (n = 24/group). KYSE150 cells (4 × 10⁶ cells/mouse), with or without L3MBTL4 expression, were inoculated subcutaneously. The tumor volume was calculated as length × width²/2. When the average tumor volume reached 100 mm³, the xenograft-bearing mice were then sub-grouped into the control (saline), cisplatin (2  mg/kg), NU7441 (10  mg/kg) and combination of cisplatin with NU7441 (2  mg/kg cisplatin + 10  mg/kg NU7441) groups. Cisplatin and NU7441 were administered intraperitoneally twice a week for 2 weeks. The tumor dimensions were measured every 3 d for 21 d. Thereafter, the mice were euthanized by cervical dislocation in strict compliance with the American Veterinary Medical Association Guidelines (2020 Edition). The animal experiments were performed under the guidance of ARRIVE. These procedures were approved by the Animal Ethics Committee of the Chinese PLA General Hospital (Approval No. 2022-X18-72).

5.10. Statistical analysis

The GraphPad Prism 8.0 (GraphPad Software Inc., CA, USA) was used for the statistical analysis. The independent experiments were repeated at least three times. The values are expressed as mean ± standard deviation (SD) of replicate measurements. The associations between L3MBTL4 methylation status and clinicopathological features were analyzed using the Chi-square test. The difference between the two experimental groups was determined by Student’s t-test, and P < 0.05 was regarded as statistically significant.

Supplementary Material

Supplementary figure and legends.docx

Supplementary figure and legends.docx

Figure S1.tif

Figure S1.tif

Supplementary_Table_revised_clean.docx

Supplementary_Table_revised_clean.docx

Funding Statement

This work was supported by grants from the National Science and Technology Major Project of China (2025ZD0544701), National Key Research and Development Program of China (2020YFC2002705), National Natural Science Foundation of China (82272632, 81672138, 82403742), Beijing Natural Science Foundation (7171008, 7254313), and Youth Innovation Science Foundation of Chinese PLA general hospital (22QNCZ027).

Disclosure statement

The authors declare no conflicts of interest.

Data availability statement

All data are available from the corresponding author upon reasonable request.

Supplemental material

Supplemental data for this article can be accessed at https://doi.org/10.1080/15384047.2026.2646393.