Biophysical Properties of CTCFL and AHR Condensate Regulate Glutathione S-Transferase Mediated Lymph Node Density in Esophageal Squamous Cell Carcinoma

doi:10.21203/rs.3.rs-4675218/v1

Download PDF

Research Article

Biophysical Properties of CTCFL and AHR Condensate Regulate Glutathione S-Transferase Mediated Lymph Node Density in Esophageal Squamous Cell Carcinoma

https://doi.org/10.21203/rs.3.rs-4675218/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background

Esophageal squamous cell carcinoma (ESCC) prognosis is closely associated with lymph node density (LND). The identification of biomarkers and regulatory mechanisms influencing LND could enhance our understanding of ESCC progression and inform therapeutic strategies.

Methods

This study analyzed 8,716 esophageal cancer patients to determine the prognostic significance of LND. Univariate and multivariate Cox regression analyses were performed to assess clinical factors. Gene expression data from The Cancer Genome Atlas (TCGA) were used to identify differentially expressed genes (DEGs) between LND < 0.12 and LND ≥ 0.12 groups. Functional enrichment, protein-protein interactions, and transcriptional regulation were investigated using advanced computational tools, immunoprecipitation, immunofluorescence, CUT&Tag sequencing, and phase separation assays.

Results

Higher LND (≥ 0.12) was associated with poorer survival outcomes. DEGs analysis revealed significant enrichment in glutathione metabolic pathways. CTCFL and AHR transcription factors were identified as key regulators of glutathione S-Transferase (GSTs) genes. These transcription factors exhibited phase separation properties, enhancing GSTs transcription. Knockdown experiments confirmed that CTCFL and AHR collaboratively regulate GSTs, affecting reactive oxygen species (ROS) levels and LND. In vivo, ESCC models demonstrated upregulation of CTCFL, AHR, and GSTs in high-LND mice, corroborating the regulatory role of these factors in tumor progression.

Conclusion

The transcription factors CTCFL and AHR regulate GST-mediated glutathione metabolism, influencing LND and ESCC progression. Targeting these regulatory pathways may offer novel therapeutic approaches for managing ESCC.

esophageal squamous cell carcinoma

lymph node density

optimal cutoff

cox regression

glutathione metabolism

phase separation

Glutathione S-Transferase

Esophageal cancer is ranking as the sixth leading cause of cancer-related deaths worldwide.^1,2 In the last 20 years, the incidence of esophageal cancer in the United States has been arising.³ It was estimated that there were approximately 455,800 new cases and 400,200 deaths for esophageal cancer in 2012 worldwide.⁴ In 2019, it was reported 17,650 new cases and 16,080 deaths esophageal cancer in US.⁵ The advances in esophageal cancer treatment, such as the application of minimally invasive esophagectomy and endoscopic treatment, can significantly reduce the incidence of complications, decrease the length of hospital stay after surgery, and improve quality of life, however, most of esophageal cancer cases are diagnosed at the late stage, thus the 5-year survival rate of esophageal cancer patients in the United States is relatively poor.^6–10 Therefore, developing a more comprehensive system of prognostic factors is important for improving the prognosis of esophageal cancer patients.

Recently, increasing evidence had identified lymph node density as an important prognostic factor for various cancer patients. For instance, it was reported that the LNR was a key prognostic factor which should be applied in stratification schemes for those clinical trials investigating adjuvant treatments in colorectal cancer patients who received curative resection.¹¹¹ Furthermore, LNR was the most important prognostic factor for overall as well as disease-free survival in rectal cancer patients even in those patients with less than lymph nodes examined.¹² Similarly, LNR was also recognized as an independent prognostic factor for oral squamous cell carcinoma patients.¹³ Here, we intend to use the SEER database, which is generated by the National Cancer Institute including 18 registries that cover 30% of the US population,^14,15 to explore the prognostic potential of lymph node density in esophageal cancer patients.

Glutathione(GSH)plays an important component of the intracellular antioxidant system.¹⁶ In cancer cells, high levels of glutathione are necessary for clearing excess reactive oxygen species (ROS) and detoxifying exogenous substances, including synthetize its product by GCLC and a conjugation reaction mediated by glutathione S-transferases (GST_S).¹⁷ GST_S are mainly expressed in the liver. However, some of GSTs are also expression in a high level in ESCC.¹⁸ Glutathione metabolism have long been associated with prognosis and process in ESCC. For example, GSTO1, GSTP1 and GST-pi have been perceived as tumor-associated antigens and act as predictor in ESCC.¹⁹ Previous research demonstrated that deletion of GSTM1 polymorphism was recognized as a risk factor in ESCC.¹⁸ In skin SCC, the glutathione metabolism pathway is specifically enriched in the TGF-β-activating cancer stem cell population and is responsible for the chemo-resistant property of these cell,²⁰ thus suggesting that glutathione metabolism might be critical for ESCC tumor progression.

2.1. Patients selection

For ESCC patients from 2004 to 2019 using SEER*Stat, version 8.3.6, Only patients over 18 years old were enrolled. All cases were stratified according to the original tumor sites: C15.0 cervical esophagus, C15.1 thoracic esophagus, C15.2 abdominal esophagus, C15.3 upper third of esophagus, C15.4 middle third of esophagus, C15.5 lower third of esophagus, C15.8 overlapping lesion of esophagus, and C15.9 esophagus, NOS. All patients were microscopically confirmed. The exclusion criteria include: patients without surgery, diagnosis, or microscopic examination; patients only have autopsy examinations result; patients with missing or inappropriate follow-up information. These exclusion criteria finally recruited 8716 eligible patients for the current study. The following information was obtained for each patient: age at diagnosis, marital status, race, sex, grade, SEER stage, surgery, radiation, chemotherapy, number of lymph nodes examined, staging according to American Joint Committee on Cancer (AJCC), year of diagnosis, survival months and vital status.

For ESCC samples with corresponding clinical data from The Cancer Genome Atlas database (TCGA, https://portal.gdc.cancer.gov/), Prior to inclusion, patients must possess comprehensive clinical data encompassing age, survival status, survival duration, lymph node density, among other pertinent factors.

2.2. Data processing

If not specifically addressed, continuous data were shown as the mean ± standard deviations, and categorical data were presented as frequencies or proportions. Cox regression was used to search independent risk factors first, only those variables with statistical significance by univariate analysis were then analyzed in multivariate cox regression. The informed consent from each patient was not required as the patients included in the SEER database were anonymized and de-identified before release.^21,22 All statistical tests were two sided, P value less than 0.05 was accepted statistical significant. All statistical analyses were performed using SPSS and R software.

To identify the differently expressed genes between LND < 0.12 and LND ≥ 0.12 in ESCC samples, We employed the Limma package (http://www.bioconductor.org/packages/release/bioc/html/limma.html) to conduct analyses utilizing the negative binomial distribution. All raw data underwent preprocessing via the Limma package, wherein genes with an average count value below 1 were excluded. Furthermore, the Limma package were utilized to ascertain differentially expressed genes based on |log2 fold change (FC)|≥1 and false discovery rate (FDR). Unless otherwise specified, continuous data were depicted as mean ± standard deviations, while categorical data were represented as frequencies or proportions.

2.3. KEGG pathway and GO enrichment analysis

The functional roles of the differentially expressed genes were thoroughly assessed through GO enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis. The GO analysis encompassed terms related to cellular component (CC), molecular function (MF), and biological process (BP). Enrichment analyses were conducted using the R package (Cluster Profiler) and the DAVID online tool. Both p-values and false discovery rate (FDR) values were found to be less than 0.05, indicating statistical significance.

2.4. PPI network construction and module screening

The differentially expressed genes identified between LND < 0.12 and LND ≥ 0.12 were queried against the STRING database (http://www.string-db.org/) to elucidate protein-protein interaction details. Subsequently, Cytoscape 3.7.0 software was employed to build and visualize the protein-protein interaction (PPI) network. Key modules and genes within this network were identified using the Molecular Complex Detection (MCODE) plug-in, with selection criteria of MCODE score and node counts exceeding 5. A significance threshold of P ≤ 0.05 was applied to denote statistically significant differences.

2.5. Plasmids, antibodies and siRNA

Plasmids containing CTCFL-Flag, AHR-Myc, CTCFL-EGFP, AHR-mCherry were constructed by PCR. These sequences were amplified by PCR(P520-01, Vazyme) using human cDNA made by priming KYSE150 total RNA, and assembled into the expression vectors (pCS2) with tags (2×Myc, 3×Flag, EGFP, mCherry). Plasmid were generated with a one-step cloning kit (Vazyme, C115-02), and the sequences were verified by colony genetic identification and Sanger sequencing. For immunoblotting and immunoprecipitation detection of endogenous CTCFL (ThermoFisher, MA5-49545), AHR (ThermoFisher, 14-9854-82), GSTA1(abcam, ab259727), GSTA1(abcam, ab211449), GSTM3(abcam, ab229858) and GSTM4(absin, abs117890) against specific to homo sapiens. The siRNA were designed from GenePharma (Suzhou, China).

2.6. CUT&Tag library preparation and data analysis

In order to find out which genes were transcriptionally activated by CTCFL or AHR, we used the CUT&Tag Assay Kit (N259, Novoprotein, China) to capture the CTCFL or AHR binding site. The experimental procedure was performed according to the manufacturer's instructions. Briefly, 1×10⁵ ESCC cells were prepared and immobilized on ConA Binding magnetic beads. The beads were incubated with anti primary antibody and then with anti-rabbit IgG secondary antibody. The cell-fixed magnetic beads were washed twice, and the pA/G-Tnp Pro transposon was added, followed by fragmentation of the target DNA. DNA was extracted and a CUT&Tag library was constructed according to the manufacturer's instructions. The analysis of CUT&Tag data followed a rigorous methodology. Initially, we conducted a comprehensive quality control assessment to ensure the integrity of the data. Subsequently, the data were aligned to the reference genome employing Bowtie2, followed by peak calling using MACS2 to pinpoint DNA binding sites accurately. Functional annotation was then performed utilizing GO and KEGG databases, followed by a differential analysis facilitated by DESeq2. The results were visually represented using IGV for clarity, aiding in the interpretation of biologically significant peaks.

2.7. In vitro droplet assay

The recombinant fusion proteins were appropriately diluted using a buffer composed of 25mM Tris-HCl and 0.5mM DTT, pH 7.4, supplemented with 5% PEG 3000. Specifically, AHR-mCherry protein at a concentration of 20µM was diluted in the buffer and applied onto a coverslip containing CTCFL-EGFP protein also at 20µM. The mixture was then incubated at 37°C for 10 minutes, and droplets were visualized using either phase contrast microscopy or confocal microscopy (LSM880, Zeiss). Quantification of droplets was performed using ImageJ software.

2.8. Flourescence recovery after photobleacing (FRAP)

For fluorescence recovery after photobleaching (FRAP) analysis of droplets within living cells, the entire droplets were designated as the region of interest (ROI) to ensure precise quantification of CTCFL-EGFP and AHR-mCherry's propensity for phase separation into droplets. CTCFL-EGFP granules were photobleached using a 488-nm laser, while AHR-mCherry granules were photobleached with a 561-nm laser, each for 1 µm at a frame rate. Subsequent frames captured after photobleaching corresponded to 3 seconds of recovery time. Each data point presented reflects the average and standard deviation of fluorescence intensities across ten unbleached (control) or ten photobleached (experimental) droplets, ensuring statistical robustness.

2.9. Immunofluorescence

4% paraformaldehyde was used to fix cells for 5 min and followed by three PBS washes. Permeabilize using 0.2% Triton X-100 for 10 minutes and rinse three times with PBS. Block with serum from the same host as the secondary antibody for 30 minutes, followed by three PBS washes. Incubate with the primary antibody overnight at 4 degrees Celsius, with subsequent three PBS washes for optimal efficacy. Apply the secondary antibody at room temperature for 2 hours or at 37 degrees Celsius for 1.5 hours, followed by three PBS washes. Next, stain the nuclei with DAPI and visualize the fluorescent signal directly. Remove PBS, seal the slide with glycerol, and protect the edges with nail polish for preservation.

2.10. Co-immunoprecipitation (co-IP) and Western blotting

Co-IP and Western blotting were performed as previously described.²³

2.11. Animals and Carcinogen Treatment.

Five-week-old female C57BL/6 mice were purchased from Gempharmatech Co., Ltd. (Jiangsu, China). All mice were fed a basal diet (Shenzhen Topbiotech Co., Ltd., Shenzhen, China) (NO. SYXK, 2020 − 0230) and allowed free access to deionized water. All mice were reared in a specific pathogen-free (SPF) laboratory, and all the animal procedures and experiments conducted in this study were approved by the Animal Ethics Committee of Shenzhen Topbiotech Co., Ltd. (Permit NO. TOP-IACUC-2022-0205). We used 4-nitroquinoline 1-oxide (4NQO) from Sigma Aldrich (St. Louis, MO, USA) to induce esophageal cancer in a working concentration of 100 µg/mL, and stored at 4°C. Drinking water containing 4NQO was freshly prepared every week in total 12 weeks. Mice were euthanatized with a 4% paraformaldehyde solution at the 18th weeks or obtain tumor tissue for organoids culture, respectively.

2.12. glutathione (GSH) and reactive oxygen species (ROS) assays

The intracellular GSH and ROS levels were assessed using a GSH colorimetric assay kit. All of these reagents were purchased from Solarbio (Beijing, China) and used according to the manufacturer’s instructions.

3.1. LND is a critical factor in determining the prognosis of ESCC

In total 8716 eligible esophageal cancer patients were analyzed. The value of consecutive lymph nodes was determined and stratified using R code. As list in Table 1, the average ages at the patients alive or dead were 62.21 ± 9.46 and 64.15 ± 10.02, respectively. Most of the patients in the present cohort were married, white, male, received surgery, radiation therapy and chemotherapy. In the LND (Stratified) < 0.12 (69.0%) subgroup, nearly half of patients were alive (2701, 45.2%), whereas in the LND (Stratified) ≥ 0.12 (31.0%) subgroup, majority of patients were dead (2279, 83.2%). The selection strategy for eligible patients was shown in Fig. 1A.

Table 1

Patient characteristics
Variables	Alive	Dead	p-value
Total	3162(36.3%)	5554(63.7%)
Age	62.21 ± 9.46	64.15 ± 10.02
Sex
Male	2597(35.9%)	4631(64.1%)	0.136
Female	565(38.0%)	923(62%)
Marital status
Married	2150(37.4%)	3606(62.6%)	< 0.001
Unmarried	455(37.6%)	756(62.4%)
SWD	444(30.3%)	1020(69.7%)
Unknown	113(39.6%)	172(60.4%)
Race
White	2854(36.6%)	4951(63.4%)	< 0.001
Black	134(28.4%)	338(71.6%)
Other	160(37.9%)	262(62.1%)
Unknown	14(82.4%)	3(17.6%)
First malignant primary indicator
Yes	2718(37.0%)	4622(63.0%)	0.001
No	444(32.3%)	93267.7%)
LND
< 0.12	2701(45.2%)	3275(54.8%)	< 0.001
≥ 0.12	461(16.8%)	2279(83.2%)
Surgery
Yes	3084(37.9%)	5045(62.1%)	< 0.001
No/Unknown	78(13.3%)	509(86.7%)
Radiation
Yes	1809(34.8%)	3388(65.2%)	0.001
No/Unknown	1353(38.4%)	2166(61.6%)
Chemotherapy
Yes	1990(34.9%)	3719(65.1%)	< 0.001
No/Unknown	1172(39.0%)	1835(61.0%)
AJCC
I	1203(56.8%)	914(43.2%)	< 0.001
II	1166(37.6%)	1935(62.4%)
III	653(23.9%)	2078(76.1%)
IV	140(18.3%)	627(81.7%)
N
N0	2204(47,7%)	2415(52.3%)	< 0.001
N1	753(20.4)	2940(79.6%)
N2	161(55.9%)	127(44.1%)
N3	44(37.9%)	72(62.1%)

According to the univariate cox analysis, we found that age, marital status, race, sex, surgery, radiation, chemotherapy, LND consequent and AJCC stage had significant impact on the survival of esophageal cancer patients. We further included all of these variables in the multivariate cox regression. We found the age of diagnosis was a significant variable in mutivariate analysis, age at diagnosis (hazard ratio [HR] = 1.020, 95% confidence interval [CI] = 1.017–1.023, p < 0.001). Thus, age might be an important factor affecting the esophageal cancer patients' survival. Indeed, marital status (unmarried vs married: HR = 1.141, 95%CI = 1.053–1.238, p = 0.001; separated, widowed or divorced [SDW] vs married: HR = 1.161, 95%CI = 1.081–1.246, p < 0.001), race (black vs white, HR = 1.278, 95%CI = 1.142–1.430, p < 0.001), first malignant primary indicator (no vs yes: HR = 1.150, 95%CI = 1.069–1.237, p < 0.001), surgery (no/unknown vs yes: HR = 1.696, 95%CI = 1.531–1.879, p < 0.001), chemotherapy (no/unknown vs yes: HR = 1.345, 95%CI = 1.257–1.439, p < 0.001), AJCC stage (II vs I: HR = 2.164, 95%CI = 1.983–2.362, p < 0.001; III vs I: HR = 3.366, 95%CI = 3.039–3.727, p < 0.001; IV vs I: HR = 3.779, 95%CI = 3.332–4.286, p < 0.001) were significantly correlated with the risk of esophageal cancer. R software was used to determine the following optimal cutoffs for the LND stratified: <0.12, ≥ 0.12 (Fig. 1B). Importantly, LND stratified was also related to an increased risk of esophageal cancer (LND ≥ 0.12 vs LND < 0.12: HR = 1.629, 95%CI = 1.521–1.745, p < 0.001)(Table 2).

Table 2

Univariate and multivariate cox regression
Variables	Univariate analysis			Multivariate analysis
Variables	HR	95%CI	p-value	HR	95%CI	p-value
Age at diagnosis	1.017	1.014–1.020	< 0.001	1.020	1.017–1.023	< 0.001
Marital status
Married	Reference			Reference
Unmarried	1.061	0.981–1.148	0.137	1.141	1.053–1.238	0.001
SWD	1.222	1.140–1.310	< 0.001	1.161	1.081–1.246	< 0.001
Unknown	1.033	0.886–1.203	0.680	1.141	0.979–1.330	0.092
Race
White	Reference			Reference
Black	1.292	1.157–1.442	< 0.001	1.278	1.142–1.430	< 0.001
Other	1.004	0.887–1.137	0.948	0.948	0.836–1.074	0.378
Unknown	0.216	0.070–0.671	0.008	0.221	0.071–0.686	0.009
Sex
Male	Reference			Reference
Female	0.970	0.904–1.041	0.398		...
First malignant primary indicator
Yes	Reference			Reference
No	1.223	1.140–1.313	< 0.001	1.150	1.069–1.237	< 0.001
Surgery
Yes	Reference			Reference
No/Unknown	2.840	2.590–3.114	< 0.001	1.696	1.531–1.879	< 0.001
Radiation
Yes	Reference			Reference
No/Unknown	0.775	0.734–0.819	< 0.001		...
Chemotherapy
Yes	Reference			Reference
No/Unknown	0.762	0.720–0.806	< 0.001	1.345	1.257–1.439	< 0.001
LND (Stratified)
< 0.12	Reference			Reference
≥ 0.12	2.648	2.507–2.797	< 0.001	1.629	1.521–1.745	< 0.001
AJCC Stage
I	Reference			Reference
II	1.982	1.831–2.144	< 0.001	2.164	1.983–2.362	< 0.001
III	3.541	3.271–3.833	< 0.001	3.366	3.039–3.727	< 0.001
IV	4.502	4.062–4.989	< 0.001	3.779	3.332–4.286	< 0.001

We further stratified the esophageal cancer patients according to LND stratification. We first analyzed all of the variables both in the univariate cox regression analysis and multivariate cox regression. As listed in (Table 3).

Table 3

Multivariate cox regression analysis in the LND Stratified
Variables	LND (Stratified) < 0.12			LND (Stratified) ≥ 0.12
Variables	HR	95%CI	p-value	HR	95%CI	p-value
Age at diagnosis	1.025	1.021–1.029	< 0.001	1.012	1.008–1.016	< 0.001
Marital status
Married	Reference			Reference
Unmarried	1.098	0.986–1.222	0.087	1.169	1.034–1.322	0.013
SWD	1.126	1.027–1.236	0.012	1.179	1.056–1.317	0.003
Unknown	1.148	0.949–1.389	0.156	1.181	0.912–1.530	0.207
Race
White	Reference			Reference
Black	1.423	1.233–1.643	< 0.001
Other	1.084	0.927–1.268	0.312		...
Unknown	0.228	0.057–0.911	0.037
Sex
Male	Reference			Reference
Female		...			...
First malignant primary indicator
Yes		Reference			Reference
No	1.278	1.166–1.401	< 0.001		...
Surgery
Yes	Reference			Reference
No/Unknown	1.747	1.412–2.163	< 0.001	1.669	1.484–1.877	< 0.001
Radiation
Yes	Reference			Reference
No/Unknown	0.787	0.695–0.892	< 0.001		...
Chemotherapy
Yes	Reference			Reference
No/Unknown	1.299	1.141–1.479	< 0.001	1.821	1.599–2.074	< 0.001
AJCC Stage
I	Reference
II	1.931	1.758–2.120	< 0.001		Reference
III	3.129	2.789–3.510	< 0.001	1.497	1.332–1.682	< 0.001
IV	2.985	2.471–3.605	< 0.001	1.766	1.533–2.035	< 0.001

In the LND < 0.12 cohort, age at diagnosis (HR = 1.025, 95%CI = 1.021–1.029, p < 0.001), marital status (unmarried vs married was no significant difference whereas widowed or divorced [SDW] vs married: HR = 1.126, 95%CI = 1.027–1.236, p = 0.012), race (black vs white, HR = 1.423, 95%CI = 1.233–1.643, p < 0.001), first malignant primary indicator (no vs yes: HR = 1.278, 95%CI = 1.166–1.401, p < 0.001), surgery (no/unknown vs yes: HR = 1.747, 95%CI = 1.412–2.163, p < 0.001), chemotherapy (no/unknown vs yes: HR = 1.299, 95%CI = 1.141–1.479, p < 0.001) and AJCC stage (II vs I: HR = 1.931, 95%CI = 1.758–2.120, p < 0.001; III vs I: HR = 3.129, 95%CI = 2.789–3.510, p < 0.001; IV vs I: HR = 2.985, 95%CI = 2.471–3.602, p < 0.001) were significantly correlated with the risk of esophageal cancer.

In the LND ≥ 0.12 cohort, age at diagnosis (HR = 1.012, 95%CI = 1.008–1.016, p < 0.001), marital status (unmarried vs married: HR = 1.169, 95%CI = 1.034–1.322, p = 0.013; separated, widowed or divorced [SDW] vs married: HR = 1.179, 95%CI = 1.056–1.317, p = 0.003), surgery (no/unknown vs yes: HR = 1.669, 95%CI = 1.484–1.877, p < 0.001), chemotherapy (no/unknown vs yes: HR = 1.821, 95%CI = 1.599–2.074, p < 0.001) and AJCC stage (III vs II: HR = 1.497, 95%CI = 1.322–1.682, p < 0.001; IV vs II: HR = 1.766, 95%CI = 1.533–2.035, p < 0.001) were significantly increased the risk of esophageal cancer. Patients in LND < 0.12 subgroup were in a lower burden of deaths and hazard ratio than in LND ≥ 0.12 subgroup deaths and hazard ratio during the study period (Fig. 1C).

3.2. Identification of Biomarkers of Lymph Node Metastasis in Esophageal Squamous Cell Carcinoma

In this study, we conducted a comprehensive analysis using advanced computational methods to examine the key roles and prognostic significance of genes between patient in N0 or NX (lymph node metastasis). We obtained ESCC datasets from TCGA, comprising 77 ESCC tumor samples. Data processing and identification of differentially expressed genes (DEGs) were performed using R software packages. Out of 16,190 genes analyzed, 382 genes met the screening criteria of this study (P < 0.05, qValue < 0.2, |log2FC>|1.0), including 168 upregulated genes and 214 downregulated genes.. (Fig. 2A). To investigate the function of the identified genes, we performed functional enrichment analysis on the differentially expressed genes (DEGs). The results indicated that pathways related to glutathione metabolism were enriched in both Gene Ontology Biological Processes (GO BP) and the Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis (Fig. 2B and 2C). Following that, we developed a genetic risk score model for predicting prognosis in lymph node metastasis samples of esophageal squamous cell carcinoma (Fig. 2D). This model was constructed using 26 candidate hub genes associated with prognosis, identified through a unique stepwise Cox regression analysis. We performed further analysis using multiple stepwise Cox regression to assess their impact on patient survival status, seven of these hub genes were found to be independent predictors of lymph node metastasis in ESCC (Fig. 2E). The survival analysis was performed to assess the predictive ability of the prognostic model, which indicated that patients in the high-risk subgroup exhibited significantly poorer overall survival (OS) compared to those in the low-risk subgroup (Fig. 2F). Subsequently, a total of 77 LUAD patients were divided into train or test groups, each group was stratified into low-risk and high-risk subgroups based on the median risk score. Consistent with previous findings, the high-risk subgroup demonstrated significantly reduced OS relative to the low-risk subgroup (Fig. 2G and 2J). To further validate the prognostic efficacy of the seven biomarkers, a time-dependent ROC analysis was conducted, revealing moderate diagnostic performance (Fig. 2H and 2K). An expression heatmap and survival status plot of patients, incorporating seven genes, were generated for both subgroups (Fig. 2I and 2L). These findings underscore that the prognostic model possesses robust sensitivity and specificity.

3.3. The transcription factors CTCFL and AHR collaboratively regulate Glutathione S-Transferase (GSTs)-mediated lymph node density.

To investigate the regulatory mechanisms underlying lymph node metastasis in esophageal cancer, we delved into the downstream pathways associated with this process. Among the seven identified hub genes, CTCFL and AHR were notable as transcription factors, implying that lymph node metastasis in esophageal squamous cell carcinoma patients may be driven by transcriptional regulatory mechanisms. (Fig. 3A). Notably, CTCFL and AHR, known for their elevated expression across various tumor types (Fig. 3B). Besides, analysis demonstrated a significant positive correlation in patients with esophageal squamous cell carcinoma (ESCC) (Fig. 3C). Given the consistent expression and functional patterns of CTCFL and AHR, we endeavored to elucidate their intrinsic relationship. To investigate potential interactions among these two transcription factors promoting lymph node density (LND), we employed a protein-protein interaction docking prediction model merging the three-dimensional structures of Alpha Fold2 and ZDOCK docking. Our findings revealed a stable interaction between CTCFL and AHR, with a binding free energy of 32.8 kcal/mol. This stability is facilitated by five hydrogen bond connections and one halogen bond connection between the two proteins (Fig. 3D).

Following that, we conducted Co-IP experiments to validate the interaction between CTCFL and AHR (Fig. 3E). Subsequently, immunofluorescence staining confirmed the colocalization of CTCFL and AHR, a finding further supported by overexpression using distinct tags (Fig. 3F). It appears that these two factors interact to execute specific biological functions. To delve into the regulatory role of their interaction in the glutathione metabolism pathway, we considered the close relationship between esophageal squamous cell carcinoma's lymph node density and GSTs expression.

Hence, we conducted CUT&Tag sequencing analysis on CTCFL, AHR, and their combined form to investigate their role in activating downstream genes within this pathway. The data revealed that CTCFL or AHR individually can bind to the transcription initiation regions of GSTA1, GSTA2, GSTM3, and GSTM4 (Fig. 3G). Surprisingly, upon interaction between CTCFL and AHR, there was a notable enhancement in their binding to the promoter regions of GSTs (Fig. 3G).

3.4. CTCFL and AHR condensates enrich the transcription apparatus to promote GSTs transactivation.

It's notable that nuclei in ESCC cells displayed a droplet-like distribution of co-localized CTCFL and AHR staining, suggesting the self-aggregation capability of these proteins. To delve into the structural basis of CTCFL and AHR condensates, we employed IUPred2 and PONDR to analyze their amino acid sequences. The results uncovered a classic intrinsically disordered region (IDR) from amino acids 37–252 within CTCFL protein, but the IDRs of AHR is not obvious (Fig. 4A and 4B). To validate the formation of CTCFL and AHR condensates, we conducted overexpression and purification of EGFP-tagged CTCFL and mCherry-tagged AHR proteins in HEK293T cells. Fluorescence phase contrast microscopy revealed that at 37°C, CTCFL exhibited protein concentration-dependent condensation-forming ability, whereas AHR did not. However, the fusion proteins of CTCFL and AHR exhibited droplet-forming ability (Fig. 4C).

To assess whether CTCFL or AHR demonstrate liquid-like properties within cells, we conducted in vivo experiments using the fluorescence recovery after photobleaching (FRAP) assay. The results indicated rapid recovery of CTCFL and the fusion proteins of CTCFL and AHR in solution, while AHR protein alone did not exhibit this ability. This suggests the significance of the intrinsically disordered region (IDR) domain for CTCFL's phase separation potential. Moreover, the fusion of CTCFL with AHR enabled the latter to recover after bleaching, emphasizing the collaborative role in this process (Fig. 4D). The findings above indicate that CTCFL and AHR interact through phase separation, with AHR's phase-separation ability contingent on CTCFL. Additionally, the fusion protein of CTCFL and AHR notably enhances their capacity to transcribe downstream genes.

We next employed siRNAs to downregulate CTCFL, AHR, and both of two TFs in KYSE150 esophageal squamous cell carcinoma cells. The combined Wound Healing and Transwell experiments demonstrated that CTCFL along with AHR collectively governs the migration capability of esophageal cancer cells (Fig. 4E and 4F).

3.5. GSTs impact LND in esophageal squamous cell carcinoma by controlling Reactive Oxygen Species (ROS) levels.

The results above indicate that CTCFL undergoes liquid-liquid phase separation, leading to the formation of droplets with AHR and the aggregation into a CTCFL-AHR complex. AHR likely functions as a transcriptional regulator of CTCFL, enhancing its transcriptional activity. For instance, AHR facilitates CTCFL in transcribing GSTs (Fig. 5A).

To validate the regulatory role of CTCFL and AHR in the transcription of GSTs, we employed siRNA to knock down CTCFL, AHR, and the combination of CTCFL with AHR in KYSE150 cells. The results revealed that reducing CTCFL expression led to decreased expression levels of GSTA1, GSTA2, GSTM3, and GSTM4 within the GST family, resulting in varying degrees of transcriptional inhibition. Knocking down AHR individually showed minimal impact. However, when both transcription factors were downregulated, there was a dramatically inhibiting in GST expression levels (Fig. 5B). The protein expression levels of GSTs corresponded with the observed transcriptional trends (Fig. 5C). To delve into the mechanism by which GSTA1, GSTA2, GSTM3, and GSTM4 regulate lymph node density (LND), we constructed a protein-protein interaction (PPI) network using Cytoscape software. This network integrated 1065 nodes sourced from the STRING database. Leveraging the MODE tool, we analyzed the co-expression network to identify potential key modules. The primary module obtained genes in pivotal subgroups enriched in processes related to glutathione metabolism and transferase pathways (Fig. 5D), Suggesting interactions between GSTA1, GSTA2, GSTM3, and GSTM4. To validate their direct interactions, we investigated the relationships between these four transsulfurases in KYSE150 cells. GSTA1-GSTM3 and GSTA2-GSTM4 exhibited co-localization with each other (Fig. 5E and 5F). Indeed, these proteins demonstrated potential co-localization and played essential roles in the transsulfurization pathway of glutathione synthesis.

Based on previous bioinformatics analyses, glutathione metabolism was notably enriched in the high LND group. GSTs serve as pivotal roles in this metabolic pathway. The levels of glutathione (GSH) and its biosynthetic capacity are influenced by the enzymatic activity of GSTs, consequently impacting tumor malignancy through the regulation of ROS levels.^24–26 Therefore, we assessed glutathione (GSH) levels (Fig. 5G) and reactive oxygen species (ROS) levels (Fig. 5H) in KYSE 150 cells by overexpressing GSTA1, GSTA2, GSTM3, and GSTM4 in AHR, CTCFL, AHR combined with CTCFL knockdown cell lines, respectively. The findings revealed that knockdown of AHR and CTCFL significantly reduced GSH levels while increasing ROS levels. Supplementing any of these four enzymes individually could partially restore GSH content and reduce ROS levels. However, supplementing a single type of GST alone was insufficient to normalize ROS levels, suggesting a comprehensive regulation of ROS levels by GSTs.

These results highlight the role of CTCFL-AHR in regulating GST expression in esophageal cancer cells and mediating ROS changes, which in turn influence the lymph node density of ESCC.

3.6. CTCFL and AHR collaboratively activate GSTA1 and GSTA2, promoting LND of ESCC in mice.

To confirm lymph node density (LND) in esophageal cancer relies on CTCFL, AHR and GSTs in vivo, we established a spontaneous ESCC model in p53-/-, ED-L2cre mice and generated organoids from mouse esophageal tumor tissues (Fig. 6A and 6B). These were then compared with organoids from low-LND mice (Fig. 6C). Remarkably, the organoids from high-LND mice exhibited clear growth advantages (Fig. 6D), accompanied by a significant upregulation in protein expression levels of CTCFL, AHR and GSTs (Fig. 6E). These indicate that the transcriptional factors CTCFL and AHR regulate GSTs expression, thereby influencing the lymph node density in ESCC.

It is well recognized that lymph node metastasis is one of the most important prognostic factors for cancer patients.²⁷ Increasing studies have shown that lymph node ratio (LNR is the ratio of the number of positive lymph nodes to the number of total lymph nodes examined) or lymph node density (LND) has been correlated with the prognosis¹³. LNR or LND has been found to be more reliable than the AJCC N stage in predicting gastric cancer prognosis.^28–31 The study by Han et al. showed that a prognostic scoring system containing lymph node ratio could predict the survival ratio of IIIA-N2 patients after surgery and postoperative chemotherapy, and the lymph node ratio might be a useful complement to TNM staging in IIIA-N2 patients.³² However, correlation between lymph node density and esophageal cancer patients' survival was still unclear. Thus, our study indicates that lymph node density is an prognostic factor for the survival of esophageal cancer patients for the first time. Therefore, our study elucidates the affect of LND in ESCC patients’ prognosis via clinical analysis and validation of molecule phenotype.

We included 8716 cases of esophageal cancer from the SEER database. The age of diagnosis was a risk factor for patients with esophageal cancer (HR = 1.020, 95%CI = 1.017–1.023, p < 0.001). In another study regarding distant organ metastasis in esophageal cancer, age was also proved to be a prognostic factor (HR = 1.016, 95%CI = 1.012–1.020, p < 0.001).³³ Esophageal cancer is a hostile disease with poor prognosis and the surgery is the best treatment option for esophageal cancer patients till now.^34,35 After we included LND as a stratified variable in the model, whether received surgery was found to be a risk factor for survival in patients with esophageal cancer (no / unknown vs yes: HR = 1.696, 95%CI = 1.531–1.879, p < 0.001). Radiotherapy plays an important role in the comprehensive treatment of esophageal cancer, and has been recognized as a treatment which can improve the therapeutic outcome.³⁴ Radiotherapy and chemotherapy can improve overall survival, yet might also enhance the risk of postoperative death in patients with locally advanced resectable esophageal cancer.^24,36 Therefore, in our study, whether received radiation therapy was excluded from multivariate analysis. Importantly, chemotherapy was found to significantly affect esophageal cancer patients' survival, failure to receive chemotherapy was dangerous for esophageal cancer patients. Finally, compared with AJCC I, the HR in higher AJCC patients was also constantly increasing, indicating that the higher the AJCC level, the poor prognosis for the patient's survival. Furthermore, we used R software to divide the continuous LND into two groups, with 0.12 as the cutoff point. Compared with the LND < 0.12 group, the group with higher lymph node density was indeed a risk factor for esophageal cancer patients (LND ≥ 0.12 vs LND < 0.12: HR = 1.629, 95% CI = 1.521–1.745, p < 0.001). So far, the prognostic significance of lymph node density in esophageal cancer has not been fully understood. Therefore, assessing the value of LND as a prognostic factor for esophageal cancer is a meaningful attempt. In addition, we further performed subgroup analysis with LND as the stratification. We also performed a subgroup analysis using LND as a stratification, and identified different risk factors for esophageal cancer patients in different LND subgroups.

Lymph node metastasis plays a crucial role in the treatment and prognosis of esophageal squamous cell carcinoma (ESCC), and how to prevent which is particularly crucial in the early intervention process of patients. The role implications of glutathione metabolism in relation to ESCC have remained unexplored thus far. In this study, we undertook a comprehensive transcriptome analysis encompassing both high-LND and low-LND samples derived from TCGA ESCC patients. The findings from this analysis have unveiled biophysical properties of CTCFL and AHR condensate as a pivotal orchestrator of lympha node density. One open question in this study pertains to the mechanism by which glutathione metabolic process, as well as the role played the specific enzymes in this process. Our investigation has demonstrated that the interation of GSTA1, GSTA2, GSTM3 and GSTM4 can heighten lymphatic metastasis risk by inhibiting ROS level across diverse ESCC models.

In conclusion, our study suggests lymph node density has a significant correlation with poor disease-specific survival for ESCC patients. Importantly, unlike the LND < 0.12 group, the group with higher lymph node density is a prognostic factor for ESCC whose LND ≥ 0.12. The dysregulated expression of CTCFL and AHR leads to decreased ROS levels mediated by GSTs, thereby expediting process of lymph node density (LND) in ESCC.

Authorship contributions: Conceptualization, H.-R.Z.; methodology, H.-R.Z. and R.W.; software, H.-R.Z. and H.-H.Z; resources, H.-R.Z. and R.W.; data curation, P.-G.G. and R.W.; writing—original draft preparation, P.-G.G..; writing-original draft, H.-R.Z. and H.-H.Z.; writing review and editing, H.-R.Z. ; supervision, H.-R.Z.. All authors have read and agreed to the published version of the manuscript.

Funding: No.

Data availability

The clinical data of ESCC patients can gain from the Surveillance, Epidemiology, and End Results (SEER) (https://seer.cancer.gov/) and the Cancer Genome Atlas (TCGA) (https://www.cancer.gov/). The RNA sequencing data of ESCC patients can gain from The Cancer Genome Atlas (TCGA) (https://www.cancer.gov/). The other datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Ethics Declarations

Ethics approval and consent to participate

This study was performed with full institutional ethical approval via the Meizhou people’s hospital Institutional Review Board. Approval reference number TOP-IACUC-2022-0205.

Competing interests.

The authors declare no competing interests.

Consent for publication statement

Not applicable.

Ai D, Chen Y, Liu Q, Deng J, Zhao K. The effect of tumor locations of esophageal cancer on the metastasis to liver or lung. J Thorac Dis. 2019; 11: 4205-10.10.21037/jtd.2019.09.67
Shi Y, Zhang B, Feng X, Qu F, Wang S, Wu L, et al. Apoptosis and autophagy induced by dvdms-pdt on human esophageal cancer eca-109 cells. Photodiagnosis Photodyn Ther. 2018; 24: 198-205.10.1016/j.pdpdt.2018.09.013
Hur C, Miller M, Kong CY, Dowling EC, Nattinger KJ, Dunn M, et al. Trends in esophageal adenocarcinoma incidence and mortality. Cancer. 2013; 119: 1149-58.10.1002/cncr.27834
Torre LA, Bray F, Siegel RL, Ferlay J, Lortet-Tieulent J, Jemal A. Global cancer statistics, 2012. Ca: A Cancer Journal for Clinicians. 2015; 65: 87-108.10.3322/caac.21262
Siegel RL, Miller KD, Jemal A. Cancer statistics, 2019. Ca: A Cancer Journal for Clinicians. 2019; 69: 7-34.10.3322/caac.21551
Biere SSAY, van Berge Henegouwen MI, Maas KW, Bonavina L, Rosman C, Garcia JR, et al. Minimally invasive versus open oesophagectomy for patients with oesophageal cancer: a multicentre, open-label, randomised controlled trial. Lancet (London, England). 2012; 379: 1887-92.10.1016/S0140-6736(12)60516-9
Maas KW, Cuesta MA, van Berge Henegouwen MI, Roig J, Bonavina L, Rosman C, et al. Quality of life and late complications after minimally invasive compared to open esophagectomy: results of a randomized trial. World J Surg. 2015; 39: 1986-93.10.1007/s00268-015-3100-y
Moss A, Bourke MJ, Hourigan LF, Gupta S, Williams SJ, Tran K, et al. Endoscopic resection for barrett's high-grade dysplasia and early esophageal adenocarcinoma: an essential staging procedure with long-term therapeutic benefit. The American Journal of Gastroenterology. 2010; 105: 1276-83.10.1038/ajg.2010.1
Zhang Y. Epidemiology of esophageal cancer. World J Gastroenterol. 2013; 19: 5598-606.10.3748/wjg.v19.i34.5598
Pech O, May A, Manner H, Behrens A, Pohl J, Weferling M, et al. Long-term efficacy and safety of endoscopic resection for patients with mucosal adenocarcinoma of the esophagus. Gastroenterology. 2014; 146: 652-60.10.1053/j.gastro.2013.11.006
Berger AC, Sigurdson ER, Levoyer T, Hanlon A, Mayer RJ, Macdonald JS, et al. Colon cancer survival is associated with decreasing ratio of metastatic to examined lymph nodes. Journal of Clinical Oncology : Official Journal of the American Society of Clinical Oncology. 2005; 23: 8706-12
Peschaud F, Benoist S, Julié C, Beauchet A, Penna C, Rougier P, et al. The ratio of metastatic to examined lymph nodes is a powerful independent prognostic factor in rectal cancer. Ann Surg. 2008; 248: 1067-73.10.1097/SLA.0b013e31818842ec
Ebrahimi A, Clark JR, Zhang WJ, Elliott MS, Gao K, Milross CG, et al. Lymph node ratio as an independent prognostic factor in oral squamous cell carcinoma. Head & Neck. 2011; 33: 1245-51.10.1002/hed.21600
Healy MA, Morris AM, Abrahamse P, Ward KC, Kato I, Veenstra CM. The accuracy of chemotherapy ascertainment among colorectal cancer patients in the surveillance, epidemiology, and end results registry program. Bmc Cancer. 2018; 18: 481.10.1186/s12885-018-4405-7
Duggan MA, Anderson WF, Altekruse S, Penberthy L, Sherman ME. The surveillance, epidemiology, and end results (seer) program and pathology: toward strengthening the critical relationship. The American Journal of Surgical Pathology. 2016; 40: e94-102
Jones DP. Extracellular redox state: refining the definition of oxidative stress in aging. Rejuvenation Res. 2006; 9: 169-81
Chatterjee A, Gupta S. The multifaceted role of glutathione s-transferases in cancer. Cancer Lett. 2018; 433: 33-42.10.1016/j.canlet.2018.06.028
Makhdoomi MA, Shah IA, Bhat GA, Amin S, Lone MM, Islami F, et al. Association between gstm1 and gstt1 polymorphisms and esophageal squamous cell carcinoma: results from a case-control study in kashmir, india. Tumour Biology : The Journal of the International Society for Oncodevelopmental Biology and Medicine. 2015; 36: 2613-9.10.1007/s13277-014-2882-0
Li Y, Zhang Q, Peng B, Shao Q, Qian W, Zhang J. Identification of glutathione s-transferase omega 1 (gsto1) protein as a novel tumor-associated antigen and its autoantibody in human esophageal squamous cell carcinoma. Tumour Biology : The Journal of the International Society for Oncodevelopmental Biology and Medicine. 2014; 35: 10871-7.10.1007/s13277-014-2394-y
Oshimori N, Oristian D, Fuchs E. Tgf-β promotes heterogeneity and drug resistance in squamous cell carcinoma. Cell. 2015; 160: 963-76.10.1016/j.cell.2015.01.043
Guan X, Hu H, Chen W, Jiang Z, Liu Z, Zhao Z, et al. Comparison of long-term outcome between hemicolectomy and partial colectomy in the elderly: a large population-based study. Oncotarget. 2017; 8: 51076-85.10.18632/oncotarget.16993
Yang J, He G, Chen S, Pan Z, Zhang J, Li Y, et al. Incidence and risk factors for suicide death in male patients with genital-system cancer in the united states. European Journal of Surgical Oncology : The Journal of the European Society of Surgical Oncology and the British Association of Surgical Oncology. 2019; 45: 1969-76.10.1016/j.ejso.2019.03.022
Wu Y, Li Q, Chen X. Detecting protein-protein interactions by far western blotting. Nat Protoc. 2007; 2: 3278-84
Jaeschke H, Adelusi OB, Akakpo JY, Nguyen NT, Sanchez-Guerrero G, Umbaugh DS, et al. Recommendations for the use of the acetaminophen hepatotoxicity model for mechanistic studies and how to avoid common pitfalls. Acta Pharmaceutica Sinica. B. 2021; 11: 3740-55.10.1016/j.apsb.2021.09.023
Mak TW, Grusdat M, Duncan GS, Dostert C, Nonnenmacher Y, Cox M, et al. Glutathione primes t cell metabolism for inflammation. Immunity. 2017; 46: 675-89.10.1016/j.immuni.2017.03.019
Tang G, Li S, Zhang C, Chen H, Wang N, Feng Y. Clinical efficacies, underlying mechanisms and molecular targets of chinese medicines for diabetic nephropathy treatment and management. Acta Pharmaceutica Sinica. B. 2021; 11: 2749-67.10.1016/j.apsb.2020.12.020
Nolan T, Hands RE, Bustin SA. Quantification of mrna using real-time rt-pcr. Nat Protoc. 2006; 1: 1559-82
Garcia B, Guzman C, Johnson C, Hellenthal NJ, Monie D, Monzon JR. Trends in lymph node excision and impact of positive lymph node ratio in patients with colectomy for primary colon adenocarcinoma: population based study 1988 to 2011. Surgical Oncology. 2016; 25: 158-63.10.1016/j.suronc.2016.05.013
Lee Y, Yang P, Zhong Y, Clancy TE, Lin M, Wang J. Lymph node ratio-based staging system outperforms the seventh ajcc system for gastric cancer: validation analysis with national taiwan university hospital cancer registry. American Journal of Clinical Oncology. 2017; 40: 35-41.10.1097/COC.0000000000000110
Rodríguez Santiago JM, Muñoz E, Martí M, Quintana S, Veloso E, Marco C. Metastatic lymph node ratio as a prognostic factor in gastric cancer. European Journal of Surgical Oncology : The Journal of the European Society of Surgical Oncology and the British Association of Surgical Oncology. 2005; 31: 59-66
You H, Yang J, Liu Q, Tang L, Bu Q, Pan Z, et al. The impact of the lymph node density on overall survival in patients with wilms' tumor: a seer analysis. Cancer Manag Res. 2018; 10: 671-7.10.2147/CMAR.S163514
Liu C, Lu P, Lu Y, Xu H, Wang S, Chen J. Clinical implications of metastatic lymph node ratio in gastric cancer. Bmc Cancer. 2007; 7: 200
Ai D, Zhu H, Ren W, Chen Y, Liu Q, Deng J, et al. Patterns of distant organ metastases in esophageal cancer: a population-based study. J Thorac Dis. 2017; 9: 3023-30.10.21037/jtd.2017.08.72
Zhang N, Zhang S. Long-term effects of radiation prior to surgery and chemotherapy on survival of esophageal cancer undergoing surgery. Medicine (Baltimore). 2019; 98: e17617.10.1097/MD.0000000000017617
Jin X, Gai W, Chai T, Li L, Guo J. Comparison of endoscopic resection and minimally invasive esophagectomy in patients with early esophageal cancer. J Clin Gastroenterol. 2017; 51: 223-7.10.1097/MCG.0000000000000560
Chan KKW, Saluja R, Delos Santos K, Lien K, Shah K, Cramarossa G, et al. Neoadjuvant treatments for locally advanced, resectable esophageal cancer: a network meta-analysis. Int J Cancer. 2018; 143: 430-7.10.1002/ijc.31312

No competing interests reported.

Download PDF

Reviews received at journal
10 Sep, 2024
Reviewers agreed at journal
08 Sep, 2024
Reviewers invited by journal
18 Jul, 2024
Editor assigned by journal
15 Jul, 2024
Submission checks completed at journal
11 Jul, 2024
First submitted to journal
02 Jul, 2024

You are reading this latest preprint version

Biophysical Properties of CTCFL and AHR Condensate Regulate Glutathione S-Transferase Mediated Lymph Node Density in Esophageal Squamous Cell Carcinoma

Status:

Version 1

Abstract

Figures

1. Introduction

2. Materials and Methods