Abstract
Background/Aim: The earliest cellular and molecular biologic changes in the esophagus that lead to esophageal cancer were evaluated in a mouse model. We correlated numbers of senescent cells with the levels of expression of potentially carcinogenic genes in sorted side population (SP) cells containing esophageal stem cells and non-stem cells in the non-side population cells in the 4-nitroquinolone oxide (NQO)-treated esophagus. Materials and Methods: We compared stem cells with non-stem cells from the esophagus of mice treated with the chemical carcinogen 4-NQO (100 μg/ml) in drinking water. We also compared gene expression in human esophagus samples treated with 4-NQO (100 μg/ml media) to non-treated samples. We separated and quantitated the relative levels of expression of RNA using RNAseq analysis. We identified senescent cells by luciferase imaging of p16+/LUC mice and senescent cells in excised esophagus from tdTOMp16+ mice. Results: A significant increase in the levels of RNA for oncostatin-M was found in senescent cells of the esophagus from 4-NQO-treated mice and human esophagus in vitro. Conclusion: Induction of OSM in chemically-induced esophageal cancer in mice correlates with the appearance of senescent cells.
Esophageal and esophagogastric cancers are one of the leading causes of death in males in the United States, and the incidence in females is increasing (1-8). There has been a 7.5-fold annual increase in incidence over the years 1973-2010 (8). The overall 5-year survival ranges from 15 to 25% (4-7). Therapeutic outcomes are superior if the disease is diagnosed at an early stage (9-12). The current evidence indicates that cancer initiating cells in the esophagus and esophagogastric region reside in the esophageal stem cell (ESC) population (13, 14).
The role of the microbiome (15-17) and specific genetic mutations (18-39) in esophagus cancer are subjects of current investigation. Oxidative stress induces cell senescence in the microenvironment of the esophagus, specifically, in the stem cell niche (39-46). The cytokines produced by senescent cells (senescence associated secretory phenotype, SASP) may have a critical role in initiating esophagus cancer (47-51).
In the present study, we evaluated the earliest cellular and molecular biologic changes that were detected in sorted/purified esophageal stem cells compared to non-stem cells in 4-Nitroquinolone oxide (4-NQO)-treated mice and correlated these changes with the appearance of senescent cells in chemical carcinogen exposed mouse esophagus. We used mouse strains that respond to oral administration of the chemical carcinogen 4-NQO, which was continuously delivered in the drinking water over 16 weeks. We correlated the first time of detection of senescent cells with both molecular changes in the esophagus and the appearance of cancer. We utilized two mouse models to quantitate the time of first appearance and numbers of senescent cells (47-50) in the esophagus. We correlated these senescent cells with the first detected up-regulation of specific RNA transcripts of cancer-associated genes in subpopulations of explanted esophageal stem cells and non-stem cells. Senescent cells in the p16+/LUC mice express luciferase, which can be used as a biomarker of senescence in vivo. Luciferase is activated by senescence associated p16 and turns the luciferin substrate fluorescent, which is detectable when imaged in live mice.
A second tdTOMp16+ mouse strain (52) displays senescent cells, which express the fluorochrome red dye (tomato) when p16 is activated.
The isolation of esophageal stem cells was done using a side population method (45, 46) where single cell suspensions of the esophagus are stained with Hoechst dye 33342 and sorted by flow cytometry into what is designated as SP (side population cells) or Non-SP (non-side population cells) (Figure 1). The SP cells appear as a hook on the main body of the cells, which are the Non-SP cells. For the purpose of the presentation of the results we refer to the SP population as stem cells, since this sorted cell fraction contains over 90% of cells which form multilineage colonies in vitro and thus fulfill the criteria of esophageal stem cells (45, 46). Not all of the SP cells have been identified as stem cells but many of the SP cells are stem cells since they have all the characteristics of stem cells. The Non-SP cells are mainly differentiated cells, which have no stem cell characteristics. Removal of the esophagus from tdTOMp16+ mice and isolation of stem cells (located in the SP cells) and non-stem cells (Non-SP cells) for TOM+ (red color positive) cells allowed separation of senescent cells from each subpopulation based on the expression of p16.
Detection of senescent cells in the esophagus of p16+/LUC mice receiving 4-NQO (100 μM) in drinking water using the Lumina XR imaging system and immunochemistry. At 14 weeks, imaging revealed luciferase positive senescent cells at the gastroesophageal junction of two representative mice ocmpared to two mice receiving regular drinking water (A). The mice were sacrificed and the esophagus was removed, fixed, sectioned and stained for senescence (Beta-Gal) and Oncostatin-M (OSM). The mice treated with 4-NQO had higher levels of senescence (green arrow) and OSM (red arrow). Senescent cells which are OSM positive are shown with a yellow arrow. Quantification of cells showed that 40.0±6.5% were β−gal+, 25.7±7.2% OSM+, and 20.9±7.5% both of 1000 cells counted (B).
The results indicated that by 14 weeks of 4-NQO treatment, senescent cells in the esophagus of p16+/LUC mice were detectable by luciferase scanning. Senescent cells detected in p16+/LUC mice were correlated with numbers of both SP and Non-SP senescent cells in explanted tdTOMp16+ mouse esophagus that was exposed to 4-NQO for the same duration. We used RNAseq to identify increased levels of gene transcripts for oncostatin-M (OSM) in senescent cells in Non-SP populations of the esophagus. Furthermore, OSM was elevated in the esophagus of human esophageal cancer patients. OSM initiates a pathway that can lead to up-regulation of the oncogene C-MYC. These studies facilitate analysis of the time course and specific molecular and cellular changes during carcinogenesis in esophageal stem cells compared to other cells of the (niche) esophageal microenvironment.
Materials and Methods
Mice and animal care. C57BL/6J, p16+/LUC (50), tdTOMp16+ (52), K14E7Fancd2−/− (53), and C57BL/6 Gfp+ (44) mice were housed at 4 mice per cage for males and 5 mice per cage for females. All protocols were approved by the University of Pittsburgh Institutional Animal Care and Use Committee (IACUC). Animals were provided with standard irradiated laboratory chow and deionized water. Veterinary care was provided by the University of Pittsburgh Division of Laboratory Animal Resources.
Preparation of Gfp+ bone marrow chimeric mice. Recipient adult female (20 to 23 g) mice were irradiated to 8 Gy total body irradiation (TBI) and 24 h later were injected through the tail vein with 1×106 male GFP+ bone marrow cells (44) isolated from male GFP+ mice and made into single cell suspension as previously described. Bone marrow chimerism was confirmed at 60 days by examination of the peripheral blood of recipients. Mice deemed to be chimeric had over 50% GFP positive cells in the peripheral blood (44).
Administration of 4-NQO. Drinking water containing 4-NQO (100 μg/ml) was made available to the mice constantly for 16 weeks according to published methods (53).
Monitoring mice for induction of senescence. P16+/LUC mice were irradiated to 20 Gy to the thoracic cavity using a Varian TrueBeam Linear Accelerated (Varian Medical Systems, Palo Alto, CA, USA) and scanned weekly following intraperitoneal injection of the mice with D-luciferin (150 mg/kg, Millipore-Sigma, Burlington, MA, USA), and imaged using a Lumina XR imaging system (Caliper Life Sciences, Perkin Elmer, Waltham, MA, USA) (51).
At the time of detection of senescence by scanning and imaging of p16+/LUC mice, we sacrificed other tdTOMp16+ mice that had been similarly treated with 4-NQO for the same time duration (52). The esophagus was removed from tdTOMp16+ mice, SP and Non-SP cell populations were separated (44), and each subpopulation was analyzed for the percent of senescent cells (red color) and those simultaneously expressing biomarkers of carcinogenesis using RNAseq.
Analysis of senescent cells and biomarker positive cells in human esophagus specimens. Human esophagus tissue was dissected from tissues that were removed at the time of esophagectomy from both cancer and non-cancer patients. Tissues were collected as part of an University of Pittsburgh Institutional Review Board (IRB) approved protocol in the registry of Thoracic Surgery at the University of Pittsburgh. Normal esophagus tissue that was adjacent to esophageal cancer in esophagectomy specimens was prepared as single cell suspensions according to a previous publication (54). Esophagus tissue from other esophagectomy patients with a non-cancerous condition (achalasia) was also prepared in tissue sections and in single cell suspensions as previously described (54).
Separation of esophageal stem cells from cells of the microenvironmental niche. The methods for sorting esophageal stem cells located in the SP cells compared to non-stem cells (Non-SP) cells from single cell suspensions of freshly explanted esophagus have been published previously (45, 46). Briefly the esophagus is removed from the mouse and made into single cell suspensions by incubating the esophagus for 1 h in 0.2% type XI collagenase, dispase (grade II, 240 units), and 0.1% trypsin at 37°C. The cells were drawn through a series of syringes beginning at 20 gauge to 28 gauge and then filtered using a 45 μM filter to remove cell clumps. The cells were then incubated for 90 min at 37°C in Hoechst dye 3342 followed by staining with an anti-CD45. The cells were sorted using a flow cytometer to obtain cells positive for Hoechst dye but CD45 negative as previously described (45, 46). The SP cells were identified as shown in Figure 1.
Histological staining. Esophagus tissue specimens from mouse and human tissues were immunohistochemically stained for the detection of OSM (green color – GFP) using an FITC labelled anti-OSM antibody (sc-374039, Santa Cruz Biotechnology, Dallas, TX, USA). Senescent cells were stained for beta-gal (red color) using Cell Event Senescence Green Flow Cytometry Assay Kit for Beta-gal (Lot # 2256815, Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA), or combined imaging (yellow color) according to published methods (51). The esophagus from 4-NQO-treated mice was evaluated for the presence of tumors by excising the esophagus, fixing in 10% paraformaldehyde (Thermo Fisher Scientific), and staining with Hematoxylin and Eosin as previously published (53).
RNAseq analysis of the relative abundance of transcripts in mouse esophageal non-stem cells (non-side population cells). Analysis of the Non-SP cell population RNA was performed using RNAseq. Non-SP cells were isolated from 5 esophagi at each of two time points (pre-treatment and 14 weeks after 4-NQO treatment). RNAseq was performed as previously described (52) by Medgenome (Wilmington, DE, USA).
RNA isolation and cDNA synthesis. Total RNA was isolated from the esophagus tissue of tdTOMp16+ and C57BL/6 mice using TRIZOL Reagent (Invitrogen, Life Technologies, Thermo Fisher Scientific) as described in the manufacturer’s instructions. The concentrations of the RNA samples were determined using an Epoch microplate Spectrophotometer (Agilent, Santa Clara, CA, USA at 260/280 ratios. Two micrograms of RNA were used to synthesize cDNAs using high-capacity RNA-to-cDNA™ Kit (Thermo Fisher Scientific) following the manufacturer’s instructions.
Real Time PCR. A BioRad CFX-connect Real-Time System instrument was used for Quantitative-Polymerase Chain Reaction (qPCR). Commercially available target probes and Master mix (all from Applied Biosystems) were used in the qPCR reactions. Cycling times were 95°C for 12 min, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. Gene expression was calculated using ΔΔCT calculation (55). Detection of mouse p16 (CDKN2A), p21 (CDKN1A), p19 Collagens (1 and 3), TGF-beta, smooth muscle actin (Acta 2), CTGF, and GAPDH were achieved using specific Taqman Gene Expression Assays (Mm00438951_m1, Mm00494449_m1, Mm04205640_g11, Mm01191861_m1, Mm01192933_g1, Mm01257348_m1, g1, respectively).
RNA library construction and next-generation sequencing. Esophageal cells from tdTOMp16+ mice were separated into three groups of 1) non-4-NQO treated cells (n=3), 2) control senescent tdTOM+ (red) cells (n=3), and 3) 4-NQO-non-senescent tdTOMp16 negative cells (n=3). RNA was isolated and RNA libraries were made for RNAseq. To prevent RNA degradation, libraries for RNA-seq were generated from ribosomal RNA depleted total RNA rather than from mRNA isolated by poly-A selection. RNA was treated with DNase1 and one microgram of RNA was used for library construction using the Illumina TruSeq Stranded Total RNA Library Prep Kit (Illumina, Inc., San Diego, CA, USA) with Ribo-Zero Gold (Illumina, Inc.) per the manufacturer’s protocol. Nine cycles of PCR were used to amplify the adaptor-ligated fragments. The quality and size of the final library preparations were analyzed on Agilent TapeStation (Agilent, San Clara, CA, USA). Sequencing was performed on the Illumina NextSeq. 500 NGS platform (Illumina Inc.), generating ∼40 million paired-end 75 bp reads for each sample. The BioSample accession code for the deposited RNA-seq data to NIH is SAMN22069450.
Differential gene expression (DEG) detection and gene set enrichment analysis (GSEA) from RNA-seq data. Raw reads of the RNA-seq data were assessed for quality using FastQC (v0.11.7). Using the reverse strand-specific mode, transcript-level abundance was quantified using Kallisto (55-60) (v0.46.1) with mouse reference assembly (GRCm38) and Gencode gene annotation (v25) and summarized into gene-level using tximport (57-59) (v1.16.1). The lowly expressed genes with counts per million reads mapped (CPM) <1 were removed, and the data was normalized across all samples using Trimmed Mean of M-values (TMM) method. Significant DEGs were identified using limma voom (61) (v3.46.01) with precision weights and filtered by fold change ≥2.0 or £−2.0 and FDR-adjusted p<0.05. The R package fGSEA (v1.14.0) was used to perform GSEA using gene sets of interest (chemotaxis, senescence, and carcinogen).
Ingenuity Pathway Analysis® (IPA) with Ingenuity Knowledge Base (QIAGEN, Germantown, MD, USA) was used for pathway enrichment of DEGs.
Statistics. A two-way ANOVA followed by post hoc t tests was performed to determine the percent of red senescent cells. Radiation dose, day after irradiation, and the interaction of these factors were considered in the analysis of data. A one-way ANOVA was used in the statistical analysis of the RT-qPCR data of gene expression. For the other two group comparisons, we used two-sample t-tests or Wilcoxon rank-sum tests, where appropriate. p-Values less than 0.05 were regarded as significant. In these exploratory analyses, we did not adjust p-values for multiple comparisons (52).
Results
Detection of senescent cells in the esophagus of mice that received 4-NQO (100 μM) in drinking water. The p16+/LUC mice that received 4-NQO in drinking water for 14 weeks were scanned using the IVIS Lumina XR imaging system and showed luciferase positive senescent cells in the esophagus, particularly at the gastrointestinal junction (Figure 1A). The esophagus was removed from other p16TOMp16+ mice that had received 4-NQO (100 μM) in drinking water in parallel for 14 weeks. The esophagi were fixed in 10% PFA, sectioned, and stained for Beta Gal and OSM. All of the senescent cells were found in the Non-SP cell population. The esophagi from mice treated with 4-NQO had 40.0±6.5% senescent cells, 25.7±7.2% OSM positive cells, and 20.9±7.5% senescent cells, which were OSM positive (Figure 1B) (51, 52).
Separation of esophageal stem cell populations by the side population SP sorting method. Single cell suspensions of both explanted mouse and explanted human esophagus from esophagectomy patients were analyzed for stem cell populations referred to as SP and Non-SP population cells using flow cytometry according to published methods (44-46). The difference between the SP and Non-SP cells is seen in Figure 2A where the majority of the cells are located in the region identified as Non-SP. The SP cells are located in the region on the right side of the Non-SP cells and appear as a hook off the Non-SP cells. The SP cells have been demonstrated to contain multilineage esophageal stem cells (45). The Non-SP cells contain the differentiated cells which have no characteristics of stem cells (45). As shown in Figure 2A, mice treated with 4NQO had more SP cells than the non-treated mice. Senescence-associated beta-gal (SA-β-GAL) staining of the single cell suspensions demonstrated increased numbers of senescent cells in the esophagus of mice treated with 4-NQO (Figure 2B).
Isolation of side population (SP) cells containing esophageal stem cells and non-side population (Non-SP) cells from explanted esophagus at 14 weeks after 4-NQO treatment of p16+/LUC mice. (A) SP and Non-SP cells from groups of 5 esophagi from controls or 4-NQO-treated mice were sorted (Methods for Hoechst, propidium iodide sorting are described in reference 44). (B) Non-sorted esophageal cells from the esophagus were stained for senescence. Esophageal cells from esophagus treated with 4-NQO had increased number of senescent cells compared to cells from non-treated cells.
Analysis of gene transcripts that are increased by 4-NQO treatment in esophageal cells identifies OSM. We next evaluated the relative abundance of specific gene transcripts in explanted Non-SP esophagus cell populations using RNAseq for the spectrum of gene transcripts detected after mice were exposed to 4-NQO. Purified and sorted Non-SP cells from tdTOMp16+ mice at 14 weeks after 4-NQO treatment, which was the time point when other p16+/LUC mice that had been treated with 4-NQO showed increased luciferase activity, which is indicative of increased senescence, were analyzed using RNAseq. The tdTOMp16+ mouse esophagus revealed prominent differences in RNA transcripts (Figure 3) and the 14-week time point correlated with increased numbers of senescent cells in the esophagus. The top 20 up-regulated transcripts in the 4-NQO esophagus Non-SP cells are shown in Table I. While OSM was not one among the top 20, it was up-regulated by 4-NQO (Table II, Figure 3B). Senescent cells show up-regulation in transcripts of potentially carcinogenic genes (Table III).
Relative levels of gene expression detected using RNAseq analysis in control mouse esophagus and 4-NQO-treated mouse esophagus cells. (A) Cell populations from mice treated with or without 4-NQO at 14 weeks after beginning treatment were sorted for SP and Non-SP cells. RNA seq was performed on the Non-SP cells and heat maps of gene expression were made (n=2,400 genes. (B) Relative expression levels of top 22 genes from Panel A showing OSM at bottom. Green=elevated level, red=decreased level compared to other groups.
Top 20 differentially expressed genes in 4-NQO-treated non-side population cells compared to control.
Fold induction of genes associated with the OSM pathway in esophageal non-side population cells from 4NQO-treated mice compared to control mice.
List of up-regulated senescence genes, some of which are potentially carcinogenic and are increased in senescent cells [these genes are included in the published list of the Senescence Associated Secretory Phenotype (SASP) (82)].
We next compared the relative abundance of RNAs in sorted Non-SP cells from control and 4-NQO mice using RNAseq. The RNAseq analysis of Non-SP cells and the 4-NQO-treated esophagus revealed a clear abundance of specific genes in both populations of cells after 4-NQO treatment. There were significant differences in the numbers of up-regulated and suppressed transcripts in the mouse esophagus (Figure 3). We analyzed a panel of 400 genes and observed profound differences (Figure 3A). We then specifically focused on 22 genes that are known to be involved in carcinogenesis (Figure 3B). This analysis revealed elevated levels of OSM (Table II). Senescent cells from tdTOMp16+ mouse esophagus showed increased OSM (Figure 1B). Increased expression of other senescence-associated genes was also found in the esophagus (Table III). These results are consistent with the immunohistochemical detection of OSM in senescent cells from the 4-NQO-treated mouse esophagus (Figure 1B).
Human esophagus contains senescent cells that are OSM positive. Esophageal cancer tissue obtained from esophageal cancer or achalasia patients who underwent esophagectomy. Tissue sections from the tumors were stained for p16 to identify senescent cells and OSM. Merged P16 (green arrow) and OSM (red arrow) images showed senescent cells positive for OSM (yellow arrow) in the human esophageal tumor sections but not in the achalasia patients (Figure 4). In addition, a patient with Fanconi anemia (FA), who had esophagectomy for a mid-esophageal squamous cell carcinoma showed increased levels of OSM in the explanted esophagus (Figure 5A and B). Thus, esophageal cells from mice and humans had increased senescent cells that were also positive for OSM following 4-NQO treatment.
Oncostatin-M (OSM) positive and senescent (β-galactosidase positive) (β-gal+) cells in human esophagus from an area 5 cm. distant from the margin of excised esophageal adenocarcinoma. In situ staining of sagittal sections of human esophageal tissue from an achalasia or esophageal cancer patients was performed (green=β-gal, red=OSM, yellow=both). Of the normal cells from the esophageal cancer patient 31.0±4.0% were positive for both β-gal and OSM. No OSM+/Beta-Gal+ cells were detected in explanted tissue from an achalasia patient specimen while numerous cells were expressing both OSM and Beta-Gal staining in the normal human tissue 5 cm from an esophageal tumor. Immunohistochemistry and antibodies are described in the Methods section.
Detection of elevated levels of Oncostatin-M (OSM) in the esophagus of a FA patient with squamous cell carcinoma of the esophagus. RNA was extracted from a tissue sample from an FA esophageal cancer patient as well as from esophageal tissue from a normal human esophagus. qPCR using primers specific for OSM was performed comparing OSM gene expression from the cancer patient with gene expression from the normal human esophagus (A). Staining for OSM expression was performed in the FA esophageal tumor and normal esophagus. The FA patient had an increased expression of OSM compared to the normal human esophagus. (B) Normal esophagus is on the top and the FA cancer patient on the bottom. OSM positive cells are red and are more numerous in the FA cancer patient. Photos on right are the insets of the photos on the left at a higher magnification.
Normal human esophagus tissue obtained from the margins of human esophagectomy specimens was incubated with and without 4-NQO (100 μg/ml) for 4 days and slides were prepared using cytospin of single cell suspension. Immunohistochemistry for p16 positive cells (green arrows) and OSM positive cells (red arrows) demonstrated that the tissue incubated with 4-NQO had increased numbers of senescent cells positive for OSM (yellow arrows) (Figure 6). RNA was extracted from the normal human esophagus treated with or without 4-NQO (100 μg/ml) for 4 days and RNAseq performed. Expression of OSM and its downstream target Stat3 were significantly increased in the tissue treated with 4-NQO compared to untreated tissue while there was no change in the expression of OSMR following treatment with 4-NQO (Figure 7).
Human esophagus cells treated with 4-NQO have increased number of senescent cells expressing Oncostatin-M (OSM) compared to non-treated esophagus samples. Cells isolated from normal human esophagus taken from the margins of an esophageal tumor were incubated with or without 4-NQO for 4 days. The cells treated with 4-NQO had increased number of senescent cells expressing OSM. Green=OSM-Gfp, Red=β-gal, Yellow=both.
Increased expression of Oncostatin-M (OSM) pathway genes in 4-NQO-treated esophagus cells in vitro. Human esophagus was treated in 4-NQO (100 μM) for 4 days, then RNA was extracted and RNAseq was carried out. Fold induction of OSM, STAT-3, and OSM receptor was calculated (n=3).
Time of detection of senescent cells in the mouse esophagus correlates with migration of bone marrow cells into the esophagus. We next analyzed non transplanted mice with sex mismatched marrow from Gfp+ mice as described in Materials and Methods. Mice that had been transplanted with Gfp+ sex mismatched marrow were proven to be chimeric at day 60 by documented evidence of marrow origin Gfp+ cells in the peripheral blood. We then administered 4-NQO (100 μg/ml) through the drinking water to the chimeric mice. At 16 weeks, we evaluated the explanted esophagus for Gfp+ cells (Figure 8). There were significant numbers of Gfp+ cells in the 4-NQO-treated mouse esophagus. There was a greater percentage of Gfp+ cells in the esophagus of 4-NQO-treated mice compared to control untreated mice (Figure 8). There were 8% β-galactoside positive senescent cells in Gfp+ marrow chimeric mice that were treated with 4-NQO compared to less than 2% in control mice. The data indicate that 4-NQO treatment resulted in migration of marrow origin cells into the esophagus.
Marrow-origin cells in the esophagus of a 4-NQO-treated C57BL/6 mice that were chimeric for Gfp+ marrow shows Gfp+ cells at 14 weeks. Mice that were maintained on 4-NQO in drinking water show significant increase in Gfp+ cells (n=5) (p>0.01).
Identification of esophagus cancer in 4-NQO treated mice. The above data indicated that senescence and OSM induction were detected in mouse esophagus after 14 weeks of exposure to 4-NQO. We next evaluated the esophagus for the appearance of cancer. K14E7Fancd2−/− mice were treated with 4-NQO (10 μg/ml) in the drinking water for 16 weeks. We had to lower the concentration in the K14E7Fancd2 mice due to toxicity of the 4NQO in this mouse strain. After 16 weeks, esophageal cancer was detected in the same regions where senescent cells were identified in the p16+/LUC mice that were treated with the original 4-NQO concentration (100 μg/ml) for the 16 weeks (Figure 9). Thus, the detection of senescent cells and OSM positive cells preceded the detection of cancer in the esophagus of 4-NQO exposed mice.
Esophagus cancer detected in K14E7Fancd2−/− mice at 22 weeks after 4-NQO treatment. (A) Unstained section of the tumor (100×); (B) Hematoxylin and eosin staining of tumor in panel A (200×).
Discussion
The present data establish a correlation between chemical carcinogen induction of senescent cells in the mouse esophagus with the detection of cells that are positive for increased abundance of OSM transcripts. OSM has been reported to be involved in the early molecular biologic changes during the process of carcinogenesis in both experimental animal and human cancer (62). We have previously published that K14E7Fancd2−/− mice, which had been administered with 4-NQO in the drinking water developed multiple tumors in both the head and neck region and in the esophagus (53). This prior study did not address the earliest time of appearance of OSM positivity or senescent cells during carcinogenesis. The present study did not determine whether the appearance of senescent cells occurs prior to or after the increase in OSM. There are multiple studies showing that the effect of toxic agents including ionizing irradiation on the mouse and human esophagus involves inflammatory changes (54). Furthermore, the effects of oncogenic viruses, as well as intrinsic genetic DNA repair defects have been shown to damage the esophagus before development of esophageal cancer (63, 64). While determining the relationship between the appearance of senescent cells in the esophagus with the detection of potential biomarkers of esophagus cancer, we discovered upregulation of OSM.
Senescent cells have been reported to have both positive and negative regulatory roles in carcinogenesis (65). Whether senescent cells precede or coincide with the appearance of OSM positive cells, and whether the distal steps in the OSM pathway including genes for STAT-3, TGF-β/SMAD-3, and C-MYC are sequentially activated during esophageal carcinogenesis is currently a subject of investigation.
There is recent evidence that esophageal stem cells interact with cells of the esophageal microenvironment (niche) during exposure to toxic agents, including ionizing irradiation (61, 65-74) and photodynamic therapy (75). This interaction between the stem cells and niche also occurs during the process of carcinogenesis (63, 64, 76, 77). The mouse esophagus represents a model system for elucidation of the sequence of events during carcinogenesis (53). DNA damage repair deficiency such as that found in FA increases the magnitude of esophagus damage by toxic agents including ionizing irradiation (63, 64). Knowledge of the sensitivity of mice with a defect in DNA repair by homologous recombination such as those in FA led to studies with K14E7Fancd2−/− mice (53, 63, 64). These mice are FA models due to absence of the Fancd2 gene; they also have cytokeratin (K14) driven expression of the human papilloma virus (HPV) E7 oncogene (64). Thus, these mice are an excellent model system to study esophageal stem cell responses to a toxin such as 4-NQ. In addition, these cells do not express the Fancd2 gene, which is required for homologous recombination repair of DNA. The role of genetics, sex, and the esophageal microenvironment relative to the appearance of senescent cells during carcinogen responses is unknown. The K14E7Fancd2−/− mouse may be an excellent model to answer these questions.
Esophageal damage has been detected during radiotherapy in both experimental animal models (53) and in the evaluation of human cancers (65). We discovered that there is increased migration of marrow cells into the esophagus of 4-NQO-treated mice. These results are consistent with both inflammation, which attracts marrow origin cells, but also with the increased number of senescent cells since senescent cells secrete chemokines as part of the SASP that recruit marrow cells (65). RNAseq analysis of senescent cells identified up-regulated transcripts for genes including those for multiple cytokines, chemokines, and metallopeptidases.
The present study indicates that the chemical carcinogen 4-NQO induces the up-regulation of several gene transcripts that may be indicators of carcinogenesis. Our discovery of increased levels of OSM suggests a specific molecular biologic pathway may be involved in esophageal cancer and merits further study.
Acknowledgements
This study was supported by a grant from the NIAID/NIH U19-AI068021.
Footnotes
Authors’ Contributions
A.M., M.E., R.F., D.S. and W.H. performed the experiments; A.V., J.U. obtained the human esophagus samples during surgery; J.G. designed cell experiments and prepared the manuscript.
Conflicts of Interest
There are no conflicts of interest to declare in relation to this study.
- Received December 9, 2022.
- Revision received January 20, 2023.
- Accepted February 1, 2023.
- Copyright © 2023, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).