The past few years have brought new advances in our understanding of the molecular mechanisms underlying the development of Barrett’s esophagus and esophageal adenocarcinoma. Although knowledge of the genetic basis for these conditions has not yet translated into clinically useful biomarkers, the current pace of biomedical discovery holds endless possibilities for molecular medicine to improve the diagnosis and management of patients with these conditions. This article provides a useful conceptual basis for understanding the molecular events involved in the making of Barrett metaplasia and in its neoplastic progression, and provides a rationale for evaluating studies on the application of molecular medicine to the diagnosis and management of patients with Barrett’s esophagus and esophageal adenocarcinoma.
Although overall cancer incidence in the United States has decreased in recent years, the number of new cases of esophageal cancer is increasing. According to American Cancer Society estimates, there were 16,470 new cases and 14,530 deaths in this country in 2009 from esophageal cancer. Esophageal cancer has 2 main histologic subtypes: squamous cell carcinoma and adenocarcinoma. In the west, the incidence of the squamous cell carcinoma has remained stable or decreased since the 1970s; the incidence of adenocarcinoma has risen steadily during the same time period. Esophageal adenocarcinoma has now become the more prevalent histologic subtype in the United States.
Esophageal adenocarcinoma typically arises in the distal one-third of the esophagus, and its main risk factors are gastroesophageal reflux disease (GERD) and Barrett’s esophagus. For patients with Barrett’s esophagus, endoscopic surveillance to detect dysplasia is the primary strategy recommended to decrease morbidity and mortality from esophageal adenocarcinoma. This strategy has not proved effective, as shown by the increasing incidence of esophageal adenocarcinoma and the results of a recent study showing that most patients with this cancer have no prior diagnosis of Barrett’s esophagus and, therefore, are not enrolled in surveillance programs.
Basic investigations that have defined the genetic events underlying colonic carcinogenesis have led to effective strategies for the management and prevention of colorectal cancer. Analogously, it is important to understand the molecular carcinogenesis of Barrett’s esophagus to identify specific targets to guide the development of effective diagnostic strategies and novel therapeutic agents. To do this, the molecular events that lead to the replacement of normal esophageal squamous cells by metaplastic Barrett cells must first be understood. Building on this understanding, we can appreciate how the genetic abnormalities acquired by metaplastic Barrett cells disrupt their normal properties so they can take on the morphologic and physiologic features of dysplasia and cancer. This report provides a conceptual basis for how normal esophageal squamous cells undergo columnar metaplasia and how metaplastic Barrett cells progress to dysplasia and carcinoma. Some of the main genetic alterations involved in the development and neoplastic progression of Barrett’s esophagus are reviewed; however, these represent a fraction of the genetic changes required for the making of Barrett metaplasia, dysplasia, and esophageal adenocarcinoma.
The making of Barrett metaplasia
Most, if not all, esophageal adenocarcinomas arise from Barrett’s esophagus, the condition in which the normal squamous cells lining the distal esophagus are replaced by intestinal-type columnar cells. Barrett’s esophagus develops through the process of metaplasia, the replacement of one adult cell type by another. Metaplasia is believed to arise as a protective response to chronic tissue inflammation, which in the esophagus is believed to be caused by GERD. Barrett metaplasia can result from either fully differentiated esophageal squamous cells changing directly into intestinal-type columnar cells or from changing the differentiation pattern of esophageal stem cells.
Metaplasia through transdifferentiation
Transdifferentiation is the switch of one fully differentiated cell type directly into another. In general, this switch occurs between cell phenotypes that were present in the organ during embryonic development. During embryogenesis, the esophagus is initially lined by ciliated columnar cells, which are replaced by stratified squamous cells as maturation proceeds ( Fig. 1 ). Data from ex vivo organ cultures of embryonic mouse esophagus demonstrate direct conversion of the columnar cells lining the esophagus into squamous cells, a process found to be independent of cell proliferation or apoptosis. In theory, a reversal of this normal developmental switch in cell phenotype may occur during the formation of Barrett metaplasia. In support of this hypothesis, studies using scanning electron microscopy have demonstrated a distinctive cell at the squamocolumnar junction in Barrett mucosa that expresses cytokeratin markers and demonstrates morphologic features of both squamous and columnar epithelium; moreover, this distinctive cell has not been detected at the squamocolumnar junction in patients without Barrett mucosa. Once Barrett metaplasia is established, the epithelium must undergo maintenance and self-renewal, processes that are not explained by the transdifferentiation hypothesis, however.
Metaplasia through transdifferentiation
Transdifferentiation is the switch of one fully differentiated cell type directly into another. In general, this switch occurs between cell phenotypes that were present in the organ during embryonic development. During embryogenesis, the esophagus is initially lined by ciliated columnar cells, which are replaced by stratified squamous cells as maturation proceeds ( Fig. 1 ). Data from ex vivo organ cultures of embryonic mouse esophagus demonstrate direct conversion of the columnar cells lining the esophagus into squamous cells, a process found to be independent of cell proliferation or apoptosis. In theory, a reversal of this normal developmental switch in cell phenotype may occur during the formation of Barrett metaplasia. In support of this hypothesis, studies using scanning electron microscopy have demonstrated a distinctive cell at the squamocolumnar junction in Barrett mucosa that expresses cytokeratin markers and demonstrates morphologic features of both squamous and columnar epithelium; moreover, this distinctive cell has not been detected at the squamocolumnar junction in patients without Barrett mucosa. Once Barrett metaplasia is established, the epithelium must undergo maintenance and self-renewal, processes that are not explained by the transdifferentiation hypothesis, however.
Metaplasia through stem cells
Stem cells can proliferate, self-renew, give rise to a variety of cell types, and regenerate tissue after injury. A stem cell origin would account for the persistence and maintenance of Barrett epithelium and could explain the predisposition of this tissue to neoplastic transformation. The stem cell for Barrett’s esophagus may reside in the esophagus itself or originate in the bone marrow. During development, tracheo-esophageal progenitor cells express p63, a homolog of p53. As the esophageal lining forms, p63+ progenitor cells differentiate into ciliated columnar cells that lack p63 expression. After stratified squamous epithelium replaces the ciliated columnar epithelium, cells in the proliferative basal layer of the squamous epithelium continue to stain strongly for p63, whereas cells in the fully differentiated more superficial layers demonstrate no p63 staining. In mice null for p63, the esophagi completely lack stratified epithelium and are lined by simple columnar epithelium, suggesting that p63+ cells are necessary to establish a stratified squamous epithelium. Barrett epithelium has been found to lack immunostaining for p63, suggesting that the Barrett stem cell differs from the p63+ embryonic esophageal progenitor cell and the adult, squamous esophageal stem cell. These findings do not eliminate the possibility that the stem cells for Barrett metaplasia reside in esophageal submucosal glands or in glands of the gastric cardia, adjacent to the gastroesophageal junction, as has been suggested by some investigators.
A second potential source of stem cells for the esophagus is the bone marrow. In mice treated with high-dose irradiation to induce esophagitis, injection of either esophageal progenitor cells or bone marrow cells was able to repair the injured esophagus through regeneration of new squamous cells. Using a rat model of severe reflux esophagitis, our group investigated the possibility that bone marrow cells can give rise to metaplastic Barrett epithelium. Female rats were lethally irradiated and then rescued with bone marrow from male donor rats. An esophagojejunostomy was then performed on the female rats to induce the reflux of both acid and bile salts. Eight weeks after surgery, the esophagi of the female rats contained squamous and metaplastic cells with a Y chromosome, suggesting that bone marrow cells can hone to the esophagus and give rise to both squamous and columnar epithelium. In humans, cells from male donors have been found within the gastrointestinal tract of female patients who have undergone bone marrow transplant.
Regardless of where the stem cell originates, it is likely that the environment in the inflamed, reflux-damaged esophagus mediates the phenotypic switch from a squamous cell to an intestinal-like columnar cell (see Fig. 1 ). This phenotypic switch presumably occurs by altering the expression of a few key master genes that regulate cell phenotype. Candidate master genes that are upregulated in Barrett’s esophagus compared with neighboring esophageal squamous epithelium include the transcription factors CDX1, CDX2, and SOX9 (see Fig. 1 ). Not only are these genes normally expressed in the intestine but target genes of these transcription factors define an intestinal phenotype.
Role for CDX1, CDX2, and SOX9 in Barrett metaplasia
Homeotic genes define the developmental pattern of an organism. Cdx1 and Cdx2 are homeobox genes that specify intestinal epithelial differentiation. Studies in mice suggest that Cdx2 is required for intestinal differentiation and that Cdx1 may specify a columnar cell. CDX1 and CDX2 mRNA and protein expression have been detected in esophageal biopsy specimens from nondysplastic Barrett metaplasia, Barrett metaplasia with dysplasia, and Barrett-associated adenocarcinomas, but not in normal esophageal squamous epithelium. Sox9 is another transcription factor, expressed by potential stem cells in intestinal crypts, that plays a role in the formation of goblet cells. Recently, SOX9 protein has been shown to be expressed in Barrett metaplasia, Barrett with dysplasia, and adenocarcinoma, but not in esophageal squamous epithelium.
How GERD may induce Barrett metaplasia
Barrett metaplasia is a sequelae of chronic GERD. Components of the refluxed gastric juice (eg, acid, bile salts) and/or the resulting esophageal inflammation (reflux esophagitis) could cause esophageal metaplasia by inducing transcription factors or activating developmental signaling pathways that determine an intestinal phenotype.
Stimulation of the CDX Transcription Factors by GERD
In mouse esophageal squamous epithelial cells, exposure to bile acids or acid activates Cdx2 expression. In human esophageal squamous cells (HET1A), exposure to a combination of acid and bile salts increases CDX2 expression and leads to squamous cells forming cryptlike structures and expressing intestinal genes such as Villin, Sucrase-isomaltase, and MUC2 In addition, data suggest that GERD-induced inflammation activates CDX expression in esophageal squamous epithelial cells. In a rodent model of esophageal intestinal metaplasia, squamous cells begin to express Cdx2 before the development of intestinal metaplasia. In human esophageal biopsies, CDX2 expression has been found in inflamed esophageal squamous epithelium, but not in noninflamed squamous epithelium. In telomerase-immortalized normal esophageal squamous cells established from patients with GERD with and without Barrett’s esophagus, exposure to acid and/or bile salts increased CDX2 expression in the squamous cells from patients with Barrett’s esophagus, but not in those from patients with GERD without Barrett’s esophagus. Inhibition of nuclear factor-κB (NF-κB), a well-established mediator of GERD-induced inflammation, prevented the increase in CDX2 expression in esophageal squamous cells from patients with Barrett’s esophagus in response to acid and/or bile salt exposure, suggesting that inflammatory signaling cascades can also activate CDX2 expression in esophageal squamous cells.
Stimulation of Developmental Signaling Pathways by GERD
An attractive hypothesis is that esophageal activation of developmental signaling pathways involved in maintaining or developing the normal intestine may lead to Barrett metaplasia. These include pathways that are required for normal intestinal development, such as Wnt and Notch, or pathways that are expressed in the embryonic esophagus to maintain a columnar phenotype, such as Hedgehog and Bone Morphogenic Protein (Bmp) 4. Wnt is required to maintain the intestinal crypt progenitor cell population and regulates the expression of Cdx1. Wnt pathway activation (as determined by nuclear β-catenin) has not been found in nondysplastic Barrett metaplasia, but has been observed in Barrett metaplasia with dysplasia and in esophageal adenocarcinomas.
The Notch pathway also participates in maintaining the intestinal crypt progenitor pool and perhaps even that of the esophagus. As intestinal cells begin to differentiate, persistent Notch signaling leads to an absorptive enterocyte fate, whereas the lack of Notch signaling leads to a secretory fate as an enteroendocrine, goblet, or Paneth cell. Unlike the other developmental signaling pathways, components of the Notch signaling pathway are present in the normal adult esophagus. As noted earlier, CDX2 overexpression in squamous HET1A cells causes the cells to form cryptlike structures. In these same cells, expression of Hes1, a downstream target of Notch, is downregulated by CDX2 overexpression, suggesting that inhibition of Notch signaling by CDX2 may play a role in metaplasia formation. Bile salt exposure has also been shown to decrease expression of Notch pathway components in esophageal adenocarcinoma cells. Recently, in an animal model of reflux and Barrett’s esophagus, inhibitors of Notch signaling caused the proliferative Barrett cells to differentiate into goblet cells.
Bmp4 is normally expressed within the stroma of the embryonic columnar-lined esophagus, but it is absent in the adult squamous-lined esophagus. In a rodent model of reflux esophagitis and Barrett’s esophagus, investigators demonstrated Bmp4 expression in the stroma underlying inflamed esophageal squamous epithelium and specialized intestinal metaplasia, but not in the stroma underlying normal esophageal squamous epithelium. When human esophageal squamous cells were treated with BMP4 in vitro, the squamous cells began to express cytokeratins characteristic of columnar cells, suggesting that stromal BMP4 expression promotes the change in the esophageal epithelium from squamous to columnar.
The Hedgehog signaling pathway likely plays a role in esophageal metaplasia. Sonic hedgehog, the most ubiquitous Hedgehog ligand, is expressed by the embryonic esophagus although it has a columnar epithelium and before it takes on a stratified squamous phenotype. Recently, Sonic hedgehog expression was observed in Barrett metaplasia, but not in normal adult esophageal epithelium. In a mouse model of reflux esophagitis and Barrett’s esophagus, Sonic hedgehog expression was found in the Barrett metaplasia as well as in esophageal squamous cells before the development of intestinal metaplasia. Bmp4 is a target of Hedgehog signaling, therefore it was not surprising that stromal BMP4 expression was seen adjacent to Barrett epithelium from esophagectomy specimens. Activation of BMP4 signaling in HET1A cells induced SOX9 expression and subsequent expression of cytokeratins characteristic of columnar cells.
The making of Barrett-associated dysplasia and adenocarcinoma
The histologic diagnoses of dysplasia and cancer are based on a compilation of morphologic features of the tissue that indicate that the cells have acquired abnormal physiologic properties. In 2000, Hanahan and Weinberg characterized 6 physiologic properties of cancer, also called hallmarks, that normal cells acquire as cancer ensues. These hallmarks include the ability of cells to provide their own growth signals, avoid growth inhibitory signals, resist apoptosis, replicate without limit, synthesize new blood vessels, and invade and metastasize ( Table 1 ). Studies have shown that these cancer hallmarks can be acquired by normal cells through disruptions in only a few key growth regulatory pathways including the p16/Retinoblastoma (Rb) and p53 pathways, the Ras signaling pathway, and the telomerase-dependent senescence pathway (see Table 1 ). Recently, cancer-related inflammation has been proposed as a seventh physiologic hallmark of cancer.
| Cancer Hallmark | Key Growth Regulatory Pathway |
|---|---|
| Provide growth signals | Ras pathway |
| Avoid growth inhibitory signals | p16/Rb and p53 pathways |
| Resist apoptosis | p53 pathway, Ras pathway |
| Replicate without limit | Telomerase-dependent senescence pathway, p16/Rb pathway, p53 pathway |
| Synthesize new blood vessels | Ras pathway |
| Invade and metastasize | ?? ?? Inflammatory microenvironment |
| Inflammatory microenvironment | ?? |
Related posts:
Barrett’s Esophagus: Clinical Issues
Barrett’s Esophagus Surveillance: When, How Often, Does It Work?
Endoscopic Evaluation and Advanced Imaging of Barrett’s Esophagus
The Role of Radiofrequency Ablation in the Management of Barrett’s Esophagus
Endotherapy for Barrett’s Esophagus: Which, How, When, and Who?
Role of Minimally Invasive Surgery in the Modern Treatment of Barrett’s Esophagus
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
