- Email: [email protected]
Expression of p53-related protein p63 in the gastrointestinal tract and in esophageal metaplastic and neoplastic disorders
Original Contributions Expression of p53-Related Protein p63 in the Gastrointestinal Tract and in Esophageal Metaplastic and Neoplastic Disorders JONATHAN N. GLICKMAN, MD, PHD, ANNIE YANG, BS, ALIAKBAR SHAHSAFAEI, MSC, FRANK MCKEON, PHD, AND ROBERT D. ODZE, MD, FRCP(C) p63 is a p53-related DNA-binding protein that helps regulate differentiation and proliferation in epithelial progenitor cells. Its expression has never been evaluated in the human gastrointestinal tract. The aim of this study was to evaluate the expression of p63 in the esophagus and related metaplastic and neoplastic disorders to gain insight into the pathogenesis of these processes. Of particular interest was the expression of p63 in Barrett esophagus (BE) and in BE-associated multilayered epithelium. Multilayered epithelium has been postulated to represent an early precursor to the development of BE primarily because it shares morphologic and immunophenotypic features of both squamous and columnar epithelium, and has been shown prospectively to be highly associated with BE. Routinely processed mucosal biopsy or resection specimens that contained normal esophageal squamous epithelium (n ⴝ 20), squamous dysplasia (n ⴝ 4), squamous cell carcinoma (n ⴝ 7), BE (n ⴝ 10), BEassociated multilayered epithelium (n ⴝ 13), esophageal mucosal gland ducts (n ⴝ 10), BE-associated dysplasia (n ⴝ 12), and BEassociated adenocarcinoma (n ⴝ 7) were immunostained for p63 to determine the extent and location of staining. p63 staining was compared with the staining patterns observed for p53, Ki 67 (proliferation marker), and cytokeratins (CKs) 13 (squamous marker), 14 (basal squamous marker), 8/18 (columnar marker), and 19 (basal/ columnar marker). Expression of p63 messenger RNA (mRNA) isoforms was also analyzed by reverse-transcription polymerase chain reaction of freshly isolated tissues. In the normal esophagus, p63 was expressed in the basal and suprabasal layers of the squamous epithelium and in basal cells that line the mucosal gland ducts but was
negative in all other epithelia of the gastrointestinal tract, including the stomach, small intestine, and colon. Similarly, p63 was not expressed in BE, but it, was present in the basal layer of multilayered epithelium in 9 of 13 cases (69%). p63-positive cells in multilayered epithelium and in the mucosal gland duct epithelium were positive for CK8/18 (100%) and CK13 (67% and 30%, respectively) and negative for CK14 (0%), in contrast to p63-positive cells in squamous epithelium, which were positive for CK14 and CK13 (100%) but negative for CK8/18. In neoplastic tissues, p63 was diffusely expressed in all cases of esophageal squamous cell dysplasia and carcinoma but was negative in all cases of esophageal and colorectal adenocarcinoma. The ⌬N isoform of p63 mRNA predominated in all benign and neoplastic squamous tissues examined. p63 may represent a marker of 2 distinct epithelial progenitor cells (basal squamous epithelium and gland duct epithelium) in the esophagus. P63 is upregulated in squamous neoplastic conditions and in this manner may play a role in squamous carcinogenesis. These data also indicate that multilayered epithelium is phenotypically similar to, and may share a lineage relationship with, mucosal gland duct epithelium. HUM PATHOL 32:1157-1165. Copyright © 2001 by W.B. Saunders Company Key words: p63, Barrett esophagus, squamous cell carcinoma, intestinal metaplasia. Abbreviations: GI, gastrointestinal; BE, Barrett esophagus; CK, cytokeratin; PBS, phosphate-buffered saline; H&E, hematoxylin and eosin; RT-PCR, reverse-transcription polymerase chain reaction.
The tumor-suppressor gene p53 is altered in a wide range of human malignancies.1 For instance, germline mutation in the p53 gene results in increased susceptibility to tumors in mice2,3 and humans.1,4,5 p53 is known to induce cell cycle arrest and/or apoptosis in response to cellular stresses such as DNA damage or hypoxia.6-8 Accordingly, loss of p53 function confers a growth ad-
vantage on cells via a mechanism that induces dysregulation of growth and resistance to apoptosis.5,8 Recently, 2 genes homologous to p53, each with distinct expression patterns and biologic activities, have been identified in mice and humans. The human p73 gene is widely expressed in benign tissues and is located on the short arm of chromosome 1, a region frequently lost in neuroblastomas.9,10 The p63 gene also shows homology to p53 in the DNA binding, transactivation, and oligomerization domains, but its expression in humans is most apparent in a small subset of benign epithelial tissues, including skin, squamous epithelia of the cervix and oral cavity, and prostatic acinar basal cells.11 Targeted disruption of the murine p63 gene results in neonatal death. Homozygotes show profound defects in limb and craniofacial development and in the differentiation of tissues with stratified epithelium, such as the skin, oral cavity, and esophagus.12,13 The restricted expression of p63 in epithelial cells,
From the Department of Pathology, Brigham and Women’s Hospital, and the Department of Cell Biology, Harvard Medical School, Boston, MA. Accepted for publication July 24, 2001. Presented in part at the United States and Canadian Academy of Pathology, Atlanta, GA, March 2001. Address correspondence and reprint requests to Jonathan N. Glickman, MD, PhD, Department of Pathology, Brigham and Women’s Hospital, 75 Francis St, Boston, MA 02115. Copyright © 2001 by W.B. Saunders Company 0046-8177/01/3211-0003$35.00/0 doi:10.1053/hupa.2001.28951
1157
HUMAN PATHOLOGY
Volume 32, No. 11 (November 2001)
and its functional similarity to p53, suggests that p63 may play a role in the regulation of proliferation and differentiation of specific epithelial cell types. However, the expression of p63 in the human gastrointestinal (GI) tract, which is lined by a variety of differentiated epithelial cell types, has not been studied. Our hypothesis in this study is that p63 is a marker of epithelial progenitor cells in the esophagus and as such plays a role in metaplastic and neoplastic disorders of this organ. In this study, we were particularly interested in the expression of p63 in Barrett esophagus (BE), squamous cell carcinoma, and adenocarcinoma. In BE, the stratified squamous epithelium of the lower esophagus is replaced by intestinal-type columnar epithelium containing goblet cells.14,15 The cell of origin of BE is unknown, but members of our group have postulated that this condition develops from a multipotential epithelial stem cell that can differentiate into either squamous or columnar epithelium.16,17 This hypothesis is supported by the finding of a distinctive type of multilayered epithelium associated with BE, which has morphologic and immunophenotypic characteristics intermediate between squamous and columnar epithelium.18-20 Clarification of the cell of origin of BE may provide insight into the pathogenesis of this condition and ultimately may lead to better forms of treatment for this disorder. Therefore, we performed this study to characterize the expression of p63 in the human GI tract and particularly to evaluate the expression of this protein in metaplastic, dysplastic, and malignant tissues of the esophagus. These studies were intended to evaluate the hypothesis that p63 is a marker of epithelial precursor cells in the esophagus. The pattern of p63 staining was compared with that of the proliferation marker Ki 67 and to cytokeratin (CK) 13 (suprabasal squamous marker), 14 (basal squamous marker), 8/18 (columnar marker), and 19 (basal/columnar marker) to determine the biologic and differentiation characteristics of p63-positive cells. Because p53 gene losses or mutations are frequently seen in GI malignancies, including esophageal squamous cell carcinoma and adenocarcinoma,1,5,15,21 we were also interested in the relationship between p63 and p53 expression. MATERIALS AND METHODS Patient Selection Study cases were selected by a retrospective review of the surgical pathology files of the Brigham and Women’s Hospital (BWH) and included patients who underwent an endoscopic biopsy or surgical resection of the esophagus, stomach, small intestine, or colon between 1994 and 1998. The patients included 7 (2 biopsies, 5 resections) with esophageal squamous cell carcinoma, including 4 with coexisting squamous cell carcinoma in situ, 7 (6 biopsies, 1 resection) with BEassociated adenocarcinoma, 10 (all biopsies) with BE (defined as replacement of the lower esophageal mucosa by intestinal-type columnar epithelium with goblet cells; average length, 7.0 cm; range, 2 to 15 cm), 13 who had multilayered
epithelium (12 biopsies, 1 resection) associated with intestinal metaplasia (defined as described previously17-19) and 3 (all resections) with colonic adenocarcinoma. In addition, esophageal mucosal and submucosal glands and ducts were present in 14 of the above-listed cases and were evaluated as well. Normal controls included biopsy specimens from patients with an endoscopically and histologically normal-appearing gastroesophageal junction (n ⫽ 3) containing both squamous and gastric cardia-type mucosa, gastric antrum (n ⫽ 4), duodenum (n ⫽ 2), jejunum (n ⫽ 3), ileum (n ⫽ 2), or colon (n ⫽ 3).
Histologic Analysis Tissues were fixed in 10% formalin, embedded in paraffin, and processed routinely. Five-micron-thick tissue sections were stained with hematoxylin and eosin (H&E) and examined for the presence of various epithelial cell types, including benign squamous epithelium, specialized-type intestinal epithelium (identified by the presence of goblet cells admixed with mucous cells), dysplastic columnar or squamous epithelium (identified as previously described22), and invasive squamous cell carcinoma or adenocarcinoma. Esophageal multilayered epithelium was identified using previously published criteria.17-19 Briefly, multilayered epithelium shows 4 to 7 layers of cells composed of compact squamous-appearing cells in the basal portion and columnar cells in the superficial layers.
Immunohistochemical Analysis Tissue blocks representative of each case were selected for immunohistochemical analysis. In cases of dysplasia or malignancy, blocks containing both tumor and adjacent nondysplastic mucosa were used. Immunohistochemical staining was performed as follows: tissue sections were incubated for 60 minutes at 60°C, deparaffinized, and rehydrated in graded ethanol solutions. Endogenous peroxidase activity was blocked by incubation of the slides in 3% H2O2 (Fisher Scientific, Pittsburgh, PA) for 5 minutes. The slides were then rinsed under running water for 5 minutes and transferred to phosphate-buffered saline (PBS; Fisher Scientific). Sections pretreated with heat-induced epitope retrieval were transferred to 10 mmol/L citrate buffer, pH 6.0, and heated for 30 minutes at 199°F in a microwave oven equipped with a temperature probe. The slides were then rinsed under running water and transferred to PBS (Fisher Scientific). The tissue sections were blocked with 2% horse serum (Vector Laboratories, Burlingame, CA) for 15 minutes and incubated with the antibodies (p63, p53, CK13, CK14, CK19, CK8/18, Ki 67) at the dilutions listed in Table 1. All antibodies were incubated for 1 hour at room temperature in a humid chamber. The slides were then washed for 5 minutes in PBS and incubated with a biotinylated secondary antibody (mouse immunoglobulin [Ig] G) and then avidin– horseradish peroxidase (Vectastain Elite ABC kit; Vector Laboratories) according to the manufacturer’s instructions. The slides were washed in PBS between incubations. The tissue sections were then developed using 3,3⬘-diaminobenzidine (DAB; Sigma) as a substrate and counterstained with Gill hematoxylin (Fisher Scientific) according to the manufacturer’s instructions. Positive staining control for p63 was normal human tonsil tissue. Other positive staining controls were used according to the recommendations of the antibody supplier. Tissue sections in which the primary antibody was substituted for one of the same IgG subclass, but different antigenic specificity, served
1158
p63 IN GI TRACT (Glickman et al)
TABLE 1. Summary of Antibodies Used in This Study Antigen
Function
Clone
Supplier
p63 p53 CK13 CK14 CK19 CK8/18 Ki 67
p53 homolog Tumor suppressor Squamous marker Squamous marker Columnar marker Columnar marker Proliferation marker
4A4 DO-1 AE8 LL002 RCK108 CAM5.2 MIB-1
Yang et al11* Biogenex (San Ramon, CA) Biogenex Novocastra (Burlingame, CA) Biogenex Becton-Dickinson (Mountain View, CA) AMAC (Westbrook, ME)
Dilution 1:50† 1:100† 1:100 1:100† 1:300† 1.5† 1:200
NOTE. All antibodies are mouse IgG *Antibody also available from Santa Cruz Biotechnology (catalog no. sc-8431). †Pretreatment with heat-induced epitope retrieval as described in Materials and Methods.
as negative staining controls. In addition, the specificity of the anti-p63 antibody has been previously demonstrated in a variety of immunoblotting experiments, and in experiments that used immunohistochemical staining of tissues from mice that contain deletion of the p63 gene.11,13 Immunohistochemical staining for each marker was scored by one of the authors (J.G.) as positive or negative in the epithelial type under study. For p63 labeling, only cells with a strong nuclear staining signal were counted as positive. The Ki 67–labeling index for each epithelial type was calculated as the percentage of cells staining positive for this marker (at least 200 cells counted).
Reverse-Transcription Polymerase Chain Reaction Analysis of p63 Expression Mucosal tissue from 6 esophagectomy specimens (2 squamous cell carcinoma, 3 BE-associated adenocarcinoma, 1 BE), 3 gastrectomy specimens (all for peptic ulcer disease), and 3 colectomy specimens (all for colorectal adenocarcinoma) was harvested within 2 hours of the procedure, in compliance with protocols approved by the BWH Human Studies Committee. Mucosal tissue was harvested by scraping the mucosal surface with a sterile scalpel from the following areas: normal squamous mucosa (n ⫽ 4), BE (n ⫽ 4), tumor tissue (2 squamous cell carcinoma, 2 adenocarcinoma), normal gastric mucosa from the corpus (n ⫽ 4) or antrum (n ⫽ 3), normal colonic mucosa (n ⫽ 3), and colonic adenocarcinoma (n ⫽ 3). The identities of the tissues sampled were confirmed by examination of formalin-fixed, paraffin-embedded H&E-stained sections of adjacent mucosa and by examination of smear preparations of the scraped tissue. Total cellular RNA was isolated using Trizol reagent (Gibco BRL, Gaithersburg, MD) according to the manufacturer’s recommendations and dissolved in diethyl pyrocarbonate–treated distilled water at a concentration of 2 g/L. For reverse-transcription polymerase chain reaction (RTPCR) analysis of p63 messenger RNA (mRNA)-isoforms, 2 g of total RNA was reverse-transcribed with the p63-specific reverse primer 5⬘-AGCTCATGGTTGGGGCAC using SuperScript II (Gibco BRL) according to the recommendations of the enzyme manufacturer. One twentieth of this reaction was used to amplify p63 isoforms using Taq polymerase (Perkin Elmer, Shelton, CT) in a Perkin Elmer 9600 thermal cycler with the following cycling parameters: 2 minutes at 95°C; 40 cycles of 15 seconds at 52°C; 45 seconds at 72°C; 15 seconds at 95°C; and 8 minutes at 72°C. Primers were 5⬘-ATTCCCAGAGCACACAG and 5⬘-AGCTCATGGTTGGGGCAC for the TA form (600 – base pair [bp] product), and 5⬘-CAGACTCAATTTAGTGAG and 5⬘-AGCTCATGGTTGGGGCAC for the ⌬N form (400-bp product). For analysis of glyceralde-
hyde-3-phosphate dehydrogenase (GAPDH) mRNA as a control for mRNA quality and content, 2 g of total RNA was reverse-transcribed with oligo-dT(12–18) and SuperScript II (Gibco BRL). This reaction was amplified with the primers 5⬘-CACATCGCTCAGACACCATG and 5⬘-GCCATGGAATTTGCCATGGG using the cycling parameters above, except that 30 cycles were used and the annealing temperature was 60°C. RNA isolated from ME180 cervical carcinoma cells11 served as positive controls. First-strand synthesis reactions run in the absence of reverse transcriptase served as negative controls. Reaction products were analyzed on 2% agarose gels, stained with ethidium bromide, and photographed using a Kodak DC120 (Kodak, Rochester, NY) digital camera and Photoshop 4.0 software (Adobe Corp, San Jose, CA).
Statistical Analysis Statistical comparisons of immunohistochemical staining results were done using the Fisher exact text and Student t test, as appropriate. A P value of ⬍ .05 was considered statistically significant.
RESULTS Expression of p63 in the GI Tract Normal esophageal squamous epithelium showed strong nuclear reactivity for p63 in all cases (100%). Staining was present in all cells of the basal layer and in the suprabasal layer of cells (Fig 1A). Reactivity was also present in the middle layers of the epithelium, in a lower proportion of cells. Rare staining was noted in the superficial squamous cells [8/20 cases (40%)]. In contrast, columnar surface and crypt epithelium in the gastric cardia, gastric antrum, duodenum, jejunum, ileum, and colon showed no reactivity for p63. In addition, p63 stained basal cells in esophageal mucosal [10/11 cases (91%)] and submucosal [3/3 cases (100%)] gland ducts, but did not stain the mucinous columnar epithelium of the glands of either type (Fig 1B). Expression of p63 in Esophageal Squamous Neoplasia Squamous mucosa from areas of active esophagitis showed an expanded zone of p63-positive cells that involved from 75% to 100% of the mucosal area (Table 2). All cases of squamous dysplasia and squamous cell
1159
HUMAN PATHOLOGY
Volume 32, No. 11 (November 2001)
FIGURE 1. Expression of p63 in the human gastrointesinal tract. (A) Esophageal squamous epithelium showing p63 nuclear reactivity in basal and suprabasal layers of the epithelium. (Original magnification ⫻200.) Esophageal mucosal gland duct showing p63 positivity in the basal layer. (Original magnification ⫻200.)
carcinoma showed strong, diffuse expression of p63 in more than 90% of cells (Fig 2A). Similarly, all cases (100%) of invasive squamous cell carcinoma showed strong diffuse expression of p63 in the tumor cells. Expression of p63 in BE, Multilayered Epithelium, and BE-Associated Adenocarcinoma Intestinal-type epithelium from BE was negative for p63 in all cases. However, BE-associated multilayered epithelium contained p63-positive cells in 9 (69%) of 13 cases. p63-positive cells were located in the basal or immediate suprabasal portions of this epithelium (Fig 3A). P63 staining was absent from all cases of BE-associated dysplasia and adenocarcinoma, as well as from 3 cases of colorectal adenocarcinoma. Comparison of p63 Staining With p53 Expression p53 nuclear staining in benign squamous epithelium was weak and limited to a minority (⬍5%) of cells in the basal layer of the epithelium (Fig 2B). However, strong nuclear reactivity for p53 was present in all
(100%) squamous dysplasia cases and in 5 (71%) of 7 squamous cell carcinomas. The staining pattern of p63 and p53 was similar in these cases (Fig 2C). Strong p53 nuclear staining was present in 4 of 13 cases of multilayered epithelium, all of which were also positive for p63. Staining for both markers was similar and was located in the basal layer of the epithelium. Although BE and BE-associated dysplasia and adenocarcinoma cases were all negative for p63, 0%, 75%, and 57% of these cases, respectively, stained positive for p53 (Table 2). Comparison of p63 Staining With Ki 67 Labeling Ki 67 is a proliferation-associated nuclear antigen expressed in the G1, S, and G2 phases of the cell cycle and in this manner labels all proliferating cells. Ki 67 staining highlighted a narrow strip of cells in the immediate suprabasal region of normal esophageal squamous epithelium (Fig 2D). These cells were also reactive for p63. The p63-positive cells in squamous dysplasia and invasive squamous cell carcinoma were also reactive with Ki 67 (not shown).
TABLE 2. Summary of p63, p53, and Ki 67 Expression in the Esophagus and Related Metaplastic and Neoplastic Disorders Marker Ki 67 Epithelium Type Squamous epithelium Normal (n ⫽ 20) Dysplasia (n ⫽ 4) Carcinoma (n ⫽ 7) BE Intestinal metaplasia (n ⫽ 10) Multilayered epithelium (n ⫽ 13) Dysplasia (n ⫽ 12) Adenocarcinoma (n ⫽ 7)
p63
p53
No. (%) Positive
Mean % of Cells Positive
20 (100%) 4 (100%) 7 (100%)
0 (0%) 4 (100%) 5 (71%)
20 (100) 4 (100) 7 (100)
50 69 72
0 (0%) 9 (69%) 0 (0%) 0 (0%)
0 (0%) 4 (31%) 9 (75%) 4 (57%)
10 (100) 13 (100) 12 (100) 7 (100)
11 12* 85 76
*Only in basal cells.
1160
p63 IN GI TRACT (Glickman et al)
FIGURE 2. Comparison of p63 expression with P53 and Ki67 in esophageal squamous cell carcinoma. (A) p63, showing diffuse reactivity in invasive squamous cell carcinoma. (Original magnification ⫻400). (B) p53, showing rare reactive cells in the basal layer of normal squamous epithelium. (Original magnification ⫻200). (C) p53, showing diffuse reactivity in invasive squamous cell carcinoma. (Original magnification ⫻400.) (D) Ki67 proliferation marker, showing reactivity in suprabasal layers of normal squamous epithelium. (Original magnification ⫻200.)
The p63-positive cells of esophageal multilayered epithelium and in the esophageal gland ducts were positive for the proliferation marker Ki 67 in 100% and 88% of cases, respectively, in a minority of cells (Fig 3B). All cases of BE, BE-associated dysplasia, and adenocarcinoma (all p63 negative) contained Ki 67–positive cells. Ki 67 staining in BE was limited to the crypt region of the epithelium, whereas in dysplasia and adenocarcinoma cases, the staining was strong and diffuse (⬎80% of cells positive). Comparison of p63 Staining With Cytokeratin Expression To further define the differentiation state of p63positive epithelial cells in the esophagus, we compared the staining pattern of this protein with that of a variety of cell type-specific CKs, including the basal squamous marker CK14,23-27 the suprabasal squamous marker CK13,24-26 the columnar markers CK8/18,23,24 and the basal squamous/columnar marker CK1924,28,29 (Table 3). In normal squamous epithelium and in esophagitis, p63positive cells were also positive for CK14 and CK19, focally weakly reactive for CK13 (100% of cases), and negative for the columnar markers CK8/18. Squamous dysplasia and squamous cell carcinoma cases were positive for CK14 and CK19 and negative for CK13 and CK8/18.
p63-positive cells in multilayered epithelium and in esophageal mucosal gland ducts were also positive for CK8/18 (Fig 3E) and CK19 in all cases. They were focally positive for CK13 (67% and 30% of cases, respectively; Fig 3C), and negative for the basal squamous CK14 (Fig 3D). The keratin immuonophenotype of p63-positive esophageal multilayered epithelium (CK8/18 and CK19 positive, CK13 focally positive, CK14 negative) differed significantly from that of p63-positive squamous epithelium (CK14 and CK19 positive, CK13 focally positive, CK8/18 negative) and from that of p63negative intestinal-type BE (CK8/18 and CK19 positive, CK13 and CK14 negative). Expression of p63 mRNA Isoforms RT-PCR analysis of p63 expression in esophageal squamous epithelium showed that the ⌬N form was the predominant mRNA isoform in all cases tested (4 of 4), and the TA form was found only occasionally (1 of 4 cases) in this type of epithelium (Fig 4). Two cases of esophageal squamous cell carcinoma also expressed the ⌬N isoform. Weak expression of both the TA and ⌬N isoforms was occasionally present in BE (1/4 cases; 25%). Neither p63 isoform was detected in gastric cor-
1161
HUMAN PATHOLOGY
Volume 32, No. 11 (November 2001)
FIGURE 3. Expression of p63 in multilayered epithelium associated with BE; comparison with CK expression. (Original magnification ⫻400.) (A) p63. Multilayered epithelium is positive in the basal layer, whereas adjacent intestinal-type epithelium (right) is negative. (B) Ki67–proliferation marker showing focal reactivity in basal cells. (C) CK13 (suprabasal squamous marker). Multilayered epithelium is positive in the basal portion of the epithelium, whereas adjacent intestinal-type epithelium (right) is negative. (D) CK14 (basal squamous marker). Multilayered eptihelium is negative. (E) CK8/18 (columnar marker) positive in multilayered epithelium and adjacent intestinaltype epithelium (right).
pus or antral epithelium, or in normal colonic epithelium or adenocarcinoma. DISCUSSION The results of this study showed that in the normal esophagus, p63 expression was restricted to the basal and suprabasal cell layers of normal esophageal squamous epithelium and the basal cells of esophageal mucosal and submucosal gland ducts. P63 expression was upregulated in esophageal squamous dysplasia and squamous cell carcinoma. p63-positive cells in squamous dysplasia and carcinoma showed an immature keratin immunophenotype (CK14 positive, CK13 negative), similar to the p63-positive cells in the basal portion of normal squamous epithelium. These results suggest that p63 may play a role in squamous carcino-
genesis in the esophagus and that squamous dysplasia and carcinoma may be derived from the basal cells of the esophageal squamous epithelium. In fact, there are other data to support this concept. For instance, p53 overexpression and/or mutation have been found in the basal cells of nonneoplastic squamous mucosa adjacent to squamous cell carcinomas in a high percentage of cases.30-32 The results of this study are consistent with previous analyses of p63 expression in murine tissues11-13 and in keratinocyte cultures33 and support the hypothesis that p63 is a marker of precursor cells and may play a role in the regulation of epithelial differentiation and proliferation in the esophagus. The severe derangements in epithelial development displayed by mice mutant for the p63 gene12,13 are also consistent with this hypothesis. Furthermore, heterozygous germline muta-
1162
p63 IN GI TRACT (Glickman et al)
TABLE 3. CK Immunophenotype of p63-Positive Epithelial Cells in Normal Squamous Epithelium, Multilayered Epithelium, and Esophageal Mucosal Gland Duct Epithelium Epithelium Type
Marker
Normal Squamous Epithelium (n ⫽ 20)
Esophageal Multilayered Epithelium (n ⫽ 9)
Esophageal Mucosal Gland Duct Epithelium (n ⫽ 10)
CK14 CK13 CK19 CK8/18
19/19 (100%) 18/18 (100%) 16/16 (100%) 0/17 (0%)
0/9 (0%)* 6/9 (67%) 9/9 (100%) 9/9 (100%)*
0/9 (0%)* 3/10 (30%)* 10/10 (100%) 10/10 (100%)*
*P ⬍ .05 compared with normal squamous epithelium.
tions in the p63 gene cause the human autosomal dominant syndrome known as EEC (Ectrodactyly, Ectodermal dysplasia, Cleft lip),34 which has limb and craniofacial defects similar to those seen in p63-null mice. The results of previous functional studies of p63 have suggested a possible mechanism by which this protein may regulate epithelial differentiation and proliferation. The p63 gene product exists in several alternatively spliced forms with distinct activities.11 The ⌬N form, lacking the transactivation domain, is the predominant form in normal squamous epithelium and in squamous cell carcinoma, as demonstrated by the results of this study. By RT-PCR analysis, the ⌬N mRNA isoform was present in 4 of 4 samples of normal squamous epithelium and in 2 of 2 samples of squamous cell carcinoma, whereas the TA mRNA isoform containing the transactivation domain was present in only 1 case of normal squamous epithelium and was absent from all squamous cell carcinomas. The ⌬N isoform is capable of interacting with wild-type p53 and inhibiting its function in a dominant negative fashion.11 Therefore, expression of this protein might be expected to block the growth-inhibitory and apoptosis-inducing activities of p53 or of signals that act through p53 and thus could help maintain the proliferative capacity of progenitor cells, as originally proposed by Yang et al.13 In esophageal squamous neoplasia, p63 expression may be upregulated secondary to failure of terminal differentiation and, as such, may have growth-promoting and/or carcinogenic activity. However, based solely on the results of our study, we cannot be certain that p63 plays a role in squamous carcinogenesis. Tissue culture or animal studies are needed to resolve this issue adequately. Although p63 and p53 have structural and functional similarity,11,13 our results, as well as those of previous studies, indicate that these 2 homologous proteins may play distinct roles in vivo. The high levels of p53 observed in many tumors are usually the consequence of inactivating mutations that result in the accumulation of nonfunctional p53 protein.1,35 In contrast, p63 is constitutively expressed in normal epithelial progenitor cells, and its overexpression in squa-
mous neoplasia may reflect the immaturity of the tumor cell lineage. An expanded zone of p63 expression in areas of esophagitis with basal cell hyperplasia is consistent with this hypothesis. In this study, p63 expression was not observed in BE or in BE-associated dysplasia or adenocarcinoma. Thus, it is unlikely that p63 is involved in BE-associated carcinogenesis. However, our finding of p53 overexpression in these neoplasms does support the wellknown role of this protein in BE-associated carcinogenesis.15 Although p63 and p53 may interact, our data suggest that the mechanism of action of p53 in BEassociated neoplasia is independent of p63. Our finding of an absence of p63 expression in colonic adenocarcinomas, as well as in a variety of other tumors,11 supports this theory. In fact, molecular studies of colon and lung carcinoma cell lines have shown only rare mutations in p63, but frequent mutations of p53.36-38 Further studies are needed to determine whether p63 mutations that occur in human tumors are clinically relevant. Another interesting result of this study was that p63 did not stain any cases of BE but was present in multilayered epithelium. Multilayered epithelium is a distinctive type of epithelium with morphologic and immunophenotypic features intermediate between squamous and columnar epithelium20 and has been shown prospectively to be highly associated with BE.39 Furthermore, multilayered epithelium has also been noted to have mucin histochemical and immunophenotypic properties similar to BE.18-20 As a result, multilayered epithelium was hypothesized by our group to represent an early or intermediate stage in the devel-
FIGURE 4. Expression of p63 mRNA isoforms. Ethidium bromide stained agarose gel of representative RT-PCR products: lane1, normal gastric mucosa; lane 2, BE; esophagus; lane 3, normal esophageal squamous epithelium; lane 4, esophageal squamous cell carcinoma; lane 5, positive control ME180 cells. Control GAPDH (192 bp, top row) is positive in all samples; p63 TA form (600 bp; middle row) is negative in all samples, except for the positive control; p63 ⌬N form (400 bp, bottom row) is positive in normal squamous epithelium and in squamous cell carcinoma.
1163
HUMAN PATHOLOGY
Volume 32, No. 11 (November 2001)
opment of BE.20 Our finding of p63-positive cells in multilayered epithelium supports this hypothesis and suggests that multilayered epithelium may contain epithelial progenitor cells capable of differentiating toward a columnar phenotype. Experimental systems using either cell culture or animal models are needed to more fully demonstrate a direct lineage relationship between BE and ME. We found that p63-positive cells in multilayered epithelium and in esophageal mucosal gland duct epithelium were phenotypically similar (CK8/18 and CK19 positive, focally CK13 positive, CK14 negative) and significantly different from normal esophageal basal squamous cells (CK14 and CK19 positive, focally CK13 positive, CK8/18 negative). These data raise the possibility that multilayered epithelium may arise from mucosal gland duct epithelium and that mucosal gland duct epithelium contains a population of p63-positive cells distinct from basal squamous epithelium. Although the cell of origin of BE is essentially unknown, some have postulated that a multipotent epithelial progenitor cell may reside either in the basal layer of normal esophageal squamous epithelium or in esophageal gland ducts.40,41 Support for these theories stems from experimental studies in dogs in which BE was found to develop from cells that are intrinsic to the esophagus.42-45 Furthermore, in 1963 Adler40 noted that mucosal gland duct epithelium often appeared to be “growing out” in an attempt to re-epithelialize eroded surface epithelium in a pathologic study of 6 patients with gastroesophageal reflux disease. Finally, previous studies on the morphologic and phenotypic properties of multilayered epithelium performed by our group have shown that multilayered epithelium may occur within esophageal gland ducts and often in continuity with columnar epithelium at the surface of the mucosa.20 The previously reported association of multilayered epithelium with BE, combined with our finding of p63 in multilayered epithelium, raises an interesting comparison between multilayered epithelium and the epithelium of the cervical transition zone. Both of these epithelial types occur at or near squamocolumnar junctions and are associated with metaplasia (eg, squamous metaplasia of the endocervical mucosa). Furthermore, the reserve cells of the cervical transition zone, like the basal cells of multilayered epithelium, have also been found to be p63 positive.46 In fact, p63 expression has been noted in benign cervical squamomucinous epithelium, an epithelium with morphologic features intermediate between squamous and columnar epithelium similar to multilayered epithelium. Increased or maintained p63 expression has also been reported in early cervical squamous neoplasia.47 In summary, in this study we characterized the expression of the p53-related protein p63 in the esophagus and in related metaplastic and neoplastic disorders. Our results indicate that p63 expression may be a marker of certain epithelial progenitor cells in the esophagus, including basal squamous epithelium and mucosal gland ducts, and suggest that this protein may
play a role in the regulation of epithelial cell differentiation. Our data also provide evidence that multilayered epithelium and mucosal gland duct epithelium are phenotypically similar and may share a lineage relationship. Finally, although p63 expression is upregulated in esophageal squamous neoplasia, p63 plays no role in BE-associated carcinogenesis. REFERENCES 1. Hollstein M, Sidransky D, Vogelstein B, et al: P53 mutations in human cancers. Science 253:49-53, 1991 2. Donehower LA, Harvey M, Slagle BL, et al: Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumors. Nature 356:215-221, 1992 3. Jacks T, Remington L, Williams BO, et al: Tumor spectrum analysis in p53-mutant mice. Curr Biol 4:1-7, 1994 4. Malkin D, Li FP, Strong LC, Fraumeni JF, et al: Germ line p53 mutations in a familial syndrome of breast cancer, sarcomas, and other neoplasms. Science 250:1233-1238, 1990 5. Levine AJ, Wu MC, Chang A, et al: The spectrum of mutations at the p53 locus. Evidence for tissue-specific mutagenesis, selection of mutant alleles, and a “gain of function” phenotype. Ann NY Acad Sci 768:111-128, 1995 6. Hartwell LH, Kastan MB. Cell cycle control and cancer. Science 266:1821-1828, 1994 7. Ko LJ, Prives C: p53: Puzzle and paradigm. Genes Dev 10: 1054-1072, 1996 8. Polyak K, Xia Y, Zweier JL, et al: A model for p53-induced apoptosis. Nature 389:300-305, 1997 9. Kaghad M, Bonnet H, Yang A, et al: Monoallelically expressed gene related to p53 at 1p36, a region frequently deleted in neuroblastoma and other human cancers. Cell 90:809-819, 1997 10. Jost CA, Marin MC, Kaelin WG: P73 is a human p53-related protein that can induce apoptosis. Nature 389:191-194, 1997 11. Yang A, Kaghad M, Wang Y, et al: p63, a p53 homolog at 3q27-29. Encodes multiple products with transactivating, death-inducing, and dominant-negative activities. Mol Cell 2:305-316, 1998 12. Mills AA, Zheng B, Wang X-J, et al: p63 is a p53 homolog required for limb and epidermal morphogenesis. Nature 398:708713, 1999 13. Yang A, Schweitzer R, Sun D, et al: p63 is essential for regenerative proliferation in limb, craniofacial and epithelial development. Nature 398:714-718, 1999 14. DeMeester SR, DeMeester TR: The diagnosis and management of Barrett’s esophagus. Adv Surg 33:29-68, 1999 15. Jankowski JA, Wright NA, Meltzer SJ, et al: Molecular evolution of the metaplasia-dysplasia-adenocarcinoma sequence in the esophagus. Am J Pathol, 154:965-973, 1999 16. Shields HM, Zwas F, Antonioli DA, et al: Detection by scanning electron microscopy of a distinctive esophageal surface epithelium at the junction of squamous and Barrett’s epithelium. Dig Dis Sci 38:97-108, 1993 17. Sawhney RA, Shields HM, Allan CH, et al: Morphological characterization of the squamocolumnar junction of the esophagus in patients with and without Barrett’s epithelium. Dig Dis Sci 41:10881098, 1996 18. Boch JA, Shields HM, Antonioli DA, et al: Distribution of cytokeratin markers in Barret’s specialized columnar epithelium. Gastroenterology 112:760-765, 1997 19. Chen YY, Wang HH, Antonioli DA, et al: Shahsafaei A, Odze RD. Significance of acid-mucin-positive nongoblet columnar cells in the distal esophagus and gastroesophageal junction. HUM PATHOL 30:1488-1495, 1999 20. Glickman JN, Chen Y, Antonioli D, et al: Characterization of a distinctive multilayered epithelium associated with Barrett’s esophagus. Am J Surg Pathol 25:569-578, 2001 21. Shi ST, Yang GY, Wang LD, et al: Role of p53 gene mutations in human esophageal carcinogenesis: Results from immunohistochemical and mutation analyses of carcinomas and nearby noncancerous lesions. Carcinogenesis 20:591-597, 1999
1164
p63 IN GI TRACT (Glickman et al) 22. Dawsey SM, Lewin KJ: Histologic precursors of squamous esophageal cancer. Pathol Annu 30:209-226, 1995 23. Moll R, Franke WW, Schiller DL, et al: The catalog of human cytokeratins: Patterns of expression in normal epithelia, tumors and cultured cells. Cell 31:111-124, 1982 24. Smedts F, Ramaekers F, Troyansovsky S, et al: Keratin expression in cervical cancer. Am J Pathol 141:497-511, 1992 25. Takahashi H, Shikata N, Sanzaki H, et al: Immunohistochemical staining patterns of keratins in normal oesophageal epithelium and carcinoma of the esophagus. Histopathology 26:45-50, 1995 26. Viaene AI, Baert JH: Expression of cytokeratin mRNAs in normal human esophageal epithelium. Anat Rec 241:88-98, 1995 27. Smedts F, Ramaekers F, Troyanovsky S, et al: Basal-cell keratins in cervical reserve cells and a comparison of their expression in cervical intraepithelial neoplasia. Am J Pathol 140:601-612, 1992 28. Kwaspen FHL, Smedts FMM, Broos A, et al: Reproducible and highly sensitive detection of the broad spectrum epithelial marker keratin 19 in routine cancer diagnosis. Histopathology 31: 503-516, 1997 29. Bosch FX, Leube RE, Achstatter T, et al: Expression of simple type cytokeratins in stratified epithelia as detected by immunolocalization and hybridization in situ. J Cell Biol 106:1635-1648, 1988 30. Wang LD, Zhou Q, Hong JY, et al: p53 protein accumulation and gene mutations in multifocal esophageal precancerous lesions from symptom free subjects in a high incidence area for esophageal carcinoma in Henan, China. Cancer 77:1244-1249, 1996 31. Shi, ST, Yang GY, Wang LD, et al: Role of p53 gene mutations in human esophageal carcinogenesis: Results from immunohistochemical and mutation analyses of carcinomas and nearby noncancerous lesions. Carcinogenesis 20:591-597, 1999 32. Tian D, Feng Z, Hanley NM, et al: Multifocal accumulation of p53 protein in esophageal carcinoma: Evidence for field cancerization. Int J Cancer 78:568-575, 1998 33. Pellegrini G, Dellambra E, Golisano O, et al: p63 identifies keratinocyte stem cells. Proc Natl Acad Sci USA 98:3156-3161, 2001 34. Celli J, Duijf P, Hamel BCJ, et al: Heterozygous germline
mutations in the p53 homolog p63 are the cause of EEC syndrome. Cell 99:143-153, 1999 35. Greenblatt M, Bennett WP, Hollstein M, et al: Mutations in the p53 tumor suppressor gene: Clues to cancer etiology and molecular pathogenesis. Cancer Res 54:4855-4878, 1994 36. Hagiwara K, McMenamin MG, Miura K, et al: Mutational analysis of the p63/p73/p51/p40/CUSP/KET gene in human cancer cell liens using intronic primers. Cancer Res 59:4165-4169, 1999 37. Osada M, Ohba M, Kawahara C, et al: Cloning and functional analysis of human p51, which structurally and functionally resembles p53. Nat Med 4:839-843, 1998 38. Tani M, Shimizu K, Kawahara C, et al: Mutation and expression of the p51 gene in human lung cancer. Neoplasia 1:71-79, 1999 39. Shields H, Rosenberg SJ, Zwas FR, et al: Prospective evaluation of multilayered epithelium in Barrett’s esophagus. Gastroenterology 116:A309, 1999 40. Adler RH: The lower esophagus lined by columnar epithelium: Its association with hiatal hernia, ulcer stricture, and tumor. J Thorac Cardiovas Surg 45:13-32, 1963 41. Meyer W, Vollmar F, Bar VW: Barrett-esophagus following total gastrectomy. Endoscopy 2:121-126, 1979 42. Gillen P, Keeling P, Byme PJ, et al: Experimental columnar metaplasia in the canine oesophagus. Br J Surg 75:113-115, 1988 43. Li H, Walsh TN, O’Dowd G, et al: Mechanisms of columnar metaplasia and squamous regeneration in experimental Barrett’s esophagus. Surgery 115:176-181, 1994 44. Bremnar CG, Lynch VP, Ellis FH: Barrett’s esophagus: Congenital or acquired? An experimental study of esophageal mucosal regeneration in the dog. Surgery, 68:209-216, 1970 45. Li H, Walsh TN, O’Dowd G, et al: Mechanisms of columnar metaplasia and squamous regeneration in experimental Barrett’s esophagus. Surgery 115:176-181, 1994 46. Park JJ, Sun D, Quade BJ et al: Stratified mucin-producing intraepithelial lesions of the cervix: Adenosquamous or columnar cell neoplasia? Am J Surg Pathol 24:1414-1419, 2000 47. Quade BJ, Yang A, Wang Y, et al: Expression of the p53 homologue p63 in early cervical neoplasia. Gynecol Oncol 80:24-29, 2001
1165
negative in all other epithelia of the gastrointestinal tract, including the stomach, small intestine, and colon. Similarly, p63 was not expressed in BE, but it, was present in the basal layer of multilayered epithelium in 9 of 13 cases (69%). p63-positive cells in multilayered epithelium and in the mucosal gland duct epithelium were positive for CK8/18 (100%) and CK13 (67% and 30%, respectively) and negative for CK14 (0%), in contrast to p63-positive cells in squamous epithelium, which were positive for CK14 and CK13 (100%) but negative for CK8/18. In neoplastic tissues, p63 was diffusely expressed in all cases of esophageal squamous cell dysplasia and carcinoma but was negative in all cases of esophageal and colorectal adenocarcinoma. The ⌬N isoform of p63 mRNA predominated in all benign and neoplastic squamous tissues examined. p63 may represent a marker of 2 distinct epithelial progenitor cells (basal squamous epithelium and gland duct epithelium) in the esophagus. P63 is upregulated in squamous neoplastic conditions and in this manner may play a role in squamous carcinogenesis. These data also indicate that multilayered epithelium is phenotypically similar to, and may share a lineage relationship with, mucosal gland duct epithelium. HUM PATHOL 32:1157-1165. Copyright © 2001 by W.B. Saunders Company Key words: p63, Barrett esophagus, squamous cell carcinoma, intestinal metaplasia. Abbreviations: GI, gastrointestinal; BE, Barrett esophagus; CK, cytokeratin; PBS, phosphate-buffered saline; H&E, hematoxylin and eosin; RT-PCR, reverse-transcription polymerase chain reaction.
The tumor-suppressor gene p53 is altered in a wide range of human malignancies.1 For instance, germline mutation in the p53 gene results in increased susceptibility to tumors in mice2,3 and humans.1,4,5 p53 is known to induce cell cycle arrest and/or apoptosis in response to cellular stresses such as DNA damage or hypoxia.6-8 Accordingly, loss of p53 function confers a growth ad-
vantage on cells via a mechanism that induces dysregulation of growth and resistance to apoptosis.5,8 Recently, 2 genes homologous to p53, each with distinct expression patterns and biologic activities, have been identified in mice and humans. The human p73 gene is widely expressed in benign tissues and is located on the short arm of chromosome 1, a region frequently lost in neuroblastomas.9,10 The p63 gene also shows homology to p53 in the DNA binding, transactivation, and oligomerization domains, but its expression in humans is most apparent in a small subset of benign epithelial tissues, including skin, squamous epithelia of the cervix and oral cavity, and prostatic acinar basal cells.11 Targeted disruption of the murine p63 gene results in neonatal death. Homozygotes show profound defects in limb and craniofacial development and in the differentiation of tissues with stratified epithelium, such as the skin, oral cavity, and esophagus.12,13 The restricted expression of p63 in epithelial cells,
From the Department of Pathology, Brigham and Women’s Hospital, and the Department of Cell Biology, Harvard Medical School, Boston, MA. Accepted for publication July 24, 2001. Presented in part at the United States and Canadian Academy of Pathology, Atlanta, GA, March 2001. Address correspondence and reprint requests to Jonathan N. Glickman, MD, PhD, Department of Pathology, Brigham and Women’s Hospital, 75 Francis St, Boston, MA 02115. Copyright © 2001 by W.B. Saunders Company 0046-8177/01/3211-0003$35.00/0 doi:10.1053/hupa.2001.28951
1157
HUMAN PATHOLOGY
Volume 32, No. 11 (November 2001)
and its functional similarity to p53, suggests that p63 may play a role in the regulation of proliferation and differentiation of specific epithelial cell types. However, the expression of p63 in the human gastrointestinal (GI) tract, which is lined by a variety of differentiated epithelial cell types, has not been studied. Our hypothesis in this study is that p63 is a marker of epithelial progenitor cells in the esophagus and as such plays a role in metaplastic and neoplastic disorders of this organ. In this study, we were particularly interested in the expression of p63 in Barrett esophagus (BE), squamous cell carcinoma, and adenocarcinoma. In BE, the stratified squamous epithelium of the lower esophagus is replaced by intestinal-type columnar epithelium containing goblet cells.14,15 The cell of origin of BE is unknown, but members of our group have postulated that this condition develops from a multipotential epithelial stem cell that can differentiate into either squamous or columnar epithelium.16,17 This hypothesis is supported by the finding of a distinctive type of multilayered epithelium associated with BE, which has morphologic and immunophenotypic characteristics intermediate between squamous and columnar epithelium.18-20 Clarification of the cell of origin of BE may provide insight into the pathogenesis of this condition and ultimately may lead to better forms of treatment for this disorder. Therefore, we performed this study to characterize the expression of p63 in the human GI tract and particularly to evaluate the expression of this protein in metaplastic, dysplastic, and malignant tissues of the esophagus. These studies were intended to evaluate the hypothesis that p63 is a marker of epithelial precursor cells in the esophagus. The pattern of p63 staining was compared with that of the proliferation marker Ki 67 and to cytokeratin (CK) 13 (suprabasal squamous marker), 14 (basal squamous marker), 8/18 (columnar marker), and 19 (basal/columnar marker) to determine the biologic and differentiation characteristics of p63-positive cells. Because p53 gene losses or mutations are frequently seen in GI malignancies, including esophageal squamous cell carcinoma and adenocarcinoma,1,5,15,21 we were also interested in the relationship between p63 and p53 expression. MATERIALS AND METHODS Patient Selection Study cases were selected by a retrospective review of the surgical pathology files of the Brigham and Women’s Hospital (BWH) and included patients who underwent an endoscopic biopsy or surgical resection of the esophagus, stomach, small intestine, or colon between 1994 and 1998. The patients included 7 (2 biopsies, 5 resections) with esophageal squamous cell carcinoma, including 4 with coexisting squamous cell carcinoma in situ, 7 (6 biopsies, 1 resection) with BEassociated adenocarcinoma, 10 (all biopsies) with BE (defined as replacement of the lower esophageal mucosa by intestinal-type columnar epithelium with goblet cells; average length, 7.0 cm; range, 2 to 15 cm), 13 who had multilayered
epithelium (12 biopsies, 1 resection) associated with intestinal metaplasia (defined as described previously17-19) and 3 (all resections) with colonic adenocarcinoma. In addition, esophageal mucosal and submucosal glands and ducts were present in 14 of the above-listed cases and were evaluated as well. Normal controls included biopsy specimens from patients with an endoscopically and histologically normal-appearing gastroesophageal junction (n ⫽ 3) containing both squamous and gastric cardia-type mucosa, gastric antrum (n ⫽ 4), duodenum (n ⫽ 2), jejunum (n ⫽ 3), ileum (n ⫽ 2), or colon (n ⫽ 3).
Histologic Analysis Tissues were fixed in 10% formalin, embedded in paraffin, and processed routinely. Five-micron-thick tissue sections were stained with hematoxylin and eosin (H&E) and examined for the presence of various epithelial cell types, including benign squamous epithelium, specialized-type intestinal epithelium (identified by the presence of goblet cells admixed with mucous cells), dysplastic columnar or squamous epithelium (identified as previously described22), and invasive squamous cell carcinoma or adenocarcinoma. Esophageal multilayered epithelium was identified using previously published criteria.17-19 Briefly, multilayered epithelium shows 4 to 7 layers of cells composed of compact squamous-appearing cells in the basal portion and columnar cells in the superficial layers.
Immunohistochemical Analysis Tissue blocks representative of each case were selected for immunohistochemical analysis. In cases of dysplasia or malignancy, blocks containing both tumor and adjacent nondysplastic mucosa were used. Immunohistochemical staining was performed as follows: tissue sections were incubated for 60 minutes at 60°C, deparaffinized, and rehydrated in graded ethanol solutions. Endogenous peroxidase activity was blocked by incubation of the slides in 3% H2O2 (Fisher Scientific, Pittsburgh, PA) for 5 minutes. The slides were then rinsed under running water for 5 minutes and transferred to phosphate-buffered saline (PBS; Fisher Scientific). Sections pretreated with heat-induced epitope retrieval were transferred to 10 mmol/L citrate buffer, pH 6.0, and heated for 30 minutes at 199°F in a microwave oven equipped with a temperature probe. The slides were then rinsed under running water and transferred to PBS (Fisher Scientific). The tissue sections were blocked with 2% horse serum (Vector Laboratories, Burlingame, CA) for 15 minutes and incubated with the antibodies (p63, p53, CK13, CK14, CK19, CK8/18, Ki 67) at the dilutions listed in Table 1. All antibodies were incubated for 1 hour at room temperature in a humid chamber. The slides were then washed for 5 minutes in PBS and incubated with a biotinylated secondary antibody (mouse immunoglobulin [Ig] G) and then avidin– horseradish peroxidase (Vectastain Elite ABC kit; Vector Laboratories) according to the manufacturer’s instructions. The slides were washed in PBS between incubations. The tissue sections were then developed using 3,3⬘-diaminobenzidine (DAB; Sigma) as a substrate and counterstained with Gill hematoxylin (Fisher Scientific) according to the manufacturer’s instructions. Positive staining control for p63 was normal human tonsil tissue. Other positive staining controls were used according to the recommendations of the antibody supplier. Tissue sections in which the primary antibody was substituted for one of the same IgG subclass, but different antigenic specificity, served
1158
p63 IN GI TRACT (Glickman et al)
TABLE 1. Summary of Antibodies Used in This Study Antigen
Function
Clone
Supplier
p63 p53 CK13 CK14 CK19 CK8/18 Ki 67
p53 homolog Tumor suppressor Squamous marker Squamous marker Columnar marker Columnar marker Proliferation marker
4A4 DO-1 AE8 LL002 RCK108 CAM5.2 MIB-1
Yang et al11* Biogenex (San Ramon, CA) Biogenex Novocastra (Burlingame, CA) Biogenex Becton-Dickinson (Mountain View, CA) AMAC (Westbrook, ME)
Dilution 1:50† 1:100† 1:100 1:100† 1:300† 1.5† 1:200
NOTE. All antibodies are mouse IgG *Antibody also available from Santa Cruz Biotechnology (catalog no. sc-8431). †Pretreatment with heat-induced epitope retrieval as described in Materials and Methods.
as negative staining controls. In addition, the specificity of the anti-p63 antibody has been previously demonstrated in a variety of immunoblotting experiments, and in experiments that used immunohistochemical staining of tissues from mice that contain deletion of the p63 gene.11,13 Immunohistochemical staining for each marker was scored by one of the authors (J.G.) as positive or negative in the epithelial type under study. For p63 labeling, only cells with a strong nuclear staining signal were counted as positive. The Ki 67–labeling index for each epithelial type was calculated as the percentage of cells staining positive for this marker (at least 200 cells counted).
Reverse-Transcription Polymerase Chain Reaction Analysis of p63 Expression Mucosal tissue from 6 esophagectomy specimens (2 squamous cell carcinoma, 3 BE-associated adenocarcinoma, 1 BE), 3 gastrectomy specimens (all for peptic ulcer disease), and 3 colectomy specimens (all for colorectal adenocarcinoma) was harvested within 2 hours of the procedure, in compliance with protocols approved by the BWH Human Studies Committee. Mucosal tissue was harvested by scraping the mucosal surface with a sterile scalpel from the following areas: normal squamous mucosa (n ⫽ 4), BE (n ⫽ 4), tumor tissue (2 squamous cell carcinoma, 2 adenocarcinoma), normal gastric mucosa from the corpus (n ⫽ 4) or antrum (n ⫽ 3), normal colonic mucosa (n ⫽ 3), and colonic adenocarcinoma (n ⫽ 3). The identities of the tissues sampled were confirmed by examination of formalin-fixed, paraffin-embedded H&E-stained sections of adjacent mucosa and by examination of smear preparations of the scraped tissue. Total cellular RNA was isolated using Trizol reagent (Gibco BRL, Gaithersburg, MD) according to the manufacturer’s recommendations and dissolved in diethyl pyrocarbonate–treated distilled water at a concentration of 2 g/L. For reverse-transcription polymerase chain reaction (RTPCR) analysis of p63 messenger RNA (mRNA)-isoforms, 2 g of total RNA was reverse-transcribed with the p63-specific reverse primer 5⬘-AGCTCATGGTTGGGGCAC using SuperScript II (Gibco BRL) according to the recommendations of the enzyme manufacturer. One twentieth of this reaction was used to amplify p63 isoforms using Taq polymerase (Perkin Elmer, Shelton, CT) in a Perkin Elmer 9600 thermal cycler with the following cycling parameters: 2 minutes at 95°C; 40 cycles of 15 seconds at 52°C; 45 seconds at 72°C; 15 seconds at 95°C; and 8 minutes at 72°C. Primers were 5⬘-ATTCCCAGAGCACACAG and 5⬘-AGCTCATGGTTGGGGCAC for the TA form (600 – base pair [bp] product), and 5⬘-CAGACTCAATTTAGTGAG and 5⬘-AGCTCATGGTTGGGGCAC for the ⌬N form (400-bp product). For analysis of glyceralde-
hyde-3-phosphate dehydrogenase (GAPDH) mRNA as a control for mRNA quality and content, 2 g of total RNA was reverse-transcribed with oligo-dT(12–18) and SuperScript II (Gibco BRL). This reaction was amplified with the primers 5⬘-CACATCGCTCAGACACCATG and 5⬘-GCCATGGAATTTGCCATGGG using the cycling parameters above, except that 30 cycles were used and the annealing temperature was 60°C. RNA isolated from ME180 cervical carcinoma cells11 served as positive controls. First-strand synthesis reactions run in the absence of reverse transcriptase served as negative controls. Reaction products were analyzed on 2% agarose gels, stained with ethidium bromide, and photographed using a Kodak DC120 (Kodak, Rochester, NY) digital camera and Photoshop 4.0 software (Adobe Corp, San Jose, CA).
Statistical Analysis Statistical comparisons of immunohistochemical staining results were done using the Fisher exact text and Student t test, as appropriate. A P value of ⬍ .05 was considered statistically significant.
RESULTS Expression of p63 in the GI Tract Normal esophageal squamous epithelium showed strong nuclear reactivity for p63 in all cases (100%). Staining was present in all cells of the basal layer and in the suprabasal layer of cells (Fig 1A). Reactivity was also present in the middle layers of the epithelium, in a lower proportion of cells. Rare staining was noted in the superficial squamous cells [8/20 cases (40%)]. In contrast, columnar surface and crypt epithelium in the gastric cardia, gastric antrum, duodenum, jejunum, ileum, and colon showed no reactivity for p63. In addition, p63 stained basal cells in esophageal mucosal [10/11 cases (91%)] and submucosal [3/3 cases (100%)] gland ducts, but did not stain the mucinous columnar epithelium of the glands of either type (Fig 1B). Expression of p63 in Esophageal Squamous Neoplasia Squamous mucosa from areas of active esophagitis showed an expanded zone of p63-positive cells that involved from 75% to 100% of the mucosal area (Table 2). All cases of squamous dysplasia and squamous cell
1159
HUMAN PATHOLOGY
Volume 32, No. 11 (November 2001)
FIGURE 1. Expression of p63 in the human gastrointesinal tract. (A) Esophageal squamous epithelium showing p63 nuclear reactivity in basal and suprabasal layers of the epithelium. (Original magnification ⫻200.) Esophageal mucosal gland duct showing p63 positivity in the basal layer. (Original magnification ⫻200.)
carcinoma showed strong, diffuse expression of p63 in more than 90% of cells (Fig 2A). Similarly, all cases (100%) of invasive squamous cell carcinoma showed strong diffuse expression of p63 in the tumor cells. Expression of p63 in BE, Multilayered Epithelium, and BE-Associated Adenocarcinoma Intestinal-type epithelium from BE was negative for p63 in all cases. However, BE-associated multilayered epithelium contained p63-positive cells in 9 (69%) of 13 cases. p63-positive cells were located in the basal or immediate suprabasal portions of this epithelium (Fig 3A). P63 staining was absent from all cases of BE-associated dysplasia and adenocarcinoma, as well as from 3 cases of colorectal adenocarcinoma. Comparison of p63 Staining With p53 Expression p53 nuclear staining in benign squamous epithelium was weak and limited to a minority (⬍5%) of cells in the basal layer of the epithelium (Fig 2B). However, strong nuclear reactivity for p53 was present in all
(100%) squamous dysplasia cases and in 5 (71%) of 7 squamous cell carcinomas. The staining pattern of p63 and p53 was similar in these cases (Fig 2C). Strong p53 nuclear staining was present in 4 of 13 cases of multilayered epithelium, all of which were also positive for p63. Staining for both markers was similar and was located in the basal layer of the epithelium. Although BE and BE-associated dysplasia and adenocarcinoma cases were all negative for p63, 0%, 75%, and 57% of these cases, respectively, stained positive for p53 (Table 2). Comparison of p63 Staining With Ki 67 Labeling Ki 67 is a proliferation-associated nuclear antigen expressed in the G1, S, and G2 phases of the cell cycle and in this manner labels all proliferating cells. Ki 67 staining highlighted a narrow strip of cells in the immediate suprabasal region of normal esophageal squamous epithelium (Fig 2D). These cells were also reactive for p63. The p63-positive cells in squamous dysplasia and invasive squamous cell carcinoma were also reactive with Ki 67 (not shown).
TABLE 2. Summary of p63, p53, and Ki 67 Expression in the Esophagus and Related Metaplastic and Neoplastic Disorders Marker Ki 67 Epithelium Type Squamous epithelium Normal (n ⫽ 20) Dysplasia (n ⫽ 4) Carcinoma (n ⫽ 7) BE Intestinal metaplasia (n ⫽ 10) Multilayered epithelium (n ⫽ 13) Dysplasia (n ⫽ 12) Adenocarcinoma (n ⫽ 7)
p63
p53
No. (%) Positive
Mean % of Cells Positive
20 (100%) 4 (100%) 7 (100%)
0 (0%) 4 (100%) 5 (71%)
20 (100) 4 (100) 7 (100)
50 69 72
0 (0%) 9 (69%) 0 (0%) 0 (0%)
0 (0%) 4 (31%) 9 (75%) 4 (57%)
10 (100) 13 (100) 12 (100) 7 (100)
11 12* 85 76
*Only in basal cells.
1160
p63 IN GI TRACT (Glickman et al)
FIGURE 2. Comparison of p63 expression with P53 and Ki67 in esophageal squamous cell carcinoma. (A) p63, showing diffuse reactivity in invasive squamous cell carcinoma. (Original magnification ⫻400). (B) p53, showing rare reactive cells in the basal layer of normal squamous epithelium. (Original magnification ⫻200). (C) p53, showing diffuse reactivity in invasive squamous cell carcinoma. (Original magnification ⫻400.) (D) Ki67 proliferation marker, showing reactivity in suprabasal layers of normal squamous epithelium. (Original magnification ⫻200.)
The p63-positive cells of esophageal multilayered epithelium and in the esophageal gland ducts were positive for the proliferation marker Ki 67 in 100% and 88% of cases, respectively, in a minority of cells (Fig 3B). All cases of BE, BE-associated dysplasia, and adenocarcinoma (all p63 negative) contained Ki 67–positive cells. Ki 67 staining in BE was limited to the crypt region of the epithelium, whereas in dysplasia and adenocarcinoma cases, the staining was strong and diffuse (⬎80% of cells positive). Comparison of p63 Staining With Cytokeratin Expression To further define the differentiation state of p63positive epithelial cells in the esophagus, we compared the staining pattern of this protein with that of a variety of cell type-specific CKs, including the basal squamous marker CK14,23-27 the suprabasal squamous marker CK13,24-26 the columnar markers CK8/18,23,24 and the basal squamous/columnar marker CK1924,28,29 (Table 3). In normal squamous epithelium and in esophagitis, p63positive cells were also positive for CK14 and CK19, focally weakly reactive for CK13 (100% of cases), and negative for the columnar markers CK8/18. Squamous dysplasia and squamous cell carcinoma cases were positive for CK14 and CK19 and negative for CK13 and CK8/18.
p63-positive cells in multilayered epithelium and in esophageal mucosal gland ducts were also positive for CK8/18 (Fig 3E) and CK19 in all cases. They were focally positive for CK13 (67% and 30% of cases, respectively; Fig 3C), and negative for the basal squamous CK14 (Fig 3D). The keratin immuonophenotype of p63-positive esophageal multilayered epithelium (CK8/18 and CK19 positive, CK13 focally positive, CK14 negative) differed significantly from that of p63-positive squamous epithelium (CK14 and CK19 positive, CK13 focally positive, CK8/18 negative) and from that of p63negative intestinal-type BE (CK8/18 and CK19 positive, CK13 and CK14 negative). Expression of p63 mRNA Isoforms RT-PCR analysis of p63 expression in esophageal squamous epithelium showed that the ⌬N form was the predominant mRNA isoform in all cases tested (4 of 4), and the TA form was found only occasionally (1 of 4 cases) in this type of epithelium (Fig 4). Two cases of esophageal squamous cell carcinoma also expressed the ⌬N isoform. Weak expression of both the TA and ⌬N isoforms was occasionally present in BE (1/4 cases; 25%). Neither p63 isoform was detected in gastric cor-
1161
HUMAN PATHOLOGY
Volume 32, No. 11 (November 2001)
FIGURE 3. Expression of p63 in multilayered epithelium associated with BE; comparison with CK expression. (Original magnification ⫻400.) (A) p63. Multilayered epithelium is positive in the basal layer, whereas adjacent intestinal-type epithelium (right) is negative. (B) Ki67–proliferation marker showing focal reactivity in basal cells. (C) CK13 (suprabasal squamous marker). Multilayered epithelium is positive in the basal portion of the epithelium, whereas adjacent intestinal-type epithelium (right) is negative. (D) CK14 (basal squamous marker). Multilayered eptihelium is negative. (E) CK8/18 (columnar marker) positive in multilayered epithelium and adjacent intestinaltype epithelium (right).
pus or antral epithelium, or in normal colonic epithelium or adenocarcinoma. DISCUSSION The results of this study showed that in the normal esophagus, p63 expression was restricted to the basal and suprabasal cell layers of normal esophageal squamous epithelium and the basal cells of esophageal mucosal and submucosal gland ducts. P63 expression was upregulated in esophageal squamous dysplasia and squamous cell carcinoma. p63-positive cells in squamous dysplasia and carcinoma showed an immature keratin immunophenotype (CK14 positive, CK13 negative), similar to the p63-positive cells in the basal portion of normal squamous epithelium. These results suggest that p63 may play a role in squamous carcino-
genesis in the esophagus and that squamous dysplasia and carcinoma may be derived from the basal cells of the esophageal squamous epithelium. In fact, there are other data to support this concept. For instance, p53 overexpression and/or mutation have been found in the basal cells of nonneoplastic squamous mucosa adjacent to squamous cell carcinomas in a high percentage of cases.30-32 The results of this study are consistent with previous analyses of p63 expression in murine tissues11-13 and in keratinocyte cultures33 and support the hypothesis that p63 is a marker of precursor cells and may play a role in the regulation of epithelial differentiation and proliferation in the esophagus. The severe derangements in epithelial development displayed by mice mutant for the p63 gene12,13 are also consistent with this hypothesis. Furthermore, heterozygous germline muta-
1162
p63 IN GI TRACT (Glickman et al)
TABLE 3. CK Immunophenotype of p63-Positive Epithelial Cells in Normal Squamous Epithelium, Multilayered Epithelium, and Esophageal Mucosal Gland Duct Epithelium Epithelium Type
Marker
Normal Squamous Epithelium (n ⫽ 20)
Esophageal Multilayered Epithelium (n ⫽ 9)
Esophageal Mucosal Gland Duct Epithelium (n ⫽ 10)
CK14 CK13 CK19 CK8/18
19/19 (100%) 18/18 (100%) 16/16 (100%) 0/17 (0%)
0/9 (0%)* 6/9 (67%) 9/9 (100%) 9/9 (100%)*
0/9 (0%)* 3/10 (30%)* 10/10 (100%) 10/10 (100%)*
*P ⬍ .05 compared with normal squamous epithelium.
tions in the p63 gene cause the human autosomal dominant syndrome known as EEC (Ectrodactyly, Ectodermal dysplasia, Cleft lip),34 which has limb and craniofacial defects similar to those seen in p63-null mice. The results of previous functional studies of p63 have suggested a possible mechanism by which this protein may regulate epithelial differentiation and proliferation. The p63 gene product exists in several alternatively spliced forms with distinct activities.11 The ⌬N form, lacking the transactivation domain, is the predominant form in normal squamous epithelium and in squamous cell carcinoma, as demonstrated by the results of this study. By RT-PCR analysis, the ⌬N mRNA isoform was present in 4 of 4 samples of normal squamous epithelium and in 2 of 2 samples of squamous cell carcinoma, whereas the TA mRNA isoform containing the transactivation domain was present in only 1 case of normal squamous epithelium and was absent from all squamous cell carcinomas. The ⌬N isoform is capable of interacting with wild-type p53 and inhibiting its function in a dominant negative fashion.11 Therefore, expression of this protein might be expected to block the growth-inhibitory and apoptosis-inducing activities of p53 or of signals that act through p53 and thus could help maintain the proliferative capacity of progenitor cells, as originally proposed by Yang et al.13 In esophageal squamous neoplasia, p63 expression may be upregulated secondary to failure of terminal differentiation and, as such, may have growth-promoting and/or carcinogenic activity. However, based solely on the results of our study, we cannot be certain that p63 plays a role in squamous carcinogenesis. Tissue culture or animal studies are needed to resolve this issue adequately. Although p63 and p53 have structural and functional similarity,11,13 our results, as well as those of previous studies, indicate that these 2 homologous proteins may play distinct roles in vivo. The high levels of p53 observed in many tumors are usually the consequence of inactivating mutations that result in the accumulation of nonfunctional p53 protein.1,35 In contrast, p63 is constitutively expressed in normal epithelial progenitor cells, and its overexpression in squa-
mous neoplasia may reflect the immaturity of the tumor cell lineage. An expanded zone of p63 expression in areas of esophagitis with basal cell hyperplasia is consistent with this hypothesis. In this study, p63 expression was not observed in BE or in BE-associated dysplasia or adenocarcinoma. Thus, it is unlikely that p63 is involved in BE-associated carcinogenesis. However, our finding of p53 overexpression in these neoplasms does support the wellknown role of this protein in BE-associated carcinogenesis.15 Although p63 and p53 may interact, our data suggest that the mechanism of action of p53 in BEassociated neoplasia is independent of p63. Our finding of an absence of p63 expression in colonic adenocarcinomas, as well as in a variety of other tumors,11 supports this theory. In fact, molecular studies of colon and lung carcinoma cell lines have shown only rare mutations in p63, but frequent mutations of p53.36-38 Further studies are needed to determine whether p63 mutations that occur in human tumors are clinically relevant. Another interesting result of this study was that p63 did not stain any cases of BE but was present in multilayered epithelium. Multilayered epithelium is a distinctive type of epithelium with morphologic and immunophenotypic features intermediate between squamous and columnar epithelium20 and has been shown prospectively to be highly associated with BE.39 Furthermore, multilayered epithelium has also been noted to have mucin histochemical and immunophenotypic properties similar to BE.18-20 As a result, multilayered epithelium was hypothesized by our group to represent an early or intermediate stage in the devel-
FIGURE 4. Expression of p63 mRNA isoforms. Ethidium bromide stained agarose gel of representative RT-PCR products: lane1, normal gastric mucosa; lane 2, BE; esophagus; lane 3, normal esophageal squamous epithelium; lane 4, esophageal squamous cell carcinoma; lane 5, positive control ME180 cells. Control GAPDH (192 bp, top row) is positive in all samples; p63 TA form (600 bp; middle row) is negative in all samples, except for the positive control; p63 ⌬N form (400 bp, bottom row) is positive in normal squamous epithelium and in squamous cell carcinoma.
1163
HUMAN PATHOLOGY
Volume 32, No. 11 (November 2001)
opment of BE.20 Our finding of p63-positive cells in multilayered epithelium supports this hypothesis and suggests that multilayered epithelium may contain epithelial progenitor cells capable of differentiating toward a columnar phenotype. Experimental systems using either cell culture or animal models are needed to more fully demonstrate a direct lineage relationship between BE and ME. We found that p63-positive cells in multilayered epithelium and in esophageal mucosal gland duct epithelium were phenotypically similar (CK8/18 and CK19 positive, focally CK13 positive, CK14 negative) and significantly different from normal esophageal basal squamous cells (CK14 and CK19 positive, focally CK13 positive, CK8/18 negative). These data raise the possibility that multilayered epithelium may arise from mucosal gland duct epithelium and that mucosal gland duct epithelium contains a population of p63-positive cells distinct from basal squamous epithelium. Although the cell of origin of BE is essentially unknown, some have postulated that a multipotent epithelial progenitor cell may reside either in the basal layer of normal esophageal squamous epithelium or in esophageal gland ducts.40,41 Support for these theories stems from experimental studies in dogs in which BE was found to develop from cells that are intrinsic to the esophagus.42-45 Furthermore, in 1963 Adler40 noted that mucosal gland duct epithelium often appeared to be “growing out” in an attempt to re-epithelialize eroded surface epithelium in a pathologic study of 6 patients with gastroesophageal reflux disease. Finally, previous studies on the morphologic and phenotypic properties of multilayered epithelium performed by our group have shown that multilayered epithelium may occur within esophageal gland ducts and often in continuity with columnar epithelium at the surface of the mucosa.20 The previously reported association of multilayered epithelium with BE, combined with our finding of p63 in multilayered epithelium, raises an interesting comparison between multilayered epithelium and the epithelium of the cervical transition zone. Both of these epithelial types occur at or near squamocolumnar junctions and are associated with metaplasia (eg, squamous metaplasia of the endocervical mucosa). Furthermore, the reserve cells of the cervical transition zone, like the basal cells of multilayered epithelium, have also been found to be p63 positive.46 In fact, p63 expression has been noted in benign cervical squamomucinous epithelium, an epithelium with morphologic features intermediate between squamous and columnar epithelium similar to multilayered epithelium. Increased or maintained p63 expression has also been reported in early cervical squamous neoplasia.47 In summary, in this study we characterized the expression of the p53-related protein p63 in the esophagus and in related metaplastic and neoplastic disorders. Our results indicate that p63 expression may be a marker of certain epithelial progenitor cells in the esophagus, including basal squamous epithelium and mucosal gland ducts, and suggest that this protein may
play a role in the regulation of epithelial cell differentiation. Our data also provide evidence that multilayered epithelium and mucosal gland duct epithelium are phenotypically similar and may share a lineage relationship. Finally, although p63 expression is upregulated in esophageal squamous neoplasia, p63 plays no role in BE-associated carcinogenesis. REFERENCES 1. Hollstein M, Sidransky D, Vogelstein B, et al: P53 mutations in human cancers. Science 253:49-53, 1991 2. Donehower LA, Harvey M, Slagle BL, et al: Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumors. Nature 356:215-221, 1992 3. Jacks T, Remington L, Williams BO, et al: Tumor spectrum analysis in p53-mutant mice. Curr Biol 4:1-7, 1994 4. Malkin D, Li FP, Strong LC, Fraumeni JF, et al: Germ line p53 mutations in a familial syndrome of breast cancer, sarcomas, and other neoplasms. Science 250:1233-1238, 1990 5. Levine AJ, Wu MC, Chang A, et al: The spectrum of mutations at the p53 locus. Evidence for tissue-specific mutagenesis, selection of mutant alleles, and a “gain of function” phenotype. Ann NY Acad Sci 768:111-128, 1995 6. Hartwell LH, Kastan MB. Cell cycle control and cancer. Science 266:1821-1828, 1994 7. Ko LJ, Prives C: p53: Puzzle and paradigm. Genes Dev 10: 1054-1072, 1996 8. Polyak K, Xia Y, Zweier JL, et al: A model for p53-induced apoptosis. Nature 389:300-305, 1997 9. Kaghad M, Bonnet H, Yang A, et al: Monoallelically expressed gene related to p53 at 1p36, a region frequently deleted in neuroblastoma and other human cancers. Cell 90:809-819, 1997 10. Jost CA, Marin MC, Kaelin WG: P73 is a human p53-related protein that can induce apoptosis. Nature 389:191-194, 1997 11. Yang A, Kaghad M, Wang Y, et al: p63, a p53 homolog at 3q27-29. Encodes multiple products with transactivating, death-inducing, and dominant-negative activities. Mol Cell 2:305-316, 1998 12. Mills AA, Zheng B, Wang X-J, et al: p63 is a p53 homolog required for limb and epidermal morphogenesis. Nature 398:708713, 1999 13. Yang A, Schweitzer R, Sun D, et al: p63 is essential for regenerative proliferation in limb, craniofacial and epithelial development. Nature 398:714-718, 1999 14. DeMeester SR, DeMeester TR: The diagnosis and management of Barrett’s esophagus. Adv Surg 33:29-68, 1999 15. Jankowski JA, Wright NA, Meltzer SJ, et al: Molecular evolution of the metaplasia-dysplasia-adenocarcinoma sequence in the esophagus. Am J Pathol, 154:965-973, 1999 16. Shields HM, Zwas F, Antonioli DA, et al: Detection by scanning electron microscopy of a distinctive esophageal surface epithelium at the junction of squamous and Barrett’s epithelium. Dig Dis Sci 38:97-108, 1993 17. Sawhney RA, Shields HM, Allan CH, et al: Morphological characterization of the squamocolumnar junction of the esophagus in patients with and without Barrett’s epithelium. Dig Dis Sci 41:10881098, 1996 18. Boch JA, Shields HM, Antonioli DA, et al: Distribution of cytokeratin markers in Barret’s specialized columnar epithelium. Gastroenterology 112:760-765, 1997 19. Chen YY, Wang HH, Antonioli DA, et al: Shahsafaei A, Odze RD. Significance of acid-mucin-positive nongoblet columnar cells in the distal esophagus and gastroesophageal junction. HUM PATHOL 30:1488-1495, 1999 20. Glickman JN, Chen Y, Antonioli D, et al: Characterization of a distinctive multilayered epithelium associated with Barrett’s esophagus. Am J Surg Pathol 25:569-578, 2001 21. Shi ST, Yang GY, Wang LD, et al: Role of p53 gene mutations in human esophageal carcinogenesis: Results from immunohistochemical and mutation analyses of carcinomas and nearby noncancerous lesions. Carcinogenesis 20:591-597, 1999
1164
p63 IN GI TRACT (Glickman et al) 22. Dawsey SM, Lewin KJ: Histologic precursors of squamous esophageal cancer. Pathol Annu 30:209-226, 1995 23. Moll R, Franke WW, Schiller DL, et al: The catalog of human cytokeratins: Patterns of expression in normal epithelia, tumors and cultured cells. Cell 31:111-124, 1982 24. Smedts F, Ramaekers F, Troyansovsky S, et al: Keratin expression in cervical cancer. Am J Pathol 141:497-511, 1992 25. Takahashi H, Shikata N, Sanzaki H, et al: Immunohistochemical staining patterns of keratins in normal oesophageal epithelium and carcinoma of the esophagus. Histopathology 26:45-50, 1995 26. Viaene AI, Baert JH: Expression of cytokeratin mRNAs in normal human esophageal epithelium. Anat Rec 241:88-98, 1995 27. Smedts F, Ramaekers F, Troyanovsky S, et al: Basal-cell keratins in cervical reserve cells and a comparison of their expression in cervical intraepithelial neoplasia. Am J Pathol 140:601-612, 1992 28. Kwaspen FHL, Smedts FMM, Broos A, et al: Reproducible and highly sensitive detection of the broad spectrum epithelial marker keratin 19 in routine cancer diagnosis. Histopathology 31: 503-516, 1997 29. Bosch FX, Leube RE, Achstatter T, et al: Expression of simple type cytokeratins in stratified epithelia as detected by immunolocalization and hybridization in situ. J Cell Biol 106:1635-1648, 1988 30. Wang LD, Zhou Q, Hong JY, et al: p53 protein accumulation and gene mutations in multifocal esophageal precancerous lesions from symptom free subjects in a high incidence area for esophageal carcinoma in Henan, China. Cancer 77:1244-1249, 1996 31. Shi, ST, Yang GY, Wang LD, et al: Role of p53 gene mutations in human esophageal carcinogenesis: Results from immunohistochemical and mutation analyses of carcinomas and nearby noncancerous lesions. Carcinogenesis 20:591-597, 1999 32. Tian D, Feng Z, Hanley NM, et al: Multifocal accumulation of p53 protein in esophageal carcinoma: Evidence for field cancerization. Int J Cancer 78:568-575, 1998 33. Pellegrini G, Dellambra E, Golisano O, et al: p63 identifies keratinocyte stem cells. Proc Natl Acad Sci USA 98:3156-3161, 2001 34. Celli J, Duijf P, Hamel BCJ, et al: Heterozygous germline
mutations in the p53 homolog p63 are the cause of EEC syndrome. Cell 99:143-153, 1999 35. Greenblatt M, Bennett WP, Hollstein M, et al: Mutations in the p53 tumor suppressor gene: Clues to cancer etiology and molecular pathogenesis. Cancer Res 54:4855-4878, 1994 36. Hagiwara K, McMenamin MG, Miura K, et al: Mutational analysis of the p63/p73/p51/p40/CUSP/KET gene in human cancer cell liens using intronic primers. Cancer Res 59:4165-4169, 1999 37. Osada M, Ohba M, Kawahara C, et al: Cloning and functional analysis of human p51, which structurally and functionally resembles p53. Nat Med 4:839-843, 1998 38. Tani M, Shimizu K, Kawahara C, et al: Mutation and expression of the p51 gene in human lung cancer. Neoplasia 1:71-79, 1999 39. Shields H, Rosenberg SJ, Zwas FR, et al: Prospective evaluation of multilayered epithelium in Barrett’s esophagus. Gastroenterology 116:A309, 1999 40. Adler RH: The lower esophagus lined by columnar epithelium: Its association with hiatal hernia, ulcer stricture, and tumor. J Thorac Cardiovas Surg 45:13-32, 1963 41. Meyer W, Vollmar F, Bar VW: Barrett-esophagus following total gastrectomy. Endoscopy 2:121-126, 1979 42. Gillen P, Keeling P, Byme PJ, et al: Experimental columnar metaplasia in the canine oesophagus. Br J Surg 75:113-115, 1988 43. Li H, Walsh TN, O’Dowd G, et al: Mechanisms of columnar metaplasia and squamous regeneration in experimental Barrett’s esophagus. Surgery 115:176-181, 1994 44. Bremnar CG, Lynch VP, Ellis FH: Barrett’s esophagus: Congenital or acquired? An experimental study of esophageal mucosal regeneration in the dog. Surgery, 68:209-216, 1970 45. Li H, Walsh TN, O’Dowd G, et al: Mechanisms of columnar metaplasia and squamous regeneration in experimental Barrett’s esophagus. Surgery 115:176-181, 1994 46. Park JJ, Sun D, Quade BJ et al: Stratified mucin-producing intraepithelial lesions of the cervix: Adenosquamous or columnar cell neoplasia? Am J Surg Pathol 24:1414-1419, 2000 47. Quade BJ, Yang A, Wang Y, et al: Expression of the p53 homologue p63 in early cervical neoplasia. Gynecol Oncol 80:24-29, 2001
1165