Prognostic value of mucins in the classification of ampullary carcinomas

Human Pathology (2006) 37, 160 – 167

www.elsevier.com/locate/humpath

Prognostic value of mucins in the classification of ampullary carcinomas Friedrich P. Paulsen MD (Professor)a,*, Deike Varoga MDb, Andreas R. Paulsen MDc, Anthony Corfield PhDd, Michael Tsokos MD (Priv-Doz)c a

Department of Anatomy and Cell Biology, Martin-Luther-University of Halle-Wittenberg, D-06097 Halle (Saale), Germany Department of Orthopedics, Christian Albrecht University of Kiel, Kiel, Germany c Institute of Legal Medicine, University Hospital Hamburg-Eppendorf, Hamburg, Germany d Department of Medical Laboratories, University of Bristol, Bristol, UK b

Received 23 June 2005; revised 30 September 2005; accepted 6 October 2005

Keywords: Hepatopancreatic ampulla; Major duodenal papilla; MUC; Prognostic factor; Secretion

Summary The ampulla of Vater is of high clinical relevance with regard to influx of chyme, ascending inflammation, intubation during diagnostic and therapeutic endoscopic investigation, therapeutic papillotomy, and especially to malignant transformation. Little is known about the distribution of mucins in the ampulla. In this study, we have investigated the mucin distribution in the normal ampulla of Vater and compared it to duodenal mucosa and Brunner glands. Expression of mucins in the ampulla of Vater and duodenum was monitored by reverse transcription–polymerase chain reaction and localization of the products by immunohistochemistry. The samples investigated originated from 30 autopsy cases. Mucins MUC1, MUC3, MUC4, MUC5AC, MUC5B, MUC6, MUC7, and MUC8 were expressed in the ampulla of Vater. Immunohistochemistry revealed production of MUC4, MUC5AC, MUC5B, and MUC6. The mucin composition varied in comparison with the duodenum referring to MUC2, MUC7, and MUC8. Detected mucins contribute to innate immunity, epithelial restitution, and protection against the aggressive secretions of the liver, gall bladder, and pancreas. By cross-linking, they influence the rheological properties of the secretions in the ampulla and facilitate unidirectional flow into the duodenum. Knowledge of their pattern of expression has prognostic value with regard to the detection of malignancy. The observed differences in the mucin distribution between the duodenum and the ampulla of Vater support the use of MUC2, MUC7, and MUC8 as useful tool in the classification of ampullary carcinomas. D 2006 Elsevier Inc. All rights reserved.

1. Introduction The ampulla of Vater, also known as the hepatopancreatic ampulla, is a dilation of the duodenal papilla that 4 Corresponding author. E-mail address: [email protected] (F.P. Paulsen). 0046-8177/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2005.10.001

forms the opening of the juncture of the common bile duct and the main pancreatic duct. It is a complex anatomical structure comprising distinct muscle fibers and special neuronal elements that serve to regulate the flow of bile and pancreatic juice [1-3]. However, the mechanisms preventing reflux from the duodenum into the common pancreatic and biliary duct are not completely understood. It

Prognostic value of mucins in the classification of ampullary carcinomas has been suggested that the mucosal surface may contribute to this function as it is composed of mucosal folds filling nearly the whole lumen of the ampulla [4-8]. In this context, epithelial secretions such as the mucins, which have a major influence on the rheology of mucus gels, are of considerable importance. Each region of the gastrointestinal tract has characteristic functional requirements, and the properties of the mucus produced at each site are adapted to cope with these functions [9-11]. Mucins, the major components of mucus, are high-molecular-weight epithelial glycoproteins with clustered O-glycans linked to threonine-, serine-, prolinerich tandem repeat peptide domains. Two structurally and functionally distinct classes of mucins have been identified: the secreted type, represented typically in man by MUC2, MUC5AC, MUC5B, MUC6, MUC7, MUC8, and MUC19, and the membrane-associated type, including MUC1, MUC3, MUC4, MUC12, MUC13, MUC15, MUC16, MUC17, and MUC20 [10 -13]. Secreted and membrane-associated forms are found as extracellular, secreted viscous fluids or viscoelastic polymer gels, or located as membrane-anchored molecules in the glycocalyx [11,12,14,15]. A variety of functions are now ascribed to mucins including the viscoelastic properties of the secreted mucous barrier. Some membrane-associated mucins have been shown to function in signaling and may be important as a sensor mechanism in response to invasion or damage of epithelia [16,17]. Other mucins are able to act as an antiapoptotic agent or participate in bacterial adhesion [18 -22]. Mucin genes are expressed throughout the human gastrointestinal tract in a site-specific manner [9 -11,23]. This pattern of expression has been studied for the entire gut with the exception of the ampulla of Vater. The presumable impact of mucins on the rheological properties of the secretions of liver, gall bladder, and pancreas and their physiological functions relevant to the surface integrity of mucous epithelia led us to a detailed analysis of mucins MUC1, MUC2, MUC3, MUC4, MUC5AC, MUC5B, MUC6, MUC7, and MUC8 in human epithelial cells of the ampulla of Vater.

2. Materials and methods 2.1. Collection of tissue samples from cadavers Thirty biopsy specimens each of ampulla of Vater and duodenum from a 6-month period in 2002 were collected at the Institute of Legal Medicine, University of Hamburg, Germany. Duodenal specimens were taken from an area located nearly 1.5 cm orally to the major duodenal papilla. Tissue specimens were obtained from nonconsecutive autopsy cases (16 men, 14 women; aged 36-78 years; mean age, 56 years) whose deaths were due to various natural and pathological causes (myocardial infarction, n = 5; cerebral hemorrhage, n = 3; drowning; n = 2; hanging, n = 2;

161

trauma/polytrauma, n = 13; gunshot wounds, n = 3; pulmonary embolism, n = 1; heroin intoxication, n = 1). All fatalities occurred outside the hospital, and none of these individuals had a medical history of gastrointestinal disease before death. After myocardial infarction, cerebral hemorrhage, drowning, hanging, trauma/polytrauma, gunshot wounding, pulmonary embolism, or intoxication with heroin, the affected people lived for a few seconds up to 1.5 hours until they died finally. In each case, preexisting pathological conditions of the gastrointestinal tract were ruled out by histological examination of various gastrointestinal (stomach, small intestine, liver, pancreas) tissue sections, and no other disease was found at autopsy except for the cause of death. The mean length of the postmortem interval of the subjects included in this study group was 24.9 F 2.4 hours.

2.2. Specimens For molecular biological assays, 10 tissue samples of both ampulla of Vater and duodenum were prepared, freed from surrounding tissue, placed in phosphate-buffered saline solution, and frozen immediately after collection at ÿ808C. For immunohistochemistry, 20 tissue samples of both ampulla of Vater and duodenum were fixed in 4% paraformaldehyde and embedded in paraffin.

2.3. Controls Control tissues taken from the pancreas, duodenum, conjunctiva, efferent tear ducts, gastric fundus, lacrimal gland, and ethmoid with previously described MUC gene expression patterns and protein production [24 -27] were included with each batch of investigation.

2.4. Total RNA purification and complementary DNA synthesis For conventional reverse transcription–polymerase chain reaction (RT-PCR), a part of each frozen sample was crushed in an agate mortar under liquid nitrogen, then homogenized in 5 mL peqgold RNA Pure solution (peqLab Biotechnologie, Erlangen, Germany) with a Polytron homogenizer, and the insoluble material removed by centrifugation (12 000g, 5 minutes, 48C). RNA was isolated as described by the manufacturer (phenol-guanidinium thiocyanate method). Crude RNA was purified with isopropanol and repeated ethanol precipitation, and contaminating DNA was destroyed by digestion with RNase-free DNaseI (27.27 Kunitz units; 20 minutes, 258C; Boehringer, Mannheim, Germany). This enzyme was heat-inactivated for 15 minutes at 658C. RNA (500 ng) was used for each reaction: complementary DNA (cDNA) was generated with 50 ng/lL (20 pmol) oligo(dT)15 primer (Amersham Pharmacia Biotech, Uppsala, Sweden) and 0.8 lL superscript RNase Hÿ reverse transcriptase (100 U; Gibco, Paisley, UK) for 60 minutes at 378C. Integrity of RNA was controlled by RT-PCR of glyceraldehyde 3-phosphate

162

F.P. Paulsen et al.

dehydrogenase (GAPDH) with intron-spanning primers. No PCR products larger than 360 base pairs were detected. These would indicate contaminating DNA.

2.5. Conventional reverse transcription–polymerase chain reaction analysis For conventional RT-PCR, 4 lL cDNA were incubated with 30.5 lL water, 4 lL 25 mmol/L MgCl2, 1 lL dNTP (Qiagen, Hilden, Germany), 5 lL 10 PCR buffer, 0.5 lL (2.5 U) platinum Taq DNA polymerase (Gibco), and 2.5 lL (10 pmol) primers (for mucin primers MUC1, MUC2, MUC4, MUC5AC, MUC5B, MUC6, MUC7, and MUC8, see Jung et al [25]; for MUC3, see Berois et al [28]). Mucin primers were designed to amplify the 3V nontandem repeat portion of the mucin messenger RNA (mRNA). The PCR cycle consisted of (1) 948C for 30 seconds; (2) 648C (MUC1), 578C (MUC2), 608C (MUC3), 648C (MUC4), 578C (MUC5AC), 64.78C (MUC5B), 648C (MUC6), 608C (MUC7), 648C (MUC8) for 30 seconds each; (3) 728C for 90 seconds; 35 cycles were performed with each primer pair (see Jung et al [25] and Berois et al [28] including product sequencing). A GAPDH (0.1 lmol/L) intron-spanning primer pair served as the internal control for cDNA. All primers were synthesized by MWG-Biotech AG, Ebersberg, Germany. The cDNA was replaced with water for a negative control reaction.

2.6. General histology and immunohistochemistry For immunohistochemistry, reactivity with 11 antibodies against mucin peptide core epitopes was followed in tissue sections (7 lm) from the ampulla of Vater and duodenum. All antibodies (Table 1) were tested with sections subject to microwave heating pretreatment for 10 minutes, with

Table 1

periodate pretreatment for 30 minutes (periodate: 1.93 g ammonium acetate in 500 mL aqua dest at pH 5 and 2.28 g periodic acid), and without pretreatment, as previously detailed [27]. For the origin/original reference of each antibody, see Table 1. It should be stated that no suitable anti-MUC3 antibodies could be found. All primary antibodies were applied overnight at room temperature. Secondary antibodies rabbit antimouse (1:200) or swine antirabbit (1:300; Dako, Glostrup, Denmark) were incubated at room temperature for at least 4 hours. Visualization was achieved with peroxidase-labeled streptavidin-biotin for at least 5 minutes. After counterstaining with hemalum, the sections were mounted in Aquatex (Boehringer). Two negative control sections were used in each case: one was incubated with the secondary antibody only, the other with the primary antibody only. Positive controls were as used for the PCR reactions. All slides were examined by microscope (Axiophot; Carl Zeiss, Oberkochen, Germany).

3. Results 3.1. Reverse transcription–polymerase chain reaction analysis For most of the specific mucin mRNAs tested, specimens of ampulla of Vater were either positive (MUC1, MUC3, MUC4, MUC5AC, MUC5B, MUC6, and MUC7) or negative (MUC2) (Fig. 1, Table 2). MUC8 mRNAs were visible in 2 of 10 samples (Table 2). Duodenal samples revealed positive mucin mRNAs for all mucins tested (MUC1, MUC2, MUC3, MUC4, MUC5AC, MUC5B, MUC6, MUC7, and MUC8) (Fig. 1, Table 2). Only MUC6 was negative in one tested sample (Table 2).

Antibodies used for immunohistochemical staining in paraffin-embedded sections

Mucin

Antibody

Species

Dilution

Pretreatment

PS

Origin/original reference

MUC1 MUC1

BC2 NCL-MUC-1

Mouse Mouse

1:1000 1:100

Microwave microwave

VNTR VNTR

MUC2 MUC4

NCL-MUC-2 1G8

Mouse Mouse

1:150 1:150

VNTR MBD

MUC5AC

NCL-MUC-5AC

Mouse

1:50

VNTR

MUC5B

LUM 5B-2

Rabbit

1:100

Non-VNTR

Wickstrfm et al [30]

MUC5B

5B III

Rabbit

1:100

Non-VNTR

Thornton et al [31]

MUC6 MUC6 MUC7 MUC8

LUM 6-3 NCL-MUC-6 LUM 7-1 Sc-16922

Rabbit Mouse Rabbit Rabbit

1:100 1:100 1:100 1:100

Microwave EDTA according to manufacturer instructions Microwave only or with periodatea Microwave and periodate Microwave and periodate Microwave Microwave Microwave Microwave

Xing et al [29] Novocastra, Newcastle upon Tyne, UK Novocastra Zytomed, San Francisco, Calif Novocastra

Non-VNTR VNTR Non-VNTR Non-VNTR

Lopez-Ferrer et al [32] Novocastra Wickstrfm et al [33] Santa Cruz, Calif

Abbreviations: PS, peptide specificity; VNTR, variable number of tandem repeats; MBD, membrane binding domain. a Reactivity with anti-MUC5AC was observed without periodate pretreatment, but was more intense after pretreatment.

Prognostic value of mucins in the classification of ampullary carcinomas

163

Table 3 Distribution of mucins in the ampulla of Vater and duodenum assessed by immunohistochemistry Mucin

Ampulla of Vater

Duodenum

Columnar Goblet Small Columnar Goblet Brunner cells cells glands cells cells glands MUC1 MUC2 MUC4 MUC5AC MUC5B MUC6 MUC7 MUC8

a a + + + a a a

a a a + + + a a

a a a + + + a a

a a + a a a + +

a + a + + a + a

a a + + + + + a

NOTE. a indicates no positive staining visible; +, positive staining.

antibody against MUC2 (Table 3). MUC4 revealed membrane-bound reactivity in the 20 samples (Fig. 2A and B, Table 3). MUC5AC and MUC5B cytoplasmic immunoreactivity was visible in columnar epithelial cells staining the supranuclear cytoplasm (Fig. 2C-E, Table 3). Moreover, the antibody against MUC5AC and MUC2 antibodies against MUC5B stained the secretory product of goblet cells in all samples (Table 3). Anti-MUC6 antibodies revealed weak

Fig. 1 RT-PCR analysis revealing MUC1, MUC2, MUC3, MUC4, MUC5AC, MUC5B, MUC6, MUC7, and MUC8 expression in 2 samples from the duodenum (lanes 1 and 2) and 2 samples from the ampulla of Vater (lanes 3 and 4). The integrity of the cDNAs was tested by amplification of the GAPDH transcript.

3.2. Immunohistochemistry Paraffin-embedded 7-lm sections from 20 ampullae of Vater revealed no reactivity with 2 antibodies against MUC1 (Table 3). In addition, no reactivity was observed with an Table 2 Expression of mucin mRNA in 20 samples of ampulla of Vater and duodenum Mucin

Ampulla of Vatera

Duodenuma

MUC1 MUC2 MUC3 MUC4 MUC5AC MUC5B MUC6 MUC7 MUC8

20/20 0/20 20/20 20/20 20/20 20/20 20/20 20/20 2/20

20/20 20/20 20/20 20/20 20/20 20/20 18/20 20/20 20/20

a

Only positive results were scored.

Fig. 2 Immunohistochemistry of the ampulla of Vater (A-F). MUC4 (A), MUC4 in higher magnification (B), MUC5AC (C), MUC5AC in higher magnification (D), MUC5B (E), and MUC6 (F) show positive red staining (arrows). Abbreviation: L, ampullary lumen (A and C, original magnification 100; B and D, original magnification 200; E and F, original magnification 75).

164 positive reactivity in mucous parts of seromucous glands and also slight staining of intraepithelial goblet cells (Fig. 2F, Table 3). No reactivity was obtained for MUC7 (Table 3) or MUC8 (Table 3). Analysis of the associated 20 duodenal samples revealed no reactivity for MUC1 (Table 3). MUC2 reactivity was observed in goblet cells of the epithelial lining of the duodenum, whereas columnar cells and Brunner glands were negative (Fig. 3A, Table 3). Anti-MUC4 antibodies showed membrane-bound and cytoplasmic staining of columnar cells and Brunner glands (Fig. 3B and C, Table 3). Positive MUC5AC and MUC5B reactivity was associated with intraepithelial goblet cells and Brunner glands, although the secretory product of goblet cells and Brunner glands was not always stained (perinuclear staining)

Fig. 3 Immunohistochemistry of duodenum (A-H). MUC2 (A), MUC4 (C, arrows), MUC7 (G), and MUC8 (H) show positive red staining in the mucosa, and MUC4 (B), MUC5AC (D), MUC5B (E), and MUC6 (F) reveal positive red staining in Brunner glands (A, B, and F-H, original magnification 75; C, original magnification 200; D, original magnification 100; E, original magnification 50).

F.P. Paulsen et al. (Fig. 3D and E, Table 3). MUC6 was strongly detectable in Brunner glands, but not in the epithelial lining of the duodenum (Fig. 3F, Table 3). MUC7 revealed weak staining of duodenal enterocytes and was present in the secretory product of goblet cells and to a lower intensity in Brunner glands (Fig. 3G, Table 3). MUC8 was associated with columnar cells, but could not be detected in goblet cells or Brunner glands (Fig. 3H, Table 3). Positive controls revealed that all antibodies used except that against MUC3 worked as expected.

4. Discussion To date, there have been very few investigations dealing with mucin distribution under physiological conditions in the gastrointestinal tract. More often, mucins are mentioned in the context of pathophysiological disorders, especially with malignancies. Most authors have studied mucin distribution of MUC1 to MUC7 in the stomach [34,35], duodenum, liver, gall bladder, and pancreas [36] by means of in situ hybridization and detected clear organ and sitespecific differences. The stomach produces mainly MUC5AC and MUC6 [35]. In addition, MUC1 mRNA is observed in gastric epithelium and glands; sporadic occurrence of MUC3 and MUC4 mRNA is seen in some gastric epithelial cells; and weak MUC2 expression occurs in glands of the cardia and antrum [35]. According to Buisine et al [36], MUC5B and MUC7 are not expressed in the stomach. In the duodenum, the major mucins are MUC2 in goblet cells, MUC3 in goblet and columnar cells, and MUC6 in Brunner glands [34,35]. Low levels of MUC1 and MUC2 expression were reported in Brunner glands, whereas MUC4, MUC5AC, MUC5B, and MUC7 mRNAs could not be detected [34,35]. The epithelium of the gall bladder revealed strong expression of MUC3 mRNA and weak MUC5B expression. Occasionally, MUC2, MUC5AC, and MUC6 mRNAs were also observed, and MUC4 and MUC7 mRNAs were always absent. The same patterns found in the gall bladder were also valid for the choledochal duct [34,37]. The excretory duct system of the pancreas revealed strongly positive reactivity for MUC3 mRNA and to a lower degree also for MUC1, MUC5B, and MUC6. MUC2, MUC4, MUC5AC, and MUC7 mRNAs were not observed in any sample [34]. Many of our present results are in accordance with these results, but interestingly, also differ with respect to the cited work at least with regard to the mucin distribution in the duodenum (Table 4). RT-PCR analysis revealed message for MUC3 in all samples of duodenum and ampulla of Vater (Fig. 1, Table 2). However, none of the commercially available antibodies tested immunohistochemically showed specificity. Membrane-bound mucin MUC1 was not detectable by immunohistochemistry in the duodenum and ampulla of Vater, although message could be detected by PCR in all samples and immunohistochemical positive

Prognostic value of mucins in the classification of ampullary carcinomas Table 4 Comparison between the results of Buisine et al [34] and the results of the present work with regard to mucin distribution in the duodenum Mucin

Buisine et al [34] Paulsen et al (present data) In situ hybridization

RT-PCR Immunohistochemistry X

Not detected

X

MUC3

X (low levels, epithelium and Brunner glands) X (major mucin in epithelium, low levels in Brunner glands) X (epithelium)

X

MUC4

Not detected

X

MUC5AC Not detected

X

MUC5B

Not detected

X

MUC6

X

MUC7

X (Brunner glands) Not detected

X (major mucin in epithelial goblet cells, not detected in Brunner glands) Antibodies worked unspecific X (epithelium and Brunner glands) X (epithelium and Brunner glands) X (epithelium and Brunner glands) X (Brunner glands)

X

MUC8

Not investigated

X

MUC1

MUC2

X (epithelium and Brunner glands) X (epithelium)

NOTE. X indicates detection of mucin by the applied method.

controls revealed the expected results. Comparable results were recently obtained by Paulsen et al [27] who were also unable to detect membrane-bound mucin MUC1 in the mucosa of the nasolacrimal ducts from cadavers. An explanation for the negative immunohistochemistry could be because the antibodies used were raised to epitopes within the variable number of tandem repeat region of the peptide core. As a result, these antibodies are likely to have varying reactivities, depending on the glycosylation of the mucins in the tissues under study. Mature mucins, with dense glycosylation, may not react with antibodies directed toward tandem repeat epitopes of the peptide core. Different glycosylation patterns may explain reactivity in the immunohistochemical positive control samples. In contrast to the in situ hybridization study of Buisine et al [34], our RT-PCR and immunohistochemical investigations reveal production of MUC4, MUC5AC, MUC5B, and MUC7, both in Brunner glands and the mucosa of the duodenum. The only identical finding with Buisine et al [34] is the detection of MUC2 in duodenal goblet cells; however, Brunner glands showed no reactivity with the antibody. The major secreted mucin genes, MUC2, MUC5AC, MUC5B, and MUC6, are all located together on chromosome 11p15 [38]. Examples for individual and linked regulation at this locus have been described on a tissue-specific basis. MUC8 can now be added to these mucins. Our investigations demonstrate production of MUC8 by columnar epithelial cells of the duodenum.

165

The epithelial lining of the ampulla of Vater produces MUC4, MUC5AC, MUC5B, and MUC6. All other mucins analyzed showed no reactivity, although message was present for MUC1, MUC3, MUC7, and, in 2 samples, also for MUC8. Apart from MUC4, these findings are in accordance with the results of Vandenhaute et al [37] and Buisine et al [34] in the epithelium of the gall bladder, although these authors detected only weak MUC5B expression and sporadic message for MUC5AC. Our results demonstrate strong production of MUC5B and lower levels of MUC5AC, MUC4, and MUC6 in the ampulla of Vater. Based on the lack of investigations regarding normal mucin distribution in the choledochal duct and pancreatic excretory system, a comparison with the ampulla of Vater is not possible at present. Moreover, we have only investigated MUC1 to MUC8, as antibodies for these mucins were available for this study. Finally, the study is based on autopsy material with a longer postmortem interval that could have introduced a significant loss of expression in the samples. Therefore, presence of additional mucins cannot be excluded. The ampulla of Vater is of great clinical relevance with regard to ascending inflammation and formation of malignancies. The epithelium of the ampulla and the folds reaching into the lumen are protected by its secretions and the secretory products of subepithelial glands. The secretions contain antimicrobial substances such as lysozyme [7] and together with mucins serve to flush aggressive duodenal contents away from the interfold crypts and help to prevent stasis of secretions and potential bacterial proliferation. In this context, the observed production of MUC5B by the epithelium of the ampulla is of interest as it has been shown that both MUC5B and MUC7 are able to bind bacteria [18,20 -22]. The ampulla of Vater connects the duodenum as well as the junction of the pancreatic and choledochal ducts and is the most frequent location for the occurrence of adenoma and carcinoma [39 -42]. In the case of familiar adenomatous polyposis, the ampulla is the principal site of carcinomas after proctocolectomy. The high frequency of carcinomas at this site is believed to be because of the permanent exposure of the junction of 2 different mucosae to secretions from the liver/gall bladder, the pancreas, and the duodenum [40]. According to histological criteria [43], pancreaticobiliary-type carcinomas can be distinguished from intestinal-type carcinomas. In recent years, the mucins, together with other factors, have gained credence as differentiation and prognostic markers for the diagnosis of ampullary carcinomas [40,44 - 52]. Thus, Matsubayashi et al [50] and also Chu et al [45] showed that initial stages of intestinal-type carcinomas are MUC2 positive and always include the ampulloduodenum and, in individual cases, may even be restricted to it. Pancreaticobiliary carcinomas are MUC2 negative and often leave free the ampulloduodenum with their intramucosal component. These tumors appear to develop in the ampullary regions near the pancreas and choledochal duct. Kitamura et al [48] found that the

166 prognosis of ampullary carcinoma was worse with stronger staining for MUC1 and weaker staining for MUC2. Gqrbqz and Klfppel [47] demonstrated that tumor classification is especially effective if a mucin/trefoil family factor (TFF) peptide combination is used. TFF peptides are small mucinassociated peptides that may change the rheological properties of mucus and have recently been classified in normal human ampulla of Vater [53]. Accordingly, in the study of Gqrbqz and Klfppel [47], adenocarcinomas of the ampulla of Vater showed positive reactivity for MUC1 in 73% of cases, 27% for MUC2, 55% for MUC5AC, 46% for MUC6, and 9% for both TFF1 and TFF2. MUC3, MUC4, MUC5B, MUC7, MUC8, and TFF3 were not investigated. Moreover, this study was limited to only 11 cases of papillary carcinoma. Finally, it was observed that MUC5ACpositive ampullary carcinomas initiate from those mucosal areas exposed to pancreatic secretions [40,50]. Our results partly support these observations with regard to carcinomas of the ampulla of Vater. MUC2 was only detectable in the duodenum, but not in the ampulla. MUC5AC and MUC6 were produced in the duodenum as well as in the ampulla of Vater. In conclusion, the observed differences in the mucin distribution between the duodenum and the ampulla of Vater support the suggestion that MUC2 is useful in the classification of ampullary carcinomas and favor the additional investigation of MUC7 and MUC8 for an ampullary carcinoma classification as both mucins are only produced in the duodenum, but not in the ampulla of Vater. Whether MUC3 is a secretory product of the ampulla of Vater remains to be determined.

Acknowledgments The study was supported in part by Deutsche Forschungsgemeinschaft grant PA 738/1-4. The authors thank Karin Stengel and Ute Beyer for excellent technical assistance; Joy Burchell, ICRF, London, UK, for SM3; Ingemar Carlstedt, University of Lund, Sweden, for LUM5B-2, LUM6-3, and LUM7-1; Michael McGuckin, University of Queensland, Australia, for BC2; and David Thornton, University of Manchester, UK, for 5B III.

References [1] Ishibashi Y, Murakami G, Honma T, et al. Morphometric study of the sphincter of Oddi (hepatopancreatic) and configuration of the submucosal portion of the sphincteric muscle mass. Clin Anat 2000; 13:159 - 67. [2] Kunath U, Hommering H. Is the duodenal papilla an autonomic sphincter? A contribution to the functional morphology. Res Exp Med 1981;178:103 - 16. [3] Oddi R. D’une disposition a´ sphincter spe´cial de l’ouverture du canal choledoque. Arch Ital Biol 1887;8:317 - 22. [4] Ffdisch HJ. Feingewebliche studien zur orthologie und pathologie der papilla Vateri. Stuttgart7 Thieme Verlag; 1972.

F.P. Paulsen et al. [5] Giermann H, Holle G. Stereoskopische und histologische untersuchungen zur pathologie des schleimhautreliefs und klappenapparates der papilla Vateri. Acta Hepatosplenol 1981;8:189 - 96. [6] Holle G. Die bauprinzipien der Vaterschen papille und ihre funktionelle bedeutung unter normalen und krankhaften bedingungen. Dtsch Med Wochenschr 1960;85:648 - 55. [7] Paulsen F, Bobka T, Tsokos M, et al. Functional anatomy of the papilla Vateri: biomechanical aspects and impact of papillotomy. Surg Endosc 2002;16:296 - 301. [8] Vater A. Novum bilis diverticulum. Hallers disputationum anatomicum selectarum. Gottingae 1748;3:259 - 73. [9] Byrd JC, Bresalier RS. Mucins and mucin binding proteins in colorectal cancer. Cancer Metastasis Rev 2004;23:77 - 99. [10] Corfield AP, Carroll D, Myerscough N, Probert CS. Mucins in the gastrointestinal tract in health and disease. Front Biosci 2001;6: D1321 - 57. [11] Hollingsworth MA, Swanson BJ. Mucins in cancer: protection and control of the cell surface. Nat Rev Cancer 2004;4:45 - 60. [12] Dekker J, Rossen JW, Buller HA, et al. The MUC family: an obituary. Trends Biochem Sci 2002;27:126 - 31. [13] Gipson IK, Hori Y, Argqeso P. Character of ocular surface mucins and their alteration in dry eye disease. Ocular Surf 2004;2:131 - 48. [14] van Klinken BJW, Dekker J, Buller HA, et al. Mucin gene structure and expression: protection vs adhesion. Am J Physiol 1995;269: G613 - 27. [15] Williams SJ, McGuckin MA, Gotley DC, et al. Two novel mucin genes down-regulated in colorectal cancer identified by differential display. Cancer Res 1999;59:4083 - 9. [16] Carraway KL, Ramsauer VP, Haq B, et al. Cell signalling through membrane mucins. Bioessays 2003;25:66 - 71. [17] Ren J, Li Y, Kufe D. Protein kinase C delta regulates function of the DF3/MUC1 carcinoma antigen in beta-catenin signaling. J Biol Chem 2002;277:17616 - 22. [18] Baughan LW, Robertello FJ, Sarrett DC, et al. Salivary mucin as related to oral Streptococcus mutans in elderly people. Oral Microbiol Immunol 2000;15:10 - 4. [19] Komatsu M, Jepson S, Arango ME, et al. Muc4/sialomucin complex, an intramembrane modulator of ErbB2/HER2/Neu, potentiates primary tumor growth and suppresses apoptosis in a xenotransplanted tumor. Oncogene 2001;20:461 - 70. [20] Liu B, Rayment SA, Gyurko C, et al. The recombinant N-terminal region of human salivary mucin MG2 (MUC7) contains a binding domain for oral streptococci and exhibits candidacidal activity. Biochem J 2000;345:557 - 64. [21] Situ H, Bobek L. In vitro assessment of antifungal therapeutic potential of salivary histatin-5, two variants of histatin-5, and salivary mucin (MUC7) domain 1. Antimicrob Agents Chemother 2000;44: 1485 - 93. [22] Thomsson KA, Prakobphol A, Leffler H, et al. The salivary mucin MG1 (MUC5B) carries a repertoire of unique oligosaccharides that is large and diverse. Glycobiology 2002;12:1 - 14. [23] Deplancke B, Gaskins HR. Microbial modulation of innate defense: goblet cells and the interstitial mucus layer. Am J Clin Nutr 2001; 73:1131S - 41S. [24] Audie´ JP, Janin A, Porchet N, et al. Expression of human mucin genes in respiratory, digestive and reproductive tracts ascertained by in-situ hybridisation. J Histochem Cytochem 1993;41:1479 - 85. [25] Jung HH, Lee JH, Kim YT, et al. Expression of mucin genes in chronic ethmoiditis. Am J Rhinol 2000;14:163 - 70. [26] Paulsen F, Langer G, Hoffmann W, et al. Human lacrimal gland mucins. Cell Tissue Res 2004;316:167 - 77. [27] Paulsen FP, Corfield A, Hinz M, et al. Characterization of mucins in human lacrimal sac and nasolacrimal duct. Invest Ophthalmol Vis Sci 2003;44:1807 - 13. [28] Berois N, Barangot M, So´n˜ora C, et al. Detection of bone marrowdisseminated breast cancer cells using an RT-PCR assay of MUC5B mRNA. Int J Cancer 2003;103:550 - 5.

Prognostic value of mucins in the classification of ampullary carcinomas [29] Xing PX, Tjandra JJ, Reynolds K, et al. Reactivity of anti–human milk fat globule antibodies with synthetic peptides. J Immunol 1989; 142:3503 - 9. [30] Wickstrfm C, Davies JR, Eriksen GV, et al. MUC5B is a major gelforming, oligomeric mucin from human salivary gland, respiratory tract and endocervix: identification of glycoforms and C-terminal cleavage. Biochem J 1998;334:685 - 93. [31] Thornton DJ, Howard M, Khan N, et al. Identification of two glycoforms of the MUC5B mucin in human respiratory mucus. Evidence for a cysteine-rich sequence repeated within the molecule. J Biol Chem 1997;272:9561 - 6. [32] Lopez-Ferrer A, de Bolos C, Baranco C, et al. Role of fucosyltransferases in the association between apomucin and Lewis antigen expression in normal and malignant gastric epithelium. Gut 2000; 47:349 - 56. [33] Wickstrfm C, Christersson C, Davies JR, et al. Maromolecular organization of saliva: identification of dinsolubleT MUC5B assemblies and non-mucin proteins in the gel phase. Biochem J 2000; 351:421 - 8. [34] Buisine M-P, Devisme L, Maunoury V, et al. Developmental mucin gene expression in the gastroduodenal tract and accessory digestive glands. I. Stomach: a relationship to gastric carcinoma. J Histochem Cytochem 2000;48:1657 - 66. [35] Van Klinken BJ, Dekker J, Buller HA, et al. Biosynthesis of mucins (MUC2-6) along the longitudinal axis of the human gastrointestinal tract. Am J Physiol 1997;273:G296 - G302. [36] Buisine M-P, Devisme L, Degand P, et al. Developmental mucin gene expression in the gastroduodenal tract and accessory digestive glands. II. Duodenum and liver, gallbladder, and pancreas. J Histochem Cytochem 2000;48:1667 - 76. [37] Vandenhaute B, Buisine MP, Debailleul V, et al. Mucin gene expression in biliary epithelial cells. J Hepatol 1997;27:1057 - 66. [38] Pigny P, Guyonnet-Duperat V, Hill AS, et al. Human mucin genes assigned to 11p15.5: identification and organization of a cluster of genes. Genomics 1996;38:340 - 52. [39] Dorandeu A, Raoul JL, Siriser F, et al. Carcinoma of the ampulla of Vater: prognostic factors after curative surgery: a series of 45 cases. Gut 1997;40:350 - 5. [40] Fischer HP, Zhou H. Pathogenese und histophathologie von adenomen und karzinomen der papilla Vateri. Pathologe 2003;24:196 - 203. [41] Kimura W, Ohtsubo K. Incidence, sites of origin, and immunohistochemical and histochemical characteristics of atypical epithelium

[42] [43]

[44]

[45]

[46]

[47]

[48]

[49]

[50]

[51]

[52]

[53]

167

and minute carcinoma of the papilla of Vater. Cancer 1988;61: 1394 - 402. Zhou H, Schaefer N, Wolff M, et al. Carcinoma of the ampulla of Vater. Am J Surg Pathol 2004;28:875 - 82. Albores-Saavedra J, Henson DE, Klimstra DS. Tumors of the gallbladder, extrahepatic bile ducts, and ampulla of Vater. In: Rosai J, Sobin L, editors. Atlas of tumor pathology. Third series, Fascicle 27. Washington (DC)7 Armed Forces Institute of Pathology; 2000. p. 259 - 316. Aihara R, Ajioka Y, Watanabe H, et al. Incidence and distribution of hybrid goblet cells in complete type intestinal metaplasia of the stomach. Pathol Res Pract 2005;201:11 - 9. Chu PG, Schwarz RE, Lau SK, et al. Immunohistochemical staining in the diagnosis of pancreatobiliary and ampulla of Vater adenocarcinoma: application of CDX2, CK17, MUC1, and MUC2. Am J Surg Pathol 2005;29:359 - 67. Dawson PJ, Connolly MM. Influence of site of origin and mucin production on survival in ampullary carcinoma. Ann Surg 1989;210: 173 - 9. Gqrbqz Y, Klfppel G. Differentiation pathways in duodenal and ampullary carcinomas: a comparative study on mucin and trefoil peptide expression, including gastric and colon carcinomas. Virchows Arch 2004;444:536 - 41. Kitamura H, Yonezawa S, Tanaka S, et al. Expression of mucin carbohydrates and core proteins in carcinomas of the ampulla of Vater: their relationship to prognosis. Jpn J Cancer Res 1996;87: 631 - 40. Kushima R, Stolte M, Dirks K, et al. Gastric-type adenocarcinoma of the duodenal second portion histogenetically associated with hyperplasia and gastric-foveolar metaplasia of Brunner’s glands. Virchows Arch 2002;440:655 - 9. Matsubayashi H, Watanabe H, Yamaguchi T, et al. Differences in mucus and K-ras mutation in relation to phenotypes of tumors of the papilla of Vater. Cancer 1999;86:596 - 607. Tamada S, Goto M, Nomoto M, et al. Expression of MUC1 and MUC2 mucins in extrahepatic bile carcinomas: its relationship with tumor progression and prognosis. Pathol Int 2002;52:713 - 23. Yonezawa S, Sato E. Expression of mucin antigens in human cancers and its relationship with malignancy potential. Pathol Int 1997;47: 813 - 30. Paulsen F, Varoga D, Paulsen A, et al. Trefoil factor peptides of normal human Vater’s ampulla. Cell Tissue Res 2005;321:67 - 74.