Secretin receptors in the human liver: Expression in biliary tract and cholangiocarcinoma, but not in hepatocytes or hepatocellular carcinoma

Journal of Hepatology 45 (2006) 825–835 www.elsevier.com/locate/jhep

Secretin receptors in the human liver: Expression in biliary tract and cholangiocarcinoma, but not in hepatocytes or hepatocellular carcinomaq Meike Ko¨rner1, Gregory M. Hayes2, Ruth Rehmann1, Arthur Zimmermann1, Arne Scholz3, Bertram Wiedenmann3, Laurence J. Miller2, Jean Claude Reubi1,* 1

Division of Cell Biology and Experimental Cancer Research, Institute of Pathology, University of Bern, Murtenstrasse 31, P.O. Box 62, CH-3010 Bern, Switzerland 2 Cancer Center, Mayo Clinic, Scottsdale, AZ, USA 3 Division of Gastroenterology and Hepatology, Charite´ Medical Center-Virchow Hospital, Medical School, Humboldt-University of Berlin, Berlin, Germany

Background/Aims: Gut hormone receptors are over-expressed in human cancer and allow receptor-targeted tumor imaging and therapy. A novel promising receptor for these purposes is the secretin receptor. The secretin receptor expression was investigated in the human liver because the liver is a physiological secretin target and because novel diagnostic and treatment modalities are needed for liver cancer. Methods: Nineteen normal livers, 10 cirrhotic livers, 35 cholangiocarcinomas, and 45 hepatocellular carcinomas were investigated for secretin receptor expression by in vitro receptor autoradiography using 125I-[Tyr10] rat secretin and, in selected cases, for secretin receptor mRNA by RT-PCR. Results: Secretin receptors were present in normal bile ducts and ductules, but not in hepatocytes. A significant receptor up-regulation was observed in ductular reaction in liver cirrhosis. Twenty-two (63%) cholangiocarcinomas were positive for secretin receptors, while hepatocellular carcinomas were negative. RT-PCR revealed wild-type receptor mRNA in the non-neoplastic liver, wild-type and spliced variant receptor mRNAs in cholangiocarcinomas found receptor positive in autoradiography experiments, and no receptor transcripts in autoradiographically negative cholangiocarcinomas. Conclusions: The expression of secretin receptors in the biliary tract is the molecular basis of the secretin-induced bicarbonate-rich choleresis in man. The high receptor expression in cholangiocarcinomas may be used for in vivo secretin receptor-targeting of these tumors and for the differential diagnosis with hepatocellular carcinoma. Ó 2006 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. Keywords: Secretin receptor; Human liver; Cholangiocellular carcinoma; Hepatocellular carcinoma; Receptor autoradiography

Received 24 March 2006; received in revised form 22 June 2006; accepted 26 June 2006; available online 28 July 2006 q The authors who have taken part in this study declared that they have no relationship with the manufacturers of the drugs involved either in the past or present and did not receive funding from the manufacturers to carry out their research. The authors received funding from NIH (grant DK46577), which enabled them to carry out their study. * Corresponding author. Tel.: +41 31 6323243; fax: +41 31 6328999. E-mail address: [email protected] (J.C. Reubi).

1. Introduction Peptide hormone receptors are often overexpressed in malignant human tumors [1]. This is of increasing clinical importance, because it allows for highly specific, receptortargeted tumor imaging and therapy with peptide hormone analogs. The peptide receptors first identified for these purposes have been the somatostatin receptors. Gastroenteropancreatic neuroendocrine tumors, for

0168-8278/$32.00 Ó 2006 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.jhep.2006.06.016

826

M. Ko¨rner et al. / Journal of Hepatology 45 (2006) 825–835

instance, express somatostatin receptors in high amounts. Thus, these tumors can very effectively be visualized scintigraphically with the radiolabeled somatostatin analog OctreoScanÒ [2], which forms an integral part of today’s routine diagnostic work-up of affected patients. Moreover, recent results from studies performing radiotherapy of these tumors with cytotoxic somatostatin analogs are promising [3,4]. Somatostatin analogs are also the drugs of first choice to control hormone-related symptoms in functionally active gastroenteropancreatic neuroendocrine tumors [5]. Based on the good experiences with somatostatin receptor targeting, it has been important to identify further peptide receptors suitable for in vivo targeting of other tumor types. The secretin receptor is a novel and promising candidate in this field. The secretin receptor was only recently identified [6]. It is a G protein-coupled receptor, belonging to the type II class of receptors. It forms the secretin receptor family together with the VIP, glucagon, and other receptors [7]. Under normal conditions, in humans, the secretin receptor has been identified in the pancreatic ducts as well as in antral and duodenal neuroendocrine (G-) cells [8] where it mediates important functions in digestion. It is mainly responsible for the bicarbonate and water secretion from the pancreas [9] and for the gastrin release from the antral and duodenal mucosa [10]. Recently, it was recognized that secretin receptors are also expressed in the neoplastic counterparts of these tissues: they were found in pancreatic duct cancers and in pancreatic gastrinomas [8,11,12]. Analogous to somatostatin receptors, the secretin receptors in neoplasia are of potential use for in vivo tumor targeting. Based on previous findings in pancreatic tumors, it appeared important to investigate the secretin receptor expression in tumors arising in other organs which physiologically respond to secretin. The liver was a primary candidate for two reasons: firstly, it is well documented that secretin stimulates hepatic bicarbonate-rich bile flow [13–17] yet hepatic secretin receptors have not been identified in humans but have been found in various animals, including the rat [16,18,19]; secondly, primary liver tumors are associated with poor prognosis because of a high rate of recurrence following surgical resection [20] and poor tumor response to chemotherapy and radiation [21]. Therefore, novel treatment strategies are needed for these tumors [22]. Consequently, it was the aim of the present study to investigate secretin receptor expression in both non-neoplastic human liver and primary liver tumors using receptor autoradiography and RT-PCR. 2. Materials and methods 2.1. Tissues For in vitro secretin receptor autoradiography, fresh frozen tissue samples were obtained from a total of 109 surgical resection and biopsy specimens from the University Hospital of Bern and the Charite´

Medical Center-Virchow Hospital Berlin. The samples comprised 19 cases of histologically normal liver resected together with a tumor; 10 cases of liver cirrhosis; and 80 cases of liver and extrahepatic bile duct tumors, including 45 hepatocellular carcinomas and 35 intrahepatic cholangiocellular carcinomas and adenocarcinomas of the extrahepatic bile ducts and gall bladder. Tumor typing and grading were performed according to the WHO guidelines [23]. The tissue was stored at 80 °C. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected by approval by the Institutional Review Boards of the Institute of Pathology, University of Bern, and Division of Hepatology and Gastroenterology, Charite´ Medical Center-Virchow Hospital, Berlin.

2.2. In vitro secretin receptor autoradiography In vitro secretin receptor autoradiography was performed as described previously [8]. Twenty-micrometer thick cryostat sections mounted on glass slides were pre-incubated in 0.01 M HEPES buffer (pH 7.4) for 5 min at room temperature. Afterwards, they were incubated for 120 min at room temperature in the incubation solution containing HEPES buffer, 1% BSA, 130 mM NaCl, 4.7 mM KCl, 5 mM (Mg(II)Cl2)4H2O, 1 mM EDTA, 1 mg/ml bacitracin, and 24,000 cpm/100 ll of the radioligand 125I-[Tyr10] rat secretin [24] (2000 Ci/mmol; Anawa, Wangen, Switzerland). Nonspecific radioligand binding was evaluated by incubating tissue sections with the incubation solution containing, in addition to 125I-[Tyr10] rat secretin, 100 nM of cold (non-labeled) human secretin which, at this concentration, completely and specifically displaces 125I-[Tyr10] rat secretin at the secretin receptors. Since VIP and glucagon receptors also bind secretin, but with lower affinity compared with secretin receptors [25], competition experiments were performed in order to differentiate secretin receptors from these other receptors. For this purpose, serial tissue sections were incubated with 125I-[Tyr10] rat secretin and increasing concentrations of one of the following cold peptides: human secretin, VIP (Bachem, Bubendorf, Switzerland), or glucagon(1–29) (Bachem). After incubation, the slides were washed five times in ice-cold HEPES containing 1% BSA and twice in ice-cold HEPES without BSA. The slides were dried for 15 min under a stream of cold air and then exposed to Kodak films Biomax MRÒ for 7 days at 4 °C. The resulting signals were analyzed in correlation with morphology using a corresponding HE stained tissue section. The receptor density was quantitatively assessed using tissue standards for iodinated compounds (Amersham, Aylesbury, UK) and a computer-assisted image processing system (Analysis Imaging System, Interfocus, Mering, Germany). In all experiments, rat pancreas was included as positive control [26].

2.3. Immunohistochemistry for cytokeratin 19 and secretin receptors In order to specifically identify biliary tree structures in the interpretation of autoradiograms, immunohistochemistry with a monoclonal anti-cytokeratin 19 (CK19) mouse antibody (DAKO, Zug, Switzerland) was performed on serial tissue sections adjacent to those used for receptor autoradiography from all secretin receptor expressing normal liver samples and from all cirrhosis samples. Cholangiocytes of bile ducts and bile ductules are positive for CK19, whereas hepatocytes are negative [27]. Briefly, 10 lm thick cryostat sections were postfixed in formalin. They were pretreated with APAAP to inhibit endogenous peroxidase. The primary antibody was applied in a 1:50 concentration. As secondary antibody, a goat anti-mouse immunoglobulin (DAKO) was used. Antibody binding was visualized using the ABComplex/ HRP (DAKO). Staining was carried out with DAB, and counterstaining with hemalum. In order to assess the cellular and subcellular distribution of secretin receptors in more detail, an anti-secretin receptor antibody (Biogenesis, Poole, UK) was tested on frozen and formalin-fixed, paraffin-embedded tissue samples of normal liver, cirrhotic liver, and cholangiocellular carcinomas. The antibody was tested in several concentrations and with various antigen-retrieval methods, as described previously [28]. Human gastrinomas, known to express a very high secretin receptor density, served as positive control [8].

M. Ko¨rner et al. / Journal of Hepatology 45 (2006) 825–835

2.4. RT-PCR analysis of secretin receptor transcripts

3. Results

The receptor autoradiography results were confirmed by RTPCR specific for secretin receptor transcripts in representative cases. Those included normal liver with bile ducts and bile ductules positive for secretin receptors by autoradiography, liver cirrhosis with proliferated bile ductules positive for secretin receptors by autoradiography, cholangiocarcinomas positive or negative for secretin receptors by autoradiography, and hepatocellular carcinomas negative for secretin receptors by autoradiography. An area of either non-neoplastic liver tissue containing bile ducts or ductules, or tumor tissue without intervening normal liver parenchyma was selected on an HE stained tissue section and cut out of the frozen tissue block with a sterile scalpel blade. Human spleen was used as negative control [26]. RT-PCR was subsequently performed to confirm the presence or absence of secretin receptor transcripts in the aforementioned samples. Briefly, the dissected frozen samples were weighed followed by grinding in liquid nitrogen. The resulting tissue powders were decanted into microfuge tubes and resuspended in Trizol reagent (Invitrogen, Carlsbad, CA, USA) to a final concentration of 50 lg tissue per 1 ml Trizol. Total cellular RNA was isolated from 250 lL of each Trizol suspension using the RNEasy Mini Kit (Qiagen, Valencia, CA, USA). Purified RNA was digested with amplification grade deoxyribonuclease I (DNAse I; Invitrogen) for 15 min at room temperature to remove any genomic DNA contamination. Subsequently, EDTA was added (2 mM final; Invitrogen) and samples heated to 65 °C for 15 min to inactivate the DNAse I. One-half microgram of each purified RNA was used to synthesize cDNA via the Reverse Transcription System (Promega, Madison, WI, USA) and approximately 25 ng of resultant cDNA was used as template for the ensuing PCRs. Amplification of human secretin receptor transcripts was performed as previously described [29] using 5 0 and 3 0 primers corresponding to nucleotides 107–127 and 1434–1413 of GenBank Accession No. U28281, respectively (sense, 5 0 -CCA TGC GTC CCC ACC TGT CGC-3 0 and antisense, 5 0 -CTC TCA GAT GAT GCT GGT CCT G-3 0 ). Control reactions were included utilizing previously described human secretin receptor isoforms (wild-type human secretin receptor and human secretin receptor-Dexon 3) cloned into the pBK-CMV vector [29] or with no template cDNA (water only). Actin PCRs were run using the primer pair: sense, 5 0 -CCA GCT CAC CAT GGA TGA TGA TAT CG-3 0 and antisense, 5 0 -GGA GTT GAA GGT AGT TTC GTG GAT GC-3 0 . PCRs were run in 50 ll volume containing 0.2 lM of both sense and antisense primers, 2 mM MgCl2 (Invitrogen), 0.2 mM dNTPs (Stratagene, La Jolla, CA, USA), and 0.5 unit of Platinum Taq Polymerase (Invitrogen). Reactions were performed in a DNA Engine (MJ Research, South San Francisco, CA, USA) using optimized cycle conditions comprised of an initial denaturing step for 2 min at 94 °C and subsequently 32 cycles of 94 °C for 30 s, 64 °C for 30 s, and 72 °C for 2 min. Fifteen microliter fractions of each amplification reaction were resolved on a 2% agarose/TAE gel along with the 1 kb Plus Ladder (Invitrogen) as marker.

3.1. Secretin receptor autoradiography in the human normal and cirrhotic liver

2.5. Statistics For the comparison of secretin receptor densities in different segments of the biliary tree, the Student’s t-test was used. p < 0.05 was considered to be statistically significant.

827

Secretin receptors were expressed in the human liver exclusively in the biliary tree, i.e., in bile ductules at the edges of the portal tract stroma as well as in interlobular and septal bile ducts within portal tracts [30], but not in hepatocytes (Table 1). This is illustrated with typical examples in Fig. 1. In the case of the upper row, the septal bile ducts in the portal field strongly express secretin receptors, whereas the hepatocytes are receptor negative. The precise identification of secretin receptor-expressing tissues was confirmed with immunohistochemistry for CK19, a marker for bile ducts and ductules, performed on serial tissue sections adjacent to autoradiograms, as shown in the example in the bottom row of Fig. 1. The autoradiographic signals correspond well with the CK19 positive interlobular bile duct in the portal field and with the CK19 positive bile ductules at the borders of the portal field. The secretin receptor expression in the biliary tree was heterogeneous. Indeed, not all of the patient samples were receptor positive; the bile ductules were most frequently receptor positive, and in the bile ducts the receptor incidence decreased with increasing duct size (Table 1). Also within an individual liver sample, the secretin receptor expression was often heterogeneous: some of the bile ducts and bile ductules were receptor positive, whereas the others were negative. The portal vein and hepatic artery branches were always negative. An important additional observation was made in cirrhotic livers, namely that the ductular reaction [31] at the edges of the cirrhotic nodules expressed secretin receptors particularly often and in high amounts. The secretin receptors were markedly up-regulated in the ductular reaction compared with the residual normal bile ducts or with bile ductules in normal liver samples (Table 1). Indeed, firstly, secretin receptors were positive in ductular reaction in all cirrhosis samples, whereas they were expressed in normal bile ducts and ductules in less than half of the cases. Secondly, secretin receptors were always homogeneously expressed in ductular reaction, while they often showed a heterogeneous distribution within normal bile ducts and ductules. Finally,

Table 1 Secretin receptor incidence and density in the normal liver (n = 19) and cirrhotic liver (n = 10)

Large bile ducts Septal bile ducts Interlobular bile ducts Bile ductules Ductular reaction Hepatocytes

Secretin receptor frequency: positive/total samples (%)

Mean secretin receptor density in receptor positive samples ± SEM (dpm/mg)

0/5 (0%) 2/10 (20%) 5/16 (31%) 9/19 (47%) 10/10 (100%) 0/29 (0%)

– 1992 ± 28 1150 ± 337 1066 ± 217 4033 ± 362 –

828

M. Ko¨rner et al. / Journal of Hepatology 45 (2006) 825–835

Fig. 1. In vitro receptor autoradiography on serial tissue sections to assess 125I-[Tyr10]-secretin binding in the normal liver. (A) HE stained tissue section showing portal tract with septal bile ducts (asterisks), portal vein and hepatic artery, surrounded by hepatocytes, and (D) CK19 immunostained section showing interlobular bile duct within a portal tract (asterisks) and bile ductules at the border of the portal tract (arrows). Bars = 1 mm. (B, E) Autoradiograms showing total binding of 125I-[Tyr10] rat secretin: strong signal in the septal bile duct (B) as well as in the CK19 positive interlobular bile duct and bile ductules (E), while hepatocytes and blood vessels are not labeled. (C, F) Autoradiograms assessing non-specific (ns) binding of 125I-[Tyr10] rat secretin in the presence of 100 nM unlabeled human secretin. [This figure appears in colour on the web.]

secretin receptors were present in significantly higher density in ductular reaction compared with normal bile ducts and ductules (p < 0.001). This high secretin receptor expression in ductular reaction resulted in a high total secretin receptor amount in cirrhosis samples compared with the normal liver. This is illustrated in Fig. 2 which shows impressively that the high secretin receptor amount in the tissue sample is mainly the result of the secretin receptor expression in the ductular reaction surrounding the cirrhotic nodules, whereas the scarce residual normal bile ducts contribute only little to the secretin receptor positivity. The precise identification of secretin receptor-positive ductular reaction was again confirmed with immunohistochemistry for CK19, as shown in Fig. 2D–F. The autoradiographic signal could specifically be attributed to the CK19 positive ductules surrounding the cirrhotic nodules. 3.2. Secretin receptor autoradiography in cholangiocarcinomas and hepatocellular carcinomas Secretin receptors were frequently identified in cholangiocarcinomas, but not in hepatocellular carcinomas.

Tables 2 and 3 provide detailed information on the secretin receptor expression in the cholangiocarcinomas. When all tumor samples were taken together, the secretin receptor incidence amounted to 63%. The secretin receptor incidence however depended on tumor grade: it was reduced almost by half in grade 3 carcinomas compared with grade 2 carcinomas (Table 2). In grade 3 tumors, in addition, the receptor distribution was more often heterogeneous, i.e., with receptor negative areas juxtaposing receptor positive ones, than in grade 2 tumors (Table 3). Of note, in the cases with heterogeneous receptor distribution, the secretin receptor expression did not depend on the size of neoplastic ducts, but secretin receptors were rather randomly distributed within the tumor sample. In terms of secretin receptor density, no difference was observed between grade 2 and grade 3 carcinomas, if one excepts a single grade 3 carcinoma with extremely high secretin receptor density. Also no differences in the secretin receptor expression were apparent between intrahepatic cholangiocellular carcinomas and extrahepatic bile duct or gall bladder carcinomas. Finally, no correlation of the secretin receptor expression with tumor stage, in particular presence

M. Ko¨rner et al. / Journal of Hepatology 45 (2006) 825–835

829

Fig. 2. In vitro receptor autoradiography on cirrhotic livers. (A) Overview of HE stained tissue section showing cirrhotic nodules surrounded by ductular reaction and fibrous tissue as well as residual normal interlobular bile ducts (arrowheads), and (D) CK19 immunostained section at higher magnification demonstrating CK19 positive ductular reaction around the cirrhotic nodules (arrows). Bars = 1 mm. (B, E) Autoradiograms showing total binding of 125 I-[Tyr10] rat secretin: strong signal surrounding the cirrhotic nodules as well as in the residual normal bile ducts (B); higher magnification reveals colocalization of the autoradiographic signal with the CK19 positive ductular reaction, but not with CK19 negative hepatocytes. (C, F) Autoradiograms showing non-specific 125I-[Tyr10] rat secretin binding in the presence of 100 nM unlabeled human secretin. [This figure appears in colour on the web.]

of lymph node metastases, could be detected. Fig. 3 shows two typical examples of secretin receptor positive cholangiocellular carcinomas. The first row depicts a poorly differentiated carcinoma with a very high and homogeneous secretin receptor expression. In the middle row, higher magnification of a grade 2 carcinoma shows that the secretin receptors are located in the neoplastic epithelial ducts. In contrast, the hepatocellular carcinoma in the bottom row expresses no secretin receptors. 3.3. Pharmacological characterization of secretin receptors Not only the secretin receptor, but also other members of the secretin receptor family, like VIP receptors

bind secretin. These other receptors, however, bind secretin with lower affinity compared with the secretin receptor [24,26,32]. In order to provide a pharmacological proof that the radioligand 125I-[Tyr10] rat secretin was specifically bound by secretin receptors but not by VIP receptors in the investigated tissues, competition experiments were performed with increasing concentrations of secretin, VIP, or glucagon(1–29) to assess their rank orders of potencies at the receptors. Fig. 4 shows representative results obtained in bile ducts, bile ductules, ductular reaction in liver cirrhosis, and cholangiocarcinomas. In all these tissues, 125 I-[Tyr10] rat secretin was displaced by nanomolar concentrations of cold secretin, but only by micromolar concentrations of VIP and not at all by glucagon(1–29).

Table 2 Secretin receptor incidence and density in liver and extrahepatic biliary tract tumors Secretin receptor frequency: positive/total samples (%)

Mean secretin receptor density of receptor positive cases ± SEM (dpm/mg)

Cholangiocarcinomas Grade 2 Grade 3 Total

15/19 (79%) 7/16 (44%) 22/35 (63%)

969 ± 106 2385 ± 993 1430 ± 339

Hepatocellular carcinomas

0/45 (0%)

–

830

M. Ko¨rner et al. / Journal of Hepatology 45 (2006) 825–835

Table 3 Secretin receptor expression and density in cholangiocarcinomas according to the tumor grade No.

Primary site

Grade

TNM

Secretin receptor density (dpm/mg) and distribution

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

Intrahepatic Intrahepatic Intrahepatic Intrahepatic Intrahepatic Perihilar Perihilar Perihilar Perihilar Perihilar Perihilar Cystic duct Unknown Unknown Unknown Unknown Unknown Unknown Unknown Intrahepatic Intrahepatic Intrahepatic Perihilar Perihilar Perihilar Perihilar Perihilar Perihilar Common bile duct Gall bladder Gall bladder Recurrent in head of pancreas Unknown Unknown Unknown

2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3

Tx T1 T3 T3 T4 T3 T3 T3 T2 T4 T3 T2 T3 T4 T3 T4 T3 T4 T3 T1 T4 T3 T2 T2 T2 T2 T3 T4 T2 T3 T3

1446 1211 1056 homogeneous 780 homogeneous – 958 615 606 – – – 1198 1710 homogeneous 1326 1273 homogeneous 1177 homogeneous 721 5511 homogeneous 145 homogeneous 8203 homogeneous 1054 – 2206 1140 homogeneous – – – – – 2209 842 – 1040 – –

The binding affinities of the receptor for the various ligands are summarized in Table 4. In all investigated tissues, the IC50 values for secretin were within the low nanomolar concentration range, whereas the IC50 values for VIP and glucagon(1–29) were more than two orders of magnitude higher, ranging in a micromolar concentration. A similar rank order of potencies was obtained in the control tissues, namely rat pancreas and human gastrinoma [8,25]. These binding affinities correspond well with the data reported for secretin receptors [25,26,32]. Therefore, the results provide strong pharmacological evidence that secretin receptors were specifically identified. 3.4. RT-PCR for secretin receptor transcripts Additional evidence for the expression of secretin receptors in the normal and cirrhotic liver as well as cholangiocellular carcinomas is supported by RT-PCR analysis of secretin receptor transcripts performed on selected tissue samples (Fig. 5). In all samples from

N0 N1 N1 M1 N1 N0 M0 N0 N0 N0 N0 M0 N0 M0 M1 M0 N1 M1 N1 M0 N0 M0 M0 N0 N1 N0 N0 N1 N1 N0 N0 N1

M0 M0 M0 M0 M0 M1

M1

T3 N1 M0 T4 M0 T3 N1 M0

normal and cirrhotic livers containing bile ducts and ductular reaction, respectively, that were shown to express secretin receptors in parallel autoradiography experiments, wild-type secretin receptor transcripts were detected. Further, in the cholangiocarcinomas expressing secretin receptors by autoradiography, wild-type secretin receptor mRNA was present. Of interest, these carcinomas expressed alternatively spliced secretin receptor transcripts, including the previously described spliceoform with exon 3 (residues 44–79) deletion [29,33] in addition to the wild-type receptor encoding mRNA. Conversely, cholangiocarcinomas exhibiting negative autoradiography results displayed no demonstrable secretin receptor transcripts. This absence of secretin receptor transcripts, either wild-type or splice variants, was demonstrated even after increasing the number of PCR cycles to achieve the highest possible sensitivity. Also in the hepatocellular carcinomas, no secretin receptor transcripts were detected. The spleen used as negative control was negative for secretin receptor transcripts.

M. Ko¨rner et al. / Journal of Hepatology 45 (2006) 825–835

831

Fig. 3. In vitro receptor autoradiography on cholangiocarcinomas (A–F) and hepatocellular carcinomas (G–I). (A, D, G) The HE stained tissue sections show a grade 3 cholangiocellular carcinoma in overview (A), a grade 2 cholangiocellular carcinoma at higher magnification (D), and a hepatocellular carcinoma (G); bars = 1 mm. (B, E, H) Autoradiograms showing total binding of 125I-[Tyr10] rat secretin: very strong and diffuse labeling of the cholangiocarcinomas (B, E), no labeling of the hepatocellular carcinoma (H). (C, F, I) Autoradiograms showing non-specific 125I-[Tyr10] rat secretin binding in the presence of 100 nM unlabeled human secretin. [This figure appears in colour on the web.]

3.5. Immunohistochemistry for secretin receptors Using various antigen retrieval methods and antibody concentrations [28], no specific immunostaining for secretin receptors could be obtained with the commercially available Biogenesis antibody; neither in tissues expressing secretin receptors in autoradiography experiments, namely gastrinomas, known to express secretin receptors in very high density [8], that were used as positive controls; nor in bile ducts of normal livers, ductular reaction in liver cirrhosis or cholangiocellular carcinomas.

4. Discussion The study describes for the first time the presence and distribution of secretin receptors in the normal human liver and various liver pathologies. The secretin receptors were detected morphologically using receptor autoradiography in the biliary tract exclusively, not in

hepatocytes. They were heterogeneously expressed in septal and interlobular bile ducts and bile ductules. Compared with normal liver, they were expressed more frequently and in higher density in ductular reaction in liver cirrhosis. Furthermore, the secretin receptor expression in liver tumors correlated with that found in the respective tissues of origin: cholangiocarcinomas were strongly positive for secretin receptors, while hepatocellular carcinomas were secretin receptor negative. Several lines of evidence were provided in this study that specific secretin receptors were identified by receptor autoradiography. Indeed, secretin is bound not only by secretin receptors with high affinity, but also by other members of the secretin receptor family with lower affinity [25,32]. Therefore, we had to exclude in our study that 125I-[Tyr10] rat secretin labeled receptors other than secretin receptors. VIP and glucagon receptors are indeed expressed in the human liver [34,35]. Therefore, pharmacological competition experiments were performed with secretin, VIP, and glucagon, which showed

832

M. Ko¨rner et al. / Journal of Hepatology 45 (2006) 825–835

Fig. 4. Representative competition experiments showing the displacement of 125I-[Tyr10]-secretin by secretin, VIP, and glucagon(1–29) in bile ductules (A), bile ducts (B), ductular reaction (C), and a cholangiocarcinoma (D). High affinity displacement of 125I-[Tyr10] rat secretin by human secretin (d), but low affinity displacement by VIP (n) and minimal or no displacement by glucagon(1–29) (m).

that the identified receptors had a high affinity for secretin, but a low affinity for VIP and glucagon. This rank order of potencies provides strong pharmacological evidence of the specificity of the autoradiography for secretin receptors. This secretin receptor specificity was further substantiated by RT-PCR analysis which demonstrated wild-type secretin receptor transcripts in tissues autoradiographically positive for secretin receptors, i.e., in bile ducts, ductular reaction, and cholangiocarcinomas. Unfortunately, as for secretin receptor immunohistochemistry, the only commercially available anti-secretin receptor antibody did not yield satisfactory results in our hands. The identification of secretin receptors in the intrahepatic biliary tree correlates well with and adds new information to the known role of secretin in liver physiology. Indeed, in vivo studies have demonstrated that secretin has an effect on choleresis in humans: intravenous secretin infusion stimulated hepatic bile flow and

bicarbonate secretion distal to the hepatocellular level [13–15]. The present study identifies secretin receptors in small and medium-sized liver bile ducts and ductules. This corresponds very well with the distribution of the Cl =HCO 3 -exchanger AE2, an enzyme responsible for cellular bicarbonate secretion, in small and mediumsized, but not large, bile ducts in the human liver [36]. Taken together, these data provide evidence at the molecular level that secretin may directly act on the human liver and that the secretin-induced choleresis observed in vivo may be the result of an interaction of secretin with secretin receptors and activation of AE2 in ductal and ductular epithelial cells. This appears in agreement with the available rat data. Indeed, in the rat, the secretin receptors are also present in bile ducts, where they co-localize with Cl =HCO 3 -exchangers [16,18,19], but not in hepatocytes. In the rat, it could be additionally shown that secretin directly stimulates the Cl =HCO 3 -exchanger activity and bicarbonate

Table 4 Binding affinities (IC50, nM; mean ± SEM, n = 3) of the secretin receptor for three ligands of the secretin receptor family, namely secretin, VIP, and glucagon(1–29), investigated in various liver tissues including intrahepatic bile ducts and cholangiocarcinomas Ligand

Bile ducts

Bile ductules

Ductular reaction

Cholangiocarcinomas

Secretin VIP Glucagon(1–29)

0.72 ± 0.09 402 ± 234 >1000

0.35 ± 0.19 504 ± 246 >1000

1.2 ± 0.37 371 ± 110 >1000

0.66 ± 0.13 1478 ± 806 >1000

Wild type hSecR control

CCC (ARG-)

HCC (ARG-)

CCC (ARG-)

CCC (ARG+++)

CCC (ARG-)

CCC (ARG+++)

B

CCC (ARG+++)

Cirrhosis (ARG+)

Cirrhosis (ARG+)

Bile ducts (ARG+)

Bile ducts (ARG+)

Bile ducts (ARG+)

H20 only control

A

833

Δexon 3 hSecR control

M. Ko¨rner et al. / Journal of Hepatology 45 (2006) 825–835

Actin

Fig. 5. RT-PCR detection of secretin receptor transcripts in normal and cirrhotic liver samples, cholangiocarcinomas (CCC), either strongly positive (ARG+++) or negative (ARG) for secretin receptors by autoradiography and hepatocellular carcinomas (HCC). Total RNA was isolated from microdissected liver specimens and subsequent RT-PCR was performed using primers able to amplify full-length secretin receptor transcripts (A) or actin (B). Control PCRs were run utilizing wild-type and Dexon 3 human secretin receptor (hSecR) isoforms cloned into the pBK-CMV expression plasmid or with no template DNA added (negative control). The 1 kb Plus ladder from Invitrogen was included as a MW marker.

secretion in bile duct epithelial cells [16,17], a finding that may well be extrapolated to man. Direct evidence of functionality of secretin receptors in human bile ducts is difficult to obtain with the investigated materials. A reliable method to assess receptor functionality in fresh frozen human tissue specimens would be functional receptor autoradiography [37] that unfortunately cannot be performed for secretin receptors, because this method requires a Gi/Go protein-coupled receptor, while the secretin receptor couples to the Gs-protein. The observation that secretin receptors, as assessed by receptor autoradiography, are often not expressed on all bile ducts in a given tissue sample may indicate that only some of the hepatic bile ducts are responsive to secretin at a time, whereas the others may be refractory to secretin stimulation due to internalized receptors; this mechanism may prevent an over-stimulation of bile flow. Alternatively, the ducts that are negative for secretin receptors by autoradiography may express very small receptor amounts that are not detected with the method. The impressively high secretin receptor expression in ductular reaction in liver cirrhosis suggests that secretin may have significant effects in the cirrhotic liver and that these effects may even be more pronounced than those triggered by secretin in the normal liver. Indeed, it was repeatedly reported that a slightly elevated basal and a markedly increased secretin-stimulated hepatic bile flow can be observed in vivo in patients with liver cirrhosis compared with test persons without liver disease [14,15,38]. Similar results were also reported in the rat ligated bile duct model, which is characterized by an intrahepatic ductular proliferation and an up-regulated secretin receptor expression in the proliferating ductules, very much alike human cirrhosis and other rat cirrhosis

models [18,39–41]. In these rat models, the basal and secretin-stimulated bile flow and bicarbonate secretion are also increased compared with normal rats [41–44]. It can be proposed that, in both humans and rats, secretin exerts secretory effects on proliferating ductules similar to those reported on normal bile ducts and bile ductules; furthermore, the increased secretin-dependent bile flow in cirrhosis is the result not only of an increased number of secretin receptor positive ductular epithelial cells, but also of an increased responsiveness of the ductular epithelium to secretin due to the increased secretin receptor amount. The cholangiocarcinoma is an additional human malignancy, apart from pancreatic carcinomas and gastrinomas [8,11,12], in which secretin receptors have been shown to be expressed. In cholangiocarcinomas, secretin receptors were often heterogeneously expressed, which was significantly more frequently observed in grade 3 than in grade 2 carcinomas. This finding may, together with the significantly lower frequency of secretin receptor expressing tumors in the grade 3 than in the grade 2 carcinoma group, be interpreted as a general reduction of the secretin receptor expression in higher grade cholangiocarcinomas, reflecting loss of differentiation. The functional significance of tumoral secretin receptors is known for gastrinomas, where secretin receptors were shown to mediate gastrin release [45]. In cholangiocarcinomas and pancreatic adenocarcinomas, the pathophysiological role of secretin receptors is more difficult to assess, especially in frozen human tissues samples, and remains at present unclear [46,47]. Could the identification of tumoral secretin receptors open onto new clinical applications for both diagnostic and therapeutic purposes? As for cholangiocarcinomas,

834

M. Ko¨rner et al. / Journal of Hepatology 45 (2006) 825–835

these tumors are at present afflicted with a poor prognosis because of late diagnosis and limited therapeutic options. Surgical resection, the only potentially curative measure, is afflicted with a considerable recurrence rate [20]. Tumor response to adjuvant chemotherapy or radiation is limited, and no established chemotherapeutical protocols exist so far [21]. Thus, strategies for early detection as well as novel treatment options are needed for cholangiocarcinomas [22]. Current investigational efforts focus for example on photodynamic therapy or targeting of intracellular signal transduction pathways [48]. Another tentative approach could be peptide hormone receptor targeting [49]. An important prerequisite for successful tumor targeting is however that the tumor expresses a high number of receptors [50]. The secretin receptors are the first peptide receptors to be identified in considerable amounts in cholangiocarcinomas [1]. A secretin receptor targeting of cholangiocarcinomas, analogous to somatostatin receptor targeting of gastroenteropancreatic neuroendocrine tumors, using radiolabeled secretin analogs might therefore be promising in tumor imaging and therapy. In contrast to cholangiocellular carcinomas, hepatocellular carcinomas do not express secretin receptors. This is important for the interpretation of secretin receptor targeted scintigraphy in liver tumors. A positive scan result would strongly favor a cholangiocellular carcinoma over a hepatocellular carcinoma. Furthermore, the secretin receptor negativity of hepatocellular carcinomas excludes these tumors from secretin receptor targeted therapy. The characterization of the secretin receptor expression in several human tumors, i.e., in cholangiocarcinomas, pancreatic carcinomas, and gastrinomas [8,29,33], allows for general statements to be postulated on secretin receptor expression in cancer. (1) The secretin receptor is expressed in selected types of cancer. (2) In neoplasia, secretin receptor binding can be preserved or even increased compared with that in the normal tissue of origin, like in cholangiocellular carcinomas and gastrinomas, or it can be reduced like in pancreatic ductal adenocarcinomas. (3) Secretin receptor binding is reduced with increasing tumor dedifferentiation. In both cholangiocellular carcinomas and pancreatic ductal adenocarcinomas, the secretin receptor incidence is significantly lower in grade 3 than in grade 1 or 2 tumors. It therefore appears that the secretin receptor expression is a marker of normal cholangiocellular and pancreatic ductal differentiation that is gradually lost with increasing tumor dedifferentiation and progression. (4) In neoplasia, secretin receptor spliced variant transcripts are expressed de novo in addition to wild-type secretin receptor transcripts. Secretin receptor spliced variant transcripts have not been found in normal tissues. (5) These spliced variants may, in specific tumor types, affect the function of the wild-type secretin receptor,

based largely on the stochiometry of expression and the dominant negative action of the spliceoform with exon 3 deletion [33]. While this was observed with both pancreatic ductal adenocarcinomas and gastrinomas, it was not observed in cholangiocarcinomas, where loss of secretin binding was apparently due to the complete absence of secretin receptors. Acknowledgement We would like to thank Dr Peter Neuhaus, Department of General, Visceral and Transplantation Surgery, Charite´, Campus Virchow, Berlin, Germany, for the generous gift of the majority of the tumors in the present study. Grant support: The study was supported by grant DK46577 from the National Institutes of Health. References [1] Reubi JC. Peptide receptors as molecular targets for cancer diagnosis and therapy. Endocr Rev 2003;24:389–427. [2] Gibril F, Reynolds JC, Doppman JL, Chen CC, Venzon DJ, Termanini B, et al. Somatostatin receptor scintigraphy: its sensitivity compared with that of other imaging methods in detecting primary and metastatic gastrinomas. Ann Intern Med 1996;125:26–34. [3] Waldherr C, Pless M, Maecke HR, Schumacher T, Crazzolara A, Nitzsche EU, et al. Tumor response and clinical benefit in neuroendocrine tumors after 7.4 GBq (90)Y-DOTATOC. J Nucl Med 2002;43:610–616. [4] Kwekkeboom DJ, Teunissen JJ, Bakker WH, Kooij PP, de Herder WW, Feelders RA, et al. Radiolabeled somatostatin analog [177Lu-DOTA0,Tyr3]octreotate in patients with endocrine gastroenteropancreatic tumors. J Clin Oncol 2005;23:2754–2762. [5] Oberg K. Established clinical use of octreotide and lanreotide in oncology. Chemotherapy 2001;47:40–53. [6] Ishihara T, Nakamura S, Kaziro Y, Takahashi T, Takahashi K, Nagata S. Molecular cloning and expression of a cDNA encoding the secretin receptor. EMBO J 1991;10:1635–1641. [7] Harmar AJ. Family-B G-protein-coupled receptors. Genome Biol 2001; 2:reviews3013.1-3013.10. [8] Ko¨rner M, Hayes GM, Rehmann R, Zimmermann A, Friess H, Miller LJ, et al. Secretin receptors in normal and diseased human pancreas: marked reduction of receptor binding in ductal neoplasia. Am J Pathol 2005;167:959–968. [9] Chey WY, Chang TM. Secretin, 100 years later. J Gastroenterol 2003;38:1025–1035. [10] Hattori Y, Imamura M, Tobe T. Gastrin release from antral G cells stimulated with secretin. Am J Gastroenterol 1992;87:195–200. [11] Tang C, Biemond I, Offerhaus GJ, Verspaget W, Lamers CB. Expression of receptors for gut peptides in human pancreatic adenocarcinoma and tumour-free pancreas. Br J Cancer 1997;75:1467–1473. [12] Tang C, Biemond I, Lamers CB. Expression of peptide receptors in human endocrine tumours of the pancreas. Gut 1997;40:267–271. [13] Boyer JL, Bloomer JR. Canalicular bile secretion in man. Studies utilizing the biliary clearance of (14C)mannitol. J Clin Invest 1974;54:773–781. [14] Turnberg LA, Grahame G. Secretion of water and electrolytes into the duodenum in normal subjects and in patients with cirrhosis: the response to secretin and pancreozymin. Gut 1974;15:273–277.

M. Ko¨rner et al. / Journal of Hepatology 45 (2006) 825–835 [15] Bode C, Zelder O, Goebell H, Neuberger HO. Choleresis induced by secretin: distinctly increased response in cirrhotics. Scand J Gastroenterol 1972;7:697–699. [16] Alpini G, Glaser S, Robertson W, Rodgers RE, Phinizy JL, Lasater J, et al. Large but not small intrahepatic bile ducts are involved in secretin-regulated ductal bile secretion. Am J Physiol 1997;272:G1064–G1071. [17] Alvaro D, Cho WK, Mennone A, Boyer JL. Effect of secretion on intracellular pH regulation in isolated rat bile duct epithelial cells. J Clin Invest 1993;92:1314–1325. [18] Farouk M, Vigna SR, McVey DC, Meyers WC. Localization and characterization of secretin binding sites expressed by rat bile duct epithelium. Gastroenterology 1992;102:963–968. [19] Alpini G, Roberts S, Kuntz SM, Ueno Y, Gubba S, Podila PV, et al. Morphological, molecular, and functional heterogeneity of cholangiocytes from normal rat liver. Gastroenterology 1996;110:1636–1643. [20] Huang JL, Biehl TR, Lee FT, Zimmer PW, Ryan Jr JA. Outcomes after resection of cholangiocellular carcinoma. Am J Surg 2004;187:612–617. [21] Hejna M, Pruckmayer M, Raderer M. The role of chemotherapy and radiation in the management of biliary cancer: a review of the literature. Eur J Cancer 1998;34:977–986. [22] de Groen PC, Gores GJ, LaRusso NF, Gunderson LL, Nagorney DM. Biliary tract cancers. N Engl J Med 1999;341:1368–1378. [23] Hamilton SR, Aaltonen LA. World Health Organization classification of tumours. Pathology and genetics of tumours of the digestive system. Lyon: IARC Press; 2000. [24] Ulrich II CD, Wood P, Hadac EM, Kopras E, Whitcomb DC, Miller LJ. Cellular distribution of secretin receptor expression in rat pancreas. Am J Physiol 1998;275:G1437–G1444. [25] Ulrich II CD, Holtmann M, Miller LJ. Secretin and vasoactive intestinal peptide receptors: members of a unique family of G protein-coupled receptors. Gastroenterology 1998;114:382–397. [26] Patel DR, Kong Y, Sreedharan SP. Molecular cloning and expression of a human secretin receptor. Mol Pharmacol 1995;47:467–473. [27] Rubin EM, Martin AA, Thung SN, Gerber MA. Morphometric and immunohistochemical characterization of human liver regeneration. Am J Pathol 1995;147:397–404. [28] Ko¨rner M, Eltschinger V, Waser B, Schonbrunn A, Reubi JC. Value of immunohistochemistry for somatostatin receptor subtype sst2A in cancer tissues: lessons from the comparison of antisst2A antibodies with somatostatin receptor autoradiography. Am J Surg Pathol 2005;29:1642–1651. [29] Ding WQ, Cheng ZJ, McElhiney J, Kuntz SM, Miller LJ. Silencing of secretin receptor function by dimerization with a misspliced variant secretin receptor in ductal pancreatic adenocarcinoma. Cancer Res 2002;62:5223–5229. [30] Nakanuma Y, Hoso M, Sanzen T, Sasaki M. Microstructure and development of the normal and pathologic biliary tract in humans, including blood supply. Microsc Res Tech 1997;38:552–570. [31] Roskams TA, Theise ND, Balabaud C, Bhagat G, Bhathal PS, Bioulac-Sage P, et al. Nomenclature of the finer branches of the biliary tree: canals, ductules, and ductular reactions in human livers. Hepatology 2004;39:1739–1745. [32] Jiang S, Ulrich C. Molecular cloning and functional expression of a human pancreatic secretin receptor. Biochem Biophys Res Commun 1995;207:883–890. [33] Ding WQ, Kuntz S, Bohmig M, Wiedenmann B, Miller LJ. Dominant negative action of an abnormal secretin receptor arising from mRNA missplicing in a gastrinoma. Gastroenterology 2002;122:500–511.

835

[34] Reubi JC, La¨derach U, Waser B, Gebbers JO, Robberecht P, Laissue JA. Vasoactive intestinal peptide/pituitary adenylate cyclase-activating peptide receptor subtypes in human tumors and their tissues of origin. Cancer Res 2000;60:3105–3112. [35] Livingston JN, Einarsson K, Backman L, Ewerth S, Arner P. Glucagon receptor of human liver. Studies of its molecular weight and binding properties, and its ability to activate hepatic adenylyl cyclase of non-obese and obese subjects. J Clin Invest 1985;75:397–403. [36] Martı´nez-Anso´ E, Castillo JE, Diez J, Medina JF, Prieto J. Immunohistochemical detection of chloride/bicarbonate anion exchangers in human liver. Hepatology 1994;19:1400–1406. [37] Shaw JL, Gackenheimer SL, Gehlert DR. Functional autoradiography of neuropeptide Y Y1 and Y2 receptor subtypes in rat brain using agonist stimulated [35S]GTPgammaS binding. J Chem Neuroanat 2003;26:179–193. [38] Turnberg LA, Grahame G. Bile salt secretion in cirrhosis of the liver. Gut 1970;11:126–133. [39] Alpini G, Ulrich 2nd CD, Phillips JO, Pham LD, Miller LJ, LaRusso NF. Upregulation of secretin receptor gene expression in rat cholangiocytes after bile duct ligation. Am J Physiol 1994;266:G922–G928. [40] Alpini G, Ulrich C, Roberts S, Phillips JO, Ueno Y, Podila PV, et al. Molecular and functional heterogeneity of cholangiocytes from rat liver after bile duct ligation. Am J Physiol 1997;272:G289–G297. [41] Knuchel J, Kra¨henbu¨hl S, Zimmermann A, Reichen J. Effect of secretin on bile formation in rats with cirrhosis of the liver: structure-function relationship. Gastroenterology 1989;97:950–957. [42] Alpini G, Lenzi R, Sarkozi L, Tavoloni N. Biliary physiology in rats with bile ductular cell hyperplasia. Evidence for a secretory function of proliferated bile ductules. J Clin Invest 1988;81:569–578. [43] Lenzen R, Alpini G, Tavoloni N. Secretin stimulates bile ductular secretory activity through the cAMP system. Am J Physiol 1992;263:G527–G532. [44] Taillandier J, Dumont M, Mesa V, Degott C, Erlinger S. An increase of choleresis in secondary biliary cirrhosis in rats is caused by bile duct secretion. Gastroenterol Clin Biol 1990;14:313–318. [45] Chiba T, Yamatani T, Yamaguchi A, Morishita T, Nakamura A, Kadowaki S, et al. Mechanism for increase of gastrin release by secretin in Zollinger-Ellison syndrome. Gastroenterology 1989;96:1439–1444. [46] Liehr RM, Melnykovych G, Solomon TE. Growth effects of regulatory peptides on human pancreatic cancer lines PANC-1 and MIA PaCa-2. Gastroenterology 1990;98:1666–1674. [47] Douziech N, Lajas A, Coulombe Z, Calvo E, Laine J, Morisset J. Growth effects of regulatory peptides and intracellular signaling routes in human pancreatic cancer cell lines. Endocrine 1998;9:171–183. [48] Sirica AE. Cholangiocarcinoma: molecular targeting strategies for chemoprevention and therapy. Hepatology 2005;41:5–15. [49] Fiebiger WC, Scheithauer W, Traub T, Kurtaran A, Gedlicka C, Kornek GV, et al. Absence of therapeutic efficacy of the somatostatin analogue lanreotide in advanced primary hepatic cholangiocellular cancer and adenocarcinoma of the gallbladder despite in vivo somatostatin-receptor expression. Scand J Gastroenterol 2002;37:222–225. [50] de Jong M, Breeman WA, Bernard BF, Bakker WH, Visser TJ, Kooij PP, et al. Tumor response after [(90)Y-DOTA(0),Tyr(3)]octreotide radionuclide therapy in a transplantable rat tumor model is dependent on tumor size. J Nucl Med 2001;42:1841–1846.