1 Introduction

Hepatocytes are highly differentiated epithelial cells that constitute approximately 80% of the liver. Under conditions of normal tissue renewal and following injury or surgical resection, hepatocytes rapidly proliferate to regenerate a normal functioning liver. However, if the extent of injury is too severe and hepatocyte regeneration is compromised, activation of a biliary stem cell compartment leads to expansion of a bipotent oval cell population with the capacity for hepatocytic and ductular pathways of differentiation (Sell and Pierce 1994; Coleman et al. 1997; Dabeva et al. 1997; Sell and Ilic 1997). Oval cells are not the only progenitor populations present in adult liver. Four classes of hepatic stem-like cells have been defined in human liver (Sell 1998), one or more of which may correspond to stem-like cells described in normal and injured rat liver (Novikoff et al. 1996; Tateno and Yoshizato 1996; Hixson and Fowler 1997; Gordon et al. 2000; Hixson et al. 2000). These findings led to the conclusion that the liver is populated with bipotent progenitor cells in various states of differentiation or maturation (Sell 1998).

Understanding the molecular basis for differences in origin, phenotype, and differentiation ability of these various progenitors will require comparative analysis of gene expression at both RNA and protein levels. Although valuable information has been obtained by transplantation analysis using fetal liver isolates or subfractions isolated by physical criteria rather than phenotype, informative comparisons at the molecular level will require phenotypically defined cell isolates. In two recent studies, this goal was reached by isolating clonal progenitor cell populations from total fetal liver isolates or isolates depleted of hematopoietic cells (12–14-day fetal livers) and demonstrating that these isolates can be induced to express ductal or hepatocyte antigens in vitro (Kubota and Reid 2000). Whether they can undergo complete ductal or hepatocyte differentiation in vivo remains to be determined. In this regard, it was reported that progenitors isolated by fluorescence-activated cell sorting (FACS) using anti-α-6 and anti-β-1 integrin antibodies underwent pancreatic, intestinal, hepatocytic, and bile ductal differentiation following transplantation into these organ sites (Suzuki et al. 2002).

A long-term goal in our laboratory has been the identification of surface markers effective for isolating distinct cell subpopulations that arise during differentiation of cholangiocytes, the cells lining the intrahepatic bile ducts. Our interest in cholangiocytes stemmed from previous reports showing that oval cells of ductular origin could undergo ductal or hepatocyte differentiation. In a recent report, we described a panel of monoclonal antibodies (MAbs) recognizing surface markers on fetal cholangiocytes showing expression patterns that defined four stages in cholangiocyte differentiation (Hixson et al. 2000). FACS analysis of cultured cholangiocytes and oval cells labeled by indirect immunofluorescence with these MAbs showed high levels of expression on the cell surface of bile duct epithelial (BDE) cells, a localization that has made this panel of MAbs invaluable for isolating bipotent fetal cholangiocytes (Simper-Ronan et al. 2006). Identification of the proteins recognized by this panel of MAbs has been a major ongoing effort in our laboratory that has focused on the identification of the antigen defined by the monoclonal antibody OC.10 (MAb OC.10). MAb OC.10 was raised by nude mouse reconstitution against injected colonies of rat oval cells and has been shown to recognize one of the earliest markers on nascent bile ducts that first appear in rat liver around embryonic day 16. The OC.10 antigen is also expressed at high levels by adult rat cholangiocytes, prostate epithelial cells, and hepatocellular carcinomas with mixed oval cell/hepatocyte phenotypes. Reactivity of MAb OC.10 has not been observed against hepatocytes (Hixson et al. 2000). In this report, we present data that identifies OC.10 as an epitope displayed by a surface form of the constitutively expressed heat shock protein, Hsc70. We also demonstrate that intrasplenic injection of MAb OC.10 strongly labels intraportal/intralobular bile ducts also positive for the ductal cytokeratin marker OV-6 (Dunsford et al. 1989; Hixson et al. 2000) and promotes morphogenesis of bile duct structures in newborn rat liver.

2 Materials and methods

2.1 In situ labeling of cholangiocytes with injected MAb OC.10

Timed pregnancy, wild-type Fisher 344 rats were purchased from Harlan (Indianapolis, IN, USA). Animals were fed standard rat chow ad libitum and were housed in an alternating 12-h light/dark cycle environment. Newborn Fisher rats were injected into the spleen at 24-h intervals for three consecutive days with 20 µg of MAb OC.10, MAb OC.5, a control IgM antibody that reacts with the cell surface of cholangiocytes (Hixson et al. 2000), or MAb BD.1, an IgM antibody against a cholangiocyte cytoplasmic protein (Yang et al. 1993a). Five hours after the third injection, livers were excised and frozen in hexane chilled with a dry ice–acetone bath. Serial frozen tissue sections were fixed in ice-cold acetone for 10 min and stained with Alexa Fluor 594 goat anti-mouse IgM secondary antibody (1:400; Invitrogen Corporation, Carlsbad, CA, USA) to observe in situ localization of injected MAb OC.10 and MAb OC.5. To identify bile ducts, serial sections from MAb OC.10- and MAb OC.5-injected animals were incubated for 30 min at room temperature with MAb OV-6 (IgG; 1:300 diluted in phosphate-buffered saline (PBS)), an antibody recognizing an epitope shared by cytokeratin 19 (CK19) and CK14 (Dunsford et al. 1989; Bisgaard et al. 1993). After several PBS washes, the sections were incubated at room temperature with goat anti-mouse IgG (Fc; Pierce Chemical Co., Rockford, IL, USA) and Alexa Fluor 594 goat anti-mouse IgM secondary antibodies. The number of OV-6-positive bile ducts with a well-defined lumen was determined for approximately 50 portal areas from each injected animal. Differences between the average number of lumen positive bile ducts (LPBD) in MAb OC.10 (group A) and MAb OC.5 (group B)-injected animals were tested for statistical significance using Student’s t test with P values less than 0.05 indicating significance. National Institutes of Health guidelines for the care and use of laboratory animals were followed throughout this investigation. All animal protocols carried out in these studies were approved by the Rhode Island Hospital Institutional Animal Care and Use Committee.

2.2 Purification of OC.10 antigen

Rat biliary tree isolated following collagenase perfusion of the liver as described by Yang et al. (1993a) was solubilized in 0.5% Triton X-100 with protease inhibitors (Roche Diagnostics Co., Indianapolis, IN, USA) for 2 h at 4°C. Clarified extract was passed over wheat-germ agglutinin coupled to Sepharose at a protein concentration of 5 mg/ml. Flow through was submitted to QA anion exchange chromatography using 20 mM Tris, pH 7.2. The unbound fraction was dialyzed against 10 mM phosphate buffer, pH 6.8 and loaded onto a hydroxyapatite column (Bio-Rad Laboratories, Inc., Hercules, CA, USA). Fractions eluted with a 0.1–0.5-M Na2PO4 gradient, pH 6.8, were resolved by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and OC.10-positive fractions identified by immunoblot analysis were pooled, dialyzed against 0.5 mM Tris pH 7.2, and loaded onto a VyDac C4 reverse phase column. Aliquots from fractions containing proteins eluted in a 0–90% gradient of acetonitrile with 0.1% trifluoroacetic acid were dialyzed against 0.5 mM Tris buffer pH 7.2 and subjected to immunoblot analysis as described below. A duplicate gel was stained with Sypro Ruby (Invitrogen Corporation) and a band with an apparent molecular weight identical to the reactive band in immunoblots was excised and submitted to the Yale Cancer Center Mass Spectrometry Resource/W. M. Keck Foundation Biotechnology Resource Laboratory (New Haven, CT, USA) where an in-gel trypsin digest was performed. Tryptic peptides resolved by reverse phase high-performance liquid chromatography (HPLC) were analyzed by electrospray ionization mass spectrometry. Masses of major tryptic peptides were compared to theoretical tryptic peptide profiles for known protein in the OWL protein database using the Profound Server.

2.3 Cloning of rat Hsc70

Reverse transcriptase-polymerase chain reaction (RT-PCR) was used to generate a cDNA corresponding to the full-length rat wild-type Hsc70. RT-PCR amplification was performed using oligo dT-primed rat liver RNA and sequence specific primers for Hsc70 containing a 5′ EcoRI site and a 3′ BamHI site. The amplified PCR products were directionally ligated into pFLAG-CMV5a (Sigma-Aldrich Co., St. Louis, MO, USA), which adds a carboxyl-terminal FLAG-tag to the expressed protein.

2.4 Cell culture and transient transfections

BDE4, a continuous line of rat liver bile duct epithelial cells, was derived and cultured as described (Yang et al. 1993b). Human 293T cells were cultured in DMEM/F12 media supplemented with 10% FBS and 50 μg/ml gentamicin (Invitrogen Corporation) and incubated at 37°C in a 5% CO2 humidified atmosphere. For heat shock experiments, BDE4 cells were incubated at 42°C for the indicated times in a 5% CO2 humidified atmosphere. Human 293T cells were transfected with Hsc70-FLAG constructs using Lipofectamine 2000 as described in the manufacturer’s instructions (Invitrogen Corporation). Transiently transfected 293T cells were harvested at 48 h and prepared for immunoprecipitations as described below. For small interfering RNA (siRNA) studies, BDE4 cells were grown to approximately 80% confluence and nucleofected with 21-nucleotide siRNAs using the Amaxa system (Amaxa Inc., Gaithersburg, MD, USA) as described in the manufacturer’s instructions using solution V and program X-005. Silencer pre-designed siRNAs targeting rat Hsc70 (heat shock protein 8) were from Ambion Inc. (Austin, TX, USA). The following sequences were used: 5′-GGAAUGUGCUCAUUUUUGATT-3′ and 5′-UCAAAAAUGAGCACAUUCCTT-3′ (siRNA ID #: 52960); 5′-GGUGGAAAUAAUUGCCAAUTT-3′ and 5′-AUUGGCAAUUAUUUCCACCTT-3′ (siRNA ID # 52780); 5′-GCUAUGUUGCUUUCACCGATT-3′ and 5′-UCGGUGAAAGCAACAUAGCTC-3′ (siRNA ID #: 201029). The Silencer negative control #1 siRNA was used to control for the effects of siRNA delivery. To observe knockdown of OC.10, 5 × 106 cells were nucleofected with siRNA, cultured for 72 h and examined for siRNA effects by immunoblot analysis. Anti-GAPDH antibody (Ambion Inc.) was used to normalize for equal loading of protein, and band densities were determined using Photoshop CS (Adobe, San Jose, CA, USA). The relative intensity for each sample band was determined by dividing the absolute intensity of each sample band by the absolute intensity of the GAPDH loading control. Statistical analysis was performed using a nonparametric, paired t test with GraphPad Prism software (GraphPad Software, Inc., La Jolla, CA, USA). P < 0.05 was considered to be statistically significant.

2.5 Immunofluorescent microscopy

BDE4 cells were seeded in two-well permanox chamber slides (Nalge Nunc International, Rochester, NY, USA) at a density of 1 × 105 cells/ml and incubated for 48–72 h in a 5% CO2 humidified chamber. Cells were washed three times in PBS and fixed and permeabilized for 15 min in ice-cold acetone. Live and acetone fixed BDE4 cells were blocked for 10 min in 1% bovine serum albumin (BSA)/PBS with 10% normal goat serum (Sigma-Aldrich Co.). Primary and secondary antibodies were diluted in 1% BSA/PBS with 10% normal goat serum and incubated sequentially with cells for 45 min at room temperature. Cells were examined by fluorescence microscopy using a Nikon Eclipse E800 microscope (Nikon Instruments, Inc., Melville, NY, USA). Antibody dilutions were as follows: MAb OC.10 (IgM), 1:500; monoclonal anti-Hsc70 (IgG; Stressgen Bioreagents, Victoria, British Columbia), 1:100; Alexa Fluor 488 goat anti-mouse IgG and Alexa Fluor 594 goat anti-mouse IgM secondary antibodies (Invitrogen Corporation), 1:400.

2.6 Immunoprecipitations

Cells were lysed in situ with NP-40 lysis buffer (1% NP-40, 50 mM Tris, pH 7.5, 150 mM NaCl, and 1 Complete Protease Inhibitor tablet [Roche Diagnostics Co.]) or with the M-PER Mammalian Protein Extraction Reagent (Pierce Chemical Co.) as described in the manufacturer’s instructions. Detergent insoluble material was removed by centrifugation at 14,000 rpm for 15 min at 4°C. For Hsc70 immunoprecipitations, 500 µg total protein was incubated with 5 µg of monoclonal anti-Hsc70 or MAb OC.10 (IgM) for 1 h at 4°C. Lysates were incubated with rec-Protein G-Sepharose 4B conjugate (Zymed Laboratories Inc., San Francisco, CA) or rat anti-mouse IgM Dynabeads M-450 (Dynal Biotech ASA, Oslo, Norway) for 1 h at 4°C. To immunoprecipitate Hsc70-FLAG tagged protein, 500 μg of total 293T cell lysate was incubated with anti-FLAG M2 agarose according to the manufacturer’s instructions (Sigma). Following five washes with RIPA buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 0.5% sodium deoxycholate, 1% Triton X-100, 0.1% SDS, and 1 Complete Protease Inhibitor tablet), the Protein G-Sepharose 4B beads were suspended in nonreducing SDS sample buffer (Bio-Rad Laboratories) and boiled for 5 min. The anti-FLAG M2 agarose was suspended in reducing SDS-PAGE sample buffer and boiled for 5 min.

2.7 Biotinylation of cell surface proteins

BDE4 cells were cultured to 60% confluence, washed with PBS, and biotinylated with 0.5 mg/ml Sulfo-NHS-LC-LC Biotin (Pierce Chemical Co.) in PBS for 30 min at 4°C. Cells were washed with 50 mM glycine/PBS for 15 min at 4°C to quench unreacted biotin. Cells harvested using nonenzymatic cell dissociation solution (Sigma-Aldrich Co.) were lysed with M-PER Mammalian Protein Extraction Reagent (Pierce Chemical Co.). Lysate was incubated with immobilized neutravidin biotin-binding beads (Pierce Chemical Co.) for 2 h at 4°C. Neutravidin-precipitated proteins released by boiling in SDS-PAGE sample buffer for 5 min were separated by SDS-PAGE and visualized using Gelcode Blue Stain Reagent (Pierce Chemical Co.). Recombinant Hsc70 protein was used as a size standard to excise a band in the range of 70 kDa. Biotinylated cell surface proteins were prepared and analyzed by liquid chromatography electrospray ionization tandem mass spectrometry (LC-ESI-MS/MS) on a QSTAR XL mass spectrometer (Applied Biosystems, Framingham, MA, USA and Sciex, Concord, Ontario, Canada) at the Center of Biomedical Research Excellence (COBRE) Proteomics Core Facility (Providence, RI, USA) as described (Clifton et al. 2006).

2.8 Western blotting

Proteins separated by 7.5% SDS-PAGE were transferred onto nitrocellulose (Trans-Blot Transfer Medium, Bio-Rad Laboratories) and immunoblotted as described (Yang et al. 1993a). Reactivity was visualized by chemiluminescence with SuperSignal West Pico chemiluminescent substrate according to the manufacturer’s instructions (Pierce Chemical Co.). Recombinant bovine Hsc70 and rat Hsp70 proteins were obtained from Stressgen Bioreagents, Victoria, British Columbia. Antibody dilutions were as follows: monoclonal OC.10, 1:1,000; monoclonal anti-Hsc70, 1:1,000; polyclonal anti-Hsp70 (IgG; Stressgen, Victoria, BC, Canada), 1:5,000; monoclonal anti-α-tubulin (IgG; Calbiochem Inc., San Diego, CA, USA), 1:1,000; polyclonal anti-actin (IgG; Sigma-Aldrich Co.), 1:500; monoclonal anti-GAPDH (IgG; Ambion Inc.), 1:50,000; monoclonal anti-FLAG M2 antibody, 1:1,000; and goat anti-mouse IgM, goat anti-mouse IgG, and goat anti-Rabbit Ig horseradish peroxidase (HRP)-conjugated secondary antibodies (Biosource, Camarillo, CA, USA), 1:25,000.

2.9 Epitope mapping using phage display and peptide arrays

Initial epitope mapping studies were performed with the Ph.D.-12 Phage Display Peptide Library Kit according to the manufacturer’s instructions (New England Biolabs Inc., Ipswich, MA, USA). Briefly, a sterile polystyrene petri dish was coated with antibody by incubating with 150 μg of purified MAb OC.10 in 0.1 M NaHCO3 (pH 8.6) overnight at 4°C, followed by blocking in 0.1 M NaHCO3 (pH 8.6) containing 5 mg/ml of BSA at 4°C for 1 h, and six washes with TBST (TBS + 0.1% [v/v] Tween 20). The phage library (4 × 1010/ml TBST) was then added to the blocked plate and incubated at room temperature for 1 h. The plate was washed ten times with TBST, the antibody-bound phage was eluted with 0.2 M glycine–HCl (pH 2.2), and the collected eluate was neutralized with 1 M Tris–HCl (pH 9.1). Following amplification according to the manufacturer’s protocol, the eluted phage was taken through three additional cycles of biopanning and amplification to enrich for the phages with the highest affinity. Ten clones were submitted for DNA sequencing to the W.M. Keck Biotechnology laboratory at Yale University (New Haven, CT, USA) to determine the nucleotide sequences encoding the phage displayed peptides. Epitope mapping was also performed using a PepSpots peptide array spanning amino acids 340–410 of rat Hsc70. Peptides reactive with MAb OC.10 were identified using the manufacturer’s protocol (JPT Peptide Technologies GmbH, Berlin, Germany). Briefly, a membrane containing 30 overlapping 13-mer PepSpots peptides was rinsed in methanol for 5 min, washed in TBS (50 mM Tris, 137 mM NaCl, 2.7 mM KCl, pH 8.0) for 10 min and blocked at room temperature for 2 h with SuperBlock Blocking Buffer in TBS (Pierce Chemical Co.). Purified MAb OC.10 was conjugated to HRP using the Lightning-Link HRP Conjugation Kit as described by the manufacturer’s instructions (Innova Biosciences Ltd., Cambridge, UK) and 1 μg/ml of labeled antibody was incubated for 2 h at room temperature with the blocked PepSpots membrane. Reactivity was visualized by chemiluminescence with SuperSignal West Pico Chemiluminescent Substrate according to the manufacturer’s protocol (Pierce Chemical Co.).

3 Results

3.1 In situ labeling of cholangiocytes with MAb OC.10 promotes ductal morphogenesis

Our laboratory has successfully used MAb OC.10 in combination with other antibodies to isolate antigenically defined subpopulations of OC.10-positive fetal cholangiocytes (Simper-Ronan et al. 2006). In the course of the present investigations, we discovered that within an hour of injection of MAb OC.10 into the spleen of newborn Fisher rats, the antibody appeared exclusively on the surface of intraportal/intralobular bile ducts where it remained for at least 5 h post-injection (not shown). Our ability to label bile ducts in situ with MAb OC.10 provided a means to determine if antibody binding would elicit a biological response. As shown in Fig. 1, injection of MAb OC.10 for three consecutive days resulted in strong and highly specific labeling of intraportal/intralobular ducts (Fig. 1a). In contrast, very few bile ducts were labeled in situ by MAb OC.5, although some larger ducts did exhibit weak reactivity (Fig. 1c). As expected, bile ducts were also negative for MAb BD.1, a cytoplasmic marker found on bile ducts but not oval cells (not shown; Yang et al. 1993b). These findings showed that the binding of MAb OC.10 was specific for the surface exposed Hsc70 epitope and was not the result of nonspecific adhesion of IgM antibody to intrahepatic bile ducts. To confirm that injected MAb OC.10 was labeling bile ducts, serial cryosections from MAb OC.10- and OC.5-injected animals were incubated with MAb OV-6, a biliary marker that reacts with CK19 and CK14 (Dunsford et al. 1989; Hixson et al. 2000). Primary staining was followed with a cocktail of anti-IgM and IgG secondary antibodies to compare the labeling reactivity of the injected IgM primary antibodies with OV-6 expression. As shown in Fig. 1b, the reactivity profile of injected MAb OC.10 was restricted to OV-6-positive ducts.

Fig. 1
figure 1

Injected MAb OC.10 promotes bile duct formation in newborn rat liver. MAb OC.10 (20 μg) or control IgM MAb OC.5 was injected into the spleen of newborn Fisher rats at 24-h intervals for 3 days. Five hours after the third injection, livers were excised, frozen, sectioned, and stained with Alexa Fluor 594 goat anti-mouse IgM secondary antibody to determine if injected antibody labeled intrahepatic ducts. As shown, injection of MAb OC.10 resulted in strong labeling of intrahepatic bile ducts (a). In contrast, MAb OC.5 did not label ductal cells (c). To view bile ducts, tissue sections were stained with the ductal-specific marker MAb OV-6 (IgG) and a cocktail of anti-IgM and IgG (Fc) secondary antibodies. Well-defined MAb OV-6+ ducts located around portal vein branches were compared between MAb OC.10- and MAb OC.5-injected liver sections. Sections from MAb OC.10-injected animals show that the OC.10 (a) and OV-6 (b) antigens co-localize and that MAb OC.10 promotes bile duct formation as evidenced by the significantly (P < 0.05) higher frequency of defined bile ducts (arrows) per portal vein area as compared to MAb OC.5-injected animals (d)

Full size image

During the early post-partum period, the biliary system in the liver continues to undergo morphological, phenotypic, and functional maturation. Bile canaliculi retain the distended morphology characteristic of late gestation fetal liver. Bile ducts often lack a clearly defined central lumen and have not yet begun to express surface markers such as OC.4 that are characteristic of mature ducts (Hixson et al. 2000). To determine if in situ labeling of ducts with MAb OC.10 would stimulate their morphological or phenotypic maturation, serial cryosections from MAb OC.10-injected newborn rats were labeled with MAb OC.4 that first appears on cholangiocytes at 2–3 weeks after birth (Hixson et al. 2000). Extensive analysis of serial frozen sections did not reveal any OC.4-positive bile ducts in newborn rats injected with MAbs OC.10, OC.5, BD.1, or uninjected controls (data not shown), indicating that MAb OC.10 did not accelerate phenotypic maturation as defined by the appearance of OC.4. To determine if injected antibodies had an effect on morphogenic maturation, serial cryosections were labeled with MAb OV-6 and the number of LPBD was determined for 252 or 169 nonoverlapping portal areas from five MAb OC.10-injected or three MAb OC.5-injected animals, respectively. Injection of MAb OC.10 significantly increased the number of well-defined LPBD per portal area (Fig. 1a, b) relative to MAb OC.5-injected control animals (Fig. 1c, d). As summarized in Table 1, this analysis yielded an average of 3 LPBD/portal area in MAb OC.10-injected animals, a frequency that was significantly higher (P < 0.05) than the average of 1.1 LPBD/portal area in MAb OC.5-injected animals. The morphogenic changes induced in situ by MAb OC.10 suggest that the reactive epitope is located in a functional domain of Hsc70 that is activated by antibody binding.

Table 1 Effect of MAb OC.10 on intrahepatic ductal morphology, where P represents the total number of portal areas counted, BD represents the total number of defined bile ducts counted, and BD/P represents the average number of bile ducts per portal area
Full size table

3.2 Mass spectrometry identified the OC.10 antigen as Hsc70

As a first step to identify the protein recognized by MAb OC.10, we initiated experiments to purify the reactive antigen. After numerous unsuccessful attempts to affinity purify OC.10, we designed a protocol based on established methods to isolate OC.10 from a Triton X-100 extract of rat biliary tree. The final protocol involved the removal of major contaminants by passing the extract through wheat-germ agglutinin and QA anion exchange columns, which do not bind OC.10, followed by binding to hydroxyapatite and elution with a sodium phosphate gradient. Fractions testing positive by Western blot were pooled and further purified by reverse phase chromatography on a C4 hydrophobic interaction column and SDS-PAGE. A Sypro Ruby band with the same mobility as MAb OC.10 in a duplicate immunoblot was excised, digested in-gel with trypsin, and analyzed by HPLC and tandem mass spectrometry (LC-ESI-MS/MS). LC-ESI-MS/MS analysis indicated that the presumptive OC.10 band contained highly purified Hsc70, a constitutively expressed member of the heat shock protein 70 (HSP70) family. Figure 2 shows the Hsc70 peptides identified from MS analysis of two separate preparations of purified OC.10. The first preparation identified 13 peptides that matched bovine Hsc70 (gi 123644) and represented 28% of the protein. The second preparation identified 17 Hsc70 peptides representing 34% of the full-length bovine Hsc70 sequence.

Fig. 2
figure 2

Matrix-assisted laser desorption ionization mass spectrometry (MALDI-MS) identified the OC.10 antigen as bovine Hsc70 protein (gi 123644). Two separate MALDI-MS analyses of trypsinized OC.10 protein obtained from Coomassie-stained gels were performed and revealed a number of Hsc70 peptide fragments. The 13 Hsc70 peptide fragments identified in MALDI-MS analysis 1 are italicized and indicated by uppercase letters. MALDI-MS analysis 2 identified 17 peptide fragments of Hsc70 that are italicized and underlined. Analysis 1 and analysis 2 covered 28% and 34% of the bovine Hsc70 sequence, respectively

Full size image

3.3 MAb OC.10 reacts with recombinant and immunoprecipitated Hsc70

To verify the identity between OC.10 and Hsc70, the reactivity of MAb OC.10 with recombinant bovine Hsc70 was confirmed by immunoblot analysis. A commercially available anti-Hsc70 antibody was used as a positive control. Because of the high level of sequence homology between Hsc70 and Hsp70 (Tavaria et al. 1996), the specificity of MAb OC.10 was also tested against recombinant Hsp70 protein. As shown in Fig. 3, MAb OC.10 and the anti-Hsc70 MAb each reacted with recombinant Hsc70 protein (Fig. 3, lanes 1–3) but showed no detectable reactivity against recombinant Hsp70 protein (Fig. 3, lanes 5–7). To further confirm the reactivity of MAb OC.10 with Hsc70, we transiently expressed a rat Hsc70-FLAG tagged expression plasmid in the human 293T cell line. As shown in Fig. 4, immunoblot analysis of the anti-FLAG immunoprecipitate showed that MAb OC.10, commercial anti-Hsc70, and anti-FLAG antibodies all displayed strong reactivity with the FLAG-tagged Hsc70 (Fig. 4, lane 2). No antibody reactivity was observed against immunoprecipitates from negative control 293T lysates (Fig. 4, lane 1).

Fig. 3
figure 3

Immunoblotted MAb OC.10 reacts with Hsc70 protein. Recombinant bovine Hsc70 (lanes 1–3) and recombinant Hsp70 (lanes 5–7) proteins were immunoblotted with MAb OC.10 (a), monoclonal anti-Hsc70 (b), and polyclonal anti-Hsp70 (c). For each protein, 50 ng (lanes 1 and 5), 100 ng (lanes 2 and 6), and 200 ng (lanes 3 and 7) were resolved on 7.5% polyacrylamide gels and transferred onto nitrocellulose. Lane 4 was empty

Full size image
Fig. 4
figure 4

MAb OC.10 reacts with Hsc70-FLAG protein over-expressed in human 293T cells. An Hsc70-FLAG expression construct was transfected into human 293T cells. Immunoblots of protein immunoprecipitated with anti-FLAG agarose identified an ~70-kDa protein that reacted with MAb OC.10, monoclonal anti-Hsc70, and anti-FLAG antibodies (lane 2). Reactivity was not observed against the immunoprecipitated fraction from nontransfected cell lysates (lane 1) demonstrating specificity for FLAG-tagged Hsc70 in lane 2

Full size image

To obtain further evidence for the specificity of MAb OC.10 for Hsc70, BDE4 lysate was immunoprecipitated with monoclonal anti-Hsc70 antibody. As shown in Fig. 5, MAb OC.10 reacted with a 70-kDa peptide in total BDE4 lysate and in anti-Hsc70 immunoprecipitates. Reactivity was not observed against fractions eluted from lysates incubated with beads alone, anti-Hsc70 antibody in PBS (i.e., no lysate), or a control anti-Ras IgG. Efficient immunoprecipitation of Hsc70 protein was confirmed when the blot was stripped and re-probed with monoclonal anti-Hsc70 antibody.

Fig. 5
figure 5

MAb OC.10 reacts with protein immunoprecipitated with anti-Hsc70 antibody. BDE4 cell extract (500 µg) prepared in NP-40 lysis buffer with 2 mM ADP was immunoprecipitated with 5 µg of rat monoclonal anti-Hsc70 (Hsc) or 5 µg of control rat IgG (anti-Ras, IgG) for 1 h at 4°C. Total cell lysate (20 μg, Lys), monoclonal anti-Hsc70 in PBS (i.e., no cell extract), and precipitate from lysate with beads alone (Bds) were loaded as additional controls. The membrane was immunoblotted with OC.10 (top panel) and stripped and re-probed with monoclonal anti-Hsc70 (lower panel)

Full size image

3.4 The OC.10 antigen behaves like Hsc70

The Hsc70 and Hsp70 proteins are differentially regulated with Hsc70 being constitutively expressed and Hsp70 induced by environmental stress (e.g., heat shock; De Maio 1999). To further demonstrate that MAb OC.10 reacted with Hsc70 and not an epitope shared by Hsc70 and Hsp70, we examined the effect of heat shock on protein levels recognized by MAb OC.10. Anti-Hsc70 and anti-Hsp70 antibodies were used to distinguish between constitutive and induced proteins. As shown in Fig. 6, heat treatment of BDE4 cells did not induce an increase in the level of protein recognized by MAb OC.10 or anti-Hsc70 following 42°C heat shock treatments of 0, 2, 6, or 24 h. In contrast, the levels of protein detected by anti-Hsp70 were significantly increased following heat shock treatment.

Fig. 6
figure 6

Heat shock does not induce OC.10 antigen expression. Extracts prepared from rat BDE4 cells that were heat stressed at 42°C for 0, 2, 6, or 24 h were immunoblotted with the indicated antibody. Anti-α-tubulin was used as a loading control for cell lysates and was observed to react at a lower molecular weight than described by the supplier for mouse (~60 kDa) and gerbil (57 kDa). OC.10 antigen shows a similar expression profile as the constitutively expressed Hsc70 protein. Hsp70 protein shows inducible expression kinetics in response to heat shock. Recombinant Hsp70 (Hsp) and Hsc70 (Hsc) protein were loaded as controls (0.1 μg/lane). Five micrograms of total cell lysate was loaded per lane

Full size image

To further show that the OC.10 antigen is Hsc70, we nucleofected BDE4 cells with rat Hsc70-specific siRNAs, prepared total cell lysates, and tested for expression of the OC.10 antigen and Hsc70 by immunoblot analysis. Of the three Hsc70-specific siRNAs tested, one demonstrated more significant knockdown of Hsc70 levels (siRNA ID #: 52960) than the other two at 72 h post-nucleofection (data not shown). Subsequent Hsc70 knockdown studies were limited to siRNA 52960. Immunoblot analysis showed that siRNA 52960 highly attenuated Hsc70 and OC.10 expression levels as compared with the mock-nucleofected control (Fig. 7), further confirming the specificity of MAb OC.10 for Hsc70 protein.

Fig. 7
figure 7

Hsc70 siRNA decreases OC.10 antigen expression levels. BDE4 cells were nucleofected with Hsc70-specific siRNAs (201029, 52780, 52960) or negative control siRNA (scrambled) using the Amaxa system as described under “Materials and methods.” The siRNA knockdown effects were examined by immunoblot analysis 72 h post-nucleofection using MAb OC.10 and anti-Hsc70 (SPA-815) antibody. Ten micrograms of total cell lysate was loaded per lane. The anti-GAPDH antibody was used as a loading control for the cell lysates. Immunoblot analysis demonstrated that Hsc70 siRNA 52960 provides highly efficient knockdown of Hsc70 and OC.10 expression levels (52960) as compared with the scrambled siRNA control (Scr). One representative experiment of three is shown. Mean±SEM of three separate experiments is graphed. *P < 0.05 versus scrambled siRNA nucleofection

Full size image

3.5 Immunofluorescent distribution of OC.10 and Hsc70 in BDE4 cell monolayers

We have shown previously that MAb OC.10 reacts specifically with ductal structures in fetal, newborn, and adult rat liver (Hixson et al. 2000). To examine more precisely the subcellular localization of protein epitopes recognized by MAb OC.10 and a commercially available monoclonal anti-Hsc70 antibody, rat BDE4 cells were examined by immunofluorescence microscopy. As shown in Fig. 8a, with acetone fixed cells, MAb OC.10 exhibited intense and uniform plasma membrane reactivity with minimal levels of reactivity against cytosolic Hsc70. In contrast, monoclonal anti-Hsc70 showed a distinct absence of detectable membrane reactivity and a pronounced cytosolic and nuclear staining, a well-known distribution pattern for Hsc70 (Tsukahara and Maru 2004). Co-localization of MAb OC.10 and anti-Hsc70 demonstrates the specificity for membrane and cytosolic epitopes, respectively. We next examined MAb OC.10 and anti-Hsc70 reactivity against live, nonpermeabilized BDE4 cells. As shown in Fig. 8b, reactivity of MAb OC.10 was limited to the leading edge of nonconfluent cell colonies where it appeared to be polarized towards the outer membrane surface. MAb OC.10 reactivity was not observed in other membrane domains and was absent or weakly expressed on cells in the interior of each colony. This was in contrast to the strong surface membrane reactivity observed on all cells following acetone fixation. Reactivity was not observed on live cells with the commercial monoclonal anti-Hsc70 antibody (not shown).

Fig. 8
figure 8

Immunofluorescent distribution of the OC.10 antigen and Hsc70 protein on cultured BDE4 cells. a Acetone-fixed BDE4 cells co-labeled with MAb OC.10 and monoclonal anti-Hsc70 antibodies. MAb OC.10 exhibited strong cell surface and low cytoplasmic reactivity while the commercially available monoclonal anti-Hsc70 antibody showed strong cytoplasmic but no detectable plasma membrane reactivity. b Immunofluorescent labeling of live, nonpermeabilized BDE4 cells with MAb OC.10. Combined fluorescence and phase contrast shows MAb OC.10 reactivity is limited to cells at the leading edge of expanding colonies and is not apparent in areas of confluence. Localization of MAb OC.10 is polarized towards the outer apical membrane surface of live BDE4 cells. Scale bars represent 25 μm

Full size image

3.6 Biotinylation of BDE4 cell surface proteins

To confirm biochemically that MAb OC.10 recognized a cell surface form of Hsc70, BDE4 surface proteins were labeled with sulfo-NHS-LC-LC biotin, a water-soluble reagent that does not cross intact plasma membranes. Biotinylated proteins immunoprecipitated with neutravidin beads were analyzed by Western blot. As shown in Fig. 9, both MAb OC.10 and anti-Hsc70 detected a biotinylated protein with a molecular weight of approximately 70 kDa in the neutravidin immunoprecipitated (IP) fraction. Cytosolic α-tubulin was detected in total lysate but was not present in the neutravidin bound fraction, demonstrating that cytosolic proteins were not biotinylated. To further confirm surface expression of Hsc70, a 70-kDa Gelcode Blue stained band of the surface biotinylated protein examined in Fig. 9 was excised, digested with trypsin, and analyzed by mass spectrometry. As shown in Table 2, LC-ESI-MS/MS analysis identified ten peptides that matched rat Hsc70 (gi 56385) with >99% confidence. In total, 17 Hsc70 peptides were identified that represented 37% of the Hsc70 protein sequence. These results demonstrate that a portion of the total pool of rat Hsc70 is expressed on the cell surface of rat BDE4 cells. In support of this finding, surface Hsc70 (gi 56385) was recently identified in purified plasma membrane preparations of the hepatocellular carcinoma cell line 253-NT-V (Lawson et al. 2006), suggesting that surface Hsc70 expression may be a more general phenomenon than previously appreciated.

Fig. 9
figure 9

Cell surface biotinylation of cultured BDE4 cells. Live BDE4 cells were incubated in PBS containing Sulfo-NHS-LC-LC biotin to label all cell surface proteins. Cells were harvested and incubated with immobilized neutravidin beads to precipitate biotinylated surface proteins. Immunoblot analysis with MAb OC.10 and anti-Hsc70 detected OC.10 antigen and Hsc70 protein, respectively in total cell extract (Lysate) and the neutravidin precipitated fraction (IP). Recombinant Hsc70 protein (Hsc70, 100 ng) was loaded as a positive control. The anti-α-tubulin immunoblot showed that α-tubulin, a known cytoplasmic protein, was not immunoprecipitated from the pool of biotinylated protein

Full size image
Table 2 Identification of Rattus norvegicus cell surface Hsc70 by mass spectrometry
Full size table

3.7 Epitope mapping of MAb OC.10

Although prior observations demonstrating cell surface localization of Hsc70 have been reported (Tsuboi et al. 1994; Hirai et al. 1998; Fishelson et al. 2001), little is known regarding the orientation of the protein at the cell surface. Furthermore, the morphogenic changes induced in situ by MAb OC.10 (Fig. 1) suggest that the reactive epitope is located in a functional domain that is activated by antibody binding. To characterize the epitope recognized by MAb OC.10, purified antibody was used to screen a phage display dodecapeptide library. After three rounds of selection and amplification, ten clones were randomly selected for DNA sequencing and the amino acid sequence of the inserts deduced. As shown in Table 3, five peptides with the sequence DYPTSSTAQPLP contained three residues (underlined) that aligned with amino acids 371 (Y), 377 (A), and 380 (L) that correspond to the N-terminal domain of Hsc70. Four additional peptides with the sequence TYCYPIHP[X]LLS contained two residues that aligned with amino acids 371 (Y) and 380 (L) of the same region. Although BLAST analysis did not identify these sequences as Hsc70, the conserved alignment of the underlined residues suggested that they might be part of the reactive epitope. To further evaluate the significance of this localization, region 340–410 of Hsc70 was reproduced by the PepSPOTs approach using overlapping arrays of 13 amino acid peptides offset by two and tested for their immunoreactivity against purified HRP-conjugated MAb OC.10. Strong reactivity mapped to peptides 1–8 and 15–17 that correspond to amino acids 340–366 and 368–384, respectively (Fig. 10 and Table 4). This is consistent with data obtained by phage display as the amino acids 371 (Y), 377 (A), and 380 (L) map to peptides 15 and 16 (Table 4). This pattern of immunoreactivity indicates that the surface specific epitope for MAb OC.10 resides in the N-terminal 340–384 amino acids of the ATPase domain of rat Hsc70.

Table 3 Amino acid similarities between phage clones and Hsc70 (amino acids 370–381)
Full size table
Fig. 10
figure 10

The MAb OC.10 epitope maps to the N-terminal ATPase domain of Hsc70. A membrane containing an overlapping series of 13 amino acid peptides that spanned region 340–410 of rat Hsc70 was used to define the epitope recognized by purified HRP-conjugated MAb OC10. MAb OC.10 reacted strongly with PepSPOTS 1–8 and 15–17 that correspond to amino acids 340–366 and 368–384, respectively

Full size image
Table 4 MAb OC.10-reactive peptide sequences corresponding to Hsc70 PepSpots membrane spanning amino acids 340–410
Full size table

4 Discussion

Previous work from our laboratory described the production of a panel of monoclonal antibodies reactive with bile duct and hepatic progenitor (oval cells in rodents) surface antigens. One such antibody, MAb OC.10, has been shown to recognize an epitope shared by fetal liver ductal cells, oval cells, mature bile duct cells (cholangiocytes), and hepatocellular carcinomas with oval cell characteristics but not by hepatocytes (Hixson et al. 2000). Because of its restricted surface expression profile, MAb OC.10 has been used in combination with other lineage-specific antibodies to isolate antigenically defined subpopulations present during liver development, renewal, and carcinogenesis (Simper-Ronan et al. 2006). In the present study, we have shown that intrasplenic injection of MAb OC.10 into newborn rat pups strongly labels intrahepatic bile duct cells also positive for the bile duct marker OV-6. Interestingly, the in situ labeling of OV-6-positive cholangiocytes by MAb OC.10 promoted the morphogenesis of well-defined bile duct structures relative to isotype-injected control animals. Since this finding was of particular interest, we designed a protocol for purifying OC.10 from isolated rat biliary tree and identified the protein recognized by MAb OC.10 as the constitutively expressed 70-kDa heat shock cognate protein, Hsc70. The reactive epitope for MAb OC.10 was mapped to the 340–384 stretch of amino acids showing that this region of the Hsc70 ATPase domain is exposed when presented on the cell surface. Moreover, we have previously reported that OC.10 is expressed by a subset of rat primary HCC that co-express other oval/ductal cell markers, a finding that suggested the OC.10 epitope was exposed on the surface of HCC derived from oval cell progenitors (Hixson et al. 2000). Taken together, these data indicate that Hsc70 has the potential to be expressed on the surface of certain tumor cells.

The Hsc70 protein, also known as heat shock cognate 73 (Hsc73), is a member of the heat shock protein 70 (HSP70) family (Sorger and Pelham 1987; De Maio 1999), a family of proteins ranging in size between 66 and 78 kDa. Members of the HSP70 family have three structural and functional domains: a 44-kDa N-terminal ATPase domain (residues 1–384), an 18-kDa protein binding domain (residues 385–543), and a 10-kDa variable domain (residues 544-646) that binds co-chaperone proteins (Mayer and Bukau 2005). Proteins of the HSP70 family, including Hsc70, have been shown to function as intracellular chaperones with essential roles in protein synthesis, folding and assembly, trafficking, and degradation (Daugaard et al. 2007). Hsc70 has also been implicated in various stages of clatherin-mediated endocytosis (Schlossman et al. 1984; Newmyer et al. 2003; Eisenberg and Greene 2007), antigen presentation (Zhou et al. 2005; Kamiguchi et al. 2008), and protection from apoptosis (de la Rosa et al. 1998; Dastoor and Dreyer 2000).

To the best of our knowledge, this is the first report to define a cell surface Hsc70 epitope exposed in situ on the surface of cholangiocytes and oval cells but not on hepatocytes. Although there have been previous reports describing cell surface expression (Tsuboi et al. 1994; Hirai et al. 1998; Fishelson et al. 2001), little is known regarding the orientation of Hsc70 in the plasma membrane or how this well-described cytoplasmic protein localizes to the cell surface. Previously, Hirai et al. (1998) identified a monoclonal antibody that recognized an epitope spanning amino acid residues 350–372 of the Hsc70 ATPase domain and confirmed by FACS analysis that this reactive epitope was expressed on the surface of Daudi cells. This is highly consistent with our current mapping data that places the reactive epitope for MAb OC.10 in a similar region that spans residues 340–384 of the Hsc70 ATPase domain. However, surface-exposed epitopes of Hsc70 may not be limited to the ATPase domain. Fishelson and colleagues used four different antibodies to define surface exposed epitopes located in the N-terminal ATPase domain and C-terminal peptide binding domain of Hsc70 on complement-stressed human K562 cells (Fishelson et al. 2001). Although the two antibodies directed against the N-terminal domain also reacted with Hsp70, the C-terminal antibodies were specific for Hsc70. Interestingly, surface reactivity with one of these C-terminal reactive antibodies, the commercially available SPA-816 (residues 556–568), exhibited primarily cytoplasmic staining when tested against our BDE cell lines (data not shown), suggesting that the presentation of extracellular Hsc70 domains may vary between different cell types.

In contrast to Hsc70, cell surface expression of Hsp70 has been well described on various tumor cell types (Ferrarini et al. 1992; Multhoff et al. 1995; Shin et al. 2003). Multhoff and colleagues have further demonstrated that the exposed extracellular domain of Hsp70 expressed on the surface of human colon carcinoma cells was situated in the C-terminal substrate binding domain and that this extracellular domain was associated with increased proliferative and cytolytic natural killer (NK) cell activity (Botzler et al. 1996, 1998; Multhoff et al. 1997, 1999). More recently, peptide scanning mapped a surface-exposed Hsp70 epitope to residues 454–461 of the C-terminal domain, and a synthetic peptide corresponding to residues 450–463 was shown to stimulate the proliferation and cytolytic activity of NK cells at concentrations similar to those described for full-length Hsp70 protein. Although the corresponding region for Hsc70 is highly homologous, Hsc70 failed to elicit a similar NK cell response (Multhoff et al. 2001).

Despite the high degree of similarity between Hsc70 and Hsp70, it has been possible to differentiate between these two proteins. It was shown by immunoblot analysis that MAb OC.10 reacted specifically with recombinant Hsc70 protein and did not show affinity for recombinant Hsp70 protein. Consistent with this specificity, heat shock treatment did not upregulate OC.10 or Hsc70 expression in treated BDE cells. Importantly, siRNA-mediated knockdown of Hsc70 expression resulted in decreased OC.10 expression in nucleofected BDE cells. The reduced reactivity of MAb OC.10 with immunoblots in comparison to the commercially available anti-Hsc70 antibody SPA-815 suggested that MAb OC.10 was either recognizing a subfraction of the total pool of Hsc70 protein or was displaying reduced affinity towards the denatured form of the protein. However, in immunofluorescence assays, MAb OC.10 reactivity was restricted to the cell surface and showed little reactivity with the more abundant pool of cytosolic Hsc70 recognized by SPA-815. Differences in immunofluorescence reactivity for MAb OC.10 were also observed between live and acetone fixed cells, with the latter displaying an abundant pool of surface Hsc70 and the former showing fluorescence limited to the plasma membrane at the outer edge of cell colonies. Although the significance of this finding is unclear at present, it suggests that the MAb OC.10 epitope in membrane areas of cell–cell contact is unmasked by acetone fixation, an indication that surface Hsc70 may be interacting directly with membrane lipids. Such an association has been previously shown for recombinant Hsc70 from its interaction with artificial phospholipid membranes that lead to the formation of nucleotide-dependent ion conducting channels (Arispe and De Maio 2000). The ability of both Hsc70 and Hsp70 to promote the aggregation of phosphatidylserine (PS) liposomes (Arispe et al. 2002) and to interact with PS on the surface of cultured PC12 cells (Arispe et al. 2004) provided further evidence for direct interactions with membrane lipids.

Hsp70 has also been shown to associate with lipid rafts (Broquet et al. 2003). These lipid-based organizational platforms, enriched in glycosphingolipids and cholesterol, are located in plasma membranes and have been implicated in a number of biological functions including membrane trafficking and signal transduction (Simons and Ikonen 1997; Hanzal-Bayer and Hancock 2007; Patra 2008). Interestingly, Multhoff and colleagues have recently shown that surface Hsp70 on human tumor cells interacts with globotriaosylceramide (Gb3), a glycosphingolipid associated with lipid rafts. Moreover, it was shown that binding of Hsp70 to artificial raft-like liposomes required the inclusion of Gb3 and further that depletion of ceramide-based glycosphingolipids, including Gb3, decreased Hsp70 surface presentation on tumor cells (Gehrmann et al. 2008). It will be interesting to determine if membrane lipid composition regulates surface Hsc70 epitope availability, as it may provide a possible explanation for the differences in MAb OC.10 reactivity between surface-reactive cholangiocytes and hepatocytes and for the heterogeneity of MAb OC.10 reactivity with hepatocellular carcinomas displaying oval cell characteristics (Hixson et al. 2000). It is also tempting to speculate that the punctate staining pattern of MAb OC.10 by immunofluorescence represents Hsc70 in lipid rafts, a targeting site previously established for Hsp70. This would provide a plausible mechanism for regulating membrane-mediated signaling events involved in the observed morphogenic stimulus provided by the injected MAb OC.10.

Although much work remains in elucidating the biological function of surface Hsc70 on oval and bile duct epithelial cells, it is possible to propose candidate functions based on known characteristics of the HSP70 proteins. For example, as a well-described chaperone, it is likely that Hsc70 transports proteins to the cell surface where it is also required for the proper folding of associated membrane proteins. It is also plausible that Hsc70 is required for recycling of surface proteins and targets them for proteosomal degradation. Hsc70 has also been implicated in various differentiation and developmental processes (de la Rosa et al. 1998; Elefant and Palter 1999; Vega-Nunez et al. 1999; Shrivastav et al. 2008), a finding consistent with its involvement as an active component of survival and proliferation pathways. As reported here, studies initially designed to label OC.10 expressing cells in situ for cell isolation schemes for lineage analysis demonstrated that binding of MAb OC.10 administered by intrasplenic injection promoted a change in bile duct morphology. Although the mechanism for this effect is at present unclear, the results suggest that binding of MAb OC.10 to surface Hsc70 is stimulating morphogenic signaling pathways. Supporting this assumption is an emerging regulatory role for Hsc70 in cell signaling, a role suggested by signaling complexes composed of Hsc70 and CXCR4 (Triantafilou et al. 2001), CXCR2 (Fan et al. 2002), and the G protein-coupled A1 adenosine receptor (Sarrio et al. 2000). Hsc70 binding has also been shown to be required for the activation of glucocorticoid receptors in the presence of steroids (Rajapandi et al. 2000).

It will be critical to determine if the reported morphogenic response to MAb OC.10 is epitope-dependent. If so, it is tempting to speculate that MAb OC.10 mimics a natural ligand that interacts with the cell surface-exposed epitope of Hsc70 to regulate signaling pathways critical to growth and survival. Dysregulation of surface Hsc70 or its natural ligands could therefore contribute to the development and progression of progenitor-derived liver cancers including hepatocellular carcinoma and cholangiocarcinomas originating from mature bile duct cells. Furthermore, the identification of normal signaling pathways in response to MAb OC.10 may contribute to our understanding and identification of parallel pathways shared by tumor cells responding to stress stimuli in the neoplastic microenvironment. In summary, establishing an identity between OC.10 and Hsc70 provides an essential step towards defining the significance of this surface marker in the differentiation and function of cholangiocytes and intrahepatic bile ducts during liver development and carcinogenesis.