Survival of docetaxel-resistant prostate cancer cells in vitro depends on phenotype alterations and continuity of drug exposure.

We evaluated in vitro the effect of paclitaxel and docetaxel on PC-3 and DU-145 prostate cancer cell lines to understand better the downstream events in drug-induced tumor cell death. Taxane treatments of DU-145 cells induced rapid cell death by apoptosis, but in PC-3 cells, treatments achieved growth arrest, followed by extensive karyokinesis resulting in multinucleation, giant-cell formation and delayed cell death. To determine if the giant multinucleated cells were able to produce proliferating and drug-resistant survivors, we first delineated the kinetics of drug activity and cytotoxic dose range. Analysis of both lines by colorimetric and cell viability assays demonstrated improved cytotoxicity of taxanes applied continuously. Selected doses and schedules of docetaxel were used to induce giant multinucleated cells that gave rise to docetaxel-resistant survivors, which remained sensitive to paclitaxel and other chemotherapeutics. Growth and morphology of the recovered clones was similar to parental cells. The resistant phenotype of these clones determined by immunofluorescence and immunoblot was associated with transient expression of the beta-tubulin i.v. isoform and was independent of P-glycoprotein, bcl-2 and bcl-xL. Resistant clones will be useful to model progression of resistance to taxanes and to identify unknown and clinically important molecular mechanisms of cell death and resistance.

Signal transduction by the CEACAM1 tumor suppressor. Phosphorylation of serine 503 is required for growth-inhibitory activity.

CEACAM1 is a cell-cell adhesion molecule that mediates homophilic cell adhesion. In addition, CEACAM1 was also shown to suppress the growth of prostate, breast, and colon tumors. Structural and functional analyses showed that the adhesion activity of CEACAM1 is mediated by its extracellular domain while its cytoplasmic domain is necessary and sufficient for growth-inhibitory activity. The signal pathways leading to CEACAM1-mediated growth suppression are not known. We studied the importance of phosphorylation of serine 503 in this growth-inhibitory signaling pathway. Full-length CEACAM1 was found to be phosphorylated in vivo in both tyrosine and serine residues. Mutation of tyrosine 488 to phenylalanine did not abolish the tumor-suppressive activity of CEACAM1, suggesting that phosphorylation at tyrosine 488 is not critical for CEACAM1's tumor-suppressive activity. Although expression of CEACAM1's cytoplasmic domain inhibited the growth of DU145 prostate cancer cells in vivo, mutation of serine 503 to alanine abolished the growth-inhibitory activity. In addition, the change of serine 503 to aspartic acid produced tumor-suppressive activity similar to that of the wild-type CEACAM1. These results suggested that phosphorylation at serine 503 is essential for CEACAM1's growth-inhibitory function in vivo.

Differentiation status of rat ductal cells and ethionine-induced hepatic carcinomas defined with surface-reactive monoclonal antibodies.

To gain further insight into the differentiation of oval cells and their role in carcinogenesis, we have generated cell surface reactive monoclonal antibodies (MAbs) by a Balb/c nude mouse (nu/nu) immunization protocol. Three MAbs designated OC.4, OC.5, and OC.10 were generated from a mouse immunized with CDE6, an oval cell line established from oval cells induced by feeding a choline-deficient diet containing 0.1% ethionine (CDE). These MAbs demonstrated stage-specific expression in fetal liver and displayed strong reactivity with oval and bile duct epithelial cells. In general, oval cells displayed a more mature phenotype than fetal ductal cells, suggesting the existence in adult liver of more primitive ductal progenitors. A fourth MAb recognized a cytoplasmic antigen (OC.6) expressed by mucus-secreting hepatic ducts induced by CDE diet. Immunocytochemical analysis indicated that OC.4, OC.5, and OC.10 were also expressed on CDE-induced, OV6+ hepatocellular carcinomas (HCC) but not on OV6+ HCC induced by the Solt/Farber protocol. In most cases, CDE-induced, OV6+ HCC expressed early ductal developmental markers such as OC.10 but lacked those expressed at later stages (OC.5, OC.4). These new MAb will be useful for characterizing HCC subpopulations with oval cell characteristics and for isolating biliary cells at antigenically defined stages during differentiation.

Liver regeneration in rats with retrorsine-induced hepatocellular injury proceeds through a novel cellular response.

The adult rodent liver contains at least two recognized populations of cells with stem-like properties that contribute to liver repair/regeneration under different pathophysiological circumstances: (i) unipotential committed progenitor cells (differentiated hepatocytes and biliary epithelial cells) and (ii) multipotential nonparenchymal progenitor cells (oval cells). In retrorsine-induced hepatocellular injury the capacity of fully differentiated rat hepatocytes to replicate is severely impaired and massive proliferation of oval cells does not occur. Nevertheless, retrorsine-exposed rats can replace their entire liver mass after 2/3 surgical partial hepatectomy through the emergence and expansion of a population of small hepatocyte-like progenitor cells that expresses phenotypic characteristics of fetal hepatoblasts, oval cells, and fully differentiated hepatocytes, but differ distinctly from each type of cell. The activation, proliferation, and complete regeneration of normal liver structure from small hepatocyte-like progenitor cells have not been recognized in other models of liver injury characterized by impaired hepatocyte replication. We suggest that the selective emergence and expansion of small hepatocyte-like progenitor cells observed in the retrorsine model reflect a novel mechanism of complete liver regeneration in the adult rat. Furthermore, we suggest that these cells may represent a novel progenitor cell population that (i) responds to liver deficit when the replication capacity of differentiated hepatocytes is impaired, (ii) expresses an extensive proliferative capacity, (iii) can give rise to large numbers of progeny hepatocytes, and (iv) can restore tissue mass.

The cytoplasmic domain of C-CAM1 tumor suppressor is necessary and sufficient for suppressing the tumorigenicity of prostate cancer cells.

We have previously shown that C-CAM1 cell adhesion molecule can suppress the growth of prostate cancer cells in vivo. In this study, we determined the minimal domain of C-CAM1 that is required for its tumor-suppressive activity. DU145 prostate cancer cells were infected with recombinant adenoviruses containing various C-CAM1 mutant genes, and the effects of the mutant C-CAM1 proteins on the growth of DU145 cells were assessed in a nude-mice xenograft model. Deletion of C-CAM1's cytoplasmic domain, which is not required for its adhesion activity, abolished the growth-suppressive activity, whereas deletion of the adhesion domain did not. This observation suggests that C-CAM1's extracellular domain may be not essential for its tumor suppressive activity. Indeed, we found that expression of the C-CAM1 cytoplasmic domain alone led to growth suppression of DU145 cells. These results suggest that the cytoplasmic domain of C-CAM1 is necessary and sufficient for its growth-suppressive function.

Expression and androgen regulation of C-CAM cell adhesion molecule isoforms in rat dorsal and ventral prostate.

C-CAM is an epithelial cell adhesion molecule with two major splice variants that differ in the length of the cytoplasmic domain. C-CAM1 (long (L)-form) strongly suppresses the tumorigenicity of human prostate carcinoma cells. In contrast, C-CAM2 (short (S)-form) does not exhibit tumor-suppressive activity. In the present study we have investigated the functional significance of L-form and S-form C-CAM in rat prostate by examining their expression and distribution in different prostate lobes and their response to androgen deprivation. RNase protection assays with a probe for both C-CAM isoforms detected high levels of C-CAM messages in the rat dorso-lateral prostate (DLP). L- and S-form proteins, localized by indirect immunofluorescence using isoform-specific antipeptide antibodies, were co-expressed on the apical surface of prostate epithelial cells in normal DLP. Androgen depletion did not significantly change the steady state levels of C-CAM message and protein expression in the DLP, although there was a change in the pattern of protein expression in these lobes. In contrast, C-CAM isoform messages and proteins were undetectable in normal ventral prostate (VP) but increased markedly in this lobe in response to castration, producing isoform ratios similar to those in DLP. These results demonstrate that coordinate expression of C-CAM isoforms is maintained in the VP following androgen depletion and suggest that androgen suppresses C-CAM expression in VP but not in DLP. These results suggest that balanced expression of L- and S-form C-CAM is important for normal prostate growth and differentiation.

C-CAM1 expression: differential effects on morphology, differentiation state and suppression of human PC-3 prostate carcinoma cells.

Studies in rat prostate and liver have suggested that C-CAM1 is involved in the formation and maintenance of histotypic associations in tissues and possibly tumors. Most recently, C-CAM1 has been shown to suppress tumorigenicity of prostate and colon carcinoma cells. However, the mechanisms whereby C-CAM1 suppresses growth and the relationship of this activity to its proposed role in histotypic interactions remain largely unknown. In the present study, we have analysed the growth, phenotypic, morphological and ultrastructural characteristics of four human PC-3 prostate carcinoma cell lines transduced with C-CAM1 retrovirus. We report that three of four lines regained their tumorigenic phenotype in vivo while maintaining high levels of C-CAM1 expression and a growth retarded phenotype in vitro. These findings suggested that high levels of C-CAM1 expression were negatively influencing recovery during reconstitution after freezing or during the latency period after subcutaneous injection and that loss of suppression resulted from changes in expression of other molecules required for full disclosure of C-CAM1 mediated growth inhibition. Results from Northern blot and immunofluorescence analyses of tumor nodules demonstrated that C-CAM1 decreased rather than enhanced phenotypic differentiation and induced ultrastructural and morphological changes that occurred independently of tumor suppression.

TA1 oncofetal rat liver cDNA and putative amino acid permease: temporal correlation with c-myc during acute CCl4 liver injury and variation of RNA levels in response to amino acids in hepatocyte cultures.

TA1 is a rat liver oncofetal cDNA and a member of an emerging family of evolutionarily conserved molecules with homology to amino acid transporters and permeases. The aim of these studies was to characterize the regulation and role of TA1 in acute rat liver injury by examining its relation to regeneration and metabolic stress. Following a single dose of CCl4, TA1 message was expressed 3-48 h. The major 3.3-kb TA1 transcript correlated temporally with c-myc expression. A novel 2.9-kb TA1 transcript was expressed more variably 24-48 h. TA1 protein was restricted to hepatocytes in G0 and G1 phases of the cell cycle. Relative to CCl4, a much smaller increase in TA1 was noted after partial hepatectomy and TA1 preceded the peak of c-myc expression. In vitro TA1 was not induced in hepatocytes by EGF or the acute-phase cytokines IL-6 and TNF-alpha, but was found to be modulated in response to amino acid availability. TA1 expression increased in media without arginine and glutamine and was repressed by total amino acid levels 5-fold over basal MEM. Together, these results contrast with the constitutive expression observed in transformed cells and suggest an adaptive role for TA1 during liver injury.

Association of an 80 kDa protein with C-CAM1 cytoplasmic domain correlates with C-CAM1-mediated growth inhibition.

Decreased expression of C-CAM, a member of the CEA family of immunoglobulin like cell adhesion molecules, occurs in carcinomas of the colon, liver and prostate. Down regulation of C-CAM during the early stages of carcinogenesis in rat liver and human prostate has also been reported. We have recently shown that restoration of the expression of the isoform with long cytoplasmic domain, C-CAM1, leads to suppression of the tumorigenicity of prostatic carcinoma cells in vivo and growth suppression in vitro. These observations suggest that C-CAM1 may play an important role in regulating cell growth in normal tissues. Previous studies have demonstrated that the function of many members of the Ig-supergene family is dependent on interactions with cytoplasmic proteins. In the present study, we have used a bifunctional cross-linker to identify cellular proteins that interact directly with C-CAM1. Immunoblot analysis of WGA bound membrane proteins crosslinked with DSS identified a 180 kDa complex composed of C-CAM and an 80 kDa protein designated CAP-80 (C-CAM Associated Protein). Immunoprecipitation with anti-C-CAM antibodies showed that CAP-80 was co-precipitated with C-CAM from detergent solubilized, WGA-purified proteins. To assess the specificity of CAP-80 binding, the ability of CAP-80 to form stable complexes with C-CAM1 mutants expressed in insect cells was tested. Deletion of the cytoplasmic domain of C-CAM1 abolished complex formation whereas deletion of the extracellular Ig domains had no effect. These results suggest that a CAP-80 homologue (ICAP-80) is present in insect cells and ICAP-80 interacts with the cytoplasmic domain of C-CAM1. Replacement of Tyr488, a residue in the cytoplasmic domain known to be phosphorylated in vivo, with Phe did not diminish the association between C-CAM1 and ICAP-80, suggesting that Tyr488 phosphorylation is not required for association. The ability of various C-CAM1 mutants to associate with ICAP-80 correlated with their growth inhibitory activities, suggesting that ICAP-80/CAP-80 may play an important role in C-CAM1-mediated growth inhibition.

Evaluation of the differentiation potential of WB-F344 rat liver epithelial stem-like cells in vivo. Differentiation to hepatocytes after transplantation into dipeptidylpeptidase-IV-deficient rat liver.

After intrahepatic transplantation into livers of adult syngeneic German-strain Fischer 344 rats that are deficient for the bile canalicular enzyme dipeptidyl peptidase IV (DPP-IV), cultured WB-F344 rat liver epithelial cells (without exogenous marker genes) integrate into hepatic plates and differentiate into hepatocyte-like cells that are morphologically and functionally indistinguishable from mature hepatocytes. In this model system, the differentiated progeny of transplanted WB-F344 cells are identified among the DPP-IV-negative host hepatocytes by their expression of bile canalicular DPP-IV enzyme activity. DPP-IV-positive hepatocyte-like cells also expressed other markers of hepatocytic differentiation, including albumin, transferrin, and alpha-1-antitrypsin, suggesting that the progeny of transplanted WB-F344 cells express a complete hepatocyte differentiation program. These results complement our previous studies indicating WB-F344 cells can serve as stem-like precursor cells for differentiated hepatocytes and strengthen the suggestion that WB-F344 rat liver epithelial cells represent the cultured counterpart of liver stem-like hepatocyte progenitor cells present in the normal adult rat liver.

Differentiation of pancreatic epithelial progenitor cells into hepatocytes following transplantation into rat liver.

The ability to identify, isolate, and transplant progenitor cells from solid tissues would greatly facilitate the treatment of diseases currently requiring whole organ transplantation. In this study, cell fractions enriched in candidate epithelial progenitor cells from the rat pancreas were isolated and transplanted into the liver of an inbred strain of Fischer rats. Using a dipeptidyl dipeptidase IV genetic marker system to follow the fate of transplanted cells in conjunction with albumin gene expression, we provide conclusive evidence that, after transplantation to the liver, epithelial progenitor cells from the pancreas differentiate into hepatocytes, express liver-specific proteins, and become fully integrated into the liver parenchymal structure. These studies demonstrate the presence of multipotent progenitor cells in the adult pancreas and establish a role for the liver microenvironment in the terminal differentiation of epithelial cells of foregut origin. They further suggest that such progenitor cells might be useful in studies of organ repopulation following acute or chronic liver injury.

Antigenic phenotypes common to rat oval cells, primary hepatocellular carcinomas and developing bile ducts.

The shared expression of monoclonal antibody-defined antigens by oval cells and by bile ducts, neoplastic nodules and primary hepatocellular carcinomas (PHC) has provided support for the ability of oval cells to undergo differentiation along ductular or hepatocyte lineages and/or to progress to hepatocellular carcinoma. With the aim of obtaining additional insight into this process, we have combined serial section and double labeling immunofluorescence analysis to determine if phenotypes expressed in vitro by four rat oval cell lines and the H5D.61 hepatocellular carcinoma cell line and in situ by ethionine-induced primary hepatocellular carcinomas reproduce antigenic patterns occurring during normal liver development. Analysis using monoclonal antibodies specific for the oval cell antigens OV6 and OC.2 and hepatocyte markers HBD.1 and H.4 defined subpopulations in four oval cell lines and neoplastic hepatocytes in PHC and H5D.61 with OC.2-/OV6+ and OC.2+/OV6+ phenotypes. Cells with an OC2+/OV6- phenotype were rarely observed in cell lines or primary tumors. In contrast, areas composed of OV6+/H.4+ cells were frequently found in PHC. Examination of fetal and neonatal rat livers demonstrated the stage-specific appearance of three of these phenotypes during liver development. The OC.2+/OV6- phenotype appeared transiently prior to embryonic day (ED) 18 in a subpopulation of HBD.1+ hepatoblasts. OV6 expression was first detected at ED18 on developing bile ducts that were negative for OC.2. These newly formed ducts rapidly acquired OC.2, starting with ducts in the hilar region and spreading outward towards the periphery. This OC.2 expression gradient persisted in the newborn rat liver but became more skewed towards doubly positive cells, with OC.2-/OV6+ cells being found primarily in the periphery. Hepatocytes expressing both OV6 and H.4 were not observed in fetal liver but appeared in neonatal liver in close proximity to OV6+ interlobular ducts. From these findings, it was concluded that oval cells and PHC display phenotypes representing normal stages in liver development, suggesting that oval cells and cells within ethionine-induced PHC are capable of initiating but are unable to complete pathways of hepatocytic or biliary differentiation.

Application of immunomagnetic beads in combination with RT-PCR for the detection of circulating prostate cancer cells.

Recently published protocols using Reverse Transcriptase Polymerase Chain reaction (RT-PCR) for prostate specific antigen (PSA) provide a sensitive means for detecting circulating prostate cancer cells. Attempts to use these assays for staging of prostate cancer have produced conflicting results. As a first step towards rectifying these discrepancies, a modified immunobead-RT-PCR assay capable of detecting as few as 10 prostate cancer cells in 8cc of blood was developed. This 10 fold increase in sensitivity was achieved in part by introducing two target cell enrichment steps. As a model system to assess sensitivity of the modified assay, template RNA was extracted from PSA positive human carcinoma cells suspended in human blood and isolated with immunomagnetic beads following incubation with an epithelium specific antibody. After 45 cycles of PCR, product from as few as 10 target cells could be readily detected when displayed on a 2% agarose gel stained with SYBR Green fluorescent dye. The identity of amplified DNA fragments was confirmed by Southern blot hybridization. When applied to blood samples from patients with proven metastatic disease, the immuno-bead RT-PCR assay was successful in detecting circulating PSA positive epithelial cells, suggesting this assay may be useful for assessment of disease progression or recurrence.

The long and the short isoform of cell-CAM 105 show variant-specific modifications in adult rat organs.

The cell adhesion molecule cell-CAM 105 is a member of the carcinoembryonic antigen (CEA)-gene family and expressed on the cell surface at least as a long (L-) and a short (S-) isoform. We have developed anti-peptide antisera to investigate the expression patterns of both cell-CAM 105 variants in rat organs. In immunohistochemistry the S- and the L-isoform were co-localized in every tissue investigated. In contrast, the quantification of both isoforms in rat organs by a sandwich-ELISA revealed several differences in their relative expression ratio if compared to the liver value. In submandibular gland and seminal vesicle a significant increase of the S-isoform was found, whereas in kidney the L-isoform was overexpressed. In immunoblot analysis of kidney and seminal vesicle the two isoforms did not appear with the Mr difference which is deduced from their primary structures, suggesting that the two cell-CAM 105 variants are subjected in these tissues to different post translational modifications.

TuAg.1 is the liver isoform of the rat colon tumor-associated antigen pE4 and a member of the immunoglobulin-like supergene family.

TuAg.1 is a tumor-associated membrane glycoprotein first identified in rat hepatocellular carcinoma by monoclonal antibodies (mAbs) 324.5 and 324.9. This oncofetal antigen is also expressed by hepatocytes in cell culture but not normal adult hepatocytes in vivo. Affinity chromatography and preparative continuous elution slab-gel electrophoresis were used to separate TuAg.1 from co-purified actin and immunoglobulin. TuAg.1 was recovered as a series of bands Mr 82,000-90,000, which were pooled and subjected to CNBr digestion for primary amino acid sequence analysis. Computer database analysis of TuAg.1 peptide sequence revealed homology to the rat colon carcinoma-associated antigen pE4, a member of the immunoglobulin gene superfamily. Oligonucleotide primers derived from sequences shared by TuAg.1 and pE4 were used in reverse transcription-PCR to amplify tumor-specific products corresponding to TuAg.1 cDNA. Northern blot analysis with one of these products confirmed the oncofetal expression of transcripts related to TuAg.1/pE4 and indicated an RNA species of different size expressed only in normal liver. Identity between TuAg.1 and pE4 was further confirmed by immunochemical analysis with mAb 324.5 and mAb E4. Both antibodies were reactive with the same protein on transplantable hepatocellular carcinoma AS30D but recognized different epitopes. The reactivity of human tumor cells with mAb 324.5 and 324.9 indicates the presence of a related TuAg.1 molecule expressed in human neoplasia as well.

Identification of a new isoform of cell-cell adhesion molecule 105 (C-CAM), C-CAM4: a secretory protein with only one Ig domain.

A series of Southern blot hybridization experiments using probes derived from different regions of the rat liver cell-cell adhesion molecule 105 (C-CAM) cDNA revealed the presence of a 9.6 kb EcoRI genomic fragment that seemed to encode a unique C-CAM isoform. An RNase protection study showed that this c-CAM transcript was expressed in placenta, spleen, lung and large intestine. In contrast, the other C-CAM isoforms, C-CAM1 and C-CAM2, are expressed in liver and small intestine. This result also suggests that the new isoform, which we named C-CAM4, was indeed encoded by a new C-CAM gene. A rat placenta cDNA library was then screened and the full-length cDNA coding for C-CAM4 was isolated. The deduced protein contained 142 amino acids and had a calculated molecular mass of 15 kDa. C-CAM4 was composed of a leader sequence and the first V-like Ig domain typical of C-CAM-family proteins. However, C-CAM4 lacked the C-like Ig domains, the transmembrane domain, and the cytoplasmic domain found in other C-CAM isoforms. Thus, C-CAM4 is different from the other known C-CAMs in that it is a secreted protein. We have previously shown that the first Ig domain of C-CAM1 is crucial for its adhesion function. The V-like Ig domain of C-CAM4 had 92% and 89% sequence identity with the corresponding regions of C-CAM1 and C-cam2 respectively. Together these results suggest that C-CAM4 may play a role in regulating the function of other C-CAM family proteins.

Interleukin-6 production by rat hepatocellular carcinoma cells is associated with metastatic potential but not with tumorigenicity.

BACKGROUND: The phenotypic characteristics that allow some tumor cells to metastasize have not been fully identified. The production and/or response of tumor cells to various growth factors have been shown to distinguish cells of differing metastatic potentials. OBJECTIVES: To determine (1) whether rat hepatocellular carcinoma cell lines produce interleukin-6 (IL-6) and (2) whether production of IL-6 correlates with either metastatic potential or tumorigenicity. METHODS: The clonal cell lines 1682.C.2.9.L0 (poorly metastatic) and 1682.C.2.9.L10 (highly metastatic) were selected from a parental hepatocellular carcinoma induced in ACI rats by feeding an ethionine-containing diet and adapted to growth in vitro. RESULTS: Both cell lines resulted in primary tumors with equal frequency and developed a 40-mm nodule in a similar period time, when an inoculum of 5 X 10(6) cells was injected subcutaneously; however, only L10 cells metastasized to the lung. These cell lines did not demonstrate differential expression of several antigens noted to correlate with metastatic potential, including CD44 variant glycoprotein, p53, transferrin receptor, and E-cadherin. In contrast, L0 cells produced less than 10 U of IL-6 per milliliter in culture (as determined by bioassay using 7TD1 cells), whereas L10 cells released more than 95 U of this cytokine per milliliter under identical culture conditions (P<.01, Student's t test). In addition, serum concentrations of IL-6 were elevated in animals bearing L10-induced primary tumors but not in those with L0-induced tumors of comparable mass. Exogenous addition of IL-6 to both tumor cell lines had no effect on the rate of growth in vitro, supporting the similar the tumorigenic potentials observed in vivo. CONCLUSION: Excess IL-6 production appears to identify cells with metastatic potential and does not appear to be essential to the establishment of a primary tumor.

Cloning of cDNAs from a mammalian expression library by a direct selection-amplification method.

A sensitive method was devised for cloning cDNAs from a mammalian expression library based on single-cell detection and selection of transfected cells. The method is applicable for cloning cell-surface or cytoplasmic proteins for which a detection assay, such as immunofluorescence or immunohistochemical reactivity, exists. The widely used eukaryotic expression vector cdm8 is utilized, and the method is demonstrated using the gene for C-CAM, a liver glycoprotein adhesion molecule. After transfection, cells are plated out on a collagen gel substrate to allow retrieval of selected cells. Simultaneous fixation and permeabilization with acetone permits immunological and histochemical detection of cell-surface and cytoplasmic proteins without loss of plasmid vector. Inserts of interest are recovered by PCR with vector primers.

Delineation of antigenic pathways of ethionine-induced liver cancer in the rat.

Although oval cell proliferation is observed prior to the appearance of hepatic nodules and hepatocellular carcinomas during ethionine-induced liver carcinogenesis in the rat, the role of these presumptive hepatic stem cells during the neoplastic process remains controversial. In order to investigate this question, we have used a panel of monoclonal antibodies against antigens associated with normal hepatocytes, oval cells and transplantable hepatocellular carcinomas (THC) to trace antigenic pathways leading to liver cancer. Male ACI rats were fed a choline-deficient diet containing 0.1% DL-ethionine for 4, 16 or 30 weeks. Immunocytochemical analysis of frozen liver sections revealed a subpopulation of hepatic nodules (7/52), carcinomas (8/15) and lung metastases (3/5) containing populations of cells expressing both oval cell, hepatocyte and neoplastic markers. Carcinomas expressing oval cell markers often appeared as a mosaic of well-defined patches composed of phenotypically distinct cells. Many of the phenotypes expressed closely mimicked patterns of expression observed in fetal and neonatal liver. THC derived from primary tumors positive for oval cell antigens (4/5) continued to express these markers. Northern blot analysis and immunocytochemical analysis revealed that 4/5 primary hepatocellular carcinomas (PHC) and THC expressed alpha-fetoprotein (AFP) and albumin transcripts and contained subpopulations expressing AFP together with hepatocyte and oval antigens. In contrast, a well-differentiated PHC and its corresponding THC lacked AFP mRNA and oval cell antigens but showed strong expression of both hepatocyte and neoplastic markers. These results demonstrate that a subpopulation of malignant and metastatic hepatocellular carcinomas are comprised of cells expressing multiple oval cell markers in this model system.

Structure and function of C-CAM1: effects of the cytoplasmic domain on cell aggregation.

C-CAMs are epithelial cell-adhesion molecules of the immunoglobulin supergene family with sequences highly homologous to carcinoembryonic antigen (CEA). C-CAMs and their human homologues, biliary glycoproteins, are unique among the CEA-family proteins in that they have cytoplasmic domains. Furthermore, alternative splicing generates C-CAM isoforms with different cytoplasmic domains, suggesting that the cytoplasmic domains of C-CAM may play important roles in regulating the function or functions of C-CAM. By using both sense and antisense approaches, we have shown that C-CAM1 is a tumour suppressor in prostate carcinogenesis. This observation raises the possibility that the cytoplasmic domain of C-CAM1 may be involved in signal transduction or interaction with cytoskeletal elements to elicit the tumour suppressor function. The cytoplasmic domain of C-CAM1 contains several potential phosphorylation sites, including putative consensus sequences for cyclic AMP-dependent kinase and tyrosine kinase. One of the potential tyrosine phosphorylation sites is located within the antigen-receptor homology (ARH) domain. The ARH domain of the membrane-bound IgM molecule is necessary for signal transduction in B-cells. These structural features suggest that the cytoplasmic domain of C-CAM1 may be important for signal transduction. To test this possibility, we generated several site-directed C-CAM1 mutants and tested their ability to support adhesion and their abilities to be phosphorylated in vivo. Results from these studies revealed that Tyr-488 is phosphorylated in vivo. However, replacing this tyrosine with phenylalanine did not significantly compromise its adhesion function. Similarly, Ser and Thr residues are phosphorylated in vivo, but deletion of the potential cyclic AMP-dependent kinase site did not significantly reduce the adhesion function. These results suggest that the kinase phosphorylation sites in the cytoplasmic domain of C-CAM1 are not required for the adhesion function. However, these phosphorylation sites are probably involved in the regulation of C-CAM-mediated signal transduction. Thus, there are probably distinct structural requirements for the adhesion and the signal transduction functions of C-CAM. Incidentally, a C-CAM1 deletion mutant containing a 10-amino-acid cytoplasmic domain was able to support adhesion activity. This is in contrast to our previous finding that a C-CAM isoform, C-CAM3, with a 6-amino-acid cytoplasmic domain could not support cell adhesion. This result indicates that the extra four amino acids, which are absent in C-CAM3 and contain a potential Ser/Thr phosphorylation site, are important for the adhesion function.