Cancer Research AMC Genetics and Biology of Brain Cancer
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Laurie, N. A.
Right arrow Articles by Hixson, D. C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Laurie, N. A.
Right arrow Articles by Hixson, D. C.
[Cancer Research 65, 11010-11017, December 1, 2005]
© 2005 American Association for Cancer Research

Cell and Tumor Biology

Carcinoembryonic Antigen–Related Cell Adhesion Molecule 1a-4L Suppression of Rat Hepatocellular Carcinomas

Nikia A. Laurie, Meghan M. Comegys, Marie P. Carreiro, Jeanne F. Brown, Donna L. Flanagan, Kate E. Brilliant and Douglas C. Hixson

Division of Hematology and Oncology, Department of Medicine, Rhode Island Hospital/Brown University Medical School, Providence, Rhode Island

Requests for reprints: Douglas C. Hixson, Department of Medicine, Rhode Island Hospital/Brown University Medical School, 593 Eddy Street, Providence, RI 02903. Phone: 401-444-8058; Fax: 401-444-8141; E-mail: dhixson{at}lifespan.org.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Carcinoembryonic antigen (CEA)–related cell adhesion molecule 1 (CEACAM1) is a member of the CEA family of immunoglobulin-like adhesion molecules with two major splice variants, CEACAM1a-4L and CEACAM1b-4S, differing in the length of their COOH-terminal cytoplasmic tail. Both forms are down-regulated in prostate and liver carcinomas relative to normal tissues. We have previously shown in a nude mouse xenograft model that restoration of CEACAM1a-4L expression in human prostate carcinoma cells (PC-3) suppresses tumorigenicity, an effect observed with carcinomas from several other tissues but never established for hepatocellular carcinomas. In this report, we have examined the effect of CEACAM1a-4L on tumorigenicity of 1682A, a rat hepatocellular carcinoma that grows on the omentum when injected into the peritoneal cavity. Results show that restoration of CEACAM1a-4L expression at levels 13- and 0.45-fold compared with negative controls or normal hepatocytes, respectively, completely suppressed the formation of 1682A tumor nodules on the omentum at 3 weeks after injection. In contrast, 1682A cells infected with CEACAM1b-4S or an empty retroviral vector formed multiple clusters of tumor nodules. Although tumor nodules of 1682A cells positive and negative for CEACAM1a-4L did not display significant differences in histologic organization, aggregates formed in vitro by 1682A-L were smaller in size and displayed enlarged intercellular spaces relative to their 1682A-V counterparts. Restoration of CEACAM1a-4L expression did not elevate levels of apoptosis but seemed to cause an increase in the length of G1. This is the first demonstration of CEACAM1a-4L–induced tumor suppression in liver carcinomas using a quantifiable i.p. syngeneic transplantation model.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Carcinoembryonic antigen (CEA)–related cell adhesion molecule 1 (CEACAM1) is a member of the CEA family of immunoglobulin-like adhesion molecules. It is the rat homologue of human biliary glycoprotein I (1, 2). In the rat, it is expressed as both a long (CEACAM1a-4L) and short (CEACAM1b-4S) isoform, which differ in size and charge (3) and in the length of their cytoplasmic domains (71 versus 10 amino acids, respectively). There are two alleles in the rat (Ceacam1a and Ceacam1b) distinguished by differences in the sequence of their NH2-terminal IgV-like domain (4).

In the rat liver, the ratio of CEACAM1a-4L and CEACAM1b-4S, the two major splice variants, varies between different hepatocellular carcinomas (5) and changes during liver development (6, 7) and regeneration (5). During murine or human carcinogenesis in the liver, prostate, colon, breast, and bladder, CEACAM1a-4L and CEACAM1b-4S were shown to undergo a dramatic down-regulation (3, 5, 8–12). This suggested that CEACAM1 might be a tumor suppressor, a possibility supported by the dramatic suppression of tumorigenicity resulting from restoration of CEACAM1a-4L expression (13–21). However, unlike other known tumor suppressors, CEACAM1 down-regulation was not the result of deletion or mutation, but instead seemed to be caused by a change in DNA methylation (22, 23).

Surprisingly, there have been no reports directly examining the ability of CEACAM1a-4L to suppress the growth of hepatocellular carcinoma in vivo. To address this issue, we have examined the proliferative, phenotypic, and morphologic consequences of restoring CEACAM1a-4L or CEACAM1b-4S expression to hepatocellular carcinomas that express very low endogenous levels of CEACAM1. Our results show that stable transduction with a CEACAM1a-4L but not with a CEACAM1b-4S or empty retroviral vector decreased aggregate size and altered morphology on soft agar in vitro and caused almost complete cessation of growth of an hepatocellular carcinoma that grows exclusively on the omentum in vivo.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Origin of 1682A hepatocellular carcinoma cell line. The parental 1682A cell line was established from a primary hepatocellular carcinoma isolated from a male ACI rat treated with ethionine in a choline-deficient diet as previously described (8). In vitro, cells were maintained in Waymouth's medium supplemented with 10% fetal bovine serum (FBS) as described previously (8).

Production of 1682A cell lines expressing CEACAM1a-4L, CEACAM1b-4S, and empty vector. Plasmid construction and production of retrovirus were done as previously described (16). CEACAM1-negative 1682A parental cells (8) were infected with amphotropic LNCX-7 retroviral vectors without a cDNA insert (vector) or with cDNAs for CEACAM1a-4L or CEACAM1b-4S. After selection in G418, drug-resistant colonies were collected and expanded as previously described (16). 1682A rat hepatocellular carcinoma clones infected with CEACAM1a-4L (1682A-L), CEACAM1b-4S (1682A-S), or empty vector (1682A-V) were screened for expression using isoform-specific antibodies.

Preparation and transplantation of hepatocellular carcinoma cells. 1682A-L, 1682A-S, and 1682A-V cells were grown in vitro in Waymouth's medium containing 10% FBS. At 80% confluency, cells were trypsinized and washed in HBSS. Male ACI rats 3 to 4 weeks of age were injected i.p. with 1 mL of cell suspension containing 5 x 106 to 3 x 107 cells/mL in HBSS. After 3 to 6 weeks, host rats were euthanized by CO2 asphyxiation and tumors present on the omentum were excised, weighed, and then frozen with or without optimum cutting temperature compound (OCT) in liquid nitrogen and stored at –80°C.

Indirect immunofluorescence. Frozen sections (5 µm) cut from OCT-embedded frozen tissues were fixed with ice-cold 100% acetone for 10 minutes, air dried, and subsequently stained by a previously described indirect immunofluorescence protocol (16). The production and characterization of polyclonal antibody 669 that recognizes both the S and L isoforms of CEACAM1 has been reported (24). Alexa Fluor 593 goat anti-rabbit (Molecular Probes, Inc., Eugene, OR) was used as secondary antibody. Normal rabbit serum (NRS) was used as a negative control for polyclonal antibody 669. Frozen sections were photographed using a Nikon Microphot FX microscope fitted with an epifluorescence condenser and Spot digital color camera.

Fluorescence-activated cell sorting analysis. Relative levels of CEACAM1 expression were determined by flow cytometry [fluorescence-activated cell sorting (FACS) analysis] using a previously described protocol (16). Ten million normal hepatocytes isolated by collagenase perfusion as previously described (3) or cells from the 1682A-L, 1682A-S, and 1682A-V cell lines were labeled in suspension with a 1:200 dilution of primary polyclonal antibody 669 or NRS in culture medium. After two washes with cold medium, cells were incubated with goat anti-rabbit FITC-conjugated secondary antibody (Cappel/Organon Teknika, Durham, NC), followed by two washes with cold PBS containing 0.1% bovine serum albumin. Cells fixed with 2% paraformaldehyde were resuspended at 1 x 107 cells/mL in PBS. Forward and side scatter, using a scale of 256 arbitrary units, were used to estimate size and granularity. Fluorescence was measured on a logarithmic scale using a band pass filter for FITC. Expression of CEACAM1 by normal hepatocytes or by 1682A-L cells was expressed as the ratio of median fluorescence for cell treated with anti-CEACAM1 polyclonal antibody 669 over the median fluorescence of the same cells treated with NRS.

Immunoblot analysis. Aliquots of Triton X-100 detergent extracts (6 µg total protein determined by Bio-Rad protein assay, Bio-Rad, Hercules, CA), prepared as previously described from normal hepatocytes isolated by collagenase perfusion or from 1682A-L or 1682A-V cell lines, were resolved by SDS-PAGE and transferred to nitrocellulose. After blocking in nonfat dry milk, CEACAM1a-4L was detected by a previously described indirect immunoperoxidase protocol using monoclonal antibody (mAb) 9.2 specific for CEACAM1a-4L (25) and chemiluminescence visualization (26). The area and intensity of the mAb 9.2 reactive band corresponding to the 4L isoform were determined by analysis of Versadoc digital reproductions of immunoblots using Quantity I software (Bio-Rad). The amount of CEACAM1a-4L was calculated from the product of the band intensity and its area minus the product of the background intensity multiplied times the area of the band. To adjust band densities to cell equivalent values, densities were divided by the number of cells in 6 µg of extract calculated from total protein in Triton X-100 extract from 1 x 106 cells (760, 500, and 650 µg for hepatocytes, 1682A-L, and 1682A-V cells, respectively).

Gene array expression analysis. Using the protocol supplied with the Super Array Oligo (dT)14 Array Grade mRNA purification kit (SuperArray Bioscience Corp., Frederick, MD), mRNA was prepared from 1682A-L and 1682A-V cells harvested at 24 or 68 hours after plating on soft agar. cRNA was prepared with the TrueLabeling-AMP Linear RNA Amplification kit (SuperArray Bioscience) according to the suggested protocol. To prepare cRNA, a cRNA amplification master mix containing RNA amplification buffer, biotin-16-UTP (Roche Diagnostics, Penzberg, Germany), and amplification enzyme mix was prepared according to the protocol of the manufacturer and added to the cRNA synthesis master mix. After incubation for 4 hours at 37°C, cRNA was purified on an RNeasy column as described in the instructions of the manufacturer. OligoGEArray Mouse Cell cycle microarrays from SuperArray Bioscience with 112 oligonucleotides representing genes associated with cell cycle regulation were hybridized overnight at 60°C with purified cRNA according to the protocol of the manufacturer. Following incubation with alkaline phosphatase–conjugated streptavidin for 10 minutes, bound target cRNAs were detected by incubation with CDP-Star chemiluminescent substrate using the protocol suggested by the manufacturer. Exposing microarrays to X-ray film captured the chemiluminescence signal. Scanned images converted to gray scale TIFF files were normalized to the signal from the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) control spot using the background subtraction and data normalization options from the GEArray Expression Analysis software provided on-line by SuperArray Bioscience. Signal intensity for each gene was expressed as a percentage of the GAPDH control.

Apoptosis assays. Flow cytometry was done using a Guava PC flow cytometer equipped with Guava Express Software (Guava Technologies, Inc., Hayward, CA). Cells labeled in suspension by indirect immunofluorescence using phycoerythrin-conjugated antimouse IgG were analyzed according to the protocol of the manufacturer. Apoptotic cells were quantitated with the Guava Multicaspase detection assay kit and then incubated in suspension for 1 hour at 37°C with the cell-permeable inhibitor sulforhodamine-valyl-alanyl-aspartyl-fluoromethylketone (SR-VAD-FMK) that becomes fluorescent when bound to activated caspases and with 7-aminoactinomycin D (7-AAD), a fluorescent dye that is excluded from intact viable cells. Percentages of live, early apoptotic, late apoptotic, and dead cells at 24 hours after subculture were based on the average of two to four separate assays using different cell preparations for each assay.

Apoptosis was also assessed by immunoblot analysis as previously described (26) using an antibody specific for an 85 kDa caspase-3 cleavage fragment of poly(ADP-ribose) polymerase (PARP), a 116 kDa enzyme involved in DNA repair (27, 28). Western blots were prepared with 6 or 12 µg of Triton X-100 lysates from rat liver, 1682A-L cells, 1682A-V cells, or lysates provided by Zymed Laboratories, Inc. (San Francisco, CA) from untreated or etoposide-treated (apoptotic) HL-60 cells. After blocking in nonfat dry milk, blots were incubated with PBS containing a mixture of ascites containing mouse mAb anti-PARP (ascites from Zymed Laboratories; 1:1,000 dilution) and affinity-purified anti-Hsc70 mAbs (1:1,000 dilution, Stressgen, Victoria BC, Canada) followed by horseradish peroxidase–conjugated goat anti-mouse IgG. Bound anti-PARP antibodies were subsequently detected by chemiluminescence as described previously (26).

Cell cycle and proliferation analysis by flow cytometry. 1682A-L, 1682A-V, and 1682A-S cells (5 x 105) were suspended in 5 mL serum-free Waymouth's medium for 48 hours. The cells were then plated into 100 mm Petri dishes coated with 10 mL of a 1:1 mixture of Waymouth's medium containing 10% FBS and 2% agarose. At 14 and 24 hours after being returned to medium containing 10% FBS, cells were trypsinized, washed once in medium with 10% FBS, and washed once in cold PBS (pH 7.4). For nuclear staining, 1 mL of PBS containing 2 x 106 cells was mixed with an equal volume of propidium iodide solution (0.5 mg/mL propidium iodide, 0.1% NP40, and 0.1% sodium citrate) and incubated at room temperature for 30 minutes in the dark. Cells were analyzed for DNA content using a Dickinson FACSVantage SE flow cytometer. Gating was adjusted to delineate fluorescence levels corresponding to subdiploid (apoptotic), 2N (G0-G1), 2N-4N (S), and 4N (G2-M) DNA content and the percentage of cells in each phase of the cell cycle was determined using Verity Mod Fit software. The results were presented as graphs showing the percentage of cells in each phase of the cell cycle.

Histologic and morphometric analysis. For histologic assessment, tumor nodules were fixed in 2% paraformaldehyde and 0.1% gluteraldehyde in PBS, dehydrated, and embedded in Spurrs medium as previously described (16, 25). Plastic sections, 1 µm in thickness and stained with methylene blue, were examined by bright field light microscopy to assess the histotypic organization of the tumor tissue. Aggregates of 1682A-L and 1682A-V cells growing on a soft agar substratum were harvested by centrifugation 48 hours after plating in Waymouth's medium with 10% FBS. Cell pellets were fixed and embedded in Spurrs medium by the same protocol used for tumor nodules. The cross-sectional area of colonies in sections stained with toluidine blue was determined from analysis of digital micrographs using Image Pro Plus software.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
1682A-L cells express CEACAM1a-4L at levels 0.45 and 13 times those on normal hepatocytes and 1682A-V cells. Expression of CEACAM1a-4L and CEACAM1b-4S by 1682A-S and 1682A-L cell lines (Fig. 1A and C) and by normal hepatocytes (not shown) was confirmed by indirect immunofluorescence assays using anti-CEACAM1 polyclonal antibody 669 or by flow cytometric analysis with polyclonal antibody 669 (Fig. 1E). Although 1682A-V (Fig. 1E, triangle line) was essentially negative (3-fold difference between polyclonal antibody 669 and NRS), both 1682A-L (Fig. 1A and E, dotted line) and isolated hepatocytes (Fig. 1E, solid line) displayed strong membrane fluorescence that was 43- and 20-fold higher, respectively, than the background fluorescence displayed by NRS negative controls (Fig. 1E, dashed line). When relative levels of expression of the L-form band in immunoblots of 1682A-L, 1682A-V, and hepatocyte extracts were determined by digital densitometry using Quant 1 software (Bio-Rad; Fig. 1F), the quantity of L-form expressed by 1682A-L cells was 0.45- and 13-fold of that expressed by isolated hepatocytes and 1682A-V cells, respectively.



View larger version (77K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Expression of CEACAM1 isoforms by hepatocellular carcinoma 1682A transduced with CEACAM1a-4L and CEACAM1b-4S retroviruses. A and C, 1682A-S and 1682A-L cells, respectively, labeled by indirect immunofluorescence with polyclonal antibody 669. Phase-contrast images of identical fields to those in (A and C), respectively, are shown in (B and D). Bar, 50 µm (B and D). E, results from flow cytometric analysis of normal hepatocytes and1682A-L and 1682A-V cells labeled by indirect immunofluorescence with polyclonal antibody 669 or NRS. Plots of fluorescence intensity (X axis, arbitrary units) versus cell number (Y axis) are shown for hepatocytes (solid line), 1682A-L (dotted line), and 1682A-V (triangle line) labeled with polyclonal antibody 669 and all three with NRS (dashed line). The percentage of positive cells with polyclonal antibody 669 and NRS were 87% and 0% for hepatocytes, 12% and 4% for 1682A-V, and 83% and 2% for 1682A-L. The median fluorescence intensities following labeling with polyclonal antibody 669 and NRS were 87 and 2 for normal hepatocytes, 5 and 3 for 1682A-V, and 59 and 3 for 1682A-L. F, Western blot from an SDS-PAGE gel loaded with 6 µg of detergent extract from 1682A-L (lane 1), 1682A-V (lane 2), 1682A-L (lane 3), and normal liver (lane 4). Blots were labeled by an indirect immunoperoxidase protocol with mAb 9.2 and visualized by chemiluminescence detected with X-ray film. The exposure time was increased to show the faint reactivity in 1682A-V extracts (lanes 1 and 2). The exposure times for the lanes 1 to 3 were identical. L and S, long and short isoforms of CEACAM1, respectively.

 
CEACAM1a-4L strongly suppresses the tumorigenicity of 1682A tumor cells. When injected into the peritoneal cavity, 1682A-V and 1682A-S cell lines formed nodules by 3 weeks that appeared as grape-like clusters growing almost exclusively on the omentum (Fig. 2A and B, respectively). As shown in Fig. 3 (animals 1-11), injection of 4 x 106 to 5 x 106 cells produced a total tumor burden for 1682A-V cells ranging from 0 to 25 g. In the experimental group composed of animals 26 to 28, the total mass of nodules formed by 1682-V cells increased dramatically (21-26 g) when tumors were allowed to grow for 30 days. In contrast, no tumor nodules were found at 3 weeks in rats injected with 1682A-L cells under the same conditions (Figs 2C and 3, animals 12-19) or even at twice the dose (1 x 107 cells; Fig. 3, animals 23-25). When the dose of cells (3 x 107) and length of time following injection (38 days) were increased, small tumor (1-3 mm in diameter) became apparent in animals injected with 1682A-L. However, the maximum tumor burden with 1682A-L cells at 38 days was still 3- to 10-fold less than 1682A-V (Fig. 3, animals 29-31). Tumorigenicity assays using higher-passage cells produced the same results, indicating that the growth-suppressed phenotype is more stable in 1682A than previously observed in PC3 cells that regained tumorigenicity after repeated passage in vitro (16). Similar results were obtained with two other rat hepatocellular carcinoma cell lines, H5D and 253T (data not shown). Following transduction by a CEACAM1a-4L retrovirus, these hepatocellular carcinoma cells lost the ability to form tumors on the omentum (data not shown), indicating that the ability of CEACAM1a-4L to suppress growth in vivo is not restricted to 1682A-L cells.



View larger version (116K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Tumor nodules formed in the peritoneal cavity of rats injected with hepatocellular carcinoma cell lines. After 3 weeks, rats injected with 1682A-V (A) or 1682A-S (B) cells had formed nodules in large grape-like clusters on the omentum. In contrast, the omentum in rats injected with 1682A-L cells (C) was free of tumor nodules. D and E, bright field micrographs of 1 µm plastic sections of 1682A-V (F, animals 20-22) and 1682A-L (F, animals 29-31) tumor nodules, at 21 and 38 days after injection, respectively. Tumor nodules from both cell lines were composed of undifferentiated sheets of cells. Bar, 50 µm (D and E).

 


View larger version (25K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Quantitative assessment of the tumorigenicity of 1682A-V and 1682A-L cells at 21 to 39 days after injection. Y axis, weight of the omentum (g). Rats 1 to 11, 20 to 22, and 26 to 28 were injected with 1682A-V cells. Rats 12 to 19, 23 to 25, and 29 to 31 were injected with 1682A-L cells. Rats 1 to 19, 20 to 28, and 29 to 38 are from three separate experimental groups.

 
Peritoneal nodules of 1682A-V and 1682A-L cells have similar histologies. Comparison of 1 µm plastic sections showed that the histologic organization of 1682A-L and 1682A-V tumor nodules was very similar (Fig. 2D and E). Cells grew as undifferentiated sheets, with no evidence of trabeculae, bile canaliculi, or glandular arrays. These results were unlike those reported previously for the PC-3 cell lines, where PC-3 infected with empty retrovirus showed a more histotypic organization than retrovirally transduced cells expressing CEACAM1a-4L (16).

1682A-L and 1682A-V show differences in growth and morphology when grown on a soft agar substratum. Rhim (29) has reported that transformed cells form larger aggregates than their normal counterparts when cultured on a soft agar substratum, a characteristic that was closely correlated with growth in soft agar and tumorigenicity. In keeping with Rhim's findings, the tumorigenic 1682A-V cells produced aggregates that were, on average, >2.5-fold larger than those formed by the nontumorigenic 1682A-L cells. When 1682A-V aggregates were cross-sectioned, they were found to be composed of tightly packed cells with no evidence of a lumen (Fig. 4B). 1682A-L aggregates also lacked a lumen but were distinguished by enlarged intercellular spaces (Fig. 4A, arrows).



View larger version (39K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. Aggregates formed by hepatocellular carcinoma cells shown growing on soft agar. Low- and high-magnification views of aggregates formed by 1682A-L (A) and 1682A-V cells (B) at 48 hours after seeding into Petri dishes coated with soft agar (SA). The more open, loosely packed appearance of 1682A-L aggregates results from the enlarged spaces between cells (A, inset, arrows). Bar, 200 µm (A and B) and 25 µm (inset). C, average area of aggregates (columns) determined from analysis of images similar to those in A and B using Image Pro Plus Software. Using an unpaired t test with a Welch correction, the difference in the mean area of 1682A-L and 1682A-V aggregates is highly significant (P < 0.0001).

 
1682A-L cells have reduced levels of cell cycle regulators involved in the G1-S transition. Table 1 lists genes showing at least a 2-fold differential in expression between 1682A-V and 1682A-L cells growing on a soft agar substratum. As shown, mRNA levels for cyclin D1, cyclin D2, Myb, and Rhou were decreased 1.5- to 4-fold in 1682A-L cells relative to CECAM1a-4L-negative 1682-V cells at both time points examined. Because all of these proteins are involved in promoting the G1-S transition, these results suggested that in vitro, CEACAM1a-4L was inhibiting progression past the G1-S checkpoint. Levels of Notch 2, a promoter of proliferation and MDM2, a p53 inhibitor, were lower in CEACAM1a-4L-positive cells at the 24-hour time point; however, by 68 hours, MDM2 levels decreased 2-fold in 1682A-V but did not change significantly in 1682A-L cells. Paradoxically, the levels of transcripts for p21, CDKi2a, p19, and Ran, four inhibitors of the G1-S transition, were 2.5- to 5-fold higher at the 24-hour time point in 1682A-V versus 1682A-L cells. At 68 hours, the levels of p21 and CDKi2A were essentially unchanged in 1682A-L but had dropped 2.5-fold in 1682A-V cells. P19, in contrast, increased 2-fold in 1682A-L cells but dropped 2.5-fold in 1682A-V cells by 68 hours. Microscopic examination of 1 µm sections from cell aggregates of 1682A-V and 1682A-L cells showed numerous mitotic figures in both cell lines at both time points.


View this table:
[in this window]
[in a new window]

  		Table 1. Microarray cell cycle gene expression analysis

 
1682A-L cells on soft agar display a prolonged G1-S progression. FACS analysis of cells labeled with propidium iodide was carried out to determine if CEACAM1a-4L expression was delaying progression through G1-S, a possibility suggested by the results from gene array analysis. Figure 5 shows that during the period from 14 to 24 hours after placing serum-starved cultures into medium with 10% FBS, the percentage of 1682A-V cells in S phase increased and those in G2-M decreased dramatically. In contrast, during the same interval, there was only a small increase in the percentage of S-phase 1682A-L cells and a much smaller decrease in cells in G2-M, suggesting CEACAM1a-4L was retarding progression through G1-S.



View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5. Cell cycle distribution determined by flow cytometry of propidium iodide–stained 1682A-V, 1682A-L, and 1682A-S cells. Columns, percentage of cells in each phase of the cell cycle at 14 and 24 hours after release from serum starvation. Subdiploid percentages were approximately equal for the three cell lines, a finding in agreement with the relative levels of apoptosis determined with caspase and PARP assays.

 
Restoration of CEACAM1a-4L expression does not produce a significant increase in apoptosis. A comparison of the caspase activity in 1682A-V, 1682A-S, and 1682A-L cells did not reveal a significant difference in the percentage of cells in early (7-AAD impermeant/caspase positive) or late (7-AAD permeable/caspase positive) apoptosis regardless of whether cells were grown in suspension on soft agar or attached to plastic. These results were consistent with the close similarity in the levels of subdiploid cells detected by FACS analysis (Fig. 6A-C) and by results from PARP cleavage assays (Fig. 6D), which showed barely detectable levels in lysates of 1682A-L (lanes 2 and 4), 1682A-V (lanes 5 and 6), and 1682A-S (lane 7) of the 89 kDa cleavage fragment present in lysates of etoposide-induced apoptotic HL-60 cells (lane 2). Neither the intact nor the 89 kDa cleavage fragment were present at significant levels in lysates of normal rat liver (lanes 3 and 4). Equal loading was indicated by the close similarity in the intensity of bands at 70 and 54 kDa, representing constitutively expressed isoforms detected with anti-Hsc70 antibodies.



View larger version (41K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6. Levels of apoptosis. A and B, representative FACS profiles of 1682A-V and 1682A-L cells maintained on plastic and stained for caspase activity with SR-VAD-FMK and 7-AAD. Bottom left, nonapoptotic cells; bottom right, early apoptotic cells; top right, late apoptotic cells; top left, dead cells. C, average levels (columns) of apoptosis from two to four separate assays using different cell preparations for each assay. Cells were grown on soft agar or plastic; bars, SD. The differences in the levels of early and late apoptosis (caspase activity) in cultures of 1682A-V, 1682A-S, and 1682A-L cells were not statistically significant (P < 0.05, unpaired t test). D, Western blot labeled with anti-PARP and anti-Hsc70 antibodies, the latter as an invariant loading control. Lanes 1, 3, 5, and 7, loaded with 6 µg of detergent lysate; lanes 2, 4, 6, 8, and 9, loaded with 12 µg of detergent lysate. Lane 1, lysate was prepared from untreated HL-60 cells; lane 2, lysate was prepared from etoposide-treated HL-60 cells; note the almost complete conversion of the 116 kDa PARP into its 89 kDa cleavage product. The remainder of the blot was prepared with lysates from hepatocytes (lanes 3 and 4), 1682A-V cells (lanes 5 and 6), 1682A-L cells (lanes 7 and 8), and 1682A-S cells (lane 9). Bottom two bands (70 and 54 kDa), two major splice variants of Hsc70.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Previous studies in our laboratory showed that CEACAM1 was lost by a high percentage of transplantable and primary rat hepatocellular carcinoma, a loss also observed in several other tumor models (13–21). Because induction of hepatocellular carcinoma in the rat liver is a widely used model for carcinogenesis, we initiated experiments to determine if restoration of CEACAM1a-4L suppressed the tumorigenicity of transplantable rat hepatocellular carcinoma, a model system that had not been examined previously in this regard. Of particular interest was the ability of CEACAM1a-4L to suppress hepatocellular carcinoma transplanted by i.p. instead of s.c. injection. Although the latter method has been used successfully in previous studies, the propensity for hepatocellular carcinoma to grow on the omentum allowed tumor burden to be estimated by simply weighing the omentum, a much easier and more accurate approach than measuring the dimension of s.c. nodules. Our results showed that CEACAM1a-4L mediated total suppression of 1682A-L cells at 21 days after i.p. injection, thereby demonstrating that hepatocellular carcinoma growing on the omentum are also susceptible to CEACAM1a-4L suppression.

In animals injected with 1682A-V cells, there were large variations in tumor burden that seemed to be independent of dose. Such variations are common in tumors transplanted s.c. and are likely to arise from relatively small differences in latency resulting from variability in the immune response of host animals, the confluency of cells at harvest, or the injection process. Although suppression was 100% at short time points (21 days), at longer time points (38 days), nodules of 1682A-L cells became apparent especially at higher doses (Fig. 3, animals 29-31). This extended latency could reflect the outgrowth of tumorigenic revertants, a phenomenon observed in PC-3 cells with high frequency. However, 1682A-L cells retain suppressor activity after long-term storage or repeated passage, suggesting the alternative possibility that these nodules result from partially suppressed cells expressing low levels of CEACAM1a-4L.

Unlike their PC-3 counterparts, 1682A-L and 1682A-V nodules did not display major changes in histology. Both tumor types grew as undifferentiated sheets of cells with no discernable differences in histologic organization. In contrast, marked differences in morphology were observed when the two cell lines were cultured on a soft agar substratum. On average, aggregates formed by 1682A-V cells were 3-fold larger in size than their 1682A-L counterparts. 1682A-L aggregates were also distinguished by enlarged intercellular spaces (Fig. 4A, inset). Rhim has previously reported a close correlation between tumorigenicity and the size of aggregates formed by cells growing in an anchorage-independent manner on a soft agar substratum, a finding consistent with the marked differences in tumorigenicity and aggregate size displayed by 1682A-L and 1682A-V cells (29). The fact that 1682A-L cells that were completely growth suppressed in vivo were able to grow in vitro suggests a role in vivo for other CEACAM1-mediated events, a likely possibility being an inhibition of angiogenesis, an activity shown by Volpert et al. (13) in prostatic carcinomas.

Differences in levels of apoptosis could also play a role in tumor suppression mediated by CEACAM1a-4L. Nittka et al. (30) noted a close correlation between the loss of CEACAM1 expression in early colorectal hyperplastic lesions and an increase in the levels of apoptosis, suggesting CEACAM1 involvement in the regulation of apoptosis. This idea was supported by the increased apoptosis observed in Jurkat cells transfected with CEACAM1 and in HT29 human colon cancer cells induced to express CEACAM1 by treatment with IFN{gamma}. Significantly, the increase in apoptosis required cross-linking of CEACAM1 using specific antibodies and was not induced by CEACAM1 expression alone, a finding in agreement with our inability to detect a significant difference in the levels of apoptosis in 1682A-L and 1682A-V cultures using three different assays. A point of apparent inconsistency in this regard is the much higher levels of apoptosis observed in the multicaspase assay relative to the PARP assay where there was little or no evidence of cleavage. It is likely that this discrepancy results from anoikis induced following a 1-hour incubation in suspension at 37°C in the assay buffer containing SR-VAD-FMK and 7-AAD. Cells in the PARP assay, in contrast, are directly solubilized in detergent without trypsinization or incubation in suspension.

Kunath et al. (21) reported that a 3.3-fold and a 4.2-fold increase in human CEACAM1a-4L relative to vector-infected controls resulted in suppression of CT51 mouse colon carcinoma cells, an effect that did not occur in cells expressing CEACAM1a-4L at levels 11.1 or 23.7 times greater than controls. These authors concluded that the high levels of CEACAM1a-4L negated the suppressive effects, perhaps as a consequence of alterations in the balance of associated proteins (21). However, in the report by Kunath et al. (21), the expression levels of CEACAM1a-4L on the CT51 cells relative to normal colonocytes was not determined, an important piece of information that could help explain the discrepancy with the results in the present study showing strong suppression of the 1682A-L cell line at CEACAM1a-4L levels less than half that in normal liver and 13-fold higher than 1682A-V controls (21). Assuming that the expression levels on CT51 and 1682A cells relative to their normal counterparts are indeed similar, this would suggest that there are species- and/or tissue-specific differences in the expression levels, regulation, or even identity of associated proteins essential for CEACAM1a-4L-mediated tumor suppression.

Analysis of cell cycle kinetics following release from serum starvation suggested that CEACAM1 expression was retarding progression through the G1-S transition. Between 14 and 24 hours after release from serum starvation, the percentage of 1682A-V and 1682A-S cells in S phase increased by 178% and 70%. In contrast, there was only a 12.5% increase in the percentage of 1682A-L cells in S during the same time interval. A prolongation of G1-S was also indicated by the results from cell cycle gene expression arrays showing significantly reduced levels of G1-S cyclins D1, D2, and E2 in 1682A-L relative to 1682A-V cells at 24 hours and, to a lesser extent, at 68 hours after seeding onto soft agar–coated dishes. Paradoxically, 1682A-V cells also showed higher levels of transcripts for three cell cycle inhibitors at the 24-hour time point. A transient increase in transcripts for several cell cycle inhibitors (p21Cip1, p15Ink4b, p16Ink4a, p18Inc4c, and p19Ink4d) was also detected by Awad et al. (31) within 6 to 24 hours after induction of hepatocyte proliferation by partial hepatectomy. p19ink4d transcripts were also present in fetal liver but were barely detectable by reverse transcription-PCR in adult liver. A possible explanation for the high levels of p19 and p21 in 1682A-L cells is suggested by recent evidence showing that cell cycle inhibitors have regulatory functions extending beyond cell cycle inhibition. Cytoplasmic p21, for example, has been shown to act as an assembly and nuclear transport factors for cyclin D1/cyclin-dependent kinase 4 complexes and as a transcriptional cofactor for nuclear factor-{kappa}B, myc, and signal transducers and activators of transcription 3 (32). In addition, Rossig et al. (33) have shown that Akt-mediated phosphorylation of p21 attenuates cell cycle inhibitory activity, a pathway that would obviate a direct correlation between the levels of p21 transcripts and p21 cell cycle inhibition. Additional functions beyond cell cycle inhibition are also likely for p19, a member of the INK4 family. This is suggested by the lack of mutated forms of p19 in human tumors, the absence of tumors in p19 knockout mice after as long as 2 years, and the ability of p19 to induce a partial differentiation response and extend survival of immature myeloid cells in the absence of hemopoietins (34, 35).

In conclusion, we have established the ability of CEACCAM1a-4L to suppress the growth of rat hepatocellular carcinoma that localize on the omentum following i.p. injection. We have further shown that suppression occurred at levels of CEACAM1a-4L that were approximately half those on normal hepatocytes and 13-fold higher than negative controls. We found that CEACAM1a-4L did not significantly increase levels of apoptosis but did seem to delay transition from G1 into S, an effect paralleled by decreased expression of three cyclins involved in G1-S progression. In addition, we showed that when the two cell lines were grown on a soft agar substratum, significant differences in the size of cell aggregates and intercellular spaces became apparent, suggesting that the effect of CEACAM1a-4L restoration on tumor cells is influenced by tissue and/or microenvironmental factors.


   Acknowledgments
 
Grant support: NIH grants F31-CA83170 (N.A. Laurie) and RO1-CA42714, and Centers of Biochemical Research Excellence grant P20RR017695 (D.C. Hixson).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received 8/17/04. Revised 8/18/05. Accepted 9/ 9/05.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Cheung PH, Culic O, Qiu Y, et al. The cytoplasmic domain of C-CAM is required for C-CAM-mediated adhesion function: studies of a C-CAM transcript containing an unspliced intron. Biochem J 1993;295:427–35.
  2. Culic O, Huang Q-H, Flanagan D, Hixson D, Lin S-H. Molecular cloning and expression of a new rat liver cell-CAM 105 isoform. Biochem J 1992;285:47–53.
  3. Hixson DC, McEntire KD, Obrink B. Alterations in the expression of a hepatocyte cell adhesion molecule by transplantable rat hepatocellular carcinomas. Cancer Res 1985;45:5049–58.
  4. Beauchemin N, Draber P, Dveksler G, et al. Redefined nomenclature for members of the carcinoembryonic antigen family. Exp Cell Res 1999;252:243–9.[CrossRef][Medline]
  5. Thompson NL, Lin SH, Panzica MA, Hixson DC. Cell CAM 105 isoform RNA expression is differentially regulated during rat liver regeneration and carcinogenesis. Pathobiology 1994;62:209–20.[CrossRef][Medline]
  6. Thompson NL, Panzica MA, Hull G, et al. Spatiotemporal expression of two cell-cell adhesion molecule 105 isoforms during liver development. Cell Growth Differ 1993;4:257–68.[Abstract]
  7. Cheung PH, Thompson NL, Earley K, et al. Cell-CAM105 isoforms with different adhesion functions are coexpressed in adult rat tissues and during liver development. J Biol Chem 1993;268:6139–46.[Abstract/Free Full Text]
  8. Hixson DC, McEntire KD. Detection of an altered form of Cell-CAM 105 on rat transplantable and primary hepatocellular carcinomas. Cancer Res 1989;49:6788–94.[Abstract/Free Full Text]
  9. Tanaka K, Hinoda Y, Takahashi H, et al. Decreased expression of biliary glycoprotein in hepatocellular carcinomas. Int J Cancer 1997;74:15–9.[CrossRef][Medline]
  10. Kleinerman DI, Troncoso P, Lin S-H, et al. Consistent expression of an epithelial cell adhesion molecule (C-CAM) during human prostate development and loss of expression in prostate cancer: implication as a tumor suppressor. Cancer Res 1995;55:1215–30.[Abstract/Free Full Text]
  11. Neumaier M, Paululat S, Chan A, Matthaes P, Wagener C. Biliary glycoprotein, a potential human cell adhesion molecule, is down-regulated in colorectal carcinomas. Proc Natl Acad Sci U S A 1993;90:10744–8.[Abstract/Free Full Text]
  12. Riethdorf L, Lisboa BW, Henkel U, et al. Differential expression of CD66a (BGP), a cell adhesion molecule of the carcinoembryonic antigen family, in benign, premalignant, and malignant lesions of the human mammary gland. J Histochem Cytochem 1997;45:957–63.[Abstract/Free Full Text]
  13. Volpert O, Luo W, Liu TJ, et al. Inhibition of prostate tumor angiogenesis by the tumor suppressor CEACAM1. J Biol Chem 2002;277:35696–702.[Abstract/Free Full Text]
  14. Estrera VT, Chen DT, Luo W, Hixson DC, Lin SH. Signal transduction by the CEACAM1 tumor suppressor. Phosphorylation of serine 503 is required for growth-inhibitory activity. J Biol Chem 2001;276:15547–53.[Abstract/Free Full Text]
  15. Estrera VT, Luo W, Phan D, et al. The cytoplasmic domain of C-CAM1 tumor suppressor is necessary and sufficient for suppressing the tumorigenicity of prostate cancer cells. Biochem Biophys Res Commun 1999;263:797–803.[CrossRef][Medline]
  16. Comegys MC, Carreiro MP, Brown JF, et al. C-CAM1 expression: differential effects on morphology, differentiation state and suppression of human PC-3 prostate carcinoma cells. Oncogene 1999;18:3261–76.[CrossRef][Medline]
  17. Hsieh JT, Luo W, Song W, et al. Tumor suppressive role of an androgen-regulated epithelial cell adhesion molecule (C-CAM) in prostate carcinoma cell revealed by sense and antisense approaches. Cancer Res 1995;55:190–7.[Abstract/Free Full Text]
  18. Luo W, Wood C, G, Earley K, Hung M-C, Lin S-H. Suppression of tumorigenicity of breast cancer cells by an epithelial cell adhesion molecule (C-CAM1): the adhesion and growth suppression are mediated by different domains. Oncogene 1997;14:1697–704.[CrossRef][Medline]
  19. Kleinerman DI, Dinney CP, Zhang WW, et al. Suppression of human bladder cancer growth by increased expression of C-CAM1 gene in an orthotopic model. Cancer Res 1996;56:3431–5.[Abstract/Free Full Text]
  20. Wang L, Lin SH, Wu WG, et al. C-CAM1, a candidate tumor suppressor gene, is abnormally expressed in primary lung cancers. Clin Cancer Res 2000;6:2988–93.[Abstract/Free Full Text]
  21. Kunath T, Ordonez-Garcia C, Turbide C, Beauchemin N. Inhibition of colonic tumor cell growth by biliary glycoprotein. Oncogene 1995;11:2375–82.[Medline]
  22. Rosenberg M, Nedellec P, Jothy S, et al. The expression of mouse biliary glycoprotein, a carcinoembryonic antigen-related gene, is down-regulated in malignant mouse tissues. Cancer Res 1993;53:4938–45.[Abstract/Free Full Text]
  23. Nedellec P, Turbide C, Beauchemin N. Characterization and transcriptional activity of the mouse biliary glycoprotein 1 gene, a carcinoembryonic antigen-related gene. Eur J Biochem 1995;231:104–14.[Medline]
  24. Hixson DC, Faris RF. Rat liver epithelial cells: when it comes to differentiation, there's no place like home. Hepatology 1994;19:1315–6.[CrossRef][Medline]
  25. Becker A, Neumeier R, Park CS, Gossrau R, Reutter W. Identification of a transformation-sensitive 110-kDa plasma membrane glycoprotein of rat hepatocytes. Eur J Cell Biol 1986;39:417–23.[Medline]
  26. Comegys MM, Lin SH, Rand D, et al. Two variable regions in carcinoembryonic antigen-related cell adhesion molecule1 N-terminal domains located in or next to monoclonal antibody and adhesion epitopes show evidence of recombination in rat but not in human. J Biol Chem 2004;279:35063–78.[Abstract/Free Full Text]
  27. Herceg Z, Wang ZQ. Failure of poly(ADP-ribose) polymerase cleavage by caspases leads to induction of necrosis and enhanced apoptosis. Mol Cell Biol 1999;19:5124–33.[Abstract/Free Full Text]
  28. Kaufmann SH, Desnoyers S, Ottaviano Y, Davidson NE, Poirier GG. Specific proteolytic cleavage of poly(ADP-ribose) polymerase: an early marker of chemotherapy-induced apoptosis. Cancer Res 1993;53:3976–85.[Abstract/Free Full Text]
  29. Rhim JS. Cell aggregation assay: a rapid means of evaluating and selecting in vitro transformed cells. Cancer Detect Prev 1983;6:381–8.[Medline]
  30. Nittka S, Gunther J, Ebisch C, Erbersdobler A, Neumaier M. The human tumor suppressor CEACAM1 modulates apoptosis and is implicated in early colorectal tumorigenesis. Oncogene 2004;23:9306–13.[CrossRef][Medline]
  31. Awad MM, Sanders JA, Gruppuso PA. A potential role for p15(Ink4b) and p57(Kip2) in liver development. FEBS Lett 2000;483:160–4.[CrossRef][Medline]
  32. Coqueret O. New roles for p21 and p27 cell-cycle inhibitors: a function for each cell compartment? Trends Cell Biol 2003;13:65–70.[CrossRef][Medline]
  33. Rossig L, Jadidi AS, Urbich C, et al. Akt-dependent phosphorylation of p21(Cip1) regulates PCNA binding and proliferation of endothelial cells. Mol Cell Biol 2001;21:5644–57.[Abstract/Free Full Text]
  34. Roussel MF. The INK4 family of cell cycle inhibitors in cancer. Oncogene 1999;18:5311–7.[CrossRef][Medline]
  35. Adachi M, Roussel MF, Havenith K, Sherr CJ. Features of macrophage differentiation induced by p19INK4d, a specific inhibitor of cyclin D-dependent kinases. Blood 1997;90:126–37.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 DevelopmentHome page
R. Simper-Ronan, K. Brilliant, D. Flanagan, M. Carreiro, H. Callanan, E. Sabo, and D. C. Hixson
Cholangiocyte marker-positive and -negative fetal liver cells differ significantly in their ability to regenerate the livers of adult rats exposed to retrorsine
Development, November 1, 2006; 133(21): 4269 - 4279.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Laurie, N. A.
Right arrow Articles by Hixson, D. C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Laurie, N. A.
Right arrow Articles by Hixson, D. C.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online