Intercellular Communication and Human Hepatocellular Carcinoma

202

Ann. N.Y. Acad. Sci. 1028: 202–212 (2004). © 2004 New York Academy of Sciences.doi: 10.1196/annals.1322.025

Intercellular Communication and Human Hepatocellular Carcinoma

GIUSEPPE CARRUBA,a LETIZIA COCCIADIFERRO,a VINCENZO BELLAVIA,b SERGIO RIZZO,a CHRISTOS TSATSANIS,c DEMETRIOS SPANDIDOS,c PAOLA MUTI,d COLIN SMITH,e PARMENDER MEHTA, f AND LUIGI CASTAGNETTAa,b

aDepartment of Experimental Oncology and Clinical Application, University Medical School, Palermo, ItalybExperimental Oncology Unit, Department of Clinical Oncology, ARNAS-Civico, “M. Ascoli” Cancer Hospital Center, Palermo, ItalycSchool of Medicine, University of Crete, Heraklion, Crete, GreecedDepartment of Social and Preventive Medicine, University at Buffalo, State University of New York, Buffalo, New York 14214, USAeSchool of Biomedical and Molecular Sciences, University of Surrey, Guildford, UKfDepartment of Biochemistry and Molecular Biology, University of Nebraska Medical Center, Omaha, Nebraska 68198, USA

ABSTRACT: We have previously reported that gap junction–mediated inter-cellular communication (GJIC) can be restored in junctionally deficienthuman prostate epithelial cells, also suggesting that GJIC activity is regulatedby estrogen. In the present work, we report studies on sex steroid regulation ofGJIC and proliferative activity in both nontumoral (Chang liver, CL) andmalignant (HepG2, Huh7) human liver cells. Junctional activity and liver cellgrowth were measured using the scrape-loading/dye-transfer (SL/DT) and theMTS assay, respectively. Using the SL/DT, only Huh7 cells exhibited a moderatedegree of junctional activity in basic conditions, while neither CL nor HepG2cells showed functional GJIC. Under exactly the same experimental approachused for prostate studies, we observed that, once again, both estrogen (eitherestradiol or estrone) and FK induce a significant increase of GJIC in Huh7cells, while exposure of HepG2 cells to FK produces only a limited rise of junc-tional activity in this cell line. However, estrogen induced a significant increaseand reduction of the proliferative activity of CL and Huh7 cells, respectively,while growth of HepG2 cells was not affected. While the above evidencesuggests that estrogens are primarily implicated in growth regulation andcommunication of both prostate and liver epithelial cells, it also implies thatcompounds able to restore GJIC in junctionally deficient cells or prevent itsdisruption in junctionally proficient cells may be used for development of new

Address for correspondence: Giuseppe Carruba, M.D., Ph.D., Department of ExperimentalOncology and Clinical Application, University Medical School, c/o “M. Ascoli” Cancer HospitalCenter, ARNAS-Civico, Via C. Lazzaro 2, 90127 Palermo, Italy. Voice: +39-091-666-4346; fax:+39-091-666-4352.

[email protected]

203CARRUBA et al.: INTERCELLULAR COMMUNICATION AND HCC

strategies in the prevention and/or treatment of several human malignancies,including hepatocellular carcinoma (HCC).

KEYWORDS: gap junction–mediated intercellular communication (GJIC);hepatocellular carcinoma (HCC); proliferation; steroid; tumor

Primary hepatocellular carcinoma (HCC) represents the sixth leading cancer and thethird most common cause of death from cancer worldwide. In 1996, mortalityattributable to HCC was estimated at 7.5% of all deaths for cancer.1 In our countryof Italy, mortality rates have been drastically increasing from 4.8 deaths/100,000 in1969 up to 11.0/100,000 in 1994, with even higher values in the south of Italy(Italian ISTAT database). In Italy, this phenomenon seems to be due to the spread ofchronic liver diseases caused by infection with hepatitis B and C viruses (HBV andHCV), which are considered among the major risk factors for the development ofHCC. Although a number of experimental observations underlie the potential forviral products to contribute to liver carcinogenesis, only a minority of patientshaving chronic viral hepatitis and cirrhosis develop HCC, suggesting that otherevents are likely to be involved in the process of neoplastic transformation.

Independent of race and geography, HCC incidence rates in men are at least twoto three times those in women, with a more pronounced sexual dimorphism in high-risk regions, including our own. This hints at a potential role for sex steroids in HCCdevelopment and progression. Based on previous evidence for estrogen receptors inprimary HCC,2 different clinical trials have investigated the potential impact of anti-estrogen (tamoxifen) treatment on the survival of patients having HCC. From thesestudies, there is evidence that tamoxifen is of no benefit in prolonging survival ofHCC patients.3,4 This, however, could be partly due to the presence of variantestrogen receptor (ER) forms, as previously described.5 Recently, the use of syn-thetic progestins (megestrol) for treatment of patients having inoperable HCC char-acterized by variant liver ER resulted in a significant increase of median survival ofpatients.6 In addition, previous studies have emphasized the potential role of alter-native, nonreceptorial mechanisms in growth control of both normal and cancerhuman liver cells based on peptide growth factors, such as EGF and TGFβ.7

It is well established that carcinogenesis for any human cancer, including humanHCC, is a multistep, multimechanism process, involving the interaction of genetic(oncogenes, tumor suppressor genes), endogenous (hormones, growth factors), andenvironmental (diet, radiation, chemical pollutants) factors. Because of its typicalnatural history, which goes from the development of chronic hepatitis, to cirrhosis,and to liver cancer, HCC can be taken as an in vivo model of the multistep process ofcancerogenesis. According to the theory of “oncogeny as partially blocked ontogeny”,the concept of carcinogenesis is viewed as a stem or a progenitor cell being subjectedto an initial stable alteration of the terminal differentiation pathway (initiation) anda subsequent clonal expansion of the initiated proliferating cells (promotion). Sinceterminal differentiation of stem cells is strictly associated with development of gapjunction–mediated intercellular communication (GJIC), it has been proposed thatdisturbance of GJIC may be implicated in the carcinogenetic process.

The gap junction channel–forming proteins, the connexins (Cx), comprise amultigene family with at least 20 members in mammals.8,9 The connexins are

204 ANNALS NEW YORK ACADEMY OF SCIENCES

assembled into hexameric complexes, forming a hemichannel unit called connexon.Two adjacent connexons from contacting cells align end to end to give rise to ahydrophilic membrane channel, through which ions and small molecules (below2000 daltons) may equilibrate the cytoplasms and synchronize the metabolic orelectrotonic activities of coupled cells.

Several reports have emphasized the assumption that disruption of GJIC may beimportant in carcinogenesis. There is consistent evidence that various oncogenes(e.g., ras, raf, neu, and src),10,11 some tumor suppressor genes (such as p53 andpRb),12 different tumor-promoting chemicals (e.g., phenobarbital and phorbolesters),13 and several antitumor-promoting agents (including retinoids and dexa-methasone),14 which are all known to inhibit or enhance apoptosis, can alsorespectively inhibit or enhance GJIC at the transcriptional, translational, or post-translational level. In this respect, GJIC may have a permissive role in apoptosis byfacilitating the information trafficking needed to trigger cell death. On the otherhand, transfection of different gap junction genes into a variety of noncommuni-cating tumor cells increases genetic stability, restores some or most growth control,and suppresses tumorigenicity of cells.15,16 All the above evidence would imply thatgap junction genes can be considered as a novel family of tumor suppressor genes.17

It could be speculated that alteration of GJIC may represent a nongenotoxicmechanism that plays a role in tumor promotion, also by inhibiting apoptosis.

An assortment of agents, including hormones and growth factors, may affectconnexin gene expression and both formation and activity of gap junction channelsin various cell systems. Unfortunately, there is little information on the potential roleof GJIC disturbances in development of human HCC.

In the present work, we have investigated GJIC activity and regulation in eithernontumoral or malignant human liver cell lines, also in relation to cell proliferationand sex steroid receptor expression.

MATERIALS AND METHODS

Cell Cultures

Either nontumoral (Chang liver, CL) or malignant (HepG2, Huh7) human livercell lines were used. The CL and HepG2 cells were purchased from the AmericanType Culture Collection (Rockville, MD), while Huh7 cells were generously providedby Marcello Cervello (IBIM, Palermo). Cells were maintained in RPMI-1640 (CLand HepG2) or DMEM (Huh7) culture medium, supplemented with 10% fetal calfserum (FCS), antibiotics (100 IU/mL penicillin, 100 µg/mL streptomycin, 0.25 µg/mLamphotericin B), and 2 mM L-glutamine or sodium pyruvate and Hepes, respectively.Cells were periodically tested for mycoplasma contamination. Cells having a narrowrange of passage number (CL, 259–263; HepG2, 78–83; Huh7, 12–19) were used forall experiments.

For GJIC studies, cells were examined under either standard conditions or follow-ing a 48-h exposure to test agents, including (a) estradiol (E2, 1 µM), (b) estrone (E1,1 µM), (c) testosterone (T, 1 µM), (d) the antihypertensive drug forskolin (FK,30 µM), and (e) lovastatin (LS, 5 µM), an antihypercholesterolemic agent. Controlcell cultures received relevant vehicle (DMSO for FK; ethanol for all othertreatments) alone (0.01% v/v).


Scrape-Loading/Dye-Transfer (SL/DT)

The scrutiny of GJIC in cultured cells was carried out using the dye transfertechnique as reported elsewhere.18 Briefly, cells were grown in 35-mm culture platesuntil 70–80% confluent. Cell monolayers were rinsed twice with PBS and then 2 mLof PBS containing 0.05% (vol/vol) Lucifer Yellow (Sigma, St. Louis, MO) was add-ed to each dish. This membrane-impermeable dye was loaded into the primary re-cipient cells by making 5–7 scrape lines on the cell monolayers using a sharpsurgical blade. After 3-min incubation at room temperature, the solution was dis-carded and the cells repeatedly washed with PBS to remove dye excess. Cells werefinally processed immediately or fixed in 4% formalin and stored at 4°C until furtheranalysis. Cultures were subjected to laser scanning analysis using a Meridian Ultimalaser cytometer (Meridian Instruments, Okemos, MI), where the fluorescent area(expressed in µm2) was taken as a measure of the diffusion of the dye, after correc-tion for background. Cells were also examined under epifluorescence microscopy(Nikon Diaphot phase-contrast; Nikon, Garden City, NY), where equal surfaceareas, equivalent to 20× objective lens field and having an average of 100 cells ineach randomly selected field, were compared and fluorescent cells counted.

Steroid Receptors

The study of expression and localization of ERα, ERβ, AR, and PgR, wasconducted through immunocytochemical assays on cells grown directly on tissueculture chamber slides (Nunc, Roskilde, Denmark). Cells were initially treated with1% to block nonspecific binding and subsequently incubated overnight at 4°C withmonoclonal antibodies directed against the human ERα (clone 1D5, Dako), AR(clone F39.4.1, Biogenex), and PgR (clone 1A6, Dako), or with polyclonal antibodyagainst ERβ (Upstate). After 2–3 washings in PBS, cells were incubated withspecies-specific biotinylated secondary antibodies and then with an ABC stainingdetection system (Santa Cruz Biotechnology). Counterstaining was obtained usinghematoxylin (Sigma). Control slides were processed under exactly the same experi-mental conditions, except nonspecific goat IgG replaced the primary antibody. Thepercent of positively stained cells (nuclei) and the intensity of stain were determinedin duplicate assays using the automated QWIN Image Analyzer (Leica, Germany)on an average of 30 randomly selected fields.

Cell Proliferation

The effects of sex steroids, alone or in combination with their respective antago-nists, and other test agents (including FK and LS), were determined using the MTSassay. Cells in exponential growth were seeded onto 96-well microplates (5 × 104

cells/well) in phenol red–free culture medium supplemented with 5% charcoal-treated FCS to remove endogenous steroids. Cells were incubated for 6 days in thepresence of estrogens (estradiol, estrone) or androgen (testosterone), alone or incombination with relevant pure antagonists (antiestrogen: ICI-182,780; antiandro-gen: casodex), all at 1 µM. After incubation, 20-µL aliquots of a solution containingMTS [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium] and phenazine methosulfate were added to each well and platesincubated at 37°C for 2–4 h in the dark. The bioreduction of MTS into its soluble


derivative formazan was finally measured in an ELISA microplate reader at 490 nm,with the formazan formation being directly proportional to the number of cells. Theresults were expressed as the percent of control samples receiving vehicle (0.01%ethanol alone), representing the mean ± SD of 3 separate experiments, eachperformed in quadruplicate.

RESULTS

GJIC Activity

Junctional activity was inspected in cultured cells using the SL/DT assay. In stan-dard conditions, only Huh7 cells exhibited a moderate degree of GJIC, while neitherCL nor HepG2 cells showed functional GJIC. Under exactly the same experimentalconditions that we have previously used for prostate studies,19 estrogen and FKinduced a significant increase of GJIC in Huh7 cells (see FIG. 1). In particular, E1and E2 produced respectively a 40% and 64% increase of junctional activity in Huh7cells, while FK caused a 3.4-fold increase of GJIC with respect to control untreatedcell cultures (FIG. 2). Conversely, exposure of HepG2 cells to FK resulted in a 56%rise of GJIC, while all other treatments were ineffective (FIG. 2). In CL cells, all testagents did not induce any significant change in junctional activity (not shown).

FIGURE 1. Induction of GJIC activity in Huh7 cells by estrogen and forskolin. The SL/DTmethod was used to assess GJIC activity in both controls and cells treated with 1 µM estrone(E1), 1 µM estradiol (E2), or 30 µM forskolin (FK). Control cell cultures (C) received vehiclealone. Cells were incubated for 48 h in the presence of relevant treatment and then junctionaltransfer was assessed by the SL/DT method, as described in MATERIALS AND METHODS. [Colorshown in online version.]


Sex Steroid Receptors

The expression and localization of ERα, ERβ, AR, and PgR in human liver celllines were assessed through immunocytochemical assay and image analysis. Over-all, the three cell lines studied displayed similar patterns, but significantly differentexpression levels of hormone receptors. In particular, all cells stained positively forboth AR and ERβ, while neither ERα nor PgR could be detected. As reported inTABLE 1, as far as ERβ is concerned, both CL and Huh7 cells showed comparable,moderate expression levels, while HepG2 cells exhibited markedly greater levels ofERβ, as reflected in both proportion of positively stained cells and intensity ofstaining. Conversely, CL and HepG2 cells expressed AR to a similar extent, whileHuh7 cells displayed a weak to moderate positivity and stain intensity for this receptor.Overall, there appears to be a clear-cut predominance of ERβ over AR in both Huh7and HepG2 cells, while CL cells present a slight prevalence of AR over ERβ.

TABLE 1. Immunocytochemical assay of sex steroid receptors in human liver cells

ERβ AR

Positive stain (%)

Stain intensity (%)

Positive stain (%)

Stain intensity (%)

Chang liver

NOTE: Data represent the average percent of positive cells and the intensity of staining fromtriplicate experiments, as determined using the automated QWIN Image Analyzer (Leica,Germany), on an average of 30 randomly selected fields. For methodological details, see text.

23.6 ± 3.1 33.3 ± 4.4 33.2 ± 2.9 40.5 ± 3.8

Huh7 25.4 ± 3.5 30.5 ± 2.1 10.5 ± 1.6 13.7 ± 1.2

HepG2 73.8 ± 8.4 87.6 ± 10.1 30.8 ± 5.2 43.7 ± 3.7

FIGURE 2. Measurement of GJIC activity in Huh7 (stippled bars) and HepG2(hatched bars) human liver cells. Junctional activity was measured using the SL/DT assay,as reported in MATERIALS AND METHODS. Results represent mean ± SD values of triplicateassays, with 4 to 5 randomly selected fields measured for each experiment, and are ex-pressed as percent values of control.


Cell Proliferation

The proliferative effects of either estrogen (E2, E1) or androgen (T), alone or incombination with their respective antagonists (ICI-182, casodex), as well as of othertest agents (FK and LS), were investigated using the MTS assay. As illustrated inFIGURE 3, estrogens (both E2 and E1) and notably T induced a significant decreaseof cell proliferation in CL cells, respectively being 20%, 30%, and >95% withrespect to control; in addition, FK and LS also reduced CL cell growth by 30% and34%, respectively (FIG. 3A). Conversely, E2 and E1 induced a consistent increase(respectively of 15% and 26%) of cell proliferation in Huh7 cells, while neither Tnor FK or LS significantly modified growth rates in this cell line (FIG. 3B). In HepG2cells, only E1 and T produced a statistically significant decrease of cell proliferation,while all the other treatments were ineffective in this respect (FIG. 3C). In separateexperiments, the combination of estrogens with the pure antagonist ICI-182,780(ICI), resulted in a greater growth inhibition (35% for E2 and nearly 50% for E1),

FIGURE 3. Growth regulation of human liver cells. Cell proliferation was assessed inCL (A), Huh7 (B), and HepG2 (C) cells exposed for 6 days to 1 µM estradiol (E2), estrone(E1), or testosterone (T), as well as to 30 µM forskolin (FK) or 5 µM lovastatin (LS). Cellgrowth rates were determined using the MTS assay, as described in MATERIALS AND METHODS.Data represent the mean ± SD of duplicate experiments, each performed in quadruplicate.


while coincubation of T with the pure antiandrogen casodex (CSD) produced aslightly lower decrease (72%) of cell proliferation in CL cells (FIG. 4A). In Huh7cells, the simultaneous addition of ICI to either E2 or E1 completely abolished theestrogen-induced growth increase, while the combination of T and CSD resulted ina marked decrease (nearly 60%) of cell proliferation (FIG. 4B). Likewise, in CLcells, the combination of either E1 and ICI or T and CSD resulted in a more pro-nounced decrease of cell proliferation in HepG2 cells, respectively being around20% and 26% (FIG. 4C).

DISCUSSION

In the present work, we reveal that estrogen regulates both GJIC and proliferativeactivities in the human HCC cells, Huh7. In particular, E1 and (to a lesser extent) E2are capable of inducing a marked increase of both junctional activity and cell

FIGURE 4. Proliferative effects of sex steroids and their antagonists in human livercells. Cell proliferation was assessed in CL (A), Huh7 (B), and HepG2 (C) cells exposed for6 days to 1 µM estradiol (E2), estrone (E1), or testosterone (T), alone or in combination withtheir respective antagonists, the antiestrogen ICI-182,780 (ICI) and the antiandrogencasodex (CSD), each at 1 µM.


proliferation in this cell line. Conversely, estrogen does not significantly affect GJ-mediated cell-cell communication in either CL or HepG2 cells, but remarkablyreduces their proliferation.

In recent years, consistent evidence has accumulated to suggest that disturbanceof GJIC is a relatively early event in human liver carcinogenesis and that alterationof the GJ proteins, connexins (Cx), notably their aberrant phosphorylation and/orabnormal regulation by proteins of the extracellular matrix and cell adhesionmolecules, may play a primary role in liver cell transformation.20

We report here that neither nontumoral CL cells nor hepatoma HepG2 cells havefunctional GJIC in standard conditions. Previous studies have revealed that HepG2cells have defective GJIC because of downregulation of Cx26 and aberrant localiza-tion of Cx32, and that Cx26 transfection results in growth inhibition and inductionof the E-cadherin cell adhesion molecule in this cell line.21 On the other hand, thehuman HCC cell line, Huh7, has been reported to express Cx43, but apparently lacksCx32.22 In our studies, we have observed that Huh7 cells contain a moderate degreeof GJIC activity in standard cell culture conditions, presumably due to Cx43 expres-sion. More importantly, we report here that GJIC can be markedly induced in Huh7cells by both estrogen (either E2 or E1) and FK, while this effect could not beobserved in either CL or HepG2 cells. Previous work by our own and other researchgroups have indicated that the Cx43 gene is regulated by estrogen, eventually lead-ing to upregulation of Cx43 expression. In particular, molecular cloning techniqueshave revealed that the Cx43 promoter region contains both AP-1 and AP-2 sites,along with a series of half-palindromic estrogen response elements.23 This evidenceresulted in the inclusion of the Cx43 gene in a class of delayed primary estrogen-responsive genes.24

There is consistent evidence suggesting that connexin genes are primarilyimplicated in cell growth regulation and that their transfection in tumorigenic, Cx-deficient cell lines may restore normal cell growth.25 However, it has been suggestedthat connexin expression itself, more than GJIC activity, may be strictly associatedwith cell growth control, supporting the assumption that connexins may have aGJIC-independent function.26 In particular, studies by Yamasaki and colleagues27

have indicated that gap junction genes may either suppress or promote liver cellproliferation in mice depending on whether excessive or insufficient cell growthrespectively occurs. In this work, we have observed that estrogen (E2 and E1) mark-edly increases GJIC activity and cell proliferation in Huh7 human HCC cells, whileeither estrogen or androgen reduces growth of CL and HepG2 cells, but does notaffect junctional activity in the two cell lines. In addition, FK induced a significantrise of GJIC activity in both Huh7 and HepG2 cells, without any change in their pro-liferative activity. The combination of sex steroids and their respective antagonistsresulted in the abolishment of the estrogen-induced growth increase seen in Huh7cells or in the enhancement of the steroid-induced growth inhibition of CL andHepG2 cells. This evidence, taken together, supports the conception that cell growthand GJIC activity can be independently regulated in a cell-specific manner and thatconnexins may differentially control proliferative and junctional activities of cells.Furthermore, the distinctive patterns of sex steroid receptors that we have observedin the three cell lines studied, with both AR and ERβ being expressed, but not ERαor PgR, would imply that gonadal steroids may diversely regulate growth and func-tion of these cells. This, along with the proliferative effects resulting from the com-


bination of both estrogen and androgen with their specific antagonists, may suggestthat both receptor and nonreceptor mechanisms are implicated in functional andgrowth control by steroids of these cell systems.

Overall, the present evidence suggests that gonadal steroids are primarily impli-cated in growth regulation and communication of liver epithelial cells. It also impliesthat compounds able to restore GJIC in junctionally deficient cells or prevent its dis-ruption in junctionally proficient cells may be used for development of new strate-gies in the prevention and/or treatment of human malignancies, including HCC.

ACKNOWLEDGMENTS

These studies were partly supported by the Italian Association for Cancer Re-search (AIRC) and the National Research Council (CNR).

REFERENCES

1. WORLD HEALTH ORGANIZATION (WHO). 1997. WHO World Health Report. WHO.Geneva.

2. NAGASUE, N., A. ITO, H. YUKAYA & Y. OGAWA. 1986. Estrogen receptors in hepato-cellular carcinoma. Cancer 57: 87–91.

3. ENGSTROM, P., B. LEVIN, C. MOERTEL & A. SCHUTT. 1990. A phase II trial of tamoxifenin hepatocellular carcinoma. Cancer 65: 2641–2643.

4. CLIP GROUP. 1998. Tamoxifen in treatment of hepatocellular carcinoma: a randomisedcontrolled trial. Lancet 352: 17–20.

5. VILLA, E., L. CAMELLINI, A. DUGANI et al. 1995. Variant estrogen receptor messengerRNA species detected in human primary hepatocellular carcinoma. Cancer Res. 55:498–500.

6. VILLA, E., I. FERRETTI, A. GROTTOLA et al. 2001. Hormonal therapy with megestrol ininoperable hepatocellular carcinoma characterized by variant oestrogen receptors.Br. J. Cancer 85: 1606–1608.

7. BARKHEM, T., C. ANDERSSON-ROSS, M. HOGLUND & S. NILSSON. 1997. Characteriza-tion of the “estrogenicity” of tamoxifen and raloxifene in HepG2 cells: regulation ofgene expression from an ERE controlled reporter vector versus regulation of theendogenous SHBG and PS2 genes. J. Steroid Biochem. Mol. Biol. 62: 53–64.

8. PAUL, D. 1995. New functions for gap junctions. Curr. Opin. Cell Biol. 7: 665–672.9. BRUZZONE, R., T. WHITE & D. PAUL. 1996. Connections with connexins: the molecular

basis for direct intracellular signaling. Eur. J. Biochem. 238: 1–27.10. KALIMI, G., L. HAMPTON, J. TROSKO et al. 1992. Homologous and heterologous gap-

junctional intercellular communication in v-raf-, v-myc-, and v-raf/v-myc-transducedrat liver epithelial cell lines. Mol. Carcinog. 5: 301–310.

11. FERNANDEZ-SARABIA, M. & J. BISHOP. 1993. Bcl-2 associates with the ras relatedprotein R-ras p23. Nature 366: 274–277.

12. LOWE, S., E. SCHMITT, S. SMITH et al. 1993. p53 is required for radiation inducedapoptosis in mouse thymocytes. Nature 362: 847–849.

13. TOMEI, L., P. KANTER & C. WENNER. 1988. Inhibition of radiation-induced apoptosis invitro by tumor promoters. Biochem. Biophys. Res. Commun. 155: 324–331.

14. ALLES, A. & K. SULIK. 1989. Retinoic acid–induced limb-reduction defects: perturba-tion of zones of programmed cell death as a pathogenetic mechanism. Teratology 40:163–171.

15. ROSE, B., P. METHA & W. LOEWENSTEIN. 1993. Gap junction protein gene suppressestumorigenicity. Carcinogenesis 14: 1073–1075.

16. ZHU, W., N. MIRONOV & H. YAMASAKI. 1997. Increased genetic stability of HeLa cellsafter connexin 43 gene transfection. Cancer Res. 57: 2148–2150.


17. TROSKO, J. & J. GOODMAN. 1994. Intercellular communication may facilitateapoptosis: implication for tumor promotion. Mol. Carcinog. 11: 8–12.

18. EL-FOULY, M., J. TROSKO & C.C. CHANG. 1987. Scrape-loading and dye transfer: arapid and simple technique to study gap junctional intercellular communication. Exp.Cell Res. 168: 422–430.

19. CARRUBA, G., M.M. WEBBER, S.T.A. QUADER et al. 2002. Regulation of cell-cellcommunication in non-tumorigenic and malignant human prostate epithelial cells.Prostate 50: 73–82.

20. YAMASAKI, H., M. MESNIL, Y. OMORI et al. 1995. Intercellular communication andcarcinogenesis. Mutat. Res. 333: 181–188.

21. YANO, T., F. HERNANDEZ-BLAZQUEZ, Y. OMORI & H. YAMASAKI. 2001. Reduction ofmalignant phenotype of HEPG2 cell is associated with the expression of connexin 26,but not connexin 32. Carcinogenesis 22: 1593–1600.

22. STOCKERT, R., D. SPRAY, Y. GAO et al. 1999. Deficient assembly and function of gapjunctions in Trf1, a trafficking mutant of the human liver-derived cell line HuH-7.Hepatology 30: 740–747.

23. CHEN, Z-Q., D. LEFEBVRE, X-H. BAI et al. 1995. Identification of two regulatoryregions within the promoter region of the mouse connexin 43 gene. J. Biol. Chem.270: 3863–3868.

24. DEAN, D. & M. SANDERS. 1996. Ten years after: reclassification of steroid-responsivegenes. Mol. Endocrinol. 10: 1489–1495.

25. TROSKO, J. & C.C. CHANG. 2001. Mechanism of up-regulated gap junctional intercellularcommunication during chemoprevention and chemotherapy of cancer. Mutat. Res.480: 219–229.

26. YAMASAKI, H., V. KRUTOVSKIKH, M. MESNIL et al. 1999. Role of connexin (gapjunction) genes in cell growth control and carcinogenesis. C. R. Acad. Sci. Ser. III322: 151–159.

27. OMORI, Y., M.Z. DAGLI, K. YAMAKAGE & H. YAMASAKI. 2001. Involvement of gapjunctions in tumor suppression: analysis of genetically manipulated mice. Mutat.Res. 477: 191–196.

Documents

Intercellular Communication and Human Hepatocellular Carcinoma