[CANCER RESEARCH 60, 3978–3984, July 15, 2000]

Cimetidine Inhibits Cancer Cell Adhesion to Endothelial Cells and PreventsMetastasis by Blocking E-selectin Expression1

Ken-ichi Kobayashi, Sumio Matsumoto, Takashi Morishima, Takumi Kawabe, and Takashi Okamoto2

Department of Molecular Genetics, Nagoya City University Medical School, Nagoya 467-8601 [K-i. K., T. M., T. K., T. O.], and Department of Surgery, 2nd Teaching Hospital,Fujita Health University Medical School, Nagoya 454-8509 [K-i. K., T. M., S. M.], Japan

ABSTRACT

Although the beneficial effect of cimetidine on survival in cancer hasbeen clinically demonstrated in colorectal cancer patients, the mode ofaction of cimetidine has not been elucidated. In this report, we havedemonstrated for the first time that cimetidine can block the adhesion ofa colorectal tumor cell line to the endothelial cell monolayer in cell cultureand that it can suppress the metastasis of the tumor cell in a nude mousemodel. We also demonstrated that these antimetastasis effects of cimeti-dine might occur through down-regulation of the cell surface expression ofE-selectin on endothelial cells, a ligand for sialyl Lewis antigens on tumorcells. We found that the cimetidine-mediated down-regulation of E-selec-tin did not involve down-regulation of E-selectin mRNA or blocking of thenuclear translocation of nuclear factor kB, a transcriptional activator ofE-selectin gene expression. Because two other histamine type 2 receptorantagonists, famotidine and ranitidine, did not show any similar effect,these actions of cimetidine probably do not occur via blocking of thehistamine receptor. These observations support the idea that cancer me-tastasis can be blocked by cimetidine administration through blocking theadhesion of tumor cells to the endothelium when an interaction betweenE-selectin and sialyl-Lewis antigens plays a role.

INTRODUCTION

Cimetidine has been shown to improve the survival of patients withcolorectal cancer, melanoma, and renal cell cancer (1–9). Although itis not clear whether this effect of cimetidine on cancer is direct orindirect, it has been proposed that cimetidine may act by enhancingthe host immune response against tumor cells (10, 11) or by blockingthe cell growth-promoting activity of histamine in colon cancer andmelanoma cell lines (9, 12–14). Other H2R3 antagonists includingranitidine and famotidine did not have such an effect on the survivalof cancer patients (13–15), indicating that the anticancer actions ofcimetidine might not be mediated via histamine antagonism. There-fore, the mechanism of action by which cimetidine prolongs thesurvival of patients with various forms of cancer remains to beclarified.

In this study, we have attempted to examine the action of cimetidineby investigating its effect on the cancer cell adhesion to endothelialcells, one of the critical steps of cancer invasion and metastasis. Wehave also applied anin vivo metastasis model to confirm the effect ofcimetidine. We demonstrated previously (16) that the induction ofE-selectin in HUVECs by IL-1b induced the adhesion of cancer cellsexpressing the E-selectin ligand sialyl Lewis antigens (17, 18) to

HUVECs and that inhibitors of NF-kB activation including pentoxi-fylline, aspirin, andN-acetyl-L-cysteine could block the cancer celladhesion by inhibiting the IL-1b-mediated induction of E-selectin.Here we attempted to examine the effect of cimetidine by applying asimilar approach, and we found that cimetidine blocked cancer celladhesion to HUVECs. We also demonstrated that cimetidine canprevent liver metastasis in a nude mouse model in which tumor cellline HT-29 was injected intrasplenically.

MATERIALS AND METHODS

Cells. HUVECs were isolated from fresh human umbilical cords by treat-ment with 1 mg/ml collagenase and dispase and cultured in RPMI 1640supplemented with 50mg/ml endothelial cell growth supplement (UBI, LakeSuccess, NY), 1mg/ml epithelial growth factor (UBI), 292mg/ml L-glutamine,100 units/ml penicillin, 100mg/ml streptomycin, 5mg/ml heparin, and 15%fetal bovine serum. HUVECs were characterized by expression of von Will-ebrand factor using immunostaining with a specific antiserum (Zymed, SanFrancisco, CA). HUVECs between passage 4 and 8 were used in this study.HUVECs obtained from three unrelated donors were assessed in this study, andwe ensured that the observed effects were not batch specific. HUVECs from asingle donor were used within each experiment. Human tumor cell line HT-29expressing sialyl Lewis antigens (X and A) and derived from well-differenti-ated adenocarcinoma of the colon (19) was a gift from Dr. M. Iigoh (NationalCancer Center, Tokyo, Japan). This cell line was maintained in McCoy’smedium supplemented with 10% fetal bovine serum.

Cytotoxicity of Compounds. To evaluate the cytotoxicity of each com-pound, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assayswere performed by incubating HUVECs or HT-29 cells with various concen-trations of each compound according to the method described previously (20).

Reagents.Recombinant IL-1b was a gift from Otsuka Pharmaceutical Co.(Tokushima, Japan). Immunostaining and flow cytometry (FACScan) of E-selectin were carried out using a peptide-specific mouse monoclonal antibodyto human E-selectin as a primary antibody (R&D Systems, Minneapolis, MN)as reported previously (16, 21). For ELISA, mouse monoclonal antibodies tohuman E-selectin (CD62E; PharMingen, San Diego, CA) and human ICAM-1(CD54; Becton Dickinson, San Jose, CA) were used. The secondary antibodiesused for immunostaining and ELISA were FITC-conjugated rabbit antimouseIgG or rhodamine-conjugated goat antirabbit IgG (Cappel, Durham, NC) andhorseradish peroxidase-conjugated sheep antimouse IgG antibody (AmershamPharmacia Biotech, Buckinghamshire, United Kingdom), respectively. For theimmunostaining of NF-kB, the peptide-specific rabbit antibody to human p65was used as a primary antibody (22). Commercial antibodies to p65 and p50(Santa Cruz Biotechnology, Santa Cruz, CA) were also used for this study. Forblocking of the HT-29 cell adhesion to HUVECs, antibodies to E-selectin(CD62E) and to sialyl LewisX (CD15s; PharMingen) were used. Pentoxifyllineand aspirin were purchased from Sigma Chemical Co. (St. Louis, MO).Cimetidine, famotidine, and ranitidine were kindly provided by Smith KlineBeecham Japan (Tokyo, Japan), Yamanouchi (Tokyo, Japan), and Sankyo(Tokyo, Japan), respectively. Aspirin was dissolved in 99% ethanol. H2Rblockers and pentoxifylline were dissolved in PBS.

Monolayer Cell Adhesion Assay.HUVECs were stimulated with 10ng/ml IL-1b for 0–4 h in a 10-cm2 plate. HT-29 cells (13 105 cells/200ml/well) were added onto a semiconfluent monolayer culture of HUVECs,incubated for 20 min at 37°C with rotation at 120 rpm, and washed extensivelyto exclude nonspecific cell attachment. The number of attached cells wascounted directly under a microscope as reported previously (16). For antibody-mediated blocking of cell adhesion, the IL-1b-stimulated HUVECs were

Received 12/14/99; accepted 5/16/00.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby markedadvertisementin accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1 Supported in part by grants-in-aids for Scientific Research from the Ministry ofEducation, Science, Sports and Culture, the Ministry of Health and Human Welfare, Japanand the Japanese Health Sciences Foundation.

2 To whom requests for reprints should be addressed, at Department of MolecularGenetics, Nagoya City University Medical School, 1 Kawasumi, Mizuho-cho, Mizuho-ku,Nagoya 467-8601, Japan. Phone: 81-52-853-8204; Fax: 81-52-859-1235; E-mail:tokamoto@med.nagoya-cu.ac.jp.

3 The abbreviations used are: H2R, histamine type 2 receptor; NF-kB, nuclear factorkB; HUVEC, human umbilical vein endothelial cell; ICAM, intercellular adhesion mol-ecule; IL-1, interleukin 1; poly(A) RNA, polyadenylated RNA; MAPK, mitogen-activatedprotein kinase.

3978

Research. on April 25, 2021. © 2000 American Association for Cancercancerres.aacrjournals.org Downloaded from

incubated with antibody to either E-selectin (CD62E) or ICAM-1 (CD54; finaldilution of 1:400 for both antibodies) for 3 h at 37°C in humidified CO2

incubator, and then HT-29 cells were added. In the case of sialyl LewisX, theantibody (CD15s) was incubated with HT-29 for 1 h and added to theIL-1b-stimulated HUVECs.

Northern Blot Analysis. Poly(A) RNA was purified from HUVECstreated with various concentrations of cimetidine in the absence or presence ofIL-1b using a commercial mRNA isolation kit (Boehringer Mannheim, Mann-

heim, Germany). Twomg of poly(A) RNA were loaded onto each lane. Theelectrophoresis and blotting to Hybond-N membrane filter (Amersham Phar-macia Biotech) were performed as recommended by the supplier. Hybridiza-tion was carried out with a specific cDNA probe for E-selectin that did nothybridize with other selectin genes. A cDNA probe for glyceraldehyde-3-phosphate dehydrogenase was used as an internal control. Radiolabeling wascarried out using a commercial kit (Random Primer Labeling Kit; Stratagene,La Jolla, CA).

Fig. 1. Inhibition of HT-29 tumor cell adhesion toHUVECs by cimetidine. HT-29 cells (13 105 cells/200 ml/well) were added onto a semiconfluent mono-layer culture of HUVECs, incubated for 20 min at37°C with rotation at 120 rpm, and washed extensivelyto exclude nonspecific cell attachment.A, various con-centrations of H2R antagonists were added, and HT-29-HUVEC interaction was examined in the presenceof IL-1b (10 ng/ml). Phase-contrast microscopic pic-tures of representative experiments are shown.B,quantitation of HT-29-HUVEC interaction. The num-ber of HT-29 cells adhering to the HUVEC monolayerwas counted. Values were obtained from the results ofthree independent experiments;bars, SD.

3979

CIMETIDINE BLOCKS TUMOR CELL ADHESION TO ENDOTHELIUM

Research. on April 25, 2021. © 2000 American Association for Cancercancerres.aacrjournals.org Downloaded from

Immunostaining, Flow Cytometry, Immunoblotting, and ConfocalLaser Microscopy. Immunostaining was performed as reported previously(16, 21). Briefly, semiconfluent HUVECs on Lab-Tek tissue culture chamberslides (Nunc, Inc., Naperville, IL) were fixed with acetone for 10 min at220°C and washed three times with PBS. They were subsequently incubatedwith the primary antibody for 1 h at 37°C. After washing three times with PBScontaining 0.05% Triton X-100, they were incubated with the secondaryantibody for 20 min at 37°C. Slides were mounted with buffered glycerol forfluorescence (BX50; Olympus Co., Tokyo, Japan) or confocal laser micro-scopic examination (MRC-600UV; Bio-Rad, Hercules, CA). Primary andsecondary antibodies were diluted 1:100 in PBS containing 3% BSA. Foranalysis of cell surface E-selectin expression, unfixed cells were stained withthe same anti-E-selectin antibody and FITC-conjugated rabbit antimouse IgG,and cytofluorometric examination was carried out as described previously(Ref. 21; FACScan; Becton Dickinson). For immunoblotting, total cell extractsprepared from HUVEC cultures were analyzed with the mouse monoclonalanti-E-selectin antibody according to the method reported previously (21).

ELISA. An ELISA method was developed, according to the work ofWellicomeet al. (23), to quantitatively measure the amount of cell adhesionmolecules, including E-selectin and ICAM-1, expressed on the cell surface ofHUVEC cultures. Basically, the ELISA was performed at room temperaturewith three washes of 0.1% BSA in PBS between each step. HUVECs wereincubated for 1 h with the primary antibody, antihuman E-selectin monoclonalantibody (CD62E) or antihuman ICAM-1 monoclonal antibody (CD54). Afterwashing, the secondary antibody, horseradish peroxidase-conjugated sheepantimouse IgG antibody, was added and incubated with HUVECs for 30 min.The enzyme substrateO-phenylenediamine (1 mg/ml) and 0.03% hydrogenperoxide in citrate-phosphate buffer (pH 5.0) were then added. Color devel-opment was stopped with 2N sulfuric acid, and the absorbance of each wellwas read at 450 nm in a Titertek ELISA plate reader (SLT Lab Instruments,Salzburg, Austria). Test and control samples were examined in triplicates foreach experiment. The degree of specific antibody binding was calculated bysubtracting the mean negative control value (without the primary antibody),and the results were expressed as the fold activation of surface E-selectinexpression (mean6 SD). In each experiment, HUVECs at the fourth or thefifth passage were used to minimize the interassay variation. AlthoughHUVEC cultures from three unrelated donors were tested, there was no

Fig. 2. Inhibition of HT-29 tumor cell adhesion to HUVECs by E-selectin, sialylLewisx, and ICAM-1 antibody.A, E-selectin, sialyl Lewisx, and ICAM-1 antibody wereadded (at a final dilution of 1:400), and HT-29-HUVEC interaction was examined in thepresence of IL-1b (10 ng/ml). Phase-contrast microscopic pictures of representativeexperiments are shown.B, quantitation of HT-29-HUVEC interaction. The number ofHT-29 cells adhering to the HUVEC monolayer was counted. Values were obtained fromthe results of three independent experiments;bars, SD.

Fig. 3. Effects of cimetidine on E-selectin gene expression.A, Northern blot analysisof the E-selectin mRNA level in HUVECs pretreated with various concentrations ofcimetidine before stimulation with IL-1b (10 ng/ml). Positions of E-selectin (closedarrowhead) and glyceraldehyde-3-phosphate dehydrogenase (open arrowhead) as aninternal control are indicated.B, effects of cimetidine on the nuclear translocation ofNF-kB in response to IL-1b. Treatment of HUVECs with H2R antagonists was asdescribed in the legends to Figs. 1 and 2. Cells were fixed after 30 min of IL-1bstimulation, and immunostaining was performed using rabbit polyclonal antibody to thep65 subunit of NF-kB as a primary antibody.

3980

CIMETIDINE BLOCKS TUMOR CELL ADHESION TO ENDOTHELIUM

Research. on April 25, 2021. © 2000 American Association for Cancercancerres.aacrjournals.org Downloaded from

significant difference in the level of E-selectin or ICAM-1 expression. Withineach experiment, cells from a single donor were used.

Nude Mouse Model of Liver Metastasis.All of the experimental proce-dures were performed with the approval of the Animal ExperimentationCommittee of Nagoya City University Medical School. Specific pathogen-freeathymic BALB/c female mice of 3–4 weeks of age were kept under sterileconditions in a laminar flow room in cages with filter bonnets and fed asterilized mouse diet and water. To induce hepatic metastasis of tumor cells,HT-29 was injected intrasplenically. The mice were anesthetized by i.p.administration of pentobarbital (12mg/g body weight), and a small leftsubcostal incision was made to expose the spleen. HT-29 cells (13 106 or1 3 107 cells in different experiments) in 100ml of PBS were injected beneaththe splenic capsule using a 27-gauge needle. To examine the effect of H2Rantagonists, the mice were treated with 200, 50, or 10 mg/kg/day cimetidine orsaline (control) i.p. for 10 consecutive days, that is, for 5 days before HT-29injection and for 5 days after HT-29 injection. Subsequently, H2R antagonistsor saline (control) was administered every other day for an additional 9 weeksand 2 days (a total of 10 weeks after HT-29 cell implantation).

Quantification of Hepatic Metastasis in Nude Mice. Ten weeks afterHT-29 cell injection, animals were sacrificed, and the status of liver metastasiswas evaluated quantitatively. The livers were excised and cut into 2–3-mm-thick slices that were then fixed in ice-cold acetone for subsequent H&Estaining. The area of tumor nodules was measured with the aid of a videoimage processor (VIP-21C; Olympus Co.), and it was expressed as a percent-age of occupancy in the liver according to the method of Hiroseet al. (24).

Statistical Analysis. Differences between the results of experimentaltreatments were evaluated by means of the two-tailed Student’st test.Fisher’s exact probability test was used to assess the difference in theincidence of metastatic liver lesions. StatView-J 4.0.2 computer softwarewas applied for these analyses.

RESULTS

Suppression of Tumor Cell Adhesion to Endothelial Cells byCimetidine. To examine the effects of cimetidine on HT-29 tumorcell adhesion to HUVECs, a monolayer cell adhesion assay wascarried out. As shown in Fig. 1A, the adhesion of HT-29 cells toHUVECs was strongly induced on stimulation with IL-1b (10 ng/ml).The maximum induction of cell adhesion by IL-1b was obtained after4 h of stimulation. Using this model, we investigated the effects ofvarious H2R antagonists at a range of noncytotoxic concentrations[from 1024 to 1028

M (9, 13); also confirmed by 3-(4,5-dimethylthia-zol-2-yl)-2,5-diphenyltetrazolium bromide assay in this study]. Cul-tured HUVECs were pretreated with cimetidine or other H2R antag-onists 2 h before the addition of IL-1b. After 4 h of IL-1b stimulation,HT-29 cells were added, and the cell adhesion assay was performed.As shown in Fig. 1, pretreatment of HUVECs with pentoxifyllineblocked the adhesion with HT-29 cells, as reported previously (16).Similarly, this adhesion of HT-29 cells to HUVECs was inhibited bycimetidine in a dose-dependent manner. However, other H2R antag-onists, famotidine and ranitidine, had no inhibitory effect.

To confirm that the HT-29 adhesion to HUVECs was due to thecognate interaction between sialyl LewisX (on HT-29) and E-selectin(induced on HUVECs by IL-1b), specific antibodies were applied toblock these molecules before the cell adhesion assay. In Fig. 2, weincubated HUVECs that had been stimulated with IL-1b with theantibody to E-selectin or ICAM-1, and the monolayer cell adhesionassay was carried out. When HUVECs were preincubated with theantibody to E-selectin, the IL-1b-induced HT-29 adhesion toHUVECs was abolished (Fig. 2,A and B). Similarly, when HT-29cells were preincubated with the antibody to sialyl LewisX, the HT-29adhesion to HUVECs was also blocked. In contrast, preincubation ofHT-29 cells with anti-ICAM-1 antibody did not significantly reducecell adhesion. There was no significant effect of antibodies to E-selectin or sialyl LewisX antigens on the basal level of HT-29 adhe-sion to HUVECs (without IL-1b stimulation; data not shown). Thesefindings were consistent with previous observations by others (18, 25,26) and indicate that both E-selectin and sialyl LewisX are primarilyinvolved in our experimental cell adhesion assay system using HT-29cells and HUVECs.

Effects of Cimetidine on E-selectin Gene Expression and IL-1b-mediated NF-kB Activation. To investigate the effect of cimeti-dine on the E-selectin mRNA level, Northern blot analysis was carriedout (Fig. 3A). Poly(A) RNA purified from HUVEC cultures treatedwith various concentrations of cimetidine in the absence or presenceof IL-1b was hybridized with the E-selectin cDNA probe. As shownin Fig. 3A, IL-1b could induce E-selectin gene expression. However,even at the highest concentration, cimetidine could not significantlyreduce the E-selectin mRNA level. We also examined the effect ofcimetidine on the IL-1b-induced nuclear translocation of NF-kB (Fig.3B). In contrast to aspirin, which effectively blocked the nucleartranslocation of NF-kB as reported previously (16, 21), cimetidine or

Fig. 4. Inhibition of E-selectin induction on HUVECs by cimetidine.A, confocal lasermicroscopic examination of HUVECs on treatment with cimetidine at the indicatedconcentrations.B, E-selectin expression on the cell surface of HUVECs by cytofluorom-etry. Unfixed cells were stained with mouse monoclonal antibody to E-selectin. Beforestimulation with IL-1b, HUVECs were treated with 1026, 1025, or 1024 M cimetidine.Fractions (%) of cells with high-level E-selectin expression are indicated.

3981

CIMETIDINE BLOCKS TUMOR CELL ADHESION TO ENDOTHELIUM

Research. on April 25, 2021. © 2000 American Association for Cancercancerres.aacrjournals.org Downloaded from

other H2R antagonists did not block the nuclear translocation ofNF-kB elicited by IL-1b (Fig. 3A).

Cimetidine Blocked Cell Surface Expression of E-selectin In-duced by IL-1b. We then examined the effect of H2R antagonists onthe E-selectin protein level in HUVECs stimulated by IL-1b usingimmunohistochemistry and cytofluorometry (Fig. 4). Maximum in-duction of E-selectin by IL-1b (10 ng/ml) was obtained after 4 h oftreatment (Fig. 4A), which coincided with the HT-29 adhesion toHUVECs. Interestingly, a significant suppression of E-selectin induc-tion was observed on pretreatment of HUVECs with cimetidine in adose-dependent manner at noncytotoxic concentrations (from 1028 to1024

M), which again correlated with the effect of cimetidine on theadhesion of HT-29 cells to HUVECs (Fig. 1). It was noted by confocalmicroscopic examination that there was no significant change in theintracellular localization of E-selectin (Fig. 4A). Neither famotidinenor ranitidine showed any inhibitory effect on E-selectin levels. Therewas no change in the expression level of sialyl Lewis antigens on theHT-29 cell surface by cimetidine (data not shown). There was nosignificant effect of cimetidine on the ICAM-1 level in HUVECs ineither the presence or absence of IL-1b (data not shown). In spite ofour repeated attempts to identify the change of E-selectin proteinlevels by Western blotting using a number of commercially availableantibodies, we were not able to detect the monospecific band corre-sponding to E-selectin, a failure we believe to be due to the lowaffinity of the antibodies in solution or the presence of interactingproteins that interfered in the antibody reaction.

Quantitation of the Cell Surface E-selectin Expression byELISA. We thus developed a cell ELISA system, as described byWellicome et al. (23), to quantitate the E-selectin level on theHUVEC cell surface. Using this system, the effects of H2R antago-nists on E-selectin and ICAM-1 expression on HUVECs with orwithout IL-1b stimulation were examined. As shown in Fig. 5A, theE-selectin level on HUVECs was augmented by 6.06 0.36-fold onstimulation with IL-1b. When HUVECs were pretreated with ci-metidine 2 h before IL-1b stimulation, the extent of E-selectin induc-tion decreased to 4.96 0.35-, 1.96 0.06-, and 1.16 0.12-fold atcimetidine concentrations of 1028, 1026, and 1024

M, respectively(P , 0.01). The cell surface ICAM-1 expression was also augmentedby 6.376 0.43-fold on IL-1b treatment (Fig. 5B). However, ICAM-1expression was not significantly inhibited by cimetidine pretreatment.Neither famotidine nor ranitidine showed any inhibitory effect (Fig. 5,C andD).

Suppression of Liver Metastasis in Nude Mice by Cimetidine.The efficacy of cimetidine in blocking E-selectin protein expressionand the subsequent HT-29 cell adhesion to HUVECs prompted us toexamine the effects of cimetidine on liver metastasis using a nudemouse modelin vivo. Mice were inoculated with HT-29 (13 106 and1 3 107 cells in experiments 1 and 2, respectively) intrasplenically,and the effects of cimetidine on the incidence and the extent of livermetastasis were evaluated. Ten weeks after the injection of HT-29cells, mice were sacrificed, and the status of liver metastasis wasexamined. The areas of metastatic nodules in the liver were measuredfrom the excised liver slices using a video image processor according

Fig. 5. Quantitation of cell adhesion moleculeson HUVECs by cell ELISA.A, HUVECs weretreated with cimetidine in the presence or absenceof IL-1b (10 ng/ml) stimulation, and E-selectinlevels were measured.Asterisksindicate that thereduction of the E-selectin level was statisticallysignificant (P, 0.01).B, similar quantitation wasperformed for ICAM-1.C andD, effects of otherH2R antagonists, famotidine (C) and ranitidine(D), were evaluated on the level of E-selectin onHUVECs stimulated with IL-1b. Values were ob-tained from the results of three independent exper-iments;bars, SD.

3982

CIMETIDINE BLOCKS TUMOR CELL ADHESION TO ENDOTHELIUM

Research. on April 25, 2021. © 2000 American Association for Cancercancerres.aacrjournals.org Downloaded from

to the method developed by Hiroseet al. (24). As demonstrated inTable 1, cimetidine prevented the incidence of liver metastasis in adose-dependent manner. At the highest dose of cimetidine (dailydoses of 200 mg/kg), liver metastasis was completely inhibited when1 3 106 HT-29 cells were injected intrasplenically (none of the ninemice had detectable liver metastasis; Table 1, experiment 1), whereasthere was no notable difference in the size of the primary tumor in thespleen (data not shown). Famotidine or ranitidine at doses equivalentto that of cimetidine had no inhibitory effect on liver metastasis (datanot shown). A repeated experiment using a greater number of HT-29cells (13 107) for inoculation demonstrated similar effects of cimeti-dine on the prevention of liver metastasis (Table 1, experiment 2).Furthermore, the extent of liver metastasis (the percentage of occu-pancy of the metastatic nodules in the liver slices) was significantlydiminished in the cimetidine-treated animals (Table 2).

DISCUSSION

The clinical benefit of cimetidine for the prolonged survival ofcancer patients has been established previously (1–9). In this study,we investigated the effects of cimetidine using anin vitro cell culturesystem and anin vivo nude mice cancer metastasis model. Theseexperiments clearly demonstrated that cimetidine effectively blockedHT-29 adhesion to HUVECs by preventing E-selectin induction andthe trans-portal liver metastasis of HT-29 cells in nude mice.

E-selectin is considered to play a primary role in initiating theadhesion of cancer cells to vascular endothelial cells through itsinteraction with its specific ligand sialyl Lewis antigens (16–18,25–28). In a previous study (16), we reported that adhesion of tumorcell line QG90 to HUVECs was dependent on E-selectin expressionon the cell surface of HUVECs and was induced by IL-1b. Similarresults were reported with HT-29 cells by Dejanaet al. (26). In thesestudies, IL-1b was used to activate E-selectin on HUVECs to mimicthe local inflammatory response in the metastasized region (29, 30),and some tumor cells were reported to produce IL-1b (31, 32). In fact,expression of E-selectin in endothelium adjacent to the metastatictumor lesion was reported by others (33, 34). It is also known thatE-selectin is induced on ischemia-reperfusion injury (35), which maybe associated with tumor embolism during metastasis. Thus, it islikely that E-selectin is induced by such mechanisms in our nude micemodel.

We found that the effect of cimetidine did not appear to be at thelevel of E-selectin gene expression because the E-selectin proteinlevel was reduced without significantly reducing its mRNA level. Thefailure of cimetidine to block NF-kB nuclear translocation also sup-ported this possibility. Previous studies have indicated a variety ofunexpected actions of cimetidine. For example, cimetidine was re-ported to have direct growth-inhibitory effects on certain cancer celllines (1, 14) and direct stimulatory effects on lymphocyte function(10–12). It was also reported that histamine was released into theblood stream during the operation, presumably from tumor cells (36)or the adjacent mast cells (37), and that histamine interfered with the

host immune system (10, 11, 38). It was then suggested that histaminemight promote the growth of tumor cells (12, 14, 39). However, wedid not observe any cytotoxic or cytostatic effects of cimetidine onHT-29. In addition, we found that other H2R antagonists did not havethe same effect as cimetidine in our experimental systems, indicatingthat this action of cimetidine is probably not mediated via the H2R. Inaddition, although some reports have indicated that cimetidine mayhave antioxidant activity (40–42), and antioxidants have been shownto block E-selectin gene expression by blocking the activation cascadeof NF-kB (16), cimetidine did not block IL-1b-induced NF-kB acti-vation (Fig. 3B).

The action of cimetidine on E-selectin expression does not appearto involve the transcription step. Therefore, the experimental obser-vations obtained in this study suggest that cimetidine may blockE-selectin expression at a step after transcription. In this regard, it maybe worth noting that other regulatory molecules such as p38 MAPK,in addition to NF-kB, were also involved in IL-1b signaling (43, 44).Because it has been implicated that E-selectin expression requires p38MAPK (43), and p38 MAPK activates the expression of a number ofgenes at the level of posttranscription (44–46), the issue of whethercimetidine but not other H2R antagonists might interfere with suchsignaling mediators should be explored.

Our observations support the previous clinical findings that cimeti-dine increased the survival of cancer patients. It is likely that ci-metidine may block cancer cell adhesion to endothelial cells and thusprevent metastasis. In fact, we found in our recent randomized clinicalstudy that cimetidine treatment significantly reduced cancer metasta-sis and recurrence, thus resulting in the prolonged survival of patientswith colorectal cancer whose tumors had high levels of sialyl Lewisantigens.4 There were no effects when the tumor had no sialyl Lewisantigens or low levels of sialyl Lewis antigens. Additional analyses onthe actions of cimetidine are needed to identify a novel therapeutictarget against cancer aggression.

ACKNOWLEDGMENTS

We thank M. Futakuchi (Department of Pathology, Nagoya City UniversityMedical School, Nagoya, Japan) for help in the quantitation of liver metastasisusing a video image processor.

REFERENCES

1. Adams, W. J., and Morris, D. L. Short-course cimetidine and survival with colorectalcancer. Lancet,344: 1768–1769, 1994.

2. Matsumoto, S. Cimetidine and survival with colorectal cancer. Lancet,346: 115,1995.

3. Svendsen, L. B., Ross, C., Knigge, U., Frederiksen, H. J., Graversen, P., Kjaergard,J., Luke, M., Stimpel, H., and Sparso, B. H. Cimetidine as an adjuvant treatment incolorectal cancer. Dis. Colon Rectum,38: 514–518, 1995.

4. Hellstrand, K., Naredi, P., Lindner, P., Lundholm, K., Rudenstam, C. M.,Hermodsson, S., Asztely, M., and Hafstrom, L. Histamine in immunotherapy ofadvanced melanoma. Cancer Immunol. Immunother.,39: 416–419, 1994.

4 S. Matsumoto, K. Kobayashi, S. Umemoto, K. Suzuki, and T. Okamoto. Effects ofcimetidine on survival after curative operation of colorectal cancer: cimetidine increasessurvival of colorectal cancer patients with high levels of sialyl Lewis-X and sialylLewis-A epitope expression on tumor cells, manuscript in preparation.

Table 1 Frequency of liver metastasis 10 weeks after intrasplenic injection of HT-29cells in nude mice

Liver foci larger than 0.1 mm in diameter were not observed in a group of animalstreated with 200 mg cimetidine/kg/day.

No. ofcells

injectedControl(saline)

Cimetidine (mg/kg/day)

10 50 200

Experiment 1 13 106 15/17 (88%) 6/9 (67%) 5/9 (56%) 0/9 (0%)a

Experiment 2 13 107 15/17 (88%) 8/10 (80%) 4/9 (44%)b 4/10 (40%)b

a P , 0.001.b P , 0.05 (compared with control).

Table 2 Percentage of occupancy of the metastatic regions in the liver

Values shown represent the means (%)6 SD.

No. of cells injected Control (saline)

Cimetidine (mg/kg/day)

10 50 200

1 3 106 71.76 8.0 61.66 5.5a 51.26 5.3b 0b

a P , 0.05.b P , 0.001 (compared with control).

3983

CIMETIDINE BLOCKS TUMOR CELL ADHESION TO ENDOTHELIUM

Research. on April 25, 2021. © 2000 American Association for Cancercancerres.aacrjournals.org Downloaded from

5. Creagan, E. T., Ahman, D. L., Green, S. J., Long, H. J., Frytak, S., and Itri, L. M.Phase II study of recombinant leukocyte A interferon (IFN-RA) plus cimetidine indisseminated malignant melanoma. J. Clin. Oncol.,3: 977–981, 1992.

6. Sagaster, P., Micksche, M., Flamm, J., and Ludwig, H. Randomised study usingIFN-a versusIFN-a plus coumarin and cimetidine for treatment of advanced renalcell cancer. Ann. Oncol.,6: 999–1003, 1995.

7. Morris, D. L., and Adams, W. J. Cimetidine and colorectal cancer—old drug, newuse? Nat. Med.,1: 1243–1244, 1995.

8. Kelly, M. D., King, J., Cherian, M., Dwerryhouse, S. J., Finlay, I. G., Adams, W. J.,King, D. W., Lubowski, D. Z., and Morris, D. L. Randomized trial of preoperativecimetidine in patients with colorectal carcinoma with quantitative assessment oftumor-associated lymphocytes. Cancer (Phila.),85: 1658–1663, 1999.

9. Sasson, A. R., Gamagami, R., An, Z., Wang, X., Moossa, A. R., and Hoffman, R. M.Cimetidine: an inhibitor or promoter of tumor growth? Int. J. Cancer,81: 835–838,1999.

10. Hansbrough, J. F., Zapata-Sirvent, R. L., and Bender, E. M. Prevention of alterationsin postoperative lymphocyte subpopulations by cimetidine and ibuprofen. Am. J.Surg.,151: 249–255, 1986.

11. Adams, W. J., Morris, D. L., Ross, W. R., Lubowski, D. Z., and King, D. W.Cimetidine preserves non-specific immune function after colonic resection for cancer.Aust. N. Z. J. Surg.,64: 847–852, 1994.

12. Adams, W. J., Lawson, J. A., and Morris, D. L. Cimetidine inhibitsin vivo growth ofhuman colon cancer and reverses histamine stimulatedin vitro and in vivo growth.Gut, 35: 1632–1636, 1994.

13. Lawson, J. A., Adams, W. J., and Morris, D. L. Ranitidine and cimetidine differ intheir in vitro and in vivo effects on human colonic cancer growth. Br. J. Cancer,73:872–876, 1996.

14. Reynolds, J. L., Akhter, J., and Morris, D. L.In vitro effect of histamine andhistamine HI and H2 receptor antagonists on cellular proliferation of human malig-nant melanoma cell lines. Melanoma Res.,6: 95–99, 1996.

15. Hahm, K. B., Kim, W. H., Lee, S. I., Kang, J. K., and Park, I. S. Comparison ofimmunomodulative effects of the histamine-2 receptor antagonist cimetidine raniti-dine, and famotidine on peripheral blood mononuclear cells in gastric cancer patients.Scand. J. Gastroenterol.,30: 265–271, 1995.

16. Tozawa, K., Sakurada, S., Kohri, K., and Okamoto, T. Effects of anti-nuclear factorkB reagents in blocking adhesion of human cancer cells to vascular endothelial cells.Cancer Res.,55: 4162–4167, 1995.

17. Majuri, M. L., Niemela, R., Tiisala, S., Renkonen, O., and Renkonen, R. Expressionand function ofa 2,3-sialyl-anda 1,3/1,4-fucosyltransferases in colon adenocarci-noma cell lines: role in synthesis of E-selection counter-receptors. Int. J. Cancer,63:551–559, 1995.

18. Srinivas, U., Pahlsson, P., and Lundblad, A. E-selectin: sialyl Lewis, a dependentadhesion of colon cancer cells is inhibited differently by antibodies against E-selectinligands. Scand. J. Immunol.,44: 197–203, 1996.

19. Thomas, D. R., Philpott, G. W., and Jaffe, B. M. Prostaglandin E (PGE) control of cellproliferation in vitro: characteristics of HT-29. J. Surg. Res.,16: 463–465, 1974.

20. Merin, J. P., Matsuyama, M., Kira, T., Baba, M., and Okamoto, T.a-Lipoic acidblocks HIV-1 LTR-dependent expression of hygromycin resistance in THP-1 stabletransformants. FEBS Lett.,394: 9–13, 1996.

21. Sakurada, S., Kato, T., and Okamoto, T. Induction of cytokines and ICAM-1 byproinflammatory cytokines in primary rheumatoid synovial fibroblasts and inhibitionby N-acetyl-L-cysteine and aspirin. Int. Immunol.,8: 1483–1493, 1996.

22. Hayashi, T., Sekine, T., and Okamoto, T. Identification of a new serine kinase thatactivates NF-kB by direct phosphorylation. J. Biol. Chem.,268:26790–26795, 1993.

23. Wellicome, S. M., Thornhill, M. H., Thomas, D. S., Pitzalis, C., Lanchbury, J. S. S.,Panayi, G. S., and Haskard, D. O. A monoclonal antibody that detects a novel antigenon endothelial cells that is induced by TNF, IL-1 or LPS. J. Immunol.,144:2558–2565, 1990.

24. Hirose, M., Hasegawa, R., Kimura, J., Akagi, K., Yoshida, Y., Tanaka, H., Miki, T.,Satoh, T., Wakabayashi, K., Ito, N., and Shirai, T. Inhitory effects of 1-O-hexyl-2,3,5-trimethylhydroquinone (HTHQ), green tea catechins and other antioxidants on2-amino-6-methyldipyrido[1,2-a:39,29-d]imidazole (Glu-P-1)-induced rat hepatocar-cinogenesis and dose-dependent inhibition by HTHQ of lesion induction by Glu-P-1or 2-amino-3,8-dimethylimidazo[4,5-f ]quinoxaline (MeIQx). Carcinogenesis (Lond.),16: 3049–3055, 1995.

25. Takada, A., Ohmori, K., Yoneda, K., Hasegawa, A., Kiso, M., and Kannagi, R.Contribution of carbohydrate antigens sialyl Lewis A and sialyl Lewis X to adhesionof human cancer cells to vascular endothelium. Cancer Res.,53: 354–361, 1993.

26. Dejana, E., Martin, P. I., Lauri, D., Bernasconi, S., Bani, M. R., Garofalo, A.,Giavazzi, R., Magnani, J., Mantovani, A., and Menard, S. Endothelial leukocyte

adhesion molecule-1-dependent adhesion of colon carcinoma cells to vascular endo-thelium is inhibited by an antibody to Lewis fucosylated typeI carbohydrate chain.Lab. Investig.,66: 324–330, 1992.

27. Takada, A., Ohmori, K., Takahashi, N., Tsuyuoka, K., Yago, K., Zenita, K.,Hasegawa, A., and Kannagi, R. Adhesion of human cancer cells to vascular endo-thelium mediated by a carbohydrate antigens, sialyl Lewis A. Biochem. Biophys. Res.Commun.,179: 713–719, 1991.

28. Kunzendorf, U., Kruger, K. S., Notter, M., Hock, H., Walz, G., and Diamantstein, T.A sialyl-Le(x)-negative melanoma cell line binds to E-selectin but not to P-selectin.Cancer Res.,54: 1109–1112, 1994.

29. Montgomery, K. F., Osborn, L., Hession, C., Tizard, R., Goff, D., Vassallo, C., Tarr,P. I., Bomsztyk, K., Lobb, R., Harlan, J. M., and Pohlman, T. H. Activation ofendothelial-leukocyte adhesion molecule 1 (ELAM-1) gene transcription. Proc. Natl.Acad. Sci. USA,88: 6523–6527, 1991.

30. Whelan, J., Ghersa, P., Huijsduijnen, R. H., Gray, J., Chandra, G., Talabot, F., andDeLamarter, J. F. An NF-kB-like factor is essential but not sufficient for cytokineinduction of endothelial leukocyte adhesion molecule 1 (ELAM-1) gene transcription.Nucleic Acids Res.,19: 2645–2653, 1991.

31. Burrows, F. J., Haskard, D. O., Hart, I. R., Marshall, J. F., Selkirk, S., Poole, S., andThorpe, P. E. Influence of tumor-derived interleukin 1 on melanoma-endothelial cellinteractionsin vitro. Cancer Res.,51: 4768–4775, 1991.

32. Kaji, M., Ishikura, H., Kishimoto, T., Omi, M., Ishizu, A., Kimura, C., Takahashi, T.,Kato, H., and Yoshiki, T. E-selectin expression induced by pancreas-carcinoma-derived interleukin-1a results in enhanced adhesion of pancreas-carcinoma cells toendothelial cells. Int. J. Cancer,60: 712–717, 1995.

33. Ye, C., Kiriyama, K., Mitsuoka, C., Kannagi, R., Ito, K., Watanabe, T., Kondo, K.,Akiyama, S., and Takagi, H. Expression of E-selectin on endothelial cells of smallveins in human colorectal cancer. Int. J. Cancer,61: 455–460, 1995.

34. Suzuki, Y., Ohtani, H., Mizoi, T., Takeha, S., Matsuno, S., and Nagura, H. Celladhesion molecule expression by vascular endothelial cells as an immune/inflamma-tory reaction in human colon carcinoma. Jpn. J. Cancer Res.,86: 585–593, 1995.

35. Stanimirovic, D. B., Wong, J., Shapiro, A., and Durkin, J. P. Increase in surfaceexpression of ICAM-1, VCAM-1 and E-selectin in human cerebromicrovascularendothelial cells subjected to ischemia-like insults. Acta Neurochir. Suppl.,70:12–16, 1997.

36. Garcia-Caballero, M., Nunezed, X., Castro, I., Kusche, J., and Vora-Thorbeck, L.Histamine metabolism in human breast and colorectal cancer: its effects on other hosttissues. Adv. Biosci.,89: 273–287, 1993.

37. Jeziorska, M., Haboubi, N. Y., Schofield, P. F., and Wooley, D. E. Mast celldistribution and tumour cell proliferation in colonic carcinoma. Gut,34 (Suppl.):S5,1993.

38. Katoh, J., Tsuchiya, K., Sato, W., Nakajima, M., and Iida, Y. Cimetidine andimmunoreactivity. Lancet,348: 404–405, 1996.

39. Weese, J., Ottery, F., and Emoto, S. Do operations facilitate tumor growth? Anexperimental model in rats. Surgery,100: 273–277, 1986.

40. Kimura, E., Koike, T., Shimizu, Y., and Kodama, M. Complexes of the histamineH2-antagonist cimetidine with divalent and monovalent copper ions. Inorg. Chem.,25: 2242–2246, 1986.

41. Uchida, K., and Kawakishi, S. Cimetidine anti-ulcer drug as a powerful hydroxylradical scavenger. Agric. Biol. Chem.,54: 2485–2487, 1990.

42. Lapenna, D., De Gioia, S., Mezzetti, A., Grossi, L., Festi, D., Marzio, L., andCuccurullo, F. H2-receptor antagonists are scavengers of oxygen radicals. Eur.J. Clin. Investig.,24: 476–481, 1994.

43. Read, M. A., Whitley, M. Z., Gupta, S., Pierce, J. W., Best, J., Davis, R. J., andCollins, T. Tumor necrosis factora-induced E-selectin expression is activated by thenuclear factor-kB and c-JUN N-terminal kinase/p38 mitogen-activated protein kinasepathways. J. Biol. Chem.,272: 2753–2761, 1997.

44. Pietersma, A., Tilly, B. C., Gaestel, M., Jong, N. D., Lee, J. C., Koster, J. F., andSluiter, W. p38 mitogen activated protein kinase regulates endothelial VCAM-1expression at the post-transcriptional level. Biochem. Biophys. Res. Commun.,230:44–48, 1997.

45. Caivano, M. Role of MAP kinase cascades in inducing arginine transporters and nitricoxide synthetase in RAW264 macrophages. FEBS Lett.,429: 249–253, 1998.

46. Miyazawa, K., Mori, A., Miyata, H., Akahane, M., Ajisawa, Y., and Okudaira, H.Regulation of interleukin-1b-induced interleukin-6 gene expression in human fibro-blast-like synoviocytes by p38 mitogen-activated protein kinase. J. Biol. Chem.,273:24832–24838, 1998.

3984

CIMETIDINE BLOCKS TUMOR CELL ADHESION TO ENDOTHELIUM

Research. on April 25, 2021. © 2000 American Association for Cancercancerres.aacrjournals.org Downloaded from

2000;60:3978-3984. Cancer Res   Ken-ichi Kobayashi, Sumio Matsumoto, Takashi Morishima, et al.   and Prevents Metastasis by Blocking E-selectin ExpressionCimetidine Inhibits Cancer Cell Adhesion to Endothelial Cells

  Updated version

  http://cancerres.aacrjournals.org/content/60/14/3978

Access the most recent version of this article at:

   

   

  Cited articles

  http://cancerres.aacrjournals.org/content/60/14/3978.full#ref-list-1

This article cites 45 articles, 12 of which you can access for free at:

  Citing articles

  http://cancerres.aacrjournals.org/content/60/14/3978.full#related-urls

This article has been cited by 16 HighWire-hosted articles. Access the articles at:

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/60/14/3978To request permission to re-use all or part of this article, use this link

Research. on April 25, 2021. © 2000 American Association for Cancercancerres.aacrjournals.org Downloaded from