1989;49:1850-1856. Published online April 1, 1989.Cancer Res Renato Garcea, Lucia Daino, Rosa Pascale, et al. Remodeling and Apoptosis-Adenosyl-l-methionine in Rat Liver Carcinogenesis: Role of

SInhibition of Promotion and Persistent Nodule Growth by   

  

Updated Version http://cancerres.aacrjournals.org/content/49/7/1850

Access the most recent version of this article at:

  

Citing Articles http://cancerres.aacrjournals.org/content/49/7/1850#related-urls

This article has been cited by 10 HighWire-hosted articles. Access the articles at:

  

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

SubscriptionsReprints and

.pubs@aacr.orgDepartment atTo order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions.permissions@aacr.orgDepartment at

To request permission to re-use all or part of this article, contact the AACR Publications

American Association for Cancer Research Copyright © 1989 on July 10, 2011cancerres.aacrjournals.orgDownloaded from

(CANCER RESEARCH 49, 1850-1856, April 1, 1989]

Inhibition of Promotion and Persistent Nodule Growth by S-Adenosyl-i -methioninein Rat Liver Carcinogenesis: Role of Remodeling and Apoptosis1

Renato Garcea, Lucia Daino, Rosa Pascale, Maria M. Simile, Marco Puddu, Serenella Frassetto, PatriziaCozzolino, Maria A. Seddaiu, Leonardo Caspa, and Francesco Feo2

Istituto dìPatologia generale dell'Università di Sassari, Via P. Manzella 4, Sassari, 07100 Italy

ABSTRACT

The resistant hepatocyte model (initiation/selection) and the triphasicmodel (initiation/selection followed by phénobarbital,for a maximum of16 weeks) were compared for their ability to generate enzyme-alteredfoci (EAF) and nodules in the liver of Wistar rats initiated by diethylni-trosamine. The effects of 5-adenosyl-L-methionine (SAM) on the development of preneoplastic tissue was tested in these experimental models.In the absence of phénobarbital(PB), EAF and early nodules (EN) wentthrough a phase of rapid growth, between 4 and 9 weeks after initiation,to a phase in which progressive decrease in number and size occurred.By the 26th week only a few remodeling EAF and nodules were found.In PB-treated rats a rapid increase in the percentage of liver occupied by

EAF and EN, up to the 9th week after initiation, was followed by aperiod of slow growth (from the 9th to the 20th week) and then, after PBwithdrawal (20th week), by a drop in the number and size of EAF andEN. However, at the 26th week actively growing nodules with a lowtendency to spontaneous remodeling (persistent nodules) developed. EAFand EN showed a high DNA synthesis 5 weeks after initiation. Thereafter, progressive decline in DNA synthesis, coupled with remodelingand decrease in number of biochemical markers, was seen both in theabsence and, even though to a lesser extent, in the presence of PB,indicating that preneoplastic lesions became increasingly insensitive toPB. Relatively few apoptotic bodies could be observed in EAF and ENduring PB treatment. After PB withdrawal, decrease in growth potentialwas coupled with increase in apoptotic bodies. In contrast, in persistentnodules relatively high apoptosis occurred which partially counterbalanced high DNA synthesis. Administration of SAM for a maximum of16 weeks, starting at the 4th week after initiation, caused a great decreasein number and size of EAF and EN, associated with inhibition of DNAsynthesis, high cell death by apoptosis, high remodeling, and loss ofbiochemical markers, in preneoplastic lesions of both PB-treated anduntreated rats. A 1-8-week SAM treatment, started after the develop

ment of persistent nodules, caused a great regression of nodular lesions,coupled with a sharp fall in DNA synthesis and increase in apoptosis. Itis suggested that inhibition by SAM of the development of preneoplastictissue is linked to a shift of the equilibrium between cell production andcell death in favor of cell death. This phenomenon and differentiation ofputative preneoplastic cells to normal appearing hepatocytes could accelerate disappearance of preneoplastic lesions and makes phenotype ofpersistent nodules highly unstable.

INTRODUCTION

Cell proliferation plays an essential role in the initiation (1-3), promotion (4-7), and progression (8) steps of chemical

carcinogenesis. The size of proliferating tissues largely dependson the equilibrium between cell growth and cell loss. Extensiveremodeling (8-10) but no or very limited cell death (11) has

been found in early liver lesions of rats subjected to initiation/

Received 8/15/88; revised 12/7/88; accepted 12/14/88.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 inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

' This research was supported by grants from the Consiglio Nazionale delle

Ricerche, Progetto Finalizzato Oncologia (Grant 87.01281.44), AssociazioneItaliana Ricerca sul Cancro, and Ministero Pubblica Istruzione (Program 60%).

3To whom requests for reprints should be addressed.

selection treatments (RH3 model) (12). Preneoplastic lesions

induced in experimental models based on prolonged administration of promoters to initiated rats exhibit very low cell deathby apoptosis (13) and remodeling (14). In contrast, a relativelyhigh necrogenic index which, however, only partially counterbalanced the high proliferative rate, has been found in PN, thedevelopment of which precedes that of hepatocarcinomas (11).

Previous work from our laboratory (15-18)has shown a greatfall in hepatic SAM levels during DENA-induced liver carcinogenesis of rats subjected to the initiation/selection treatments(12) followed or not by a 16-week treatment with PB. Thereconstitution of the SAM liver pool, by injection of exogenousSAM, is associated with a fall in the amount of EAF and earlyreversible nodules and the prevention of the development oflate, PN, and hepatocellular carcinomas (16, 17). This effect islargely linked to inhibition of DNA synthesis in preneoplasticlesions (15-18). However, no data are presently available onthe phenotypic changes of putative preneoplastic hepatocytesduring the promotion and progression steps of liver carcinogenesis as a consequence of SAM treatment. The present study isconcerned with remodeling, instability of the biochemical phenotype, and apoptosis, in early lesions (EAF and EN) as wellas in late PN, in the liver of SAM-treated rats. A preliminaryaccount of some of the results in this paper has already beengiven (19).

MATERIALS AND METHODS

Animals and Diets. Male Wistar rats (bred in our laboratory) andmale Fischer 344 rats (purchased from Charles River; The InternationalStandards, Calco, Como, Italy) were used. The rats (160-180 g at thebeginning of the experiment) were housed, three per cage, in suspendedwire-bottomed cages, in a constant temperature (22°C)and humidity

(55%) environment, with a 6 a.m. to 6 p.m. photoperiod. They wereplaced on a standard 26% casein diet (Piccioni, Brescia, Italy) withwater ad libitum. Rats received a single 150-mg/kg i.p. dose of DENA(initiation) and, 2 weeks later, were fed for 2 weeks a standard dietcontaining 0.03% (Wistar rats) or 0.02% 2-AAF (F344 rats), and apartial hepatectomy was performed at the midpoint of this time period(selection). Preliminary experiments showed that 150 mg/kg of DENAare necrogenic for the liver of Wistar rats. In addition, DNA synthesisof EAF was poorly affected by a 7-day feeding of 0.03% 2-AAF, whichinstead largely inhibited DNA synthesis in surrounding hepatocytes.At the end of 2-AAF feeding, Wistar rats were divided into 6 groups(groups 1-6 of Fig. 1), and F344 rats were divided into 2 groups (groups7 and 8). Groups 1, 2, 7, and 8 were given a basal diet, while groups 3-6 were fed a basal diet containing 0.05% PB (triphasic model) (20), fora maximum of 16 weeks, before being placed on basal diet. All ratgroups were killed as indicated in Fig. 1. Rat groups 1, 3, 5, and 7 werecontrols; all other groups were treated with SAM. SAM treatmentswere started at the end of 2-AAF feeding for Wistar rat groups 2 and4 and was continued for a maximum of 16 weeks. They were started at

3The abbreviations used are: RH, resistant hepatocyte; 2-AAF, 2-acetylami-nofluorene; AB, apoptotic body; DENA, diethylnitrosamine; EAF, enzyme-altered foci; EN, early nodules; GGT, -y-glutamyltranspeptidase; G6Pase, glucose-6-phosphatase; G6PD, glucose ft-phosphate dehydrogenase; LI, labeling index;5'-MTA, S'-methylthioadenosine; ODC, ornithine decarboxylase; PN, persistentnodules; PB, phénobarbital;SAM, S-adenosyl-L-methionine; SAH, 5-adenosyl-homocysteine; H & E, hematoxylin and eosin.

1850

American Association for Cancer Research Copyright © 1989 on July 10, 2011cancerres.aacrjournals.orgDownloaded from

INHIBITION OF PROMOTION AND NODULE GROWTH BY SAM

•¿�EUSI 1 2 11 11 U 11 II 11 H 11 1»21 14 If H »T H 14

' ' I

Fig. I. Schematic representation of experimental model. A, time of sacrifice.Numbers in parentheses, number of animals/group. PH, partial hepatectomy; •¿�2-AAF; G, basal diet; g, basal diet + SAM; B. PB; Q, PB + SAM.

the 26th and 12th weeks for rat groups 6 and 8, respectively, and werecontinued for a maximum of 11 weeks. SAM, in the relatively stableform (21) of sulfate ^-toluene sulfonate (BioResearch, Liscate, Milan,Italy), was injected i.m. as freshly prepared solutions containing 0. l MSAM, 0.045 MNaOH, and 0.4 Mlysine (final pH 6.9). Controls receivedthe same amounts of a solution containing 0.045 M NaOH and 0.4 Mlysine, brought to pH 6.9 with an equimolar mixture of sulfuric acidand /7-toluenesulfonic acid. Thin layer chromatography analysis (15)revealed that SAM preparations were free of methionine. High performance liquid chromatography analyses (see below) showed that SAH, 5'-

MTA, and other unidentified contaminants did not exceed 1.7%. Allrat groups received 6 daily SAM doses of 64 ^mol/kg. No differencesin food intake were observed among the various groups of animals(range for all rat groups, 10.2-12.8 g/100 g/day).

When used for analytical determinations, PN were separated fromsurrounding liver by isolating liver cells with a collagenase method (22)and collecting nodules on a nylon filter (23). By persistent nodules(hyperplastic nodules, neoplastic nodules) we mean focal proliferating,GGT-positive lesions, with a diameter of 2-5 mm and a low tendencyto spontaneous regression (24). These were distinguished from carcinoma according to the published criteria (25). Lesions smaller than aliver lobule, mainly visible microscopically, are referred to as EAF,while lesions about 0.5-2 mm in diameter, developing in the tirsi weeksafter initiation (see "Results"), are referred to as EN.

Histology and Histochemistry. Small pieces of liver taken from eachlobe were frozen in isopentane at -147°C, embedded in Paraplast, and

then cut serially in a cryostat into 5-Mm-thick sections. The sectionswere fixed in cold acetone and then used for histochemistry. The firstsection was assayed for GGT, the second for G6PD, and the third forG6Pase as described (26). Alternatively, acetone-fixed material wasprocessed for embedding in paraffin, serially sectioned, and used for H& E staining and for GGT histochemistry. Morphometric analysis wascarried out by scanning 9-10 liver sections per rat with a Leitz Diaplanmicroscope connected, by a telecamera, with a MOP Videoplan computerized image analyzer (Kontron Electronic, Echig, Federal Republicof Germany). Remodeling EAF and nodules were identified as areaslacking uniformity for GGT histochemical reaction and exhibitingirregular boundaries with surrounding liver and relatively low LI (8-10). To determine LI the rats were given [3H]dThd (90 Ci/mmol;

Amersham International Pic., Amersham, United Kingdom) i.p., at thedose of 0.5 fiCi/g body weight every 6 h for 48 h before killing. Theslices were fixed and processed for histochemistry and then coated withKodak NTB2 emulsion and stored for 5 weeks in the dark at 4°C.After

development the hepatocytes were counterstained with hematoxylin.In order to evaluate phenotypic complexity of EAF and EN, the

transections were examined on the monitor of the image analyzer. Asheet of vellum was placed onto the monitor surface, and the transectionoutlines and the EAF and EN, revealed by the histochemical testrepresented by each slide, were traced onto the vellum sheet, using adifferent color for each tracing. The superimposed images of the serialsections stained for individual markers were then digitized by an imagedigitizer connected with the image analyzer, equipped with a computerprogrammed to yield the foci number and surface area for a givenphenotype or for individual foci independently of the phenotypes represented (see below). The biochemical characteristics of each focus ornodule may not quantitatively involve all the cells within the focus (27).In Fig. 4, the data under "any" represent data obtained by the markers

listed regardless of whether or not other markers also scored in thesame transection. The other columns of data refer to the presence of 1,2, or 3 markers. However, when a biochemical alteration occurredwithin the cell population of larger foci or nodules the surface area ofeach cell population was considered as a fraction of the total population,carrying the same biochemical alteration, in the total slide. In contrast,in Fig. 5 the percentage of liver occupied by cells carrying the variousbiochemical alterations has been calculated independently of the surfacearea covered, in each lesion, by each biochemical marker.

Sections stained with H & E were used for counting ABs. Clear cell,eosinophilic, and mixed cell foci were identified according to publishedcriteria (28); their correspondence with GGT-positive foci was assessedin the serial sections.

Enzyme Assay. ODC activity was determined in 30,000 x g super-natants of liver and PN homogenates by measuring the release of MCO2from [l-'4C]ornithine as published previously (15).

Analytical Methods. High performance liquid chromatography determination of SAM, SAH, and 5'-MTA was performed as published

previously (17) on HC1Û4extracts of liver or PN. Their quantities weredetermined by comparing the area of the peak in the tissue extract withthat of standard solutions. Proteins were determined as publishedpreviously (29).

RESULTS

Development of Putative Preneoplastic Lesions. EAF (0.1 -0.2mm in diameter) were first observed, in Wistar rats subjectedto the initiation/selection treatments, as early as 3.5 weeks afterinitiation (Fig. 2). Later the foci increased in number andsurface area up to the 5th week and thereafter in surface areaalone (not shown). At the 5th week the first nodules appeared.The percentage of liver parenchyma occupied by GGT-positivefoci and EN progressively increased, reaching its maximum(22%) at the 9th week, after which EAF and EN exhibited aprogressive decrease in number and surface area. By the 26thweek, the percentage of liver occupied by small GGT-positivelesions (mostly EAF) decreased to about 9%. PB administration, at the end of 2-AAF feeding, greatly enhanced the development of EAF and EN, which increased in size, not in number,so that about 35% of liver was occupied by these lesions at 9weeks after initiation. Thereafter, liver occupied by EAF andEN did not increase significantly, and a sharp decrease occurredupon withdrawal of PB. At the 26th week EAF occupied only

ïÃ3̄0

S? 20

ë§

II"

i s

7 9TIME («eeki)

Fig. 2. Development of putative preneoplastic lesions in liver of Wistar ratssubjected to initiation/promotion treatments and SAM. Animals of groups 1-4were used for these experiments. At least SO lesions in 3 different slides havebeen counted for each point. Data of the 26th week do not consider persistentnodules. Each point represents mean ±SD (bars) of S rats. Absolute values(number/mmVsurface area in mm2). Without PB, 5 weeks, 1.3 ±0.04/0.3 ±0.1;9 weeks, 1.2 ±0.2/0.6 ±0.1; 26 weeks, 1.1 ±0.03/0.15 ±0.06. Without PB,with SAM: 5 weeks, 0.6 ±0.03/0.07 ±0.02; 9 weeks, 0.8 ±0.06/0.2 ±0.06; 26weeks, 0.4 ±0.02/0.1 ±0.01. With PB: 5 weeks, 1.3 ±0.04/0.6 ±0.1; 9 weeks,1.3 ±0.02/1.2 ±0.3; 26 weeks, 1.1 ±0.02/0.6 ±0.03. With PB plus SAM: 5weeks, 0.5 ±0.02/0.1 ±0.00; 9 weeks, 1.2 ±0.03/0.68 ±0.08; 26 weeks, 0.8 ±0.01/0.4 ±0.01. A, rat group 1 (without PB); D, rat group 3 (with PB); A, ratgroup 2 (without PB, with SAM); •¿�rat group 4 (with PB and SAM). Arrows onleft and on right, start and end of PB and/or SAM treatments, respectively, t test:without PB versus with PB and without SAM versai with SAM, at least P < 0.05for all time points.

1851

American Association for Cancer Research Copyright © 1989 on July 10, 2011cancerres.aacrjournals.orgDownloaded from

INHIBITION OF PROMOTION AND NODULE GROWTH BY SAM

13% of liver, while numerous PN were seen (see below). SAMadministration caused a great decrease in the percentage ofGGT-positive foci, in either the presence or the absence of PB.This decrease was seen in both the number and surface area ofGGT-positive lesions.

Remodeling. The data in Fig. 3 show that in the absence ofPB, remodeling, which is very low at the 5th week after startingthe experiment, progressively increased thereafter. At the 20thand 26th weeks 40-50% of EAF and EN showed a nommiformpattern and irregular outlines. PB partially inhibited remodelingof GGT-positive foci. Data relative to the 26th week, in PB-treated rats, concern only microscopic lesions, not PN, whichunderwent no or very low remodeling (see below). SAM highlystimulated remodeling both in PB-treated and untreated rats.Remodeling inhibition by PB and its activation by SAM werealso observed in foci or EN expressing high G6PD or low orno G6Pase (data not shown).

The data in Fig. 3 also show that 5 weeks after initiation,relatively high DNA synthesis characterized uniform and non-uniform GGT-positive lesions in both PB-treated and untreatedrats. However, LI was significantly lower in nonuniform thanin uniform lesions. As expected, PB stimulated DNA synthesisof both uniform and nonuniform lesions. A large drop in LIoccurred in uniform and nonuniform lesions 2 weeks later (7thweek), followed by further relatively low decreases in the following weeks. A slower decrease occurred in PB-treated rats.However, LI remained slightly but significantly lower in non-uniform than in uniform lesions throughout the duration of theexperiment, both in the absence and in the presence of PB. Atthe 26th week, GGT-positive cells of both PB-treated anduntreated rats exhibited very low DNA synthesis, which was,however, still 2.5-3-fold higher than in surrounding parenchyma [LI in surrounding liver, 0.28 ±0.05 (SD); n = 5]. SAMinhibited LI of both uniform and nonuniform foci, especiallybetween the 5th and the 9th week, in PB-treated rats and at the5th week in untreated rats.

Levels of Phenotypic Complexity. In some experimental

IMS

'5*ii

50

ii2,1U

NWithoutSAM nCWith

SAM •¿�[1

rfae•1•fj•fa4•Hiìf

Fig. 3. Effect of SAM on remodeling of GGT-positive liver lesions. The samerat groups and conditions as in Fig. 2. Nonuniform lesions have been calculatedas a percentage of total GGT-positive EAF plus EN. Labeling index was determined by counting 8000 hepatocytes/liver in GGT-positive lesions. Data aremeans ±SD (bars) of 5 rats for each time point, t test: remodeling: uniform (Õ')

versus nonuniform (NU) (with/without PB) at least P < 0.05. LI: SAM-treatedversus untreated, without PB, at least /' < 0.01 between weeks 5 and 18; uniformversus nonuniform, with PB, at least P < 0.05 at all weeks. Left without PB; right,with PB.

models, putative preneoplastic lesions exhibiting a relativelyslow growth rate are generally associated with low levels ofphenotypic complexity (number and identities of biochemicalmarkers per focus) (27, 30). In Fig. 4 the percentages of putativepreneoplastic tissue exhibiting individual biochemical markersor combinations of different markers have been considered asfractions of total liver cell population carrying the same biochemical alteration (see "Materials and Methods"). Between

the 5th and the 26th week, the preneoplastic lesions exhibitinga high GGT, either as the only marker or associated with othermarkers (Column I under "any"), clearly occupied larger ex

tents of liver parenchyma, in both PB-treated PB-treated anduntreated rats, than the lesions expressing the other biochemical markers (Columns 2 and 3 under "any"). After 9 weeks

putative preneoplastic cells exhibiting biochemical alterationsof GGT/G6PD and GGT/G6PD/G6Pase increased, with respect to the 5th week, more than those showing only onebiochemical marker. This behavior was particularly evident inPB-treated rats where the cells exhibiting altered GGT/G6PDor GGT/G6PD/G6Pase occupied the largest percentage of liverparenchyma. In animals not treated with PB, about 8% of liverwas occupied by EAF and EN exhibiting only one marker andabout 15% by lesions exhibiting 2-3 markers. In PB-treatedrats these figures were 12 and 24%, respectively. A similarbehavior was observed at the 20th week (data not included inFig. 4). However, at this time the part of the liver occupied bylesions with one or more than one marker was 10 and 12%,without PB, and 18% and 21% in its presence. At the 26thweek, the tendency to a preferential development of liver cellswith higher biochemical complexity levels was less evident,particularly in the rats that did not receive PB. A great decreasein putative preneoplastic cells with more than one biochemicalmarker occurred in SAM-treated rats, particularly at the 9thweek (both with or without PB), a period of time which coincides with maximal biochemical heterogeneity of preneoplasticlesions. The same period of time was taken into considerationto evaluate if phenotypic heterogeneity could be influenced byremodeling (Fig. 5). Complexity level was lower in nonuniformthan in uniform lesions, especially in the absence of PB. Thedisappearance of biochemical markers in nonuniform foci followed the sequence G6Pase, G6PD, GGT, so that the majorityof nonuniform lesions exhibiting only one marker was GGTpositive. SAM caused a large decrease in the percentage of bothuniform and nonuniform lesions with three markers and anincrease in those exhibiting only one marker. No significantchanges or very little change occurred in the lesions showingtwo biochemical markers. This could reflect an equilibriumbetween the fluxxomplexity level 2 to complexity level 1, and

S S? a

I

e.

Bao

I/I AC BCABC

_ Q ta - aC «/IAC I/O/B/C

_ _C »B «C B CAB C

ÌÙQitB C*Un,.>

iC *a

= 0c=IC ABHC..ODI

C 6 B A/C_

aBCA BC_0B

C A B C

ÙB fi __>•£)a _ caC A B C A B >/C B CA B C

Fig. 4. Effect of SAM on the development of putative preneoplastic lesionsexhibiting different biochemical alterations. The same rat groups and conditionsas in Fig. 2. Data are means ±SD (bars) of 5 rats for each week. A, GGT; B,G6PD; C, GoPase. In A, B, and C, the ANY columns represent all the lesionsscored by GGT (A), G6PD (B), and GoPase (C) alone or in combination. / test:with SAM (•)irra/v without SAM (II) (with/without PB) at least P < 0.05.

1852

American Association for Cancer Research Copyright © 1989 on July 10, 2011cancerres.aacrjournals.orgDownloaded from

INHIBITION OF PROMOTION AND NODULE GROWTH BY SAM

Without SAM

With SAM

MlDm

Without PI With PB

23 1

COMPLEXITY LEVELS

Fig. 5. Effect of SAM on the relative frequencies of putative preneoplasticlesions with different complexity levels (number of markers expressed per lesion).The same rat groups and conditions as in Fig. 2. Rats were killed 5 weeks afterinitiation. The number of lesions with different complexity levels have beencalculated independently of the surface area covered by each marker. Data,expressed as percentages of total focal population of each experimental group,are means ±SD (bars), t test: with PB versus without PB: uniform ({/), with/without SAM, P < 0.001 for complexity levels 1 and 3; nonuniform (NU), with/without SAM, P < 0.001 for all complexity levels. With SAM versus withoutSAM (with/without PB): at least P < 0.05 for complexity levels 1 and 3 ofuniform lesions, and for complexity levels 1, 2, and 3 of nonuniform lesions.

that:complexity level 3 to complexity level 2, in SAM-treatedrats.

Effect of SAM on Nodule Growth. PN appeared 22-26 weeksafter initiation in Wistar rats treated up to the 20th week withPB. Very few remodeling nodules were found in those rats nottreated with PB (data not shown) while numerous PN developedin F344 rats 8-12 weeks after initiation in the absence of PB.All these nodules exhibited relatively high DNA synthesis(Table 1) even in the absence (F344 rats) or after interruption(Wistar rats) of PB treatment (persistent nodules). They expressed a high GGT activity; 86% of them also exhibited a highG6PD activity, while the reaction for G6Pase was low or absent(not shown). Most of these nodules showed no evidence ofremodeling for at least 8-11 weeks (Fig. 6/1). When Wistar ratswere subjected to SAM treatment for 1-8 weeks, starting thetreatment 26 weeks after initiation, PN remodeled to variousextents as could be seen from the lack of uniformity of GGThistochemistry and the irregular outlines (Fig. 6, B-D). Remodeling appeared as early as 1 week after starting SAM injection.After 8 weeks of SAM treatment, PN started showing a decreasein volume. Most of the nodular cells lost GGT activity (Fig. 6,C and D) as well as other markers (not shown), and a few PNwere still recognizable microscopically by H & E staining.Similar behavior was observed in F344 rats subjected to SAMfor 11 weeks (not shown).

Nodule regression was also documented by a decrease in PN

number and surface area in SAM-treated rats, coupled with alarge fall in DNA synthesis (Table 1). These effects were correlated in Wistar rats with the duration of SAM treatment: agreat decrease occurred after SAM injection for 8 weeks, whileno effects were recorded after a 1-week treatment. In F344 ratsa dramatic decrease in all the parameters tested occurred afteran 11-week SAM treatment. At this time PN were present onlyin two of five rats.

The development of EAF is coupled with a fall in liver SAMcontent, high ODC activity, and polyamine synthesis (15-18).

SAM inhibits this synthesis and causes a reconstitution of theSAM liver pool and accumulation of 5'-MTA, an inhibitor of

polyamine synthesis and growth (17, 31-33). In order to assessif similar variations occur when SAM treatment is started afterdevelopment of PN, the above parameters were determined inPN various times after starting this treatment. PN of bothWistar and F344 rats showed low SAM and 5'-MTA contents,

low SAM/SAH ratio, and high ODC activity (Table 2). InWistar rats a 1-week treatment with SAM caused a 42-44%increase in SAM and 5'-MTA contents, coupled with a 13%

fall in ODC activity with respect to normal liver. SAM contentincreased 95% in PN after an 8-week treatment with SAM.This treatment also caused a 124% rise in 5'-MTA content and

a 42% fall in ODC activity. No variations in SAH content wereobserved during SAM treatments; thus SAM/SAH ratio variedconcurrently with SAM. The effect of SAM on the aboveparameters could not be assessed in F344 rats due to the smallamount of nodular tissue present after 11 weeks under SAMtreatment (cf. Table 1).

Effect of SAM on Apoptosis. The data in Table 3 indicatethat relatively few ABs were present in EAF and EN, 7 weeksafter initiation, in Wistar rats subjected to the initiation/selection treatments without PB. ABs further decreased in PB-treated rats. However, they increased by about 10-fold 72 hafter arresting PB treatment. SAM caused a 1.5-2-fold increase

in ABs in all conditions tested. Very low percentages of ABswere observed in surrounding liver (0.03-0.09%) which were

not modified by SAM (not included in Table 3).Table 4 shows that a relatively high percentage of ABs was

present in PN. This was particularly evident in the liver of F344rats which exhibit a percentage of ABs 1.6-2-fold higher thanthat of Wistar rats. A 1-week SAM injection caused a 1.7-fold

rise in ABs in Wistar rats; however, longer treatments had anopposite effect. SAM did not modify the relative percentagesof intracellular and extracellular ABs in foci and nodules (notshown). This could indicate that SAM-induced modificationsof the apoptotic process do not depend on a SAM effect on theendocytic phase of this process (34).

Table 1 Effect of SAM on the development of persistent nodulesNodules*Rat

RatstraingroupWistar

56F344

7gTreatmentNoneSAM

(1)SAM(8)None

SAM (11)Time"(wk)27

3427

342323Body

wt*387

±7410±12386

±11362 ±16294

±8279±24RLW*

c5.98

±0.116.02 ±0.86.34

±0.064.81±0.285.62

±0.125. 10 ±0.05No./liver31

±731±632

±821±4¿31

±34'Surface

area(mm2)5.49

±2.06.32 ±1.45.91

±2.72.62 ±1.5*4.43

±2.21.57'LI»5.57

+ 0.816.81±0.635.18

±0.442.83 ±0.83"7.64

±0.680.30*

°Period of time after initiation. SAM treatments were started 1-11 weeks (as indicated in parentheses) before killing.* Data are means ±SD of 5 rats for each time for groups 5, 6, and 8 and of 4 rats for group 8.' Relative liver weight in g/100 g body weight.d t lest: different from the appropriate control for P < 0.001.' Visible nodules were present only in 2 rats.

1853

American Association for Cancer Research Copyright © 1989 on July 10, 2011cancerres.aacrjournals.orgDownloaded from

INHIBITION OF PROMOTION AND NODULE GROWTH BY SAM

Fig. 6. GGT-positive persistent nodules in the liver of Wistar rats subjectedto the initiation/selection treatments plus PB.. f. nodule 26 weeks after initiationshowing a uniform pattern of GOT histochemistry; B, nodule 1 week after startingSAM treatment. Note the absence of GGT activity in the center of the noduleand the irregular nodule outlines. C and /). nodules 8 weeks after starting SAMtreatment. Note that the large nodule in C is not easily recognizable by GGThistochemistry; GGT activity is present only in limited areas scattered throughoutthe nodule. In D, a small nodule, recognized in a serial section by H & E, isalmost completely GGT negative, x 16.

DISCUSSION

Cell Growth and Remodeling in the Triphasic Model. F344rats, treated according to the RH model of experimental carci-nogenesis, undergo rapid development of a high number ofEAF and nodules ( 12). The same treatment was less efficaciousin Wistar rats, in which putative preneoplastic tissue developedrapidly between the 5th and the 9th week after initiation, but itnever occupied more than 22% of liver surface. EAF and ENshowed relatively high LI and relatively low remodeling. However, LI progressively decreased and remodeling increased during the period of time considered, at the end of which EAF andEN exhibited a high degree of biochemical complexity. Theslow development of EAF and EN after the 9th week wasassociated with a further rise in remodeling and a drop inphenotypic complexity and LI. PB, given to rats after the endof the selection period, significantly increased size and biochemical complexity, but not the number, of putative preneoplasticlesions (15, 20). It cannot be excluded, however, that rise inphenotypic complexity levels of early lesions in PB-treated ratsdepends on induction by PB of GGT (35) and G6PD (36)activities. Differently to EAF generated by models based onprolonged PB administration to initiated rats (34,37,38), those

generated by triphasic model do not undergo continuous increase in EAF number throughout the duration of PB treatment. In the present study the PB effect has been estimated asearly as 1 week after the end of 2-AAF feeding, which corresponds to the presence of the maximum number of foci in thelivers of Wistar rats not treated with PB (15). Thus, in the RHmodel there probably occurs, at least in Wistar rats, expressionof all initiated cells after the application of a powerful promoting stimulus such as partial hepatectomy.

Progressive loss of growth capacity by putative preneoplasticcells and their differentiation to normal appearing hepatocytesproceeded to a lesser extent in the presence of PB. Thus,progressive differentiation of EAF and EN does not merelydepend on the relatively rapid exhaustion of proliferative stimulus (reparative growth), behavior that distinguishes the RHmodel from experimental models based on prolonged administration of PB. Under the conditions used, many putativepreneoplastic cells appeared to be programmed to progressivelylose, at least partially, the capacity to overrespond to proliferative stimuli (14). This agrees with the suggestion (39) that onlyrare initiated cells probably retain the capacity of high DNAsynthesis for a long period of time and give rise to relativelyfew PN.

Apoptosis. Cell loss of apoptosis has been suggested as playing a crucial role in the development of putative preneoplasticliver lesions (13, 40). Low apoptosis has been found in EAFand EN. This agrees with the observation of Rotstein et al. (Il)of a low cell loss during the development of these lesions. Incontrast, in the animals subjected to a 3-week treatment withPB, these lesions showed, 72 h after stopping PB ingestion,relatively high apoptosis (cf. Réf.13). These observations couldindicate that in the presence of a proliferative stimulus, eitherreparative (RH model) or hyperplastic (PB), low cell deathoccurs in early preneoplastic lesions. However, a relatively highfrequency of occurrence of ABs has been found in liver nodulesand hepatocarcinomas generated by prolonged orotic acid feeding in 1,2-dimethylhydrazine-initiated rats (41). In PN relatively high cell loss partially counterbalances an elevated growthrate and has recently been suggested as being responsible fortheir slow growth (11). Accordingly, we have found relativelyhigh apoptosis, associated with high DNA synthesis in PN.This apparently contrasts with the presence of relatively fewABs in EAF and EN during PB treatment and could reflect abasic difference between PN and the lesions which developduring PB treatment and depend on PB to grow. Continuousstimulation of DNA synthesis and inhibition of cell death byPB could largely affect homeostatic control of cell growth inEAF and EN. After interruption of PB treatment, homeostaticcontrol should be established again and high cell death appears,in EAF, in the attempt to reestablish original liver mass (cf.Ref. 42). Differently to EAF and EN, high cell death is consti-

Table 2 Effect of SAM on SAM, SAH, and 5'-MTA content and ODC activity of persistent nodules

RatstrainWistarF344RatgroupControl

556

6Control7TreatmentNone

NoneNoneSAM (1)SAM(8)NoneTime"

(wk)27

34273423SAM*fag/g)21.05

±1.109.11 ±0.96'9.65 ±0.29e

13.19 ±1.14'

18.81 ±1.2923.44

±1.189.57 ±0.37'SAH*(cg/g)10.12

±0.5710.00±0.4310.16±0.079.60 ±0.149.57 ±0.2810.01

±0.3310.31±0.11SAM/SAH*2.08

±0.080.93 ±0.02'0.95 ±0.03'1.06±0.02'1.44±0.09'2.32

±0.080.93 ±0.03'5'-MTA*(Mg/g)1.02

±0.300.43 ±0.02'0.45 ±0.04'0.64 ±0.03'

1.01±0.100.98

±0.270.41 ±0.03'ODC

activity*

(pmol COz/h, mgprotein)34.25

±2.12103.44 ±8.97'101.05 ±11.48'87.56 ±8.61'58.16±8.36'33.23

±3.55122.40 + 21.15'

" Period of time after initiation. SAM treatments were started 1-11 weeks (as indicated in parentheses) before killing.* Data are means ±SD of 5 rats for control (normal rats) and for groups 5 and 6 and of 4 rats for group 7.' t test: groups 5 and 7 versus respective control and group 6 versus group 5, different for P< 0.001.

1854

American Association for Cancer Research Copyright © 1989 on July 10, 2011cancerres.aacrjournals.orgDownloaded from

INHIBITION OF PROMOTION AND NODULE GROWTH BY SAM

Table 3 Effect of SAM on the incidence ofapoptotic bodies in EAF and EN

Ratgroup"123434TreatmentNoneSAMPBPB

+SAMPBPB

+ SAMTime*

(h)00007272ABs/100hepatocytesc0.61

+0.241.47±0.15''0.41

±0.160.88+0.18''3.51

±0.46''5.42±0.34''

" Rats were killed 7 weeks after initiation or, when indicated, 72 h after the

arrest of PB treatment. This treatment was started 4 weeks after initiation andwas continued up to the end of the 7th week. SAM treatment (6 daily doses of64 umol/kg) was started 4 weeks after initiation and was continued up to sacrifice.

Period of time after the end of PB treatment.' Clear cells, eosinophilic foci, and EN were identified and nuclear changes

representing apoptosis and ABs were determined by scoring 5000 focal cells/liver. Data are means ±SD of 5 rats for groups 1 and 2 and of 10 rats for groups3 and 4. ABs in surrounding liver: 0.03-0.08 for all rat groups.

- / test: different from the appropriate control for P < 0.001.

Table 4 Effect of SAM on the incidence ofapoptotic bodies in persistent nodules

RatstrainWistarF344Ratgroup567TreatmentNoneSAM

(1)SAM(8)NoneTime"

(wk)26

273427

341223ABs/100

hepatocytes*1.73

±0.061.78 ±0.101.75±0.102.87

+ 0.11''0.22 ±0.00'2.75

±0.643.43 ±0.45

8 SAM (11) 23 0.70±0.18C

" Period of time after initiation. SAM treatment was started 1-11 weeks (as

indicated in parentheses) before killing. It consisted of 6 daily doses (64 /¿mol/kg) of SAM.

* Number of ABs was determined by scoring 3500 nodular cells/liver. Data

are means ±SD of 5 rats for groups 5, 6, and 8 and 4 rats for group 7.c t test: different from the appropriate control for at least P < 0.002.

tutively coupled with high DNA synthesis in PN (11), wherebasic homeostatic control mechanisms for maintenance of livermass operate even though with some alterations responsible forprogressive growth (11). Thus, high cell death in PN has beeninterpreted as a response to cell proliferation by these primaryhyperplastic (neoplastic) lesions (42).

Effect of SAM on Promotion and Nodule Growth. SAMtreatment causes a great decrease in size and number of EAFand EN. When SAM is given for 16 weeks after selection,development of PN and hepatocellular carcinomas is completely prevented (17, 18). These results, taken together, indicate that in the liver of SAM-treated rats, there are no cellscapable of progressing to PN and carcinomas, even long afterthe cessation of SAM injection. Interestingly, a supply of relatively high doses of methionine, a SAM precursor, to rats,partially prevents growth and progression of ethionine-inducedliver carcinomas (43) and benzo(a)pyrene-induced skin carcinomas (44). However, supplementation of lower doses of methionine in the diet, during promoting treatments with PB orl,l-bis(/>-chlorophenyl)-2,2,2-trichloroethane, does not modifythe incidence of hepatocarcinomas in DENA-initiated rats (45).The anti-promotion effect of SAM is linked to inhibition ofDNA synthesis, increase in apoptosis, and increase in remodeling which proceeds concurrently with loss of biochemicalmarkers. This loss could be one aspect of the apparently irreversible differentiation process induced by SAM. It should beconsidered that compensatory cell growth, in limited areasaround ABs, has been suggested to help growth and progressionof nodules (41). However, in SAM-treated rats, increase inapoptosis cannot be counterbalanced by (compensatory) cellproliferation. This could contribute to progressive decrease inEAF and nodule size and number in SAM-treated rats.

Another aspect of peculiar interest is the inhibition of PNgrowth, associated with a sharp decrease in their histochemicaluniformity and in size and number in Wistar rats exposed for8 weeks to SAM. In F344 rats an 11-week SAM treatment ledto a complete disappearance of PN. Our data, however, do notpermit a comparison between the two strains for their sensitivityto SAM, due to the differences in the length of SAM treatmentas well as in the developmental stages of PN at the time of thistreatment. Numerous observations indicate that PN are precursors of hepatocarcinomas in the rats and maybe in humans[reviewed by Farber (24)]. Thus, inhibition of PN growth andenhancement of their regression by SAM could be importantfor cancer prevention, taking into account that PN are easilyrecognizable and have a low tendency to regress spontaneously.The SAM effect on nodule development is partially linked toan alteration of the equilibrium between cell production andcell loss in favor of cell loss by apoptosis. However, apoptosiswas low in nodules, after a prolonged SAM treatment. It shouldbe noted that phenotypic stability was lost by those noduleswhich underwent a drop in DNA synthesis and loss of biochemical markers. This led to the appearance of liver cells with anapparently normal phenotype, which are expected to exhibitlow cell production and low cell loss. Our data do not, however,clarify if these apparently redifferentiated hepatocytes were stillinitiated cells.

Mechanisms of the SAM Effect. SAM is a nontoxic andnonmutagenic substance (46) which enters liver cells in vitro(47-51) and in vivo (47, 52). Indirect evidence (19) suggestsSAM uptake by hepatocytes, without previous splitting to aden-

osine and methionine. SAM treatment causes accumulation inrat liver of 5'-MTA, a SAM catabolite which inhibits polyaminesynthesis and growth (17, 18, 31-33). 5'-MTA accumulation,

inhibition of ODC activity, a key enzyme for polyamine synthesis, and of DNA synthesis have also been observed in PN asa consequence of SAM treatment. Thus, growth inhibition by5'-MTA could be envisaged as a mechanism of the SAM effect

on the development of preneoplastic tissue. In PB-treated ratsSAM could modulate PB effects on cell production and celldeath, at least in the early stages of promotion. It should alsobe noted, however, that the SAM/SAH ratio is low in PN. Alow SAM/SAH ratio may be associated with DNA under-methylation (53). This has indeed been observed in PN (19).4

When SAM treatment is started after the appearance of PN,inhibition of c-Ha-ros, c-Ki-ras, and c-myc protooncogeneexpression takes place, which seems to correlate better withnodular DNA methylation than with accumulation of SAMcatabolites which inhibit growth (19).4 The role of protoonco

gene expression in tumor promotion and progression has notyet been clarified. However, a role of these phenomena on cellgrowth and differentiation has been stressed (54). The possibility that interference of a high methylating environment withgene expression is one of the mechanisms influencing cellproduction and cell loss is an attractive hypothesis and iscurrently under study.

REFERENCES

1. Cayama, E., Tsuda, II.. Sarma, D. S. R., and Farber, E. Initiation of chemicalcarcinogenesis requires cell proliferation. Nature (Lond.), 275: 60-61, 1978.

2. Emmelot, P., and Scherer, E. The first relevant cell stage in rat livercarcinogenesis. A quantitative approach. Biochim. Biophys. Acta, 605: 247-304, 1980.

' R. Garcea, L. Daino, R. Pascale, M. N. Simile, M. Puddu, M. E. Ruggii),

M. A. Seddaiu, G. Satta, M. J. Sequenza, and F. Feo. Protooncogene methylationand expression in regenerating liver and preneoplastic liver nodules induced inthe rat by diethylnitrosamine. Effect of variations of 5-adenosyl-methionine:5-adenosylhomocysteine ratio, submitted for publication.

1855

American Association for Cancer Research Copyright © 1989 on July 10, 2011cancerres.aacrjournals.orgDownloaded from

INHIBITION OF PROMOTION AND NODULE GROWTH BY SAM

3. Pilot, H. C., and Sirica. A. E. The stages of initiation and promotion inhepatocarcinogenesis. Biochim. Biophys. Acta. 605: 191-215, 1980.

4. Las, J. O. Tissue regeneration and tumour development. Br. J. Cancer. 13:669-674, 1959.

5. Pound, A. W., and McGuire, L. J. Repeated partial hepatectomy as apromoting stimulus for carcinogenic response of liver to nitrosamines in rats.Br. J. Cancer, 37: 585-594, 1978.

6. Farber, E., and Cameron, R. The sequential analysis of cancer development.Adv. Cancer Res., 31: 125-226, 1980.

7. Schuhe-Hermann, R., Schuppler, J., Timmermann-Trosiener, I., Ohde, G.,Bursch, W., and Berger, H. The role of growth of normal and preneoplasticcell populations for tumor promotion in rat liver. Environ. Health Perspect.,50: 185-194. 1983.

8. Enomoto, K., and Farber. E. Kinetics of phenotypic maturation of remodelingof hyperplastic nodules during liver carcinogenesis. Cancer Res., 42: 2330-2335, 1982.

9. Tatematsu, M.. Nagamine, Y.. and Farber. E. Redifferentiation as a basis forremodeling of carcinogen-induced hepatocyte nodules to normal-appearingliver. Cancer Res., 43: 5049-5058, 1983.

10. Moore, M. A., Hacker, H.-J., and Kunnasch. P. Phenotypic instability infocal and nodular lesions induced in a short term system in the rat liver.Carcinogenesis (Lond.), 4: 595-603. 1983.

11. Rotstein, J., Sarma, D. S. R., and Farber, E. Sequential alterations in growthcontrol and cell dynamics of rat hepatocytes in early precancerous steps inhepatocarcinogenesis. Cancer Res., 46: 2377-2385, 1986.

12. Soil, D. B., Mediine, A., and Farber, E. Rapid emergence of carcinogen-induced hyperplastic lesions in a new model for sequential analysis of livercarcinogenesis. Am. J. Pathol., 88: 595-618, 1977.

13. Bursch, W., Lauer, B., Timmermann-Trosiener, I., Bartel, G., Schuppler, J.,and Schulte-Hermann, R. Controlled death (apoptosis) of normal and putative preneoplastic cell in rat liver following withdrawal of tumor promoters.Carcinogenesis (Lond.), 5:453-458, 1984.

14. Schulte-Hermann, R. Tumor promotion in the liver. Arch. Toxicol., 57:147-158. 1985.

15. Feo, F., Garcea, R., Daino, L., Pascale, R., Pirisi, L., Frassetto, S., anRuggiu, M. E. Early Stimulation of polyamine biosynthesis during promotionby phénobarbitalof diethylnitrosamine-induced rat liver carcinogenesis. Theeffects of variations of the 5-adenosyI-L-melhionine cellular pool. Carcinogenesis (Lond.), 6: 1713-1720, 1985.

16. Feo, F., Garcea, R., Pascale, R., Pirisi, L., Daino, L., and Donaera, A. Thevariations of 5-adenosyl-L-melhionine content modulate hepatocyte growthduring phénobarbitalpromotion of diethylnitrosamine-induced rat liver carcinogenesis. Toxicol. Pathol., Õ5:109-114, 1987.

17. Garcea, R., Pascale, R., Daino, L., Frassetto, S., Cozzolino, P., Ruggiu, M.E., Vannini, M. G., Gaspa, L., and Feo, F. Variations of ornithine decarbox-ylase activity and 5-adenosyl-L-methionine and 5'-methylthioadenosine contents during the development of diethylnitrosamine-induced liver hyperplastic nodules and hepatocellular carcinoma. Carcinogenesis (Lond.), H: 653-658, 1987.

18. Feo, F., Garcea, R., Daino, L., and Pascale. R. Mechanism of the inhibitionof liver hepatocarcinogenesis by 5-adenosyl-L-methionine. In: M. Roberfroidand V. Préat(eds.). Experimental Hepatocarcinogenesis, pp. 195-207. NewYork: Plenum Publishing Corp., 1988.

19. Feo, F., Garcea, R., Daino, L., Pascale, R., Frassetto, S., Cozzolino, P.,Vannini, M. G., Ruggiu, M. E., Simile, M., and Puddu, M. 5-Adenosylme-thionine antipromotion and antiprogression effect in hepatocarcinogenesis.Its association with inhibition of gene expression. In: F. Feo, P. Pani, A.Columbano, and R. Garcea (eds.), Chemical Carcinogenesis: Models andMechanisms, pp. 407-423. New York: Plenum Publishing Corp., 1988.

20. Lans, M., deGerlache, J., Taper, H. S., Préat,M., and Roberfroid, M. B.Phénobarbitalas a promoter in the initiation/selection process of experimental rat hepatocarcinogenesis. Carcinogenesis (Lond.), 4: 141-144, 1983.

21. Fiecchi, A. (May 1976) US Patent No. 3,954,726.22. Pascale, R., Pirisi, L., Daino, L., Zanetti, S., Salta, A., Battoli, E., and Feo,

F. Role of phosphalidylelhanolamine melhylalion in the synthesis of phos-phatidylcholine by hepatocytes isolated from choline-defícient rats. FEBSLell., 145: 293-298, 1982.

23. Seglen, P. O., Schwarze, P. E., and Saeter, G. Nuclear alterations in livercarcinogenesis: the role of non-polyploidizing growth. In: F. Feo, P. Pani, A.Columbano, and R. Garcea (eds.), Chemical Carcinogenesis: Models andMechanisms, pp. 435-445. New York: Plenum Publishing Corp., 1988.

24. Farber, E. Cellular biochemistry of the slepwise developmenl of cancer withchemicals: G. H. A. Clowes Memorial Lecture. Cancer Res., 44:5463-5474,1984.

25. Squire, R. A., and Levilt, M. H. Report of a workshop on classification ofspecific hepatocellular lesions in rats. Cancer Res., 35: 2314-2323, 1975.

26. Garcea, R., Daino. L., Frassetto, S., Cozzolino, P., Ruggiu, M. E., Vannini,M. G., Pascale, R., Lenzerini, L., Simile, M. M., Puddu, M., and Feo, F.Reversal by ribo- and deoxyribonucleosides of dehydroepiandrosterone-in-duced inhibition of enzyme allered foci in the liver of rats subjected to theinitialion-seleclion process of experimenlal carcinogenesis. Carcinogenesis(Lond.), 9:931-938, 1988.

27. Goldsworthy, T. L., and Pitol, H. C. The quanlilative analysis and slabilityof histochemical markers of altered hepatic foci in ral liver following initiation by diethylnitrosamine administration and promotion with phénobarbital.Carcinogenesis (Lond.), 6: 1261-1269, 1985.

28. Institute of Laboratory Animal Resources. Histológica!typing of liver lumors

of the rat. J. Nati. Cancer Insl., 64: 179-206, 1980.29. Feo, F., Pirisi, L., Pascale, R., Daino, L., Frasselto, S., Garcea, R., and

Gaspa, L. Modulatory effect of glucose-6-phosphate dehydrogenase deficiency on benzo(a)pyrene loxicily and transforming aclivity for in vitro-cultured human skin fibroblasts. Cancer Res., 44: 3419-3425, 1984.

30. Peraino, C., Staffeldl, E. F., Carnes, B. A., Ludeman, V. A., Blomquisl, J.A., and Vesselinoviich, S. D. Characterization of histochemically delectablealtered hepatocyte foci and their relationship to hepatic tumorigenesis in ratstreated once wilh diethylnilrosamine or benzo(a)pyrene within one day afterbirth. Cancer Res., 44: 3340-3347, 1984.

31. Pegg, A. E., Bnrcliardt. R. T., and Coward, J. K. Effecl of inhibilors ofspermidine and spermine synlhesis on polyamine concentrations and growthof mouse transformed fibroblasts. Biochem. J., 194: 79-89, 1981.

32. t'ajula. R. L., and Raina, A. Methyllhioadenosine, a polent inhibilor ofspermine synlhase from bovine brain. FEBS Lett., 99: 343-345, 1979.

33. Scalabrino, G., Ferioli, M. E., and Candiani, R. New insights into iheregulalion of 5-adenosyl-L-methionine decarboxylase in normal and regenerating ral liver. In: F. Feo, P. Pani, A. Columbano, and R. Garcea (eds.),Chemical Carcinogenesis: Models and Mechanisms, pp. 459-465. New York:Plenum Publishing Corp., 1988.

34. Schulte-Hermann, R., Schuppler, J., Ohde, G., and Timmermann-Trosiener,I. Growth control in cells of pillativi- preneoplastic islands in rat liver andthe effect of tumor promoters. In: P. Pañi,F. Feo, and A. Columbano (eds.),ReceñíTrends in Chemical Carcinogenesis, pp. 101-115. Cagliari, Italy:ESA, I981.

35. Roomi, M. W., and Goldberg, D. M. Comparison of -y-glutamyltransferaseinduclion by phénobarbitalin Ihe ral, guinea pig and rabbit. Biochem.Pharmacol., JO: 1563-1571, 1981.

36. Garcea, R., Daino, L., Pascale, R., Frasselto, S., Cozzolino, P., Ruggiu, M.E., and Feo, F. Inhibition by dehydroepiandrosterone of liver preneoplasticfoci formation in rats after initiation-selection in experimental carcinogenesis. Toxicol. Pathol., IS: 164-169, 1987.

37. Schulle-Hermann, R., Timmermann-Trosiener, I., and Schuppler, J. Response of liver foci in rats to hepalic lumor promoters. Toxicol. Pathol., 10:63-68, 1982.

38. Goldsworthy, T., Campbell, H. A., and Pilot, H. C. The natural hislory anddose-response characteristics of enzyme-altered foci in rat liver followingphenobarbilal and diethylnilrosamine administration. Carcinogenesis(Lond.), 5:67-71, 1984.

39. Farber, E. Some emerging general principles in the pathogenesis of hepatocellular carcinoma. Cancer Surv., 5: 695-718, 1986.

40. Schulte-Hermann, R., Bursch, W., Fesus, L., and Kraupp, B. Cell death byapoptosis in normal, preneoplastic and neoplastic tissue. In: F. Feo, P. Pani,A. Columbano, and R. Garcea (eds.). Chemical Carcinogenesis: Models andMechanisms, pp. 263-276. New York: Plenum Publishing Corp., 1988.

41. Columbano, A., Ledda-Columbano, G. M., Rao, P. M., Rajalakshmi, S., andSanila. D. S. R. Occurrence of cell dealh (apoptosis) in preneoplastic andneoplaslic liver cells. Am. J. Pathol., 116:441-446, 1984.'

42. Farber, E., Rotstein, J., Harris, L., Lee, G., and Chen, Z-Y. Cell proliferationand cell loss in progression in liver carcinogenesis: a new hypothesis. In: F.Feo, P. Pani, A. Columbano, and R. Garcea (eds.), Chemical Carcinogenesis:Models and Mechanisms, pp.167-172. New York: Plenum Publishing Corp.,1988

43. Brada, Z., Altman, N. H., Hill, M., and Bulba, S. The effect of methionineon the progression of hepatocellular carcinoma induced by ethionine. Res.Commun. Chem. Pathol. Pharmacol., 38: 157-160, 1982.

44. Brada, Z., Hillova, J., Hill, M., Altman, N. H., and Bulba, S. Effect ofmelhionine on development of benzopyrene (BP) induced sarcomas. Proc.Am. Assoc. Cancer Res., 27: 121, 1986.

45. Shivapurkar, N., Hoover, K. L., and Poirier, L. A. Effect of methionine andcholine on liver tumor promotion by phenobarbilal and DDT in diethylnilrosamine initiated rals. Carcinogenesis (Lond.), 7: 547-550, 1986.

46. Pezzoli, C., Galli Kicnlc, M., and Strumentinoli. G. Lack of mutagenicactivily of adomethionine in vitro and in vivo. Arzneim.-Forsch., 37: 826-829, 1987.

47. Stekol, J. A., Anderson, E. L., and Weiss, S. 5-Adenosyl-L-methionine in thesynthesis of choline, creatinine, and cysteine in vivo and in vitro. J. Biol.Chem., 233:425-429, 1958.

48. Zappia, V., Galletti, P., and Porcelli, M. Uptake of adenosylmelhionine andrelated sulfur compounds by isolated rat liver. FEBS Lett., 90: 331-335,1978.

49. Pezzoli, C., Stramentinoli, G., Galli-Kienle, M., and Pfaff, E. Uplake andmetabolism of 5-adenosyl-L-methionine by isolated rat hepatocytes.Biochem. Biophys. Res. Commun., 85: 1031-1038, 1978.

50. Farooqui, J. Z., Lee, H. W., Kim, S., and Paik, W. K. Studies on compari-mentation of 5-adenosyl-L-methionine in Saccharomyces cerevisiae and isolated hepatocytes. Biochim. Biophys. Acta, 757: 342-351, 1983.

51. Engstrom, M. A., and Benevenga, N.J. Rales of oxidation of the methionineand 5-adenosylmethionine methyl carbons in isolated rat hepatocytes. J.Nutr., 117:1820-1826, 1987.

52. Giulidori, P., Galli-Kienle, Catto, E., and Stramentinoli, G. Transmethyla-lion, iranssulfuration, and aminopropylation reaclions of .S'-adcnosyl-i.-nie

lliioni ne in vivo. ¡.Biol. Chem., 259:4205-4211, 1984.53. Chiang, P. K., and Cantoni, G. L. Perturbation of biochemical transmethyl-

ations by 3-deazaadenosine in vivo. Biochem. Pharmacol., 2K: 1897-1901,1979.

54. Kahn, P., and Graf, T. (eds.). Oncogenes and Growth Conlrol. Berlin:Springer-Verlag, 1987.

1856

American Association for Cancer Research Copyright © 1989 on July 10, 2011cancerres.aacrjournals.orgDownloaded from