Download - ICANCERRÃœSI-ARCHÃ“S,3i5-3622.Augustis. i a-Fetoprotein … · ICANCERRÃœSI-ARCHÃ“S,3i5-3622.Augustis. i a-Fetoprotein (AFP) Expression in Clones of McA-RH

ICANCERRÃœSI-ARCHÃ•S,3<>i5-3622.Augustis. i

a-Fetoprotein (AFP) Expression in Clones of McA-RH 7777 Rat Hepatoma:

Correlation with the Occurrence of Homogeneously StainingRegions on Chromosome 14 '

Leila Khamzina,2 Tatjana Eraiser, and Pierre Borgeat

Centre Je Recherche en Rhumatologie el Immunologie. Centre tie Recherche Ju Centre Hospitalier de l'UniversitÃ©Laval, UniversitÃ©Laval, QuÃ©bec,Canada, HIV 4G2

ABSTRACT

a-fetoprotein (AFP) is a well-established cell differentiation and tumormarker. We showed previously that McA-RH 7777 hepatoma cells are

heterogeneous in terms of their AFP cellular expression. In the presentstudy, we developed stable and unstable 7777 hepatoma clones in terms oftheir AFP phenotype: AFP-producing (AFP+) or AFP-nonproducing

(AI-T ) clones, and investigated in these clones (a) AFP phenotype related

to protein and mRNA levels; (/Â»cellular morphology; and in expressionof several liver-specific markers. Our results demonstrated that a-albu-min expression paralleled that of AFP, from the absence of a-albuminmessage in All' clones to high expression in Ml'1 clones, suggesting

that common mechanisms control the expression of both proteins in thishepatoma cell population. In addition, the karyotypes of the McA-RH

7777 hepatoma cell line and its 15 generated clones were analyzed andcorrelated to their AFP phenotypes. Only the stable AFP+ clones showed

homogeneously staining regions on the chromosome carrying the Ml'

gene. These results strongly suggest that amplification of either structuralor regulatory sequences of the .1/7' gene is involved in maintaining its

high expression.

INTRODUCTION

AFP3 is a well-known oncodevelopmental protein. Actively tran

scribed by fetal and postnatal hepatocytes, AFP is almost undetectablein adult liver parenchymal cells; however, its production can resumein the liver under pathological conditions such as primary hepatocel-lular carcinoma (1-5). Therefore, this protein is considered a pathol

ogy marker and is used in clinical diagnosis (6). AFP is a member ofa multigene family (albumin family) that includes ALB, DBP, and therecently discovered a-ALB (7-9). All four of these proteins are

synthesized in mammalian liver parenchymal cells and share a similarsecondary folding structure with three related domains (10, 11). Theirgenes remain in a single copy per haploid genome and have beenmapped to chromosome 14 in rats (12).

The disappearance of AFP postnatally and its reappearance inmalignancy reflect sequential gene activation and repression. Thehepatoma-derived cell lines, therefore, provide a very interesting

model for studying AFP gene regulation during normal liver development, differentiation, and oncogenesis. The McA-RH 7777 cell line(13) is a well-established, chemically induced rat hepatoma, which we

adopted for our studies. Using several clonal approaches, we haveshown (14, 15) that individual AFP-producing (AFP+) and

Received 4/3/95; accepled 6/19/95.The costs of publication of this article were defrayed in part by the payment of page

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

1This work was supported by Medical Research Council of Canada grants. This work

is the first in the series: a-Fetoprotein (AFP) Expression in Clones of McA-RH 7777 Rat

Hepatoma.2 To whom requests for reprints should be addressed, at Centre dc Recherche en

Rhumatologie et Immunologie, Local 9800, CHUL, 2705 boul. Laurier, Ste-Foy, Quebec,Canada, G1V 4G2.

' The abbreviations used are: AFP. a-fetoprotein: AFP ', a-fetoprotein-producingphenotype: AFP , a-fetoprotein-nonproducing phenotype; AFP', phenotype with mixed

a-fetoprotein production: ALB, albumin; a-ALB, a-albumin; DBP. vitamin D-binding

protein; TAT, tyrosine aminotransferase; TO, tryptophan oxygcnase; dsr. differentialstaining regions; hsr. homogeneous staining regions.

AFP-nonproducing (AFP ) cells give rise to AFP+ and AFP clones,

respectively. However, clones of alternative AFP phenotypes weresometimes generated with a frequency that far exceeded that of genemutation. In addition, these studies established a method of stabilizingcloning (7-step selection) that allows development of both stable and

unstable clones by selective pressure in function of the AFPphenotype (AFP f or AFP ).

The present study introduces a unique model system of 7777hepatoma clones. We describe the isolation and characterization ofeight new clones (AFP+ and AFP ) derived from the McA-RH 7777

cell line, their phenotype analysis, and the regulatory mechanisms thatlead to their generation. We also analyzed the expression of AFP andother liver-specific markers. Our results demonstrated an absence ofa-ALB gene expression in AFP" clones and significant high expres

sion in AFP+ clones that reached maximal levels in stable clones.

Moreover, because the malignant phenotype often arises as a result ofconsistent chromosome aberrations (16), we examined whether AFPproduction could be associated with specific changes at the chromosome level. The chromosome constitution of the 7777 hepatoma cellline and 15 generated clones were analyzed and correlated to the AFPphenotypes. High-level AFP stable expression correlated significantly

with specific gene amplification characteristics.

MATERIALS AND METHODS

Cell Culture and Cloning. McA-RH 7777 rat hepatoma cells were ob

tained from the American Type Culture Collection (ATCC, Rockville, MD)and grown as described elsewhere (13). The method of stabilizing cloningconsisted of seven cycles and was performed according to Eraiser andMi.-nii/iiu (14, 15). Each cycle spaned over al least 4 weeks and included the

following steps: (n) mass hepatoma culture cell cloning; (/;) random selectionof 10-20 clones on days 7-10 of culture; (<â€¢)isolation and maintenance of the

clones with simultaneous immunohistochcmical analysis of subclones andcellular monolayers; and ((/) selection of the most phenotypically homogeneous clone, after which the next cloning cycle started. All experiments wereconducted using cells in logarithmic growth phase.

Immunohistochemistry. Indirect staining was performed on cells grownin Petri dishes or on glass coverslides. Medium was removed by aspiration, andcells were washed twice with 10 ITIMof PBS, pH 7.4. Fixation was performedin cold methanol ( â€”20Â°C)for 30 min, and then the cells were rinsed twice with

PBS. The appropriate dilution of anti-rat AFP or anti-rat ALB mouse mAhs

(Kirkegaard & Perry Laboratories, Gaithershurg, MD) was applied for 1 or 2h. Localization of the antibodies was visualized with avidin-hiotin complexsolution (Vector Corp., Burlingame, CA) and 3,3'-diaminohen/idine-H2O,

(Pierce Chemical Co., Rockford, IL).Northern Blot Analysis. Total RNA was extracted from the cells as

described by Chomczynski and Sacchi (17), clectrophoresed on 1% formal-dehyde-agarose gel, and transferred to Hybond-/V membranes (Amersham,

Ontario, Canada). The integrity of the RNA and equal loading were verified byethidium bromide staining and hybridization with a 28S rRNA probe. Themembranes were then hybridized with different random primer 12P-labeled

cDNA probes (Amersham). Our study was carried out using the followingplasmids: pRAF 87 and pRSA 57 (8), pHDQ 105 and pHDO 833 (18), DS 3(X)(19), pSP65rDBP7ft(> (20), pSP6-cTAT-3" (21), and pSP(TO),., (22). All blots

were analyzed by densitometric scanning, and values were corrected accordingto RNA loading.

3615

Research. on August 27, 2021. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

AFP EXPRESSION CORRELATES WITH HSR ON CHROMOSOME 14

Cytogenetic Analysis. Colchicine (1 ngjmi) was added to the culturemedium for 2 h, and then the cells were trypsinized, centrifuged, treated withhypotonie KG (75 mM), and fixed with methanol:acetic acid (v/v; 3:1). Theslides were prepared by air-drying technique. Chromosomes were banded by amodification of the trypsin-Giemsa method of Seabright (23). For each cell

population, 50 metaphases were counted, photographed, and analyzed. Thestandard karyotype for trypsin-Giemsa-banded rat chromosomes was used

(24), and the nomenclature for the bands was derived from Levan (25).

RESULTS AND DISCUSSION

The stabilizing cloning was started from the mixed (AFP21) clone D7and carried out in two contrasting phenotypic directions: AFP+ and

AFP . The selected clones underwent a maximum of six culture passagesbefore starting the next cloning cycle. Stability-gaining selection throughthe stabilizing cloning was quite efficient allowing us to obtain AFP+ andAFP" clones with different levels of AFP phenotype stability. From the

panel of generated clones, eight were chosen for additional characterization: mixed (AFP^ clone D7; the unstable AFP" clones, F4 and Hll;the unstable AFP+ clones, A3 and G6; the stable AFP" clone 7H10; and

the stable AFP4 clones, 7G3 and 7G4. The filiation of these clones is

schematically summarized in Fig. 1.Analysis of AFP Phenotype in McA-RH 7777 Hepatoma Clones.

The AFP phenotype of each clone was characterized by two independent and complementary methods: immunocytochemical staining,which localizes the intracellular protein and reflects the homogeneityof the population, and Northern blot analysis, which estimates theactual AFP biosynthesis. As shown in Fig. 2, the parental 7777 cellsconsisted of a mixture of stained and nonstained cells that yieldedthree types of clones: AFP+, AFP^ and AFP". Staining appeared as

discrete brown granules within the cytoplasm, with stronger stainingnear the perinuclear space. For the AFPÂ±phenotype, some variations

in staining intensity were noted from one cell to another (Fig. 3A).AFP+ clones had staining profiles that were much more homogene

ous, with hardly visible variations whether stable or unstable (Fig. 3,B and C). On the other hand, AFP" clones did not show any staining

(Fig. 3, D and E). Growing our stable clones for more than 6 monthsshowed that cells with alternative phenotypes appear within eachpassage with a frequency not exceeding 0.01-0.005% but did not

form clones with selective advantages. This latter observation is inagreement with our previous results (14, 15).

The results of our Northern blot analysis demonstrated that the absenceof AFP production in AFP" clones correlated with the absence of the

corresponding mRNA. As illustrated in Fig. 4, a single band of 2.2 kbwas observed in the parental McA-RH 7777 cell line and its AFP+

clones, which corresponds to the reported size of AFP mRNA in rat liver(26). Densitometric analysis of the bands showed that the level of AFPmRNA of the stable AFP+ clones was twice that of unstable clones and

20 times that of the parental line. The reporter chloramphenicol acetyl-transferase gene driven by 7 kb of the 5 ' flanking region of the AFP gene,which was active only in the AFP+ clones (27), indicated a transcrip-

tional control of AFP gene expression in our system.Clones of 7777 hepatoma vary not only in the function of their AFP

phenotype but also morphologically (Fig. 2A). Some had fibroblast-

like morphology with regular epithelial form and were well spread onthe substrate. Another morphological type was the round cells, whichtended to form tightly packed colonies. However, none of thesemorphological types defined a specific AFP phenotype. In the courseof the stabilizing cloning procedure, no correlation was observedbetween the morphology and the phenotype of the clones during theinitial cloning cycles. However, at the final cycle, a remarkablechange in the morphology of the clones was visible. Fig. 3 shows thatthe stable AFP~ clone 7H10 was composed of round cells tending to

form trabecular-like structures as in mature hepatocytes, whereas thestable AFP+ clone 7G3 consisted mainly of elongated bipolar cells

similar to the embryonal morphological type.Our approach to the study of cellular differentiation consisted of

analyzing stable clones isolated from cellular populations as a function of the presence or absence of certain differentiation markers.Such clones are presumed to have undergone a stable change at one orseveral regulatory sites responsible for the expression of the differentiation markers. Identification of the involved regulatory mechanisms and their levels of action will permit reconstitution of thenormal chain of regulatory events underlying the expression of suchmarkers. A similar clonal approach has been established previously byWeiss and collaborators (28, 29) in studies of the heterogeneousnature of ALB expression in cells of Reuber H35 rat hepatoma. Aseries of differentiated (ALB producing) and dedifferentiated (ALBnonproducing) clones were isolated; the extinguishing of ALB production in dedifferentiated clones showed simultaneous extinguishingof the entire group of hepatocyte-specific markers, which were active

McA-RH 7777 cells

Unstable UnstableAFP" AFP-

cloneHll clone F4

UnstableAFP*

clone A3

UnstableAFP*

clone G6

UnstableAFP-

clonc TB6

UnstableAFP-

clone TB9

UnstableAFP +

clone A3A

Unstable UnstableAFP - AFP +

clone B3 clone E3

StableAFP +clone7G4

StableAFP +clone 7B7

StableAFP +clone 7B9

Stable

clone 7H10

Fig. 1. The filiation of clones generated by two independent stabilizing clonings from McA-RH 7777 rat hepatoma cell line.

3616




Fig. 2. AFP phenotype of McA-RH 7777 ral hepaloma cell line and its clones as determined by immunohistochemical staining. A, McA-RH 7777 cell monolayer with typicalcell-to-cell heterogeneity of AFP expression. Three types of clones of McA-RH 7777 cells: B, AFP+ clone; C, AFP- clone; D, AFP" clone.

in their differentiated mother clones. To assess whether such phenomenon occurs in our model system, we investigated the expression ofthe well-established differentiation marker AFP and of other in vivo

simultaneously expressed markers (ALB, aALB, DBF, TAT, and TO)in stable clones isolated from 7777 hepatoma.

TAT and TO are gluconeogenetic enzymes implicated in hepatic

functions (21, 22). Similar to AFP, they are developmentally regulatedand modulated by glucocorticoids. The mRNA levels of theseliver-specific markers as well as those of ALB, DBP, and a-ALB,

which are known to be members of the same multigene family asAFP, have been evaluated. DBP, TAT, and TO messages were notdetectable in either AFP" and AFP+ clones, which may reflect the

3617




Fig. 3. AFP phenotype of clones isolated from McA-RH 7777 rat hepatoma cell line. A, AFP* clone D7 with typical mix of stained and unstained cells. Variations in stainingintensity were noted from one cell to another. B, unstable AFP* clone G6. C, stable AFP+ clone 7G4. D, unstable AFP~ clone Hll. E, stable AFP" clone 7H10.

involvement of different mechanisms in the regulation of these markers and AFP. Immunocytochemical staining using anti-rat ALB anti

bodies demonstrated the presence in all clones of synthesized ALBwith fairly homogeneous staining patterns. Thus, the hepatocyte-

specific marker ALB was detected in the studied clones at the proteinlevel, although its message was not detectable, which may be due tothe limited sensitivity of the Northern analysis. Alternatively, thisobservation may reflect a greater ALB protein half-life than that of its

message in these clones, resulting in its permanent expression evenwhen mRNA is not detectable. The expression of AFP and ALB doesnot appear to be correlated, suggesting that these two markers areunder different regulatory controls.

Interestingly, a particular relationship was found between AFP andthe recently reported a-ALB genes (19). A coordinate expression ofthese two genes was evident in our system: the absence of a-ALBmessage in AFP" clones and high expression of this gene in AFP"1"

clones (Fig. 4). In addition, the transition from an unstable to a stableAFP+ clone was accompanied by a similar increase (about 2-fold) in

the activity of both proteins, suggesting that common mechanismscontrol the expression of AFP and a-ALB. It should be noted that wedid not detect any a-ALB gene expression in our AFPÂ±cell populations. Whether this gene is completely silent in AFPÂ±population, asin AFP" clones, or the absence of its message is due to the detection

limit of the Northern blot assay is not yet clear. The concept of3618



AFP EXPRESSION CORRELATES WITH HSR ON CHROMOSOME H

123456

AFP

ocALB

28S

Fig. 4. Northern blot analysis of endogeneous AFP mRNA and a-ALB mRNA levelsin total cellular RNA isolated from the McA-RH 7777 cell line (Lane 3) and its clones:stable AFP" clone 7H10 (Lane /), unstable AFP clone Hl l (Lane 2), unstable AFP*clone G6 (Lane 4), stable AFP* clones 7G3 (Lane 5), and 7G4 (Lane 6). Integrity of the

RNA and equal loading were verified by hybridization with 28S rRNA probe.

coordinate expression is supported by recent sequence data on theseproteins. Comparison of the conserved regions of the amino acidsequences of the proteins showed the a-ALB human homologue to

have 60.4% homology with AFP, whereas it has 54.8 and 41.2%homology with ALB and DBF, respectively (30). Nucleotide sequence comparisons of the a-ALB gene revealed 33% overall identity

with the rat AFP sequence versus 29 and 19% with the sequence of ratALB and DBF, respectively (19), suggesting that a-ALB genes are

phylogenetically closer to AFP.AFP expression during normal hepatocyte development (Fig. 5) and

the data obtained in the present study lead to the following conclusions. By selective pressure of the stabilizing cloning, the 7777 cellswere directed to either the AFP+ or AFP' differentiation pathway.

Continuous selection (seven cycles) maintained this particular differentiated state of 7777 cells in their clones and led to generation ofstable AFP+ and stable AFP" clones, corresponding to the final steps

of each differentiation pathway. The phenotype of stable AFP+ clones

was similar to that of embryonal hepatocytes, whereas the phenotypeof stable AFP' clones corresponded to that of adult hepatocytes. The

unstable clones reflected intermediate steps during the transition ofthe hepatocyte from the embryonal to the adult differentiated state,that is, AFP+ to AFP' order of differentiation. The hypothesis that

7777 hepatoma cells follow this order of differentiation is supportedby the data reported here as well as by other observations: (a) theactivation of embryonal synthesis is a feature of low-level differentiation, as reflected by its AFP+ phenotype; (b) the frequency of AFP~

phenotype clones in the 7777 population is much less (14%) than thatof AFP+ clones (51%; Refs. 14,15); (c) the transition from the AFP~

phenotype to the AFP+ phenotype is easier to achieve than vice versa',

(d) the morphological characteristics, i.e., the embryonal type ofelongated bipolar cells of AFP+ clones and trabecular-like structureformed by AFP' clones, correlate with the AFP"1"to AFP order of

differentiation. The pattern of a-ALB expression during normal de

velopment does not follow that of AFP but rather that of ALB and

DBP (Fig. 5; Ref. 19). Our results with the 7777 hepatoma clonesdemonstrated an a-ALB pattern of expression that corresponds well to

that proposed for AFP. Hence, this differentiation status of neoplasticcells might be attributable to control mechanisms different from thosepresent in the normal hepatocytes and, therefore, not active in normaldevelopment. Thus, 7777 hepatoma clones are a unique model systempresenting many advantages: (a) the clones have a similar origin; (b)each clone has its particular differentiation status; (c) each clone hasits pattern of expression of hepatocyte-specific differentiation mark

ers; and (d) the system by itself is complete providing the possibilityto analyze the AFP expression as internal controls.

Cytogenetic Analysis of AFP"+ and AFP~ McA-RH 7777 Hep

atoma Clones. From the panel of stable and unstable clones generated by two independent stabilizing clonings (herein and Ref. 15), thefollowing 15 clones were chosen for the present cytogenetic studies:mixed (AFP"1) clone D7; unstable AFP clones Hll, F4, TB6, andTB9; stable AFP~ clones 7H10, 7E7, and 7E10; unstable AFP+

clones G6, A3, and A3A; and stable AFP+ clones 7G3, 7G4, 7B7, and

7B9 (Fig. 1). The AFP phenotype of these clones and the parental7777 cell line were verified at the moment of metaphase preparationsby immunocytochemical staining (Fig. 3). As summarized in Table 1,the results of our cytogenetic studies clearly show that all examinedclones, as well as the parental 7777 hepatoma cell line, had thenear-tetraploid modal chromosome number with a great spread of

their chromosomes around their modes. Some clones such as TB9,7B7, and 7B9, showed more than one modal chromosome number(Table 1). These striking numerical aberrations of chromosomes wereaccompanied by multiple structural aberrations; an example of such acomplex karyotype is given in Fig. 6. Certain aberrations (Table 1)were commonly found: abnormal chromosome 3; interstitial deletion

123

- AFP

- Â«ALB

-ALB

-DBP

Fig. 5. Northern blot analysis of endogeneous mRNA levels of AFP, cr-ALB, ALB, andDBP in total cellular RNA isolated from 16-day-old fetal rat liver (Lane I), newborn ratliver (Lane 2), and adult rat liver (Lane 3).

3619




Table ! Cylugenetii- characteristics of McA-RH7777 cell line and its clones

AFPphenotypcAFP*UnstableAFP"StableAFPUnstableAFP+StableAFP*Clone7777'D7F3HllTB6TB97E77E107H10A3A3AG67B77B97G37G4Spread"64-11666-10262-11257-9854-7853-10569-7757-7961-7869-10257-9262-9866-13967-8458-11561-107ModalrÃ¤nge76-8472-8072-7874-7968-7673-7969-7770-7971-7872-8076-8268-7968-7873-7869-7771-80Mode(s)807875777675-7774757576807873,7673-757675%of cells atmode34252021185030222322182042524339StructuralCommon

'+

1P++

t(2;6)+

t(l;3)+l(2;6)t(2;6)+

t(l;3)+

lq++

t(9;15)16q++

t(8;?)+18q++

iv+

iq+t(l,3)+lq*+

Vaberrations

hsr ordsrParticulart(2;6)

1(8;?)13q*7p+16q+13q

+t(3;5)

7p+1(8;?)der(13)

16q+4p+6p+ 7q"*"+16q+

+1(8;?)

+" Chromosome number in 50 metaphase cells; counted consecutively.'Common aberrations included der(3), del(5), t(ll;?), 14q4, 15qf, and mars.' McA-RH 7777 rat hepatoma cell line.

of chromosome 5; translocations: t(ll;?), 14q+, and 15q+; and one or

more marker chromosomes of unknown origin. In addition, each clonepresented particular aberrations but none correlated to a specific AFPphenotype. Thus, all analyzed populations had a relatively uniformchromosome pattern, with marked numerical and structural aberrations. However, this pattern differed significantly from the near-

diploid 7777 hepatoma pattern reported earlier (31). Such karyotypicevolution is often provoked by continuous in vitro culturing.

Because AFP production is often linked to malignant transformation, the putative relationship between AFP production and chromo

somal compositions of hepatomas has been addressed by severalprevious studies (31-36). Becker et al. (32), using transplantable rat

hepatoma, reported a reverse relationship between karyotype and AFPproduction; rapidly growing tumors with aneuploid numbers demonstrate intense production of normal proteins and AFP, whereas near-

diploid tumors demonstrate little or no production of these proteins.These data as well as those obtained by others (34, 36) support theconclusion that aneuploid tumors produce excess AFP. However, nocorrelation between the rate of AFP production and chromosomenumber was found in rat ascites AH66 hepatoma clones producing

Â«ft 1 Â«\ f * s* t> Â»Â£

t(2;6)

vt i

der(5)

MÂ« M* flCdel(5) t(9;15)

9 10

!> â€¢' ;;Â«i* H i

11 12 16

18 20 mars

Fig. 6. Karyotype of a trypsin-Giemsa-banded cell from the stable AFP clone 7E10. The karyotype displays multiple structural and numerical aberrations.

3620




JÃ§Â»

Fig. 7. Examples of the Air-containing chromosome 14 found only in the stable AFP4

clones; arrows, hsr.

high and low amounts of AFP (35). Our present data are in agreementwith this conclusion because both AFP+ and AFP" 7777 hepatoma

clones presented patterns with marked aneuploidy and many structuralaberrations. However, our data pointed out interesting cytogeneticchanges related to AFP function. We found that only in stable AFP+

clones about one-half of the analyzed metaphases had either dsr

(clones 7G4 and 7B7) or hsr (clones 7G3 and 7B9) on the short armof chromosome 14. Several examples of hsr-containing chromosome

14 are given in Fig. 7. One cell of the 200 analyzed cells had dsr onboth chromosomes 14 and 7 (Fig. 8). This rare case is probably the

result of dsr break damages with their following attachments to otherchromosomes.

Thus, our approach (stabilizing cloning) to the study of the regulatorymechanisms leading to the malignant AFP phenotype generation isunique. Indeed, the analysis of stable clones presumed to have undergonea stable change at one or several regulatory sites responsible for AFPexpression allowed us to explore the origin of these changes. The changein our stable AFP+ clones manifested itself functionally by high AFP

production and cytogenetically by the occurrence of dsr and hsr on thechromosome carrying the genes of the ALB family. We herein demonstrated that ALB and DBF mRNA were not detectable in AFP * 7777

hepatoma clones, which indicates that their amplification is not likely. Incontrast, amplification of the a-ALB gene (a novel member of the ALBgene family) is possible, given the absence of its message in AFP" clones

and its high expression in AFP+ clones. Moreover, amplification is one

of the mechanisms by which cellular oncogenes may be altered in theirfunction, possibly leading to neoplastic transformation (37). Amplification of proto-oncogenes has been found as an occasional feature ofdiverse tumors (38, 39) and as a recurrent abnormality of specific proto-oncogenes, in particular, tumors (40-48). For example, N-myc amplifi

cation is a common event in untreated human neuroblastomas and wasshown to reside hsr (42, 47). Thus, our data support the hypothesis thatamplification of either structural or Â«.v-/rran.v-regulatorysequences of the

AFP gene is involved in maintaining its high expression; under the termrrans-regulatory sequences, we imply genes that can contribute in a

dominant manner to the malignant AFP phenotype.AFP is not only a tumor marker but also a developmental marker. We

concluded earlier that the phenotype of stable AFP* clones was similar

to that of embryonal hepatocytes, whereas the phenotype of stable AFPclones corresponded to that of adult hepatocytes. It is known that theestablishment and maintenance of the differentiated state by eukaryotic

. \

if

Â«! ({K, Â»!Â»=55? Â«"del(5)

der(3)

- idsr

C. Â»

11

17

12

ITI'

18

8

ili13 14

â€¢viif* *â€¢;â€¢â€¢â€¢

10t(1;10)

dsr

16

,8

19 20 mars

Fig. 8. Karyotype of a trypsin-Giemsa-banded cell from the stable AFP ' clone 7B7. The karyotype displays multiple structural and numerical aberrations; </.sr-containing

chromosomes 7 and 14.

3621



AFP EXPRESSION CORRELATES WITH HSR ON CHROMOSOME U

cells is under complex genetic control. Genetic events such as genemutation, chromosome aberrations, and gene amplification can affectnormal cell differentiation and modulate the cell phenotype. We presumed that these genetic events could lead to the reactivation of AFP+

phenotype and to the heterogeneity of 7777 hepatoma cells. The implication of gene mutation is unlikely because a nonmutational origin of theAFP phenotype heterogeneity was established in our earlier studies (14,15). We demonstrated herein that this heterogeneity is not associated withconsistent chromosome aberrations, either structural or numerical; however, the mechanism of control of the degree of gene amplificationprobably plays a major role in the generation of different levels of AFPexpression in these hepatoma cells. It is not yet known whether similargenetic events may control AFP expression in embryonal hepatocytes,the phenotype of which mimics our stable AFP * clones. Although the

case of gene amplification in mammalian cells and the mechanisms bywhich it occurs remain unclear, the phenomenon has generally beenfound in cells that have taken at least some of the steps toward neoplasticgrowth (49).

ACKNOWLEDGMENTS

We gratefully acknowledge Dr. Luc BÃ©langerfor pHDQ 105, pHDQ 833,and DS300; Dr. Nancy Cook for pSP65rDBP760; Dr. Tomas Sargent forpRAF 87 and pRSA 57; and Dr. Gunter SchÃ¼tzfor pSP6-cTAT-3" and

pSP(TO)2a. We also wish to thank Dr. Eric Krump and Dr. Carl SÃ©guinforcritical reading of the manuscript. We are especially grateful to Dr. ReemAl-Daccak for helpful discussion and critical review of the manuscript and to

Dr. Eugenia Kakpakova for performing our cytogenetic experiments.

REFERENCES

1. Abelev, G. I., Perova, S. D., Khramkova, N. I.. Poslnikova. Z. A., and Irlin, I. S.Production of embryonal a-globulin by transplantablc mouse hepatoma. Transplantation, /: 174-180, 1963.

2. Hirai. H., Nishi, S., Watabe, H.. and Tsukada, Y. Some chemical, experimental, andclinical investigations of a-fetoprotein. Gann, 14: 19-34, 1973.

3. Nechaud, B., and Uriel, J. AntigÃ¨nescellulaires transitories du fois de rat. I. Secretionet synthÃ¨se des protÃ©inesseriques foetospecifiques au cours du dÃ©veloppementet dela regeneration hÃ©patiques.Int. J. Cancer, 8: 71-80, 1971.

4. Rouslahti, E., and Seppala, M. a-fetoprolein in cancer and fetal development. Adv.Cancer Res., 29: 275-346, 1979.

5. Sell, S., Becker, F. F., Leffert, H., and Watabe, H. Expression of an oncodcvelop-mental gene product (a-fetoprotein) during fetal development and adult oncogenesis.Cancer Res., 36: 4239-4249, 1976.

6. Tatarinov, Y. S. New data of the embryo-specific antigenic components of humanblood serum. Fed. Proc., 24: T916-T918, 1965.

7. Gorin, M. B., Cooper, D. L., Eiferman, F.. Rijn. P., and Tilghman, S. M. Theevolution of a-fetoprotein and albumin. J. Biol. Chem., 256: 1954-1959, 1981.

8. Jagodzinski, L. L., Sargent, T. D.. Yang, M., Glackin, C., and Bonner, J. Sequencehomology between RNAs encoding rat a-fetoprotein and rat serum albumin.Biochemistry, 78: 3521-3525, 1981.

9. Law, S. W., and Dugaiczik, A. Homology between the primary structure of a-fetoprotein, deduced from a complete cDNA sequence, and serum albumin. Nature(Lond.), 291: 201-205, 1981.

HI. Brown, J. R. Structural origins of mammalian albumin. Fed. Proc., 35: 2141-2144,

1976.11. Eiferman, F. A., Young, P. R., Scot!, R. W., and Tilghman. S. M. Intragenic

amplification and divergence in the mouse a-fctoprotein gene. Nature (Lond.), 294:713-718, 1

12. Szpirer, J., Levan, G., Thorn. M., and Szpirer, C. Gene mapping in the rat bymouse-rat somatic cell hybridization: synleny of the albumin and a-fctoprotein genesand assignment to chromosome 14. Cytogenet. Cell Genet., 38: 142-149, 1984.

13. Becker, J. E., de Nechaud, B., and Potter. V. R. Two new rat hepatoma cell lines forstudying the unbalanced blocked ontogeny hypothesis. In: W. H. Ftshman and S. Sell(eds.), Oncodevelopmental Gene Expression, pp. 259-270. New York: AcademicPress, Inc., 1976.

14. Eraiser, T. L, and Khamzina, L. S. Phenotypic variability of cultured rat hepatoma cellpopulation in respect to a-fetoprotein synthesis. Int. J. Cancer, 42: 633-637, 1988.

15. Khamzina, L. S. Heterogeneity of a-fctoprotein expression in hepatomas. Annu. Rev.Med. USSR, 18: 32-37, 1987.

16. Solomon, E.. Borrow, J., and Goddard, A. D. Chromosome aberrations and cancer.Science (Washington DC), 254: 1153-1161, 1991.

17. Chomczynski. P.. and Sacchi, N. Single-step method of RNA isolation by acidguanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem., 762:156-159, 1987.

18. Turcotte, B., Guertin, M., Chevrette, M., and BÃ©langer,L. Rat al-fetoprotein messenger RNA: 5'-end sequence and glucocorticoid suppressed liver transcription in an

improved nuclear run-off assay. Nucleic Acids Res., 13: 2387-2398, 1985.19. BÃ©langer,L., Roy, S.. and Allard, D. New albumin gene 3' adjacent to the al-

feloprotein locus. J. Biol. Chem., 269: 5481-5484, 1994.20. Ray, K., Wang, X.. Zhao, M., and Cook, N. E. The rat vitamin D-binding protein

(Gc-Globulin) gene. J. Biol. Chem., 266: 6221-6229, 1991.

21. Sehinomiya. T.. Scherer, G.. Schmid, W., Zentgraf, H., and SchÃ¼tz,G. Isolation andcharacterization of the rat tyrosine aminotransfera.se gene. Proc. Nati. Acad. Sci.USA, 81: 1346-1350, 1984.

22. Schmid, W., Scherer, G., Danesch, U., Zentgraf, H.. Matthias, P.. Strange, C. M.,RÃ¶wekamp, W., and SchÃ¼tz,G. Isolation and characterization of the rat tryptophanoxygenase gene. EMBO J., /: 1287-1297, 1982.

23. Seabright, M. A. Rapid banding technique for human chromosomes. Lancet, 1:971-972, 1971.

24. Commitlee for a standardized karyotype of Rauus norvegicm. Standard karyotype ofthe Norway rat, Rallas norvegicus. Cytogenet. Cell Genet., 12: 199-205, 1973.

25. Levan, G. Nomenclature for G-bands in rat chromosomes. Hereditas, 77: 37-52,

1974.26. ChiÃ¹,J. F., Decha-Umphai, W., and Commer, P. al-Fetoprotein mRNA of rat yolk

sac and hepaloma. Nucleic Acids Res., 7: 239-249, 1979.

27. Eraiser, T. L., and Lazarjevitch, N. New highly efficient method of transfection. Mol.Biol. (Russia), in press, 1995.

28. Deschatrette, J., Moore, E. E., Dubois, M., and Weiss, M. C. Dcdifferentiated variantsof a rat hepatoma: reversion analysis. Cell, 19: 1045-1051, 1980.

29. Moore, E. E.. and Weiss, M. C. Selective isolation of stable and unstable dediffer-entiated variants from a rat hepatoma cell line. J. Cell. Physiol., ///: 1-8, 1982.

30. Lichenstein, H. S., Lyons, D. E., WÃ¼rfel,M. M., Johnson, D. A., McGinley, M. D.,Leidli, J. C, Trollinger, D. B., Mayer, J. P., Wright, S. D., and Zukowski, M. M.Afamin is a new member of the albumin, a-fetoprotein, and vitamin D-bindingprotein gene family. J. Biol. Chem., 269: 201-205, 1994.

31. Wolman, S. R., Cohen, T. I., and Becker, F. F. Chromosome analysis of hepatocel-lular carcinoma 7777 and correlation with a-fetoprotein production. Cancer Res., 37:2624-2627, 1977.

32. Becker, F. F., Wolman, S. R., Asofsky, R., and Sell, S. Sequential analysis oftransplantable hepatoccllular carcinomas. Cancer Res., 35: 3021-3026, 1975.

33. Isaka, H., Umehara, S., Hirai, H., and Tsukada, Y. Development of variant tumor cellsof Yoshida ascites sarcoma producing a-fetoprotein. Gann, 6.Ã•:63-71, 1972.

34. Sell, S., and Morris, H. P. Relationship of rat a-fetoprotein to growth rate andchromosome composition of Morris hepatoma. Cancer Res., 34: 1413-1416, 1974.

35. Tsukada. Y., Mikuni. M.. and Hirai. H. In vitnt cloning of a rat ascites hepatoma cellline. AH66, with special reference to a-fctoprotcin synthesis. Int. J. Cancer, 13:196-202, 1974.

36. Yoo, T-J., Kuo, C. Y., Patii, S. R., Kim, U., Cancilla, P., and Ackerman, L. R. Lossof a-fetoprotein in rat hepatoma culture cells. Int. J. Cancer, 24: 184-192, 1979.

37. Bishop. J. M. The molecular genetics of cancer. Science (Washington DC), 2.Ã•5:305-311, 1987.

38. Alitalo, K., and Schwab, M. Oncogene amplification in tumor cells. Adv. CancerRes., 47: 235-281, 1986.

39. Varmus, H. E. The molecular genetics of cellular oncogenes. Annu. Rev. Genet., 18:553-612, 1984.

40. Alitalo, K., Schwab, M., Lin, C. C., Varmus, H. E., and Bishop, J. M. Homogeneouslystaining chromosomal regions contain amplified copies of an abundantly expressedcellular oncogene (c-mvf) in malignant neurocndocrinc cells from human coloncarcinoma. Proc. Nati. Ã€cad. Sci. USA, 80: 1707-1711, 1983.

41. Escol, C., Theillet, C., Lidereau, R., Spyralos, F., ChÃ¡mpeme, M-H., Gest, J., andCallahan, R. Genetic alteration of the c-mvr protooncogene (MYC) in human primarybreast carcinomas. Proc. Nati. Acad. Sci! USA, 83: 4834-4838, 1986.

42. Kohl, N. E., Kanda, N., Schreck, R. R., Bruns, G., Lall, S. A., Gilbert, F., and Alt.F. W. Transposition and amplification of oncogcne-relatcd sequences in humanneuroblastomas. Cell, 35: 359-367, 1983.

43. Libermann, T. A.. Nusbaum, H. R., RazÃ³n, N., Kris, R., Lax, I., Soreq, H., Whittle,N., Waterfield, M. D., Ullrich, A., and Schlessinger. J. Amplification, enhancedexpression and possible rearrangement of EGF receptor gene in primary human braintumours of glial origin. Nature (Lond.), 313: 144-147, 1985.

44. Little, C. D., Nau, M. M., Carney, D. N., Gazdar. A. F., and Minna, J. D. Amplification and expression of the c-mvr oncogene in human lung cancer cell lines. Nature(Lond.). 306: 194-196, 1983.

45. Schwab, M.. Alitalo. K., Varmus, H. E., Bishop, J. M., and George, D. A cellularoncogene (c-Ki-ra.v) is amplified, overexpressed, and located within karyotypicabnormalities in mouse adrenocortical tumour cells. Nature (Lond.), 303: 497-501,1983.

46. Schwab, M., Alitalo, K., Klempnauer, K-H., Varmus, H. E., Bishop, J. M., Gilbert, F.,

Brodeur, G.. Goldstein, M., and Trent, J. Amplified DNA with limited homology tomvc cellular oncogene is shared by human neuroblastoma cell lines and neuroblastoma tumor. Nature (Lond.), 305: 245-248, 1983.

47. Schwab, M., Varmus, H. E., Bishop, J. M., Grzeschik, K-H., Naylor, S. L., Sakaguchi,A. Y., Brodeur, G., and Trent, J. Chromosome localization in normal human cells andneuroblastomas of a gene related to c-mvr. Nature (Lond.), 308: 288-291, 1984.

48. Schwab, M., Ramsay, G., Alitalo, K., Varmus, H. E., Bishop. J. M., Martinsson, T.,Levan, G., and Levan. A. Amplification and enhanced expression of the c-mvroncogene in mouse SEWA tumor cells. Nature (Lond.), 315: 345-347, 1985.

49. Stark G. R., and Wahl, G. M. Gene amplification. Annu. Rev. Biochem., 53:447-491, 1984.

3622



1995;55:3615-3622. Cancer Res Leila Khamzina, Tatjana Eraiser and Pierre Borgeat Staining Regions on Chromosome 14Hepatoma: Correlation with the Occurrence of Homogeneously

-Fetoprotein (AFP) Expression in Clones of McA-RH 7777 Ratα

Updated version

http://cancerres.aacrjournals.org/content/55/16/3615

Access the most recent version of this article at:

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

Subscriptions

Reprints and

[email protected] 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/55/16/3615To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

