Target - PNAS · 6738 Thepublication costsofthis article weredefrayedin part bypagecharge payment. Thisarticle musttherefore beherebymarked"advertisement" Downloaded at Microsoft

Proc. Natl. Acad. Sci. USAVol. 81, pp. 6738-6742, November 1984Cell Biology

Target polypeptide of a carcinogen is associated with normalmitosis and carcinogen-induced hyperplasias in adult hepatocytes

(liver/cytoplasln/nuclei/N-2-fluorenylacetamide/azo dye)

R. PHILIP CUSTER AND SAM SOROF*Institute for Cancer Research, Fox Chase Cancer Center, Philadelphia, PA 19111

Communicated by Allan H. Conney, July 11, 1984

ABSTRACT Normal rat liver cytosol was found previous-ly to contain a 14,000-dalton polypeptide that is the principaltarget of the carcinogen N-2-fluorenylacetamide (2-acetyl-aminofluorene) early during hepatocarcinogenesis. By usingantiserum that identifies the 14,000-dalton polypeptide in livercytosol, an immunologically related 17,500-dalton polypeptidewas shown to be present in isolated normal liver nuclei, tightlybound to chromatin. We report here that the 14,000-daltonpolypeptide is associated with cell multiplication in normaladult hepatocytes. Of 150 hepatocytes in five stages of mitosis,all invariably displayed a greatly increased concentration ofthe 14,000-dalton polypeptide in cytoplasm, compared to he-patocytes in interphase, by immunostaining with peroxidase,antiperoxidase complex. In contrast, mitosis did not appear toaffect the level of the 17,500-dalton polypeptide in nuclei. Asmall population of high-polyploid hepatocytes with large nu-clei had elevated intensities of cytoplasmic immunostpin thatapproached that in mitotic cells. The increased level of dis-cernible 14,000-dalton polypeptide target of the carcinogen incytoplasim is a marker of the rare mitotic hepatocytes in nor-mal adult rat liver. Ingestion of the liver carcinogen N-2-flu-orenylacetamide for 5 wk or an azocarcinogen, 3'-methyl-4-dimethylaminoazobenzene, for 4 wk brought about early fociof hyperplastic bepatocytes in which there was a great over-load of detectable 14,000-dalton target polypeptide in their cy-toplasm and a near absence of the 17,500-dalton polypeptide inmost nuclei. In contrast, the cytoplasm and nuclei in sur-rounding morphologically nontransformed hepatocytes of liv-ers of the carcinogen-fed rats immunostained to a much small-er degree than in normal adult hepatocytes. Removal of theazo-dye carcinogen for 4 wk resulted in the disappearance ofmost hyperplastic foci. A few foci did persist and had similarapparent abnormal levels of the two polypeptides. The highconcentration of discernible 14,000-dalton target polypeptidein the cytoplasm of hepatocytes in the hyperplastic foci corre-lates with their known proliferation, whereas the low cytoplas-mic level in the surrounding hepatocytes is consistent with theknown inhibition of their mitosis by these carcinogens. Thecollected evidence appears to join together a chemical carcino-gen, a cytoplasmic principal target polypeptide, a chromatin-bound polypeptide, mitosis in normal adult hepatocytes, earlyand persistent hyperplastic foci caused by carcinogens, and theinhibition of mitosis by carcinogens in surrounding parenchy-ma in preneoplastic livers.

Early events in chemical carcinogenesis involve the interac-tions of carcinogens with macromolecules in susceptible or-gans (1, 2). Several chemical carcinogens react preferentiallywith single target proteins in organs undergoing carcinogene-sis. Four such principal target proteins have been purifiedand characterized. All are different, basic, and found only inorgan cytosols (discussed in ref. 3). In addition, nuclear non-

histone proteins and histones react with carcinogens to smalldegrees (4-6). The functional significance of almost all ofthese interactions in oncogenesis is unknown.A 14,000-dalton polypeptide in normal liver of rats is the

principal cytosolic target protein of the carcinogen, N-2-fluorenylacetamide (FAA; 2-acetylaminofluorene) early dur-ing hepatocarcinogenesis (7, 8). Short-term ingestion of anyof the carcinogens, the aromatic amide FAA, the aminoazodye 3'-methyl-4-dimethylaminoazobenzene, or the aminoacid analog ethionine causes a marked loss of the ability ofliver to make the carcinogen-polypeptide complex fromFAA and a great reduction in the content of the target poly-peptide itself (7, 8). The target polypeptide from normal liverwas purified to homogeneity according to molecular size,molecular charge, and immunogenicity and was character-ized (3, 9). Specific rabbit antiserum was prepared that re-acted with only that polypeptide in liver cytosol (3) and withboth nuclei and cytoplasm in hepatocytes in histological sec-tions of normal adult liver (9). An immunologically relatedform of the cytosolic 14,000-dalton target polypeptide wasfound to exist in liver nuclei of normal adult rats as a 17,500-dalton polypeptide that was firmly bound to chromatin (9).These findings suggest a connection between a principalpolypeptide target of a carcinogen in cytoplasm and a poly-peptide in chromatin, where carcinogens are generallythought to act in oncogenesis (1, 2). We report here that anelevated level of detectable cytoplasmic 14,000-dalton targetpolypeptide of the liver carcinogen FAA is associated withmitosis in normal adult hepatocytes and with early and per-sistent hyperplasias caused by liver carcinogens.

MATERIALS AND METHODSRats and Diets. Normal rats (162-187 g) of both sexes of

the Fischer 344 CDF strain (Charles River Breeding Labora-tories) were fed a commercial stock diet (Wayne Lab-Blox,Allied Mills, Chicago, IL) and acidified tap water ad lib.They were housed in screen-bottomed cages at 229C with a 6a.m./6 p.m. light cycle. Male CDF rats (108-257 g) were fedalso the liver carcinogen FAA at a level of 0.03% in a graindiet (10). Male rats (190-246 g) of the Sprague-Dawley SDstrain (Charles River Breeding Laboratories) were likewisegiven the hepatocarcinogen 3'-methyl-4-dimethylaminoazo-benzene at a concentration of 0.058% in a synthetic diet (diet3 in ref. 11). All rats were sacrificed rapidly in the animalquarters with minimal prior excitement.Immunohistochemical Analyses. Immediately after animal

sacrifice, thin slices of livers were fixed in 10% buffered fqr-malin for 1 day. Paraffin blocks and successive 5-,um sec-tions were prepared and stored desiccated at 1-4°C. Sectionswere deparaffinized and exposed to specific rabbit antiseraagainst the cytosolic 14,000-dalton target polypeptide of thecarcinogen FAA (3, 9), followed by staining with peroxidase-

Abbreviations: FAA, N-2-fluorenylaretamide (2-acetylaminofluor-ene); PAP, peroxidase-antiperoxidase.*To whom reprint requests should be addressed.

6738

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

Dow

nloa

ded

by g

uest

on

Dec

embe

r 9,

202

0

Proc. Natl. Acad. Sci. USA 81 (1984) 6739

antiperoxidase (PAP) complex (DAKO; Accurate Chemicaland Scientific, Westbury, NY) and 3,3'-diaminobenzidinetetrahydrochloride (12). Control sections that were treatedwith respective preimmune rabbit sera were virtually un-

stained. The immunostain in hepatocytes thereby detectedthe 14,000-dalton polypeptide in cytoplasm and the 17,500-dalton polypeptide in nuclei (3, 9). Immunohistochemicalevaluations involved comparisons between three types ofadjacent serial sections-i.e., those stained conventionallywith hematoxylin and eosin, those stained immunologicallyonly, and those stained both immunologically and with he-matoxylin in order to visualize nuclei and background struc-tures. Counterstaining with hematoxylin was also necessaryfor detection of the rare mitotic hepatocytes in liver sections.The heavily immunostained cytoplasm in mitotic hepato-cytes obscured the nuclei in the absence of counterstain.

'ai

WF

'Jjj ' '

,- s * - t >

e

FIG. 1. Normal adult hepatocytes. Marked elevations in the lev-el of detectable 14,000-dalton target polypeptide in the cytoplasm ofnormal adult mitotic hepatocytes and in the cytoplasm of most high-polyploid hepatocytes, relative to adjacent interphase hepatocytes.Shown are the intense cytoplasmic immunoreactions that invariablyoccur during the stages of mitosis-prophase (a), metaphase (b),anaphase (c), and telophase (d)-and in most high-polyploid hepato-cytes (e). The cytoplasm of low-polyploid cells reacts only slightly.Dense immunostaining with the PAP complex and coarse granulari-ty blur the sharpness of the mitotic figures. The brown immunostainof mitotic nuclei, evident in color but not in black and white becauseof necessary counterstain with hematoxylin, indicates that the levelof the 17,500-dalton polypeptide in nuclei of dividing hepatocytes isessentially like those in interphase. (PAP-hematoxylin; x600.)

RESULTS14,000-Dalton Target Polypeptide and Mitosis in Normal

Adult Hepatocytes. Cytoplasm of mitotic hepatocytes con-tains the highest concentration of discernible 14,000-daltontarget polypeptide in normal adult rat liver. Without excep-tion, of 150 mitotic hepatocytes, all displayed markedly high-er intensities of the granular cytoplasmic immunostain thandid the cytoplasm of interphase hepatocytes. The elevatedlevel of cytoplasmic immunostaining was invariably ob-served in five phases of mitosis-i.e., prophase, metaphase,anaphase, telophase (Fig. 1 a-d), and cytokinesis (notshown), regardless of the presence or absence of the nuclearmembrane. In contrast, mitosis had no apparent effect onthe concentration of immunohistochemically detectable nu-clear 17,500-dalton polypeptide (Fig. 1 a-d; see the summa-ry in Table 1).

14,000-Dalton Polypeptide in Polyploid Hepatocytes. Theonly other liver cells with intense staining in cytoplasm wererelatively few hepatocytes of high polyploidy. Nuclei of he-patocytes in normal liver exhibit a range of diameters reflect-ing their increased ploidy with the age of the rat (13). Thehepatocytes with intermediate-sized nuclei, presumably tet-raploids, had cytoplasmic staining levels of the polypeptidesimilar to diploids at the same intralobular locations. On theother hand, the cytoplasm of certain hepatocytes with stilllarger nuclei, presumably octaploids, had elevated concen-trations approaching that in mitotic cells (Fig. le). In con-trast, the nuclei themselves of all polyploidy had detectableconcentrations of the chromatin-bound 17,500-dalton poly-peptide like those in diploid nuclei of the same intralobularregion, sometimes even less (Fig. le; see the summary inTable 1).

14,000-Dalton Polypeptide in Early Hyperplastic Foci.Short-term ingestion of liver carcinogen brings about earlyhyperplastic foci in which the hepatocytes display greatly al-tered levels of the two polypeptides. Feeding the carcinogenFAA for 5 wk or 3'-methyl-4-dimethylaminoazobenzene for4 wk results in foci of proliferating hepatocytes (8, 14-16).The morphologically nontransformed cells in the surround-ing parenchyma exhibited little or no cytoplasmic or nuclear

Table 1. Presence of the 14,000-dalton and 17,500-daltonpolypeptides in rat hepatocytes

Hepatocytes Cytoplasm* Nucleus*Normal adult

Perilobular diploidsCentrilobular diploidsMitoticLow polyploidsHigh polyploidsPolypeptides

Moderate (9)Low (9)HighLike diploidsMostly highCytosolic 14-kDa

target of FAA(3, 7, 8)

Very high (9)Low (9)HightLike diploidsLike diploidsChromatin-bound

17.5-kDaantigen (9)

Morphologically nontransformed (carcinogen-exposed)Short term FAA (5 wk) Generally low Generally lowShort term azo dye Generally low Generally low

(4 wk)fHyperplasticFAA early foci (5 wk) High Mostly absentAzo-dye early foci Very high Mostly absent

(4 wk)t High Mostly absentAzo-dye persistent foci

(4 wk of carcinogen, 4wk of no carcinogen)*

*Numbers in parentheses are reference citations.tThe brown immunostain of mitotic nuclei was evident in color butnot in black and white photographs due to the hematoxylin counter-stain that was necessary to detect mitotic cells.tAzo dye, 3'-methyl-4-dimethylaminoazobenzene.

Cell Biology: Custer and Sorof

Oki

LA

Dow

nloa

ded

by g

uest

on

Dec

embe

r 9,

202

0

6740 Cell Biology: Custer and Sorof

immunoreactivity, indicative of apparent loss of the twopolypeptides (Fig. 2 b and d). In contrast, the early hyper-plastic foci consisted of hepatocytes with abundant cyto-plasm that intensely immunoreacted. Fig. 2 a-f shows adja-cent serial sections that were stained either with the specificantiserum against the 14,000-dalton polypeptide or with he-matoxylin and eosin. The morphological outlines of the dark-ly immunostained cell clusters matched those of the hyper-plastic foci. The high intensity of the cytoplasmic immuno-stain at least equaled and usually exceeded that of normalmitotic hepatocytes. In contrast to the cytoplasm, most ofthe nuclei within the hyperplastic foci were not stained bythe antiserum. Thus, as summarized in Table 1, the two liver

I. ,.F

te'' A

A.-..w., s;_4. eUflo'

Proc. NatL. Acad. Sci. USA 81 (1984)

carcinogens produce early islands of hyperplastic hepato-cytes in which the cytoplasm is apparently heavily over-loaded with 14,000-dalton polypeptide and most nuclei virtu-ally lack the 17,500-dalton polypeptide, all surrounded bycells that have little of each polypeptide.

14,000-Dalton Polypeptide in Persistent Hyperplastic Foci.Chemically induced liver tumors appear to arise from withinpersistent foci of hyperplastic hepatocytes that remain afterthe feeding and then long-term removal of carcinogen (14,15, 17). The question arose of whether these more pro-gressed hyperplastic hepatocytes continue to be demonstra-bly abnormal in regard to the two polypeptides. Rats werefed the diet containing the azocarcinogen 3'-methyl-4-di-

A- ; ~~~~~~,.$;,S4->~~~A".. ,.,C ; i;t* M '*-t.,;)4 . p S-

FXU.¢ e

e.

I, mC

C .4E.

FIG. 2. Early hyperplastic and polypeptide foci of hepatocytes in rats fed carcinogens. Immunohistochemical detection of increased con-centrations of cytoplasmic 14,000-dalton target polypeptide and loss of the 17,500-dalton nuclear polypeptide in hepatocytes of early hyperplas-tic foci in rats fed carcinogens. Immunochemistry used the antiserum against the 14,000-dalton target polypeptide and the PAP complex. Bileductular proliferation and hepatocytic hyperplastic buds and foci are early changes caused by the hepatocarcinogens 3'-methyl-4-dimethyl-aminoazobenzene and FAA, the former acting more quickly and aggressively. (a) Many hyperplastic foci of hepatocytes are strongly immuno-reactive in a liver section from a male Sprague-Dawley rat fed the 3'-methyl-4-dimethylaminoazobenzene diet for 4 wk. The interveningparenchyma has been vastly disrupted and largely replaced by oval cells (compare with c). (PAP; x 50.) (b) A small early focus of hyperplastichepatocytes with heavily immunostained cytoplasm and mostly unstained nuclei in the adjacent serial section. Nuclei of surrounding oval cellsare barely visible. Structures of a portal triad (p) are completely unstained. (PAP; X190.) (c) The same hyperplastic focus in the adjacent serialsection stained by hematoxylin and eosin. (hematoxylin/eosin; x 190.) (d) Strongly immunoreactive hepatocytes within hyperplastic buds in aliver section from a male CDF rat fed the FAA diet for 5 wk. Liver lobules around the central veins (c) are better preserved than with 3'-methyl-4-dimethylaminoazobenzene. Immunoreactive hyperplastic hepatocytes surround portal triads (p). In contrast, the surrounding hepatocytes inlobular midzones and centers are poorly immunostained, and the triad structures themselves are not immunostained. (PAP; x50.) (e) Aperiportal cluster of large hyperplastic hepatocytes with densely immunostained cytoplasm. Most nuclei are unstained. Many intensely stainedcells are sectioned tangentially outside the nuclear plane. (PAP; x 190.) (f) Buds of proliferative hepatocytes of a periportal region in an adjacentserial section stained with hematoxylin and eosin. The hyperplastic hepatocytes are large with abundant cytoplasm. A few cells have polyploidnuclei. (hematoxylin/eosin; x 190.)

Dow

nloa

ded

by g

uest

on

Dec

embe

r 9,

202

0

Proc. Natl. Acad. Sci. USA 81 (1984) 6741

lily./A.

aa

VV

FiG. 3. Persistent hyperplastic and polypeptide foci of hepato-cytes. Elevated concentration of the cytoplasmic 14.000-daltonpolypeptide and near absence of nuclear 17,500-dalton polypeptidedetected in persistent foci of hyperplastic hepatocytes. A maleSprague-Dawley rat was fed 3'-methyl-4-dimethylaminoazobenzenediet for 4 wk, followed by the control diet (no carcinogen) for 4 wk.Much of the liver structure reverted to normal, except for few per-sistent foci of hyperplastic hepatocytes whose cytoplasm wasstained intensely by PAP immunohistochemistry. (a) A persistentfocus of hyperplastic hepatocytes. (hematoxylin/eosin; x80.) (b)Retention of strong immunoreactivity in the cytoplasm of most he-patocytes in the persistent hyperplastic focus in an adjacent serialsection. In contrast, the nuclei are virtually unstained. The sur-rounding parenchyma reacted unevenly, except for intense cyto-plasmic staining of some large high-polyploid hepatocytes. The com-pressed portal triads at the focal margin failed to stain. (PAP; x80.)

methylaminoazobenzene for 4 wk as above, followed by thesame diet without carcinogen for an additional 4 wk. Mosthepatocytes in the few persistent hyperplastic foci displayednearly the same high intensity of immunostain in their abun-dant cytoplasm and almost none in their nuclei (Fig. 3). Incontrast, the morphologically nontransformed cells in thesurrounding parenchyma were unevenly immunostained incytoplasm and nuclei (Fig. 3). Thus, as summarized in Table1 and Fig. 4, like the hepatocytes in the early hyperplasticfoci, the hepatocytes in the persistent hyperplastic foci ex-

hibited the same overload of detectable target polypeptide incytoplasm and the near absence in nuclei, while the sur-rounding hepatocytes appeared to be partially recovered.

DISCUSSIONThe 14,000-dalton polypeptide that is the principal target ofthe carcinogen FAA appears to be associated with cell multi-plication of normal adult hepatocytes. The cytoplasm of he-patocytes in all stages of mitosis stands out as apparentlycontaining highly elevated levels of the protein. It is note-worthy that at any one time, fewer than 1 in 1000 hepato-cytes are dividing in normal adult rat liver (13). Thus, thepresence of exceptionally high concentration of detectabletarget polypeptide in cytoplasm is a marker of the rare mitot-ic hepatocytes in normal adult liver. However, mitosis doesnot perceptibly alter the level of the nuclear 17,500-daltonpolypeptide. Similar concentrations of this polypeptide ap-pear to be present in nuclei throughout all phases of mitosis,even when the nuclear membrane disappears, consistentwith the tight affinity of that polypeptide with liver chroma-tin (9).Some, but not all, polyploid hepatocytes with large nuclei,

presumably octaploids, have cytoplasmic concentrations ofthe 14,000-dalton target polypeptide that approach those ofmitotic hepatocytes in normal adult liver. Nevertheless, oc-taploid hepatocytes are not able to undergo cell division (13).One can speculate that genic derangements accompanyinghigh polyploidy in some hepatocytes may lead to elevatedlevels of the polypeptide as a result of more gene copies,block(s) in premitosis, enhanced synthesis, or lessened deg-radation.The presence of the 14,000-dalton polypeptide is respon-

sive to the state of differentiation in hepatocytes. Normaladult liver differs from neonatal liver and regenerating liverafter partial hepatectomy in many regards (18). Many but notall isozymes and mRNAs in regenerating liver are reportedlycharacteristic of embryonic liver (18, 19). In fact, hepato-cytes of neonatal rat liver have none of the two polypeptides,and hepatocytes of regenerating liver have only low discern-ible levels, even in the presence of high mitotic activity (un-published data). The 14,000-dalton polypeptide, therefore,appears not to be necessary for cell multiplication in allstages of differentiation of hepatocytes. Rather, the polypep-tide is associated specifically with mitosis of hepatocytes innormal adult liver and with carcinogen-induced hyperpla-sias, perhaps acting as a modifier of mitosis in those states ofdifferentiation in hepatocytes.Changes in the immunostaining of liver sections have been

interpreted in terms of the concentrations of the two poly-peptides. It seems plausible that the observed differences

FOCAL HYPERPLASIA;I:..,.::. 4K-::::::::***:::..........-

NORMALINTERPHASE

p

NO'ARCINOGEN

....... ........

. ...........CARCINOGENSHORT TERM

MITOSIS FAA, 3'-Me-DAE3XMORPHOLOGICALLYNON-TRANSFORMED

PARENCH YMA

'ERSISTENT FOCALHYPERPLASIA

14K...'

REVERSEDPARENCHYMA

FIG. 4. Representation of the discerned levels of the cytoplasmic 14,000-dalton target polypeptide of the carcinogen FAA and the nuclear17,500-dalton polypeptide in normal interphase and mitotic adult hepatocytes, in morphologically nontransformed hepatocytes exposed to

carcinogens, and in hepatocytes of early and persistent hyperplastic foci caused by carcinogens. The apparent concentrations of the twopolypeptides are depicted as indicated in Table 1. 3'-Me-DAB, 3'-methyl-4-dimethylaminoazobenzene.

Cell Biology: Custer and Sorof

Dow

nloa

ded

by g

uest

on

Dec

embe

r 9,

202

0

6742 Cell Biology: Custer and Sorof

may also reflect in part alterations in the accessibility of theantibodies to the polypeptides-e.g., the extent of complex-ation with nucleic acids, as in a recent study (20). Thatchanges in the concentrations of the two polypeptides mayactually be involved is supported, first, by our earlier dem-onstration of the depression in the amount of chemically de-tected 14,000-dalton polypeptide in rat liver cytosols afteringestion of three carcinogens (7); second, by the use ofpolyclonal antisera that presumably detect many antigenicdeterminants of both polypeptides (refs. 3 and 9 and this re-port); and possibly third, by the previous finding by Gronowand Thackrah of the disappearance of a chemically detectednonhistone 17,500-dalton polypeptide from the euchromatinof livers of rats administered the hepatocarcinogen diethylni-trosamine (21). Nevertheless, it is important to note that,even if alterations in accessibility and complexation are alsoinvolved, the essentials as outlined in Table 1 and Fig. 4 re-main the same-only the nature of the underlying changesmay in part be different.Hepatocytes in early hyperplastic foci caused by the liver

carcinogens, FAA and 3'-methyl-4-dimethylaminoazoben-zene, contain a marked overload of perceptible 14,000-dal-ton target polypeptide in their cytoplasm and a near absenceof the 17,500-dalton polypeptide in most nuclei. This imbal-ance of the two polypeptides constitutes a marker of the hy-perplastic hepatocytes. In contrast, the morphologicallynontransformed hepatocytes in the surrounding parenchymahave little or none of the two polypeptides. Long-term with-drawal of the azocarcinogen results in the reversal of thesealterations in most hepatocytes. Importantly, the few hyper-plastic foci that do persist have measurably abnormal levelsof the two polypeptides like those of the early hyperplasticfoci. It seems reasonable to deduce from the reversibility ofmost early polypeptide foci that genic changes are not at thebasis of those aberrations. However, the finding of apparent-ly irreversible, persistent polypeptide foci suggests that he-reditable alterations are involved in their perpetuation.Farber and his co-workers have shown that chemically in-duced liver tumors appear to originate from within persistenthyperplastic foci and have inferred a direct lineage betweenthem (14, 15, 17). Similarly, the present finding of an overlapbetween the persistent hyperplastic foci and polypeptide is-lands suggests a connection between the persistent foci ofthe two polypeptides and neoplasia. Such a relationship issupported by our recent observations of close physical con-tacts between the polypeptide foci and early liver neo-plasms.The 14,000-dalton target polypeptide of the liver carcino-

gen acts in a manner that reflects the growth activity of hepa-tocytes in response to the two carcinogens (Table 1 and Fig.4). High levels of that molecule in cytoplasm are associatedwith mitosis in normal adult hepatocytes and with the carcin-ogen-induced hyperplasias. In contrast, markedly depressedconcentrations of the polypeptide are present in the cyto-plasm of the surrounding morphologically nontransformedhepatocytes in the carcinogen-fed rats. Many investigatorspreviously have implicated the inhibition of mitosis by car-cinogens in the morphologically nontransformed cells andthe lack of such inhibition in initiated cells as being responsi-ble for the focal regeneration, focal transformation, and focalneoplastic progression in liver oncogenesis (reviewed in refs.14 and 15). Accordingly, Solt, Farber, and their colleagues

have employed FAA, 3'-methyl-4-dimethylaminoazoben-zene, and ethionine to differentially inhibit mitosis in theirmodel of rapid production of hyperplastic nodules in adultrat liver (reviewed in refs. 14 and 15). Although the 14,000-dalton polypeptide has not been experimentally linked to anygrowth-stimulating or cell-transforming activity or to a prod-uct of any cellular oncogene, the reported inhibitions of mi-tosis in the morphologically nontransformed hepatocytes bythese carcinogens and the proliferation of hepatocytes in hy-perplastic foci are consistent with the observed differencesin their apparent content of the 14,000-dalton polypeptide.Thus, the collected evidence appears to join together achemical carcinogen, a cytoplasmic target polypeptide, achromatin-bound polypeptide, mitosis in normal adult hepa-tocytes, early and persistent hyperplastic foci caused by car-cinogens, and the inhibition of mitosis by carcinogens in thesurrounding parenchyma in preneoplastic livers.

We thank Bernice B. Althouse, Marian J. Mastrangelo, and GraceKroetz for excellent assistance. This study was supported in part byNational Institutes of Health Grants CA-05945, CA-30036, CA-06927, and RR-05539 and by an appropriation from the Common-wealth of Pennsylvania.

1. Miller, E. C. & Miller, J. A. (1981) Cancer 47, 1055-1064.2. Miller, E. C. & Miller, J. A. (1981) Cancer 47, 2327-2345.3. Blackburn, G. R., Schnabel, S. J., Danley, J. M., Hogue-An-

geletti, R. A. & Sorof, S. (1982) Cancer Res. 42, 4664-4672.4. Allfrey, V. G. & Boffa, L. C. (1979) in The Cell Nucleus, ed.

Busch, H. (Academic, New York), Vol. 7, pp. 521-562.5. Zytkovicz, T. H., Moses, H. L. & Spelsberg, T. C. (1979) in

The Cell Nucleus, ed. Busch, H. (Academic, New York), Vol.7, pp. 479-517.

6. MacLeod, M. C., Pelling, J. C., Slaga, T. J., Noghrei-Nik-bakht, P. A., Mansfield, B. K. & Selkirk, J. K. (1983) Prog.Nucleic Acid Res. Mol. Biol. 29, 111-115.

7. Blackburn, G. R., Andrews, J. P., Rao, K. V. K. & Sorof, S.(1980) Cancer Res. 40, 4688-4693.

8. Blackburn, G. R., Andrews, J. P., Custer, R. P. & Sorof, S.(1981) Cancer Res. 41, 4039-4049.

9. Vinores, S. A., Churey, J. J., Haller, J. M., Schnabel, S. J.,Custer, R. P. & Sorof, S. (1984) Proc. Natl. Acad. Sci. USA81, 2092-2096.

10. Miller, E. C., Miller, J. A., Sandin, R. B. & Brown, R. K.(1949) Cancer Res. 9, 504-509.

11. Miller, E. C., Miller, J. A., Kline, B. E. & Rusch, H. P. (1948)J. Exp. Med. 88, 89-98.

12. Boenisch, T., ed. (1980) Reference Guide Series 1: PAP/Im-munoperoxidase (DAKO Corp., Santa Barbara, CA), pp. 3-18.

13. Grisham, J. W. (1969) Recent Results Cancer Res. 17, 28-43.14. Farber, E. & Cameron, R. (1980) Adv. Cancer Res. 31, 125-

226.15. Farber, E., Cameron, R. G., Laishes, B., Lin, J.-C., Medline,

A., Ogawa, K. & Solt, D. B. (1979) in Carcinogens: Identifica-tion and Mechanisms of Action, eds. Griffin, A. C. & Shaw,C. R. (Raven, New York), pp. 319-335.

16. Teebor, G. W. & Becker, F. F. (1971) Cancer Res. 31, 1-3.17. Solt, D. B., Medline, A. & Farber, E. (1977) Am. J. Pathol. 88,

595-618.18. Tsanev, R. (1975) in Cell Cycle and Cell Differentiation, eds.

Reinert, J. & Holtzer, H. (Springer, Berlin), Vol. 7, pp. 197-248.

19. Fausto, N. & Shank, P. R. (1983) Hepatology 3, 1016-1023.20. Rodman, T. C., Pruslin, F. H. & Allfrey, V. G. (1984) Exp.

Cell Res. 150, 269-281.21. Gronow, M. & Thackrah, T. (1974) Eur. J. Cancer 10, 21-25.

Proc. NatL Acad Sd USA 81 (1984)

Dow

nloa

ded

by g

uest

on

Dec

embe

r 9,

202

0