[CANCER RESEARCH 45, 5845-5850, November 1985]

Keratin Gene Expression in Mouse Skin Tumors and in Mouse Skin Treatedwith 12-O-Tetradecanoy Iphorbol-13-acetate

Rune Toftgard,1 Stuart H. Yuspa, and Dennis R. Roop2

Laboratory of Cellular Carcinogenesis and Tumor Promotion, National Cancer Institute, NIH, Bethesda, Maryland 20205

ABSTRACT

Alterations in the pattern of epidermal differentiation and proliferation occur during mouse skin carcinogenesis. We have usedcDNA clones corresponding to the major keratin subunits synthesized in differentiating epidermal cells (M, 67,000 and 59,000)and in proliferating epidermal cells (M, 60,000, 55,000, and50,000) to study changes in keratin gene transcript levels inmouse epidermis exposed to tumor promoters. The same probeswere used to characterize the keratin expression patterns inbenign and malignant skin tumors. A single topical treatmentwith 12-O-tetradecanoylphorbol-13-acetate caused a rapid initial

decrease in the epidermal transcript levels corresponding to theM, 67,000 and 59,000 keratin subunits. By 48 h the transcriptlevel for the M, 67,000 keratin subunit was restored to controlvalues, whereas the transcript levels for the M, 59,000 subunitreturned to control at a slower rate. In contrast, the transcriptlevel for the M, 55,000 subunit was increased substantially 12-

48 h after treatment, the M, 50,000 subunit transcript increasedto a lesser extent, and the M, 60,000 subunit message wastransiently decreased at 12 h but returned to the level of solvent-

treated skin by 24 h. Single exposure to the incompletetumor promoters 4-O-methyl-12-0-tetradecanoylphorbol-13-

acetate, the ionophore A23187, and mezerein induced changesin keratin gene transcripts similar to those of 12-O-tetradeca-noylphorbol-13-acetate. The antipromoter fluocinolone aceton-ide, administered with 12-O-tetradecanoylphorbol-13-acetate,

partially inhibited the decrease in the M, 59,000 and 67,000transcripts and completely inhibited the increase in the M, 55,000transcript. In skin papillomas produced by initiation and promotion, keratin gene expression was similar to normal skin, withthe exception of a two-fold increase in the transcript levels for

the M, 55,000 keratin subunit. However, in carcinomas, thetranscript levels for the M, 67,000 and 59,000 subunits wereonly 1-3% of those observed in untreated mouse epidermis. In

concert with other data, the rapid and selective loss of transcriptsfor differentiation-related keratins after exposure to both com

plete and incomplete tumor promoters is most consistent withan accelerated rate of maturation in differentiating keratinocytes,resulting in the rapid production of transcript-depleted fully ma

ture squames. The enhanced level of M, 55,000 transcriptssuggests a concomitant increase in the number of all cells or asubset of cells in the proliferative compartment. Benign tumors,characterized by an increased cellular proliferation rate, may alsobe enriched in cells transcribing the Mr 55,000 keratin gene, but

1Recipient of fellowships from The Swedish Work Environment Health Fund

and O. E. and Edla Johanssons Vetenshapliga stiftelse. Present address: Department of Medical Nutrition, Karolinska Institute, Huddinge University Hospital F69,S-14186 Huddinge, Sweden.

2To whom requests for reprints should be addressed, at Building 37, Room

3B08, Laboratory of Cellular Carcinogenesis and Tumor Promotion, National Cancer Institute; NIH, Bethesda, MD 20205.

Received 5/14/85; revised 8/6/85; accepted 8/9/85.

the differentiating component of papillomas expresses a normaldifferentiation program with regard to keratins. In carcinomasthe cells are blocked in expression of transcripts for keratinsubunits associated with differentiation.

INTRODUCTION

Chemical induction of skin tumors in mice can be operationallysubdivided into at least three different stages, i.e., initiation,promotion, and malignant conversion (1, 2). Tumor promotion isthought to involve the clonal expansion of initiated epidermalcells, resulting in the formation of primarily benign tumors, papillomas (3). The conversion of epidermal papillomas to carcinomas requires additional cellular changes that are most likely tobe genetic in nature (2, 4).

Alterations of both epidermal cellular differentiation and skinproliferation patterns are thought to constitute essential eventsduring multi-stage mouse skin carcinogenesis (5-7). The major

differentiation products in mouse epidermal cells are keratins, afamily of related structural proteins of molecular weight 40,000-

70,000, which make up the subunits of the intermediate filamentsin epithelial cells (8, 9). Different subsets of keratins are preferentially expressed in basal and differentiating cell layers of newborn mouse epidermis (10-13), and differential expression has

been demonstrated in primary epidermal cell culture and inmalignantly transformed epidermal cells (14). Since individualkeratins are synthesized from different mRNAs (11,15-18), the

transcript levels of specific keratins reflect changes in the expression of genes closely regulated during differentiation.

Previous studies of keratin protein patterns during skin carcinogenesis have shown that major changes in the amount andsynthesis of keratin subunits are apparent in mouse epidermisafter treatment with tumor promoters and in papillomas andcarcinomas (19-23). The interpretation of these results with

regard to dynamic changes in skin, however, has been hamperedby the high stability and slow turnover of keratin proteins (8, 24)and by differential resistance toward proteolytic modificationamong the subunits (24-26). In the present study, we used

cDNA clones corresponding to the major keratins expressed inthe differentiating layers of mouse epidermis [M, 59,000 and67,000; Nos. 10 and 1 according to the catalogue of Moll et al.(9)] and to keratins predominantly expressed in the proliferatingbasal cells [M, 50,000, 55,000, and 60,000; Nos. 16, 14, and 5according to Moll ef a/. (9)] to characterize the changes in keratingene expression occurring after treatment of mouse skin withtumor promoters and in papillomas and carcinomas.

MATERIALS AND METHODS

Animal Treatments. Female SENCAR mice 7-9 weeks old, which arepredisposed to tumor development (2), were used. The backs of the

CANCER RESEARCH VOL. 45 NOVEMBER 1985

5845

Research. on November 11, 2020. © 1985 American Association for Cancercancerres.aacrjournals.org Downloaded from

KERATIN GENE EXPRESSION DURING SKIN CARCINOGENESIS

mice were shaved 24 h prior to topical treatment with 5 ng of TPA,3 250

ng of 4-O-methyl-TPA, 2 ¿igof mezerein, 250 /¿gof the ionophore

A23187, or 5 ^g of TPA in combination with 2 ^g of fluocinoloneacetonide. The doses of tumor promoters were chosen to give a comparable hyperplastic response (27-29) and were applied in 200 n\ ofacetone over an area of approximately 8 cm2. Mice were killed at different

time points after treatment, and total RNA was isolated. Repeatedtreatments with TPA, 2 ^g twice weekly for 4 weeks with or withoutprior initiation with 20 t¡gof 7, 12-dimethylbenz[a]anthracene were

carried out according to a schedule used in tumor induction experiments.Papillomas were induced by initiation with 20 ^g of 7, 12-dimethyl-

benz[a]anthracene followed by promotion with 2 HQof TPA twice weeklyfor 12 weeks, and RNA was prepared from pools of 6-12 papillomas.

Carcinomas were derived from papillomas induced as above but subsequently treated with a carcinogen (20 mg of urethane i.p. once weekly)to accelerate malignant conversion (2). Carcinomas were also producedby subcutaneous implantation of malignantly transformed mouse epidermal cells (cell line Pam 212) (30). RNA was isolated from pools of 3-6

carcinomas.Isolation of RNA. The epidermis was separated from the dermis by

flotation on 1% trypsin at 37°for 1 h, and total RNA was prepared from

groups of 5-15 mice, essentially as described by Chirgwin ef al. (31).

The amount of polyadenylic acid containing RNA was determined byhybridization to 3H-polyuridylic acid (32), and the integrity of each prep

aration was checked by subjecting a 10-^g portion of the RNA to

electrophoresis in a 1.1% formaldehyde/agarose gel, transfer of the RNAto nitrocellulose paper (33), followed by hybridization to cDNA probescorresponding to the transcripts of either the M, 67,000 or the M, 60,000keratin subunit.

RNA Blot Analysis. RNA was in some cases fractionated on 1.1%formaldehyde/agarose gels as described above, or slot-blots were prepared using the Mini-fold II (Schleicher and Schull, Dassel, Federal

Republic of Germany). Briefly, equal amounts of polyadenylic acid containing RNA from each sample were dissolved in 100 n\ of H2O and 300P\ of 6.15 M formaldehyde, 10 x standard saline citrate (1.5 M NaCI, 0.15M sodium citrate), incubated at 65°for 15 min, and loaded onto the slot

blotter in two- or three-fold dilutions (sample corresponding to 0.1-1 ng

of polyadenylic acid). On each blot, 10 ^g of SENCAR spleen DNA wereapplied as an internal standard. The nitrocellulose filters were then baked,prehybridized, and hybridized as described (11). The following nick-translated 32P-labeled cDNA probes were used: p4-2 (M, 67,000), 2.4

kilobases; p4-35 (M, 59,000), 2.0 kilobases; p5-23 (M, 60,000), 2.1kilobases; p15-86 (M, 55,000), 1.6 kilobases; and p8-14 (M, 50,000),1.5 kilobases (11 ).4The size of the corresponding keratin subunit is given

in parentheses, and the size of their mRNAs is given kilobases. ThecDNA clones were isolated from cDNA libraries prepared to total polyadenylic acid RNA isolated from newborn mouse epidermis [M, 59,000and 67,000 cDNA clones (11)] and from primary mouse epidermal cellcultures [M, 50,000, 55,000, and 60,000 cDNA clones].4 The cDNA

probes used were shown to hybridize specifically with individual keratinmRNAs by Northern RNA blot analysis (11).4 The blots were exposed to

X-ray film (Kodak, XAR-5) in the presence of intensifying screens (Cronex,Lightning Plus) at -70°C. For relative quantitation, autoradiograms were

scanned densitometrically (Gilford Scanning Densitometer), and the peakareas were calculated. The concentrations of RNA and exposure timesused were shown in separate control experiments to give a linearstandard curve.

RESULTS

Modulation of Keratin Gene Expression by Treatment withTumor Promoters in Vivo. A single topical application of TPA (5^g) to mouse epidermis caused a rapid and dramatic decrease

3The abbreviations used are: TPA, ^-O-tetradecanoylphorboMS-acetate; FA,fluocinoloneacetonide.

"D. R. Roop, C. K. Chung, R. Toftgard. and S. H. Yuspa, unpublished obser

vations.

in the relative level of transcripts corresponding to the M, 67,000and 59,000 keratin subunits (Fig. 1, lane 2, and Chart 1,4).Twelveh after treatment, the transcript levels were only 5% of that inuntreated epidermis. The level of transcripts for these two differentiation keratins remained low for an additional 12-24 h. By 48h after TPA treatment, the transcript levels for both genes wereincreasing so that the M, 67,000 transcripts were more abundantthan in control epidermis, while the M, 59,000 transcript was stillslightly below control values. Seven days after a single exposure,the transcript levels for both genes were essentially back tocontrol values. Similar results were obtained in two independentexperiments.

In contrast to the results with the differentiation-related kera

tins, the transcript levels corresponding to the M, 55,000 and50,000 keratin subunits, preferentially synthesized in proliferatingbasal cells, showed no such decrease after TPA exposure. Therewas a slight (50%) decrease in the transcript level correspondingto the M, 60,000 keratin subunit at 12 h only (Chart 18). Asubstantial increase in the transcript level for the M, 55,000keratin subunit was observed at 24 and 48 h after treatment(Chart 1B).

Other agents which are active as promoters but without thecomplete promoting activity of TPA were tested by a singletopical application. 4-O-methyl-TPA (250 /ig), the ionophore

123 456 78 9 10 11

2.4 kb

Fig. 1. Analysis of mRNA corresponding to the M, 67,000 keratin subunit inadult mouse epidermisafter single treatments with completeand incompletetumorpromoters. Twenty ng of total RNA were electrophoresedon a 1.1% agarose gel,transferred to a nitrocellulosesheet, and subsequently hybridizedto a cDNA clonespecific for the M, 67,000 keratin subunit. Lane 1, control; lane 2, 12 h after TPAtreatment; lane 3, 12 h after combined treatment with TPA and fluocinoloneacetonide; lane 4, 72 h after TPA treatment; lane 5, 7 days after TPA treatment;lane 6, 12 h after 4-O-methyl-TPAtreatment; lane 7, 48 h after 4-O-methyl-TPAtreatment; lane 8, 12 h after treatment with the ionophore A23187; lane 9, 48 hafter A23187 treatment; lane 10,12 h after mezereintreatment; lane 11, 48 h aftermezereintreatment.

CANCER RESEARCH VOL. 45 NOVEMBER 1985

5846

Research. on November 11, 2020. © 1985 American Association for Cancercancerres.aacrjournals.org Downloaded from

KERATIN GENE EXPRESSION DURING SKIN CARCINOGENESIS

150

12 24 48TIME <hr)

72 168

Chart 1. Relative quantitation of keratin transcript levels in adult mouse epidermis after a single treatment with TPA. Total epidermal RNA was prepared fromcontrol mouse epidermis, and at various time points after TPA treatment, slot-blotswere prepared. After hybridization to the different keratin cDNA clones, the resultingautoradiograms were scanned densitometrically, and the area under each peakwas calculated. The data are expressed relative to the transcript levels observedwith control mouse epidermal RNA. A. transcript levels corresponding to the M,67,000 and 59,000 keratin subunits preferentially expressed in differentiatingepidermal cells; B, transcript levels corresponding to the M, 60,000, 55,000, and50,000 keratin subunits preferentially expressed in proliferating basal cells.

A23187 (250 ^g), and mezerein (2 ng) caused essentially thesame pattern of changes in keratin transcripts as treatment withTPA (Chart 2). In all cases, a marked decrease in the M, 67,000and 59,000 transcripts was observed at 12 h. However, onlyTPA showed the increase above control value in M, 67,000expression at 48 h, while this message returned more slowlyafter treatment with the other agents. This can be seen also onthe Northern blots of mRNA from epidermis treated with theseagents (Fig. 1). In the case of A23187, the messages for thedifferentiation-related keratins remained depressed for longer

than 48 h. TPA appeared to have the most pronounced andselective effect of any agent tested on the increase in the levelsof the proliferation-associated M, 55,000 transcripts, but all of

the agents studied produced changes similar to TPA in pattern.Tumor promotion requires multiple exposures for the biological

response of tumor formation. When TPA exposure was repeatedtwice weekly for 4 weeks and the keratin transcript levels measured 72 h after the last exposure, the pattern was similar tothat seen at 72 h after a single exposure. Results were identicalin initiated skin or skin not exposed to an initiator (Table 1).

FA is a potent antipromoting agent in mouse skin. When FA

150

8•5100

SK

3o.

50

D 67 kd

g 59 kd

1EZEREITPA 4-0-METHYL- A23187 MEZEREIN TPA 4-0-MEÕHYL A23187 MEZEREIN3ïâ li TPA

12hr 48hr

B300||•5200

#S5^

100so.Ü|

55kdESOkd~•'\••\-r!-|!"¡"E1-fi1:;

TPA 4-0 METHYL A23187 MEZEREINTPA

4-0 METHYL A23187TPA

12hr i,Bhr

Chart 2. Relative quantitation of keratin transcript levels in adult mouse epidermis 12 and 48 h after single treatments with complete and incomplete tumorpromoters. Autoradiograms of slot-blots were scanned densitometrically, and theareas under each peak were calculated. The data are expressed relative to thetranscript levels observed with control mouse epidermal RNA. A, transcript levelscorresponding to the M, 67,000 and 59,000 keratin subunits; B, transcript levelscorresponding to the M, 60,000, 55,000, and 50,000 keratin subunits.

was applied simultaneously with TPA, there was a partial inhibition of the loss of the M, 67,000 and 59,000 transcripts at 12 h(Table 2 and Fig. 1, lane 3). Most striking was the markeddepression of the M, 55,000 transcript level in the group treatedwith FA and TPA, while the expected increase in this transcriptat 12 h was observed in the group treated with TPA alone. Theexpected increase in M, 50,000 transcript was also suppressedby FA treatment, but the suppression was less striking than thatfor the M, 55,000 transcript.

Keratin Gene Transcripts in Papillomas and Carcinomas.As shown in Table 1, the relative transcript levels correspondingto the keratin subunits are very similar in papillomas and normaladult epidermis, with the exception of a substantial increase inthe transcript level corresponding to the M, 55,000 keratin inpapillomas. In contrast, the relative transcript levels corresponding to the keratin subunits preferentially synthesized in differentiating epidermal cells (M, 67,000 and 59,000) are very low incarcinomas. The loss of these transcripts is a consistent findingin both primary carcinomas and in a transplantable carcinomaresulting from injection with cell line Pam 212. The presence and

CANCER RESEARCH VOL. 45 NOVEMBER 1985

5847

Research. on November 11, 2020. © 1985 American Association for Cancercancerres.aacrjournals.org Downloaded from

KERATIN GENE EXPRESSION DURING SKIN CARCINOGENESIS

Table 1

Relative quantitation ot keratin gene mRNAs in mouse skin after repeated treatment with TPA and inpapillomas and carcinomas

Total RNA was purified from mouse epidermis 72 h after the last treatment with TPA and from pools ofpapillomasor carcinomas.A constant amount of mRNA, as measuredby polyadenylicacid content, was usedfor slot-blot analysis.The autoradiogramswere scanneddensitometrically,and the area under each peak wascalculated. Data obtained when analyzing RNA prepared from untreated epidermis were used as control.

Peak area (% of control) for corresponding keratin sub-units:

TreatmentTPA

(4 weeks. 2 >igtwice weekly)DMBA (20 M9once) TPA (4

weeks, 2 ^g twice weekly)Papillomas8Carcinomas IoCarcinomas II6M,

67,000196164

10411M,59,0009688

6613M,

60,0008393

592571M,

55,000156105252

5879M,

50,0008864

74132105

8The data represent the mean of two separate preparations of RNA from pools of 6-12 papillomas.6 The carcinomas were derived by subcutaneous implantation of malignantlytransformed Pam 212 cells (I)

or were induced chemicallyby topical applications to mouse skin as described in "Materials and Methods" (II).RNA was purified from pools of 3-6 carcinomas.

Table 2Effect of the anti-promoter fluocinolone acetonide on changes in keratin mRNA

levels induced by TPA

Total RNA was purified from mouse epidermis 12 h after a single treatment withTPA or TPA and FA. RNA prepared from untreated epidermis was used as acontrol, and results were analyzed as in Table 1.

Peak area (% of control) for corresponding keratin subunits:

TreatmentM, M, M, M, M,

67,000 59,000 60,000 55.000 50,000

TPA (5 Mg)TPA (5 „g)+ FA (2 „g)

1132

812

8150

14566

14093

absence of the M, 67,000 and 59,000 keratin subunit proteins inpapillomas and carcinomas, respectively, was also evident whenepidermal keratins were resolved using one-dimensional poly-

acrylamide gel electrophoresis (data not shown). In carcinomas,transcripts corresponding to the M, 60,000 keratin subunit aregenerally less abundant than in normal cells, while the M, 55,000and 50,000 transcripts are found in quantities more similar tothose of normal cells.

DISCUSSION

Previous studies designed to analyze the acute and chronicchanges in the profile of mouse skin keratins following exposureto tumor promoters (19-21 ) relied on electrophoretic separation

of the proteins. Because of the high stability of the proteins oncesynthesized and variations in gel systems which resulted ininconsistent protein identification from laboratory to laboratory,it has not been possible to evlauate the observed changes as adynamic process. Post-translational modification of the proteins,particularly phosphorylation, further complicated the interpretation of the results. Laskin ef al. (19) used in vitro radiolabeling ofexpiants from dorsal mouse skin exposed to phorbol esters invivo to analyze protein synthesis changes 3 and 24 h afterexposure and concluded that both decreases and increases insynthesis of specific proteins occurred, and many of theseproteins were presumed to be keratins. In general, all studieshave found similar protein changes after treatment with TPA and

.mezerein, while results with 4-O-methyl-TPA and A23187 have

varied.

The availability of specific cDNA probes for individual epidermalkeratin genes has provided the capability to study changesassociated with promoter exposure at both the transcriptionaland post-transcriptional levels. Previous studies with theseprobes (11)4 have indicated that the differentiation-associated

keratins are expressed in suprabasal cells, while the proliferation-

associated keratins are expressed in basal cells. The currentstudies reveal that TPA exposure causes a marked and rapiddecrease in messages for the differentiation-related keratins (M,

67,000 and 59,000). The rapid decrease in these transcripts(subsequent experiments have shown a decrease to 25% ofcontrol values by 4 h after exposure), its specificity [other keratinmRNAs, calmodulin mRNA, and certain oncogene transcripts(34) are not affected], and its occurrence prior to any change inepidermal proliferation or overall cell number (35) suggest aspecific effect on cells synthesizing these messages. The rapiddecrease in these transcripts is likely to involve both cessationof transcription and an acceleration of mRNA degradation inaffected cells. In contrast to its effects on the differentiation-

related keratins, TPA causes a substantial increase in transcriptsfor the proliferation-related M, 55,000 keratin, but this is notobserved until 12-24 h after treatment, and it persists for 48-

72 h. Other changes include an increase in the M, 50,000transcript which parallels the M, 55,000 change but is generallysmaller and a transient decrease in the M, 60,000 transcript witha rapid recovery.

The acute changes in keratin transcripts were similar afterexposure to a variety of agents which induce hyperplasia andhave some activity as partial or complete promoters in SENCARmice. Thus, the specificity of these changes for complete promotion remains to be determined. Several incomplete promotersproduced effects sufficiently different from those of TPA to benoteworthy. 4-O-methyl-TPA failed to increase the M, 55,000

transcript to the level of the other agents. A23187 caused amuch prolonged loss of the M, 67,000 and 59,000 keratins, andexposure to incomplete promoters never resulted in the elevatedsteady-state level of the M, 67,000 transcript found after 48 h

with TPA. Furthermore, TPA was the only agent which specifically increased expression of the M, 55,000 transcript to a muchgreater extent than the M, 50,000 transcript.

Several interpretations regarding the acute response to TPA

CANCER RESEARCH VOL. 45 NOVEMBER 1985

5848

Research. on November 11, 2020. © 1985 American Association for Cancercancerres.aacrjournals.org Downloaded from

KERATIN GENE EXPRESSION DURING SKIN CARCINOGENESIS

and other agents are possible. TPA may directly or indirectlyregulate transcription of the keratin genes in specific cells. Thiswould be consistent with previous interpretations from proteinsynthesis studies (19, 36), in which induction of new synthesiswas inferred. Alternatively, these results could be related topromoter-induced changes in relative abundance of epidemial

subpopulations, each with its own profile of keratin expression.We have shown that epidermal subpopulations are heterogeneous in their TPA responses (37, 38). Certain cells are inducedto differentiate, while others are stimulated to proliferate. Furthermore, we have demonstrated that the response pattern isdependent on the state of epidermal maturation at the time ofTPA exposure (39). In vivo, TPA causes a rapid loss of certainbasal cells into the suprabasal compartments and an increase inthe thickness of the stratum corneum within a few hours (40,41). At the ultrastructural level, there is disruption of the normalcytoplasmic organization in epidermal cells in the more superficiallayers and some fragmentation of the bundles of tonofilaments(35). The rapid loss of transcripts for the M, 67,000 and 59,000keratins (localized in differentiating cells) could be accounted forby phorbol ester-mediated accelerated maturation of viable su

prabasal cells to the fully differentiated state. Since maturesquames do not transcribe RNA and are high in nuclease activity,the rapidity of message loss is most consistent with this explanation. The concomitant decrease of the M, 60,000 keratintranscripts suggests that these transcripts may exist in the samecells expressing the M, 59,000 and 67,000 gene or in thesubclass of basal cells which undergoes accelerated differentiation. Since the M, 60,000 decrease is only partial, other cells arelikely to be making this transcript as well. For example, the M,60,000 gene may be initially expressed in cells within the basallayer and continue to be expressed in cells that are committedto differentiate or the M, 60,000 mRNA may persist longer in thecommitted cells. In situ hybridization experiments localizingmRNAs within individual cells within the epidermis, and nuclearrun-off transcription experiments on purified cell populations are

required to distinguish between these possibilities. The specificincrease in Mr 55,000 message following TPA exposure couldresult from an increase in a subpopulation expressing this gene.Specific increases in a dark cell population (35, 42) is a knownresponse to promoter exposures. Alternatively promoters couldshorten the cell cycle in the proliferating epidermal cell pool (43),and this could alter keratin gene expression to favor the M,55,000 gene. Analysis of carcinomas failed to show an enhancedlevel of Mr 55,000 transcripts, which suggests that rapid proliferation alone is not sufficient to favor expression of the M, 55,000transcript.

Differential cellular responses to TPA as the basis for ourresults is supported by the studies with fluocinolone acetonide.This synthetic glucocorticoid antipromoter is known to be apotent inhibitor of the inflammatory and proliferative responsesof keratinocytes, but it only partially inhibits the differentiativeresponses (44). In the current study FA only partially inhibitedthe loss of the M, 67,000 and 59,000 transcripts (presumeablyassociated with accelerated differentiation) but completely inhibited the increase in the M, 55,000 message.

Previous reports based on protein studies have suggestedthat chronic TPA exposure results in keratin profiles similar tolate effects of a single exposure, and both are characteristic ofthe neonatal skin keratin pattern (20, 21). Our studies at the

RNA level are consistent with the similarities of transcript levelsfor the keratin genes at 72 h following acute or chronic exposure.However, the measured values deviate only slightly from relativetranscript levels in adult untreated controls, although significanthyperplasia was observed in the treated groups. In newbornepidermis we have observed 6-10-fold higher relative transcript

levels for the M, 67,000 and 59,000 keratin subunits from thoseof adult skin. Thus, at the RNA level there may be importantdifferences between newborn skin and chronic TPA treatmentwhich could be related to message synthesis or degradation.

The keratinocytes present in chemically induced benign skintumors, papillomas, are characterized by an increased proliferative activity which includes the capacity to divide after leavingthe basement membrane (5). However, at the same time theyretain the ability to terminally differentiate. Consistent with thismaturation competence, near-normal transcript levels for the M,

67,000 and 59,000 keratin subunits were found, whereas anincreased transcript level was observed for the M, 55,000 sub-

unit. The selection of a M, 55,000 rich cell type in papillomas isa potential explanation for this finding.

Epidermal squamous cell carcinomas show marked deviationsfrom normal epidermis and papillomas with regard to keratinprotein pattern (22, 23). The almost total loss of the differentiation-associated keratins may reflect the loss of a normal differ

entiation pattern encountered in these cells. It has been reportedthat carcinomas may contain masked transcripts for the M,67,000 keratin subunit (23). Our results from analysis of carcinomas induced chemically or derived by implantation of malignantly transformed epidermal cells do not support this but indicate that carcinomas contain very low levels of this transcript.Therefore, there appears to be a transcriptional block or aselective rapid degradation of the message precluding the synthesis of the M, 67,000 and 59,000 keratin subunits.

Our results on the differential expression of keratin genes intumors could be used to discriminate between benign and malignant epidermal tumors. Recently, Klein-Szanto ef al. (45) havesuggested that loss of differentiation-associated keratins is anearly marker of malignant conversion. The use of specific im-munological probes for screening keratin changes could providea relatively easy method to monitor conversion. We have recentlyproduced antisera that are monospecific for the differentiation-

associated keratins (M, 59,000 and 67,000) using syntheticpeptides corresponding to unique C-terminal sequences of these

keratin subunits (13). We selected this approach for antibodyproduction after an examination of amino acid sequence data,which were deduced from the nucleotide sequences of cDNAclones for these keratins, revealed that the carboxyl-terminal

sequences were unique. We are currently screening biopsies oftumors, which were produced by protocols which acceleratemalignant conversion (2), to determine if the conversion of papillomas to carcinomas can be detected in early stages.

REFERENCES

1. Shubik, P. The growth potentialitiesof inducedskin tumors in mice.The effectsof different methods of chemicalcarcinogenesis.Cancer Res., 10: 713-717,1950.

2. Hennings,H., Shores, R., Wenk, M. L., Spangler,E. F., Tarane, R., and Yuspa,S. H. Malignant conversion of mouse skin tumors is increased by tumorinitiators and unaffected by tumor promoters. Nature (Lond.), 304: 67-69,1983.

3. Yuspa, S. H., Hennings, H., Kulesz-Martin, M., and LJchti,U. The study of

CANCER RESEARCH VOL. 45 NOVEMBER 1985

5849

Research. on November 11, 2020. © 1985 American Association for Cancercancerres.aacrjournals.org Downloaded from

KERATIN GENE EXPRESSION DURING SKIN CARCINOGENESIS

tumor promotion in a cell culture model for mouse skin: a tissue that exhibitsmultistage carcinogenesis in vivo. In: E. Hecker, N. E. Fusenig, W. Kunz, F.Marks, and H. W. Thielman(eds.),Cocarcinogenesisand BiologicalEffects ofTumor Promoters, pp. 217-230. New York: Raven Press, 1982.

4. Scribner, J. D., Scribner, N. K., McKnight, B., and Mottet, N. K. Evidence fora new model of tumor progression for carcinogenesis and tumor promotionstudies with 7-bromomethylbenz(a)anthracene.Cancer Res., 43: 2034-2041,1983.

5. Burns, F. J., Vanderlaan,M., Sivak, A., and Albert, R. A. Regression kineticsof mouse skin papillomas.Cancer Res., 36: 1422-1427,1976.

6. Rheinwald,J. G. and Beckett, M. A. Defectiveterminal differentiation in cultureas a consistent and selectable character of malignant human keratinocytes.Cell, 22:629-632,1980.

7. Yuspa, S. H. and Morgan, D. L. Mouse skin cells resistant to terminaldifferentiationassociated with initiationof carcinogenesis.Nature (Lond.),293:72-74,1981.

8. Lazarides, E. Intermediate filaments: a chemically heterogeneous, develop-mentally regulated class of proteins. Ann. Rev. Biochem.,51: 219-250,1982.

9 Moll, R., Franke, W. W., Schiller,D. L., Geiger,G., and Krepier, R. The catalogof human cytokeratins: patterns of expression in normal epithelial tumors andcultured cells. Cell, 31: 11-24,1982.

10. Woodcock-Mitchell, J., Eichner, R., Nelson, W. G., and Sun, T-T. Immunoto-calization of keratin polypeptides in human epidermis using monoclonal antibodies. J. Cell Biol.. 95: 580-588, 1982.

11. Roop, D. R., Hawley-Nelson,P., Cheng, C. K., and Yuspa, S. H. Keratin geneexpression in mouse epidermis and cultured epidermalcells. Proc. Nati. Acad.Sci. USA,80: 716-720. 1983.

12. Schweizer,J., Kinjo, M., Furstenberger,G., and Winter, H. Sequentialexpression of mRNA-encodedkeratin sets in neonatal mouse epidermis: basal cellswith properties of terminally differentiating cells. Cell, 37: 159-170, 1984.

13. Roop, D. R., Cheng, C. K., Titterington, L., Meyers, C. A., Stanley, J. R.,Steinert, P. M., and Yuspa, S. H. Synthetic peptides corresponding to keratinsubunits elicit highly specific antibodies. J. Biol. Chem., 259: 8037-8040,1984.

14. Roop, D. R., Hawley-Nelson,P., Cheng, C. K., and Yuspa, S. H. Expressionof keratin genes in mouse epidermis and normal and malignantly transformedepidermalcells in culture. J. Invest. Dermatol.. 87: 144s-149s, 1983.

15 Fuchs, E. and Green, H. Multiple keratins of cultured human epidemial cellsare translated from different mRNA molecules.Cell, 17: 573-582,1979.

16. Schweizer, J., and Goerttter, K. Synthesis in vitro of keratin polypeptidesdirected by mRNA isolated from newborn and adult mouse epidermis. Eur. J.Biochem., 112: 243-249, 1980.

17. Kim, K. H., Rheinwald, J. D., and Fuchs, E. V. Tissue specificity of epithelialkeratins: differential expression of mRNAs from two multigene families. Mol.Cell. Biol.. 3: 495-502, 1983.

18. Magin, T. M., Jorcano, J. L., and Franke, W. W. Translation products ofmRNAs coding non-epidermalcytokeratins. EMBO J., 2: 1387-1392,1983.

19. Laskin, J. D., Mufson, A. R., Piccini,L., Engelhardt, D. L., and Weinstein, I. B.Effects of the tumor promoter 12-O-tetradecanoylphorbol-13-acetateon newtysynthesized proteins in mouse epidermis. Cell, 25: 441-449, 1981.

20. Schweizer, J. and Winter, H. Changes in regional keratin polypeptide patternsduring phorbol ester-mediatedreversibleand permanentlysustainedhyperpla-sia of mouse epidermis. Cancer Res., 42: 1517-1529,1982.

21. Nelson, K. G., Stephenson, K. B., and Slaga, T. J. Protein modificationsinduced in mouse epidermis by potent and weak tumor-promoting hyperpla-siogenic agents. Cancer Res., 42: 4164-4175,1982.

22. Nelson, K. G. and Slaga, T. J. Keratin modifications in epidermis, papillomas,and carcinomas during two-stage carcinogenesis in the SENCAR mouse.Cancer Res., 42: 4176-4181. 1982.

23. Winter, H. and Schweizer, J. Keratin synthesis in normal mouse epithelia andin squamous cell carcinomas: evidence in tumors for masked mRNA speciescoding for high molecular weight keratin polypeptides. Proc. Nati. Acad. Sci.USA,80: 6480-6484, 1983.

24. Breitkreutz, D., Bowden, P. E., Quinlan, R., Franke, W. W., and Fusenig, N.E. Precursor-product relationship between prekeratin and keratin in mouse

and humanepidermis.J. Invest. Dermatol.,80; 334,1983.25. Schweizer, J. and Winter, H. Keratin polypeptide analysis in fetal and in

terminallydifferentiatingnewborn mouseepidermis.Differentiation,22:19-24,1982.

26. Sun, T-T., Eichner,R., Nelson, W. G., Tseng, S. C. G., Weiss, R. A., Jarvinen,M., and Woodcock-Mitchell,J. Keratinclasses:molecularmarkers for differenttypes of epithelialdifferentiation. J. Invest. Dermatol.,81:109s-115s, 1983.

27. Marks, F., Bertsch, S., and Furstenberger,G. Omithinedecarboxylaseactivity,cell proliferation, and tumor promotion in mouse epidermis in vivo. CancerRes., 39: 4183-4188,1979.

28. Marks, F., Furstenberger, G., and Kownatzki, E. Prostaglandin E-mediatedmitogenic stimulation of mouse epidermis in vivo by divalent cation ionophoreA23187 and by tumor promoter TPA. CancerRes., 41: 696-702,1981.

29. Mufson, R. A., Fischer, S. M., Verma, A. K., Gleason,G. L., Slaga, T. J., andBoutweii. R. K. Effectsof 12-O-tetradecanoyl-phortx>l-13-acetateandmezereinon epidermalomithine decarboxylase activity, ¡soproterenol-stimulatedlevelsof cyclic adenosine 3', 5'-monophosphate, and induction of mouse skintumors. Cancer Res., 39: 4791-4795,1979.

30. Yuspa, S. H., Hawley-Nelson,P., Koehler, B., and Stanley, R. J. A survey oftransformation markers in differentiatingepidermalcell lines in culture. CancerRes., 40.:4694-4703,1980.

31. Chirgwin, J. M., Przybyla,A. E., MacDonald.R. J., and Rutter, W. J. Isolationof biologically active nbonucleic acid from sources enriched in ribonudease.Biochemistry, 18: 5294-5299,1979.

32. Bishop, J. O. and Rosbash. M. Polynucleotidesequences in eukaryotic DNAand RNA that form ribonuclease-resistantcomplexes with polyundylicacid. J.Mol. Biol., 85:75-86,1974.

33. Thomas, P. Hybridizationof denatured RNA and small DNA fragments transferred to nitrocellulose.Proc. Nati. Acad. Sci. USA, 77: 5201-5202, 1980.

34. Toftgard, R., Roop, D. R., and Yuspa,S. H. Proto-oncogeneexpression duringtwo-stage carcinogenesisin mouse skin. Carcinogenesis(Lond.),6:655-657,1985.

35. Raick, A. N. Ultrastructural, histológica!,and biochemicalalterations producedby 12-O-tetradecanoyl-phorbol-13-acetateon mouse epidermis and their relevance to skin tumor promotion. Cancer Res., 33: 269-286,1973.

36. Cabrai, F., Gottesman, M. M., and Yuspa, S. H. Induction of specific proteinsynthesis by phorbol esters in mouse epidermalcell culture. Cancer Res., 41:2025-2031,1981.

37. Yuspa, S. H., Ben, T., Hennings, H., and üchti,U. Divergent responses inepidermalbasalcells exposed to the tumor promoter 12-O-tetradecanoylphor-bol-13-acetate.Cancer Res., 42: 2344-2349,1982.

38. Yuspa, S. H., Kulesz-Martin,M., Ben, T., and Hennings,H. Transformationofepidermalcells in culture. J. Invest. Dermatol.,81:162s-168s, 1983.

39. Yuspa, S. H., Ben, T., and Hennings, H. The induction of epidermal trans-glutaminase and terminal differentiation by tumor promoters in cultured epidermalcells. Carcinogenesis(Lond.),4: 1413-1418,1983.

40. Argyris, T. S. The nature of the epidemial growth produced by the firstapplicationof 12-O-tetradecanoyl-phorbol-13-acetateon the skin of mice initiated with dimethylbenzanthracene.J. Invest. Dermatol., 77: 230-234, 1981.

41. Astrup, E. G. and Iversen, 0. H. Cell population kinetics in hairless mouseepidermis following a single topical application of 12-0-tetradecanoylphorbol-13-acetate II. Virchows Arch. (Cell Pathol.),42:1-18,1983.

42. Klein-Szanto, A. J. P. and Slaga, T. J. Numerical variation of dark cells innormal and chemically induced hyperplastic epidermis with age of animal andefficiency of tumor promoter. Cancer Res., 41: 4437-4440,1981.

43. Morris, R. and Argyris, T. J. Epidermal cell cycle and transit times duringhyperplasticgrowth inducedby abrasionor treatmentwith 12-O-tetradecanoyl-phorbol-13-acetate.Cancer Res., 43: 4935-4942,1983.

44. Yuspa, S. H., Ben, T., Hennings, H., and üchti,U. Phorbol ester tumorpromoters induceepidermaltransglutaminaseactivity. Biochem.Biophys.Res.Commun., 97: 700-708, 1980.

45. Klein-Szanto,A. J. P., Nelson, K. G., Shah, Y., and Slaga, T. J. Simultaneousappearance of keratin modifications and -y-glutamyltransferaseactivity asindicatorsof tumor progression in mouse skin papillomas.J. Nati. Cancer Inst.,70: 161-168,1983.

CANCER RESEARCH VOL. 45 NOVEMBER 1985

5850

Research. on November 11, 2020. © 1985 American Association for Cancercancerres.aacrjournals.org Downloaded from

1985;45:5845-5850. Cancer Res   Rune Toftgard, Stuart H. Yuspa and Dennis R. Roop 

-Tetradecanoylphorbol-13-acetateOSkin Treated with 12-Keratin Gene Expression in Mouse Skin Tumors and in Mouse

  Updated version

  http://cancerres.aacrjournals.org/content/45/11_Part_2/5845

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

  .pubs@aacr.orgDepartment at

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

  Permissions

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

.http://cancerres.aacrjournals.org/content/45/11_Part_2/5845To request permission to re-use all or part of this article, use this link

Research. on November 11, 2020. © 1985 American Association for Cancercancerres.aacrjournals.org Downloaded from