Strain differences in mouse skin tumorigenesis by the promotion of 12-o-tetradecanoylphorbol-13-acetate or benzoyl peroxide

Strain difference of mouse skin tumorigenesis was studied with 12-O-tetradecanoylphorbol-13-acetate (TPA) and benzoyl peroxide (BzP) both in 8-week-old hypocatalasemic C3H mice (C-s(b)) and in normal C3H/HeNCrj mice (C3H). Two weeks after initiation by 4 mu mole MNNG, both strains of female mice were promoted by 10 nmole TPA or 10 mg, 20 mg BzP twice weekly for 30 and 52 weeks. The incidence of skin tumors in C-b(s) was significantly increased as compared to that in C3H mice promoted by TPA for 30 weeks but after 52 weeks of TPA promotion there was no difference between the 2 strains. By the promotion with 10 mg or 20 mg BzP, the incidence of skin tumor in C3H was 5% at both levels after 30 weeks of promotion and 19% and 38%, respectively, after 52 weeks of promotion. No skin tumors appeared in C-b(s) mice with BzP treatment after 30 weeks of promotion. After 52 weeks of promotion by 10 mg BzP, skin tumors were induced in only 15% of C-b(s) mice. The results show that genetic factors, especially radical scavenging enzymes, controlled susceptibility to skin tumorigenesis by promotion with either TPA or BzP.     

Sulforaphane prevents mouse skin tumorigenesis during the stage of promotion

Sulforaphane (SF), a natural product from broccoli, is known to enhance detoxification of carcinogens and block initiation of chemically-induced carcinogenesis in animal models. Cell culture and xenograft studies suggest additional roles for SF, inhibiting growth of tumors, arresting the cell cycle and enhancing apoptosis. As currently reported, topical SF (1, 5 or 10 micromol/mouse) significantly inhibited 7,12-dimethylbenz(a)anthracene/12-O-tetradecanoylphorbol 13-acetate (TPA)-induced mouse skin tumorigenesis, using either an anti-promotion protocol (SF from 1 week after carcinogen until the end of the study) or a combined anti-initiation, anti-promotion protocol (SF 7 days prior to carcinogen until the end of the study). Surprisingly, no significant effect was observed in an anti-initiation protocol (SF from 7 days prior to 7 days after carcinogen). Separately, SF inhibited TPA-induced ornithine decarboxylase activity in mouse skin, an obligate step in TPA-induced promotion of carcinogenesis. These data link this molecular mechanism to SF-dependent inhibition of the promotion of tumorigenesis.     

Influence of diet and calorie restriction on the initiation and promotion of skin carcinogenesis in the SENCAR mouse model

Diets were restricted to 60% of the intake of the control mice by feeding less diet (total diet restriction, TDR) or by feeding fewer calories from fat and carbohydrate (calorie restriction, CR) during the initiation or promotion phases of skin tumorigenesis in female SENCAR mice. Skin cancer was initiated by topical treatment with 10 nmol of 7,12-dimethylbenzanthracene in acetone and promoted by twice weekly treatments with 12-O-tetradecanoylphorbol-13-acetate in acetone for 20 wk. Dietary restriction preceding and during 7,12-dimethylbenzanthracene treatment did not influence skin papilloma or carcinoma yield. Papilloma incidence and the number of papillomas per effective mouse were reduced in mice restricted by both TDR and CR protocols during and following promotion with 12-O-tetradecanoylphorbol-13-acetate. Papilloma size was reduced at experimental wk 16 and 20 in both TDR and CR groups fed these diet regimens during promotion. However, by wk 28 and 32, papilloma sizes were similar in the control and TDR groups, and smaller papillomas were observed only in the CR group. The average carcinoma latency was extended by 26% in the groups restricted during promotion, and incidence was reduced in both groups. The reduction, however, was statistically significant only in the CR group. Body weight gain was reduced during the times when dietary restriction was enforced, and in a short-term study, both restricted diet treatments reduced the percentage of carcass protein.     

Circadian rhythm of tumor promotion in the two-stage model of mouse tumorigenesis

The question of whether cancer risk is influenced by time-of-day exposure to potentially carcinogenic agents was approached in this study by exposing mouse skin to a single initiating dose of 7,12-dimethylbenz [-]anthracene, followed by a 12 week regime of bi-weekly skin treatments with the tumor promoter, 12-0-tetradecanoyl-phorbol acetate (TPA), given at four different circadian clock times (CCTs). Tumor incidence, average number of tumors per mouse and tumor size showed a dominant circadian component with an acrophase occurring at 23:00 h CCT. Pre-treatment with all trans-retinoic acid, prior to bi-weekly TPA promotion, reduced tumor incidence, average number and size of tumors per animal by greater than 80%, but did not suppress the underlying circadian rhythm of sensitivity to TPA-induced tumor formation.     

Sequential study of gamma-glutamyltransferase during complete and two-stage mouse skin carcinogenesis

The histochemical pattern of gamma-glutamyltransferase (GGT) was studied in benign and malignant tumors produced by two different experimental protocols, two-stage carcinogenesis and complete carcinogenesis. Six percent of all papillomas produced by two-stage carcinogenesis were GGT positive, whereas 14% of benign tumors produced by complete carcinogenesis exhibited GGT-positive areas. The incidence of GGT-positive papillomas in the two-step carcinogenesis protocol increased up to wk 28 of treatment. After 32 wk, the incidence decreased abruptly, coinciding with an abrupt increase in the incidence of squamous cell carcinomas. On the other hand, the incidence of GGT-positive benign tumors produced during the course of complete carcinogenesis increased gradually up to wk 32 of treatment, coinciding with the increased incidence of squamous cell carcinomas. The incidence of GGT-positive keratoacanthomas and GGT-positive papillomas produced with the complete carcinogenesis protocol exhibited different patterns, suggesting different histogenesis and biological behavior of these two types of tumors. In addition, the labeling index of GGT-positive areas was lower (17 +/- 3%) than that of the GGT-negative areas (41 +/- 0.18%) of the same papillomas, indicating that the presence of GGT may be related to abnormal keratocyte differentiation rather than to proliferative changes.     

Tumorigenesis and carcinogenesis in mouse skin treated with hyperthermia during stage I or stage II of tumor promotion

The action of hyperthermia treatments on tumor promotion separatedinto a two-stage protocol has been investigated. 7,12-Dimethylbenz[a]anthracene(DMBA) initiated dorsal skin of female SENCAR mice was promotedwith either H₂O₂ or 12-O-tetradecanoylphorbol-13-acetate (TPA)(4 applications, 2 times/week) as the first stage of promotion,followed by promotion with mezerein (28 applications, 2 times/week)as the second stage. Hyperthermia (44°C, 30 min) treatmentof the skin at the time of stage II promotion only (just beforeeach mezerein application) suppressed 100% of papillomas whenH₂O₂ was used as a first-stage promoter and 96% when TPA wasused as the first stage, as compared to unheated control animals.The same hyperthermia treatment given only at stage I of promotionhad similar results. Hyperthermia treatments just before stageI TPA promotion (4 treatments only) followed by mezerein asthe second stage reduced papilloma formation by 92%. When H₂O₂was used as the first stage promoter and again mezerein as thesecond, papilloma frequency was reduced by 74%, as comparedto unheated controls. This antipromotion activity of hyperthermiacould not be linked to an inhibition of skin protease activity.Although papilloma frequency was markedly suppressed by hyperthermiaduring stage I promotion only, carcinoma formation was not.A similar number of carcinomas appeared in the groups of micereceiving hyperthermia with either H₂O₂ or TPA as first-stagepromoters, as in comparable groups receiving no hyperthermia.In contrast, when hyperthermia treatments were given duringstage II promolion with mezerein (using either H₂O₂ or TPA asstage I promoters), carcinomas (as well as papillomas) weremarkedly reduced. The results suggest that DMBA initiation createstwo types of promotion-dependent cells, a majority with relativelylow progression probability and a minority with relatively highprogression probability. The former require both stage I andII promotion while the latter require only stage II promotionto form tumors. Hyperthermia treatments given during stage IIpromotion protected against promotion and progression of bothtypes of initiated cells, but similar treatments only duringstage I did not protect against promotion and progression ofthe latter. Although promotion was required for expression,relative progression probability appeared linked to initiationand not promotion events. These findings suggest that hyperthermiatreatment of persons exposed to tumor-promoting agents may reducethe risk of induced tumorigenesis.     

Metastasis from squamous cell carcinomas of SENCAR mouse skin produced by complete carcinogenesis

The incidence of metastasis was evaluated in female SENCAR mice after induction of squamous cell carcinomas by repetitive applications of either benzo [a] pyrene (B [a] P) or N-methyl-N'-nitro-N-nitrosogaunidine (MNNG). Between 41 and 50 weeks 50% of the animals with carcinomas in the B [a] P group had metastases, whereas 20% had metastases in the MNNG group. Very few metastases were observed before 40 weeks of treatment. The major site of metastasis was the lungs; however, metastatic tumors were also found in lymph nodes, adrenal glands and kidneys.     

Dietary fat effects on the initiation and promotion of two-stage skin tumorigenesis in the SENCAR mouse

We investigated the influence of dietary corn oil on initiation of skin tumors in SENCAR mice with 7,12-dimethylbenz(a)anthracene (DMBA) (10 nmol at 8 to 9 wk of age) and the promotion of these tumors with 12-O-tetradecanoylphorbol-13-acetate (TPA) (3.2 nmol twice weekly for 20 wk). Diet high in corn oil (24.6%) was fed, in comparison with control diet (5%), to mice during two time schedules: (a) high-fat diet was fed preceding and for 1 wk following DMBA to assess the effects of high corn oil diet on initiation; and (b) high-fat diet was fed starting at the time of the first TPA treatment (1 wk following DMBA initiation) until the end of the experiment to assess effects of high corn oil diet on promotion. Mice were trained to consume equivalent caloric allotments of the low- and high-fat diets to ensure that the observed effects on tumor development were for dietary fat at constant calorie intake. Feeding high corn oil diet during DMBA treatment did not influence the incidence of skin papilloma or carcinoma, but the number of papillomas per effective mouse was reduced in mice fed the high-fat diet during initiation. Consumption of the high corn oil diet during and following TPA treatment resulted in an increase in the incidence of papillomas up until Wk 14 of the experiment, an increase in the number of papillomas per effective mouse throughout the experiment, and an increase in the number of carcinomas per effective mouse during Wk 25 to 34. However, cumulative carcinoma yield (Wk 25-44) did not differ between the diet groups. Dietary treatment did not influence food consumption, body weight, or survival in the mice treated with DMBA and TPA. Northern blot hybridization studies were carried out on RNA purified from tumors of high- and low-fat mice to determine if diet influenced the pattern of Ha-ras oncogene expression. The results of this experiment indicated that elevated levels of Ha-ras-specific RNA, in comparison with normal epidermal RNA, were present in papillomas and carcinomas from DMBA-initiated, TPA-promoted mice irrespective of the diet the mice were fed.     

Malignant progression of mouse skin papillomas treated with ethylnitrosourea, N-methyl-N''-nitro-N-nitrosoguanidine, or 12-O-tetradecanoylphorbol-13-acetate

Mouse skin tumors were induced by a single topical application of 7,12-dimethylbenzanthracene (DMBA), followed by biweekly promotion with 12-O-tetradecanoylphorbol-13-acetate (TPA). After 20 weeks of promotion, mice were treated twice weekly for 2 weeks with either ethylnitrosourea (ENU), N-methyl-N′-nitro-N-nitrosoguanidine (MNNG) or TPA. Thereafter all groups were treated biweekly with TPA. The ENU-treated group had a higher percentage of animals with carcinomas and developed 217% more cumulative carcinomas per group than TPA-treated controls. The percentage of mice with carcinomas and the cumulative number of carcinomas per group in MNNG-treated mice was higher than TPA-treated controls but was less than ENU-treated mice. The ratio of cumulative carcinomas to cumulative papillomas in ENU treated, MNNG-treated and TPA-treated mice was 16%, 9% and 6%, respectively. Histological examination of tumors remaining at the termination of the experiment revealed the presence of keratoacanthomas, some of which stained positive for γ-glutamyltransferase (GGT), in the ENU-treated and MNNG-treated, but not the TPA-treated groups. The fact that no new papillomas developed during the progression stage indicated that enhanced carcinogenesis resulted from the progression of pre-existing tumors. Enhanced progression of benign skin tumors in mice by only a few treatments of an agent may serve as a potential model for studies into the mechanisms and the inhibition of malignant progression. The model also allows for a comparison of the potency of agents in enhancing malignant progression.     

Gomisin A inhibits tumor promotion by 12-O-tetradecanoylphorbol-13-acetate in two-stage carcinogenesis in mouse skin.

Gomisin A, isolated from the fruits of Schisandra chinensis, is one of the dibenzocyclooctadiene lignans. Application of 12-O-tetradecanoylphorbol-13-acetate (TPA, 1 microgram/ear), a tumor-promoting agent, to the ears of mice induces inflammation. Among seven dibenzocyclooctadiene lignans assayed, gomisin A, gomisin J, and wuweizisu C inhibited the inflammatory activity induced by TPA in mice. The ED50 of these compounds for TPA-induced inflammation was 1.4-4.4 mumol. Gomisin A, with an ED50 of 1.4 mumol, showed the strongest inhibitory effect. Furthermore, at 5 mumol/mouse, it markedly suppressed the promotion effect of TPA (2.5 micrograms/mouse) on skin tumor formation in mice following initiation with 7,12-dimethylbenz[a]anthracene (50 micrograms/mouse). It is assumed that the inhibition of tumor promotion by gomisin A is due to its anti-inflammatory activity.     

Effects of a biomimetic superoxide dismutase on complete and multistage carcinogenesis in mouse skin

The effects of a selective detoxifier of the proximate oxygenradical, superoxide anion, on the induction of tumors in theskin of CD-1 mice by either the initiation-promotion regimenor the complete carcinogenesis process were investigated. Theprinciple agent of interest, copper (II) (3,5-diisopropylsalicylate)₂(CuDIPS), is a low molecular weight, lipophilic copper coordinationcomplex that catalytically disproportionates superoxide anionat a rate comparable to native CuZn superoxide dismutase (SOD).The protocols used to elicit tumors were: (i) a single applicationof 0.2 µmol of 7,12-dimethylbenz[a]anthracene (DMBA) followedby twice-weekly applications of 12-O-tetradecanoylphorbol-13-acetate(TPA) in an initiation-promotion study, and (ii) either a singleapplication of 3.6 µmol DMBA followed by no further treatmentor weekly applications of 0.2 µmol DMBA in complete carcinogenesisprotocols. Application of 2 µmol CuDIPS 15 min prior tothe initiating dose of DMBA was without significant effect ontumor yield or incidence, whereas application prior to eachdose of TPA substantially reduced tumor incidence and yield.This anti-promoting property of CuDIPS can be attributed toits SOD-mimetic activity in as much as the corresponding zinccoordination complex lacking in SOD activity, zinc (II) (3,5-diisopropylsalicylate)₂,was non-inhibitory. Significant reductions in tumor yield werealso observed when CuDIPS was applied prior to DMBA in eitherof the complete carcinogenesis protocols. Additionally, covalentbinding of [³H]DMBA to epidermal DNA was markedly reduced byCuDIPS pre-treatment, suggesting that the anti-carcinogenicproperties may reflect a perturbation in superoxide anion-dependentmetabolic activation of DMBA. The induction by DMBA of ornithinedecarboxylase activity, a biochemical marker of tumor promoteractivity, was not affected by CuDIPS; however, induction ofornithine decarboxylase by TPA was potently blocked. Collectively,these effects of a biomimetic SOD further implicate reactiveoxygen species at multiple stages in chemical carcinogenesis.     

Dietary retinoids are essential for skin papilloma formation induced by either the two-stage or the complete tumorigenesis model in female SENCAR mice.

Our previous work has shown that dietary retinoic acid (RA) is necessary for skin tumor formation induced by the two-stage protocol with the initiator 7,12-dimethylbenz[a]anthracene (DMBA) and the promoter 12-O-tetradecanoyl phorbol-13-acetate (TPA) (De Luca et al., Cancer Res., 36 (1976) 2334-2339). Here we report that retinoids are required for tumorigenesis by the two-stage as well as by the complete tumorigenesis protocol. Mice were treated with a single dose of DMBA (20 micrograms), followed by 20 applications of TPA (2 micrograms), or by 20 applications of DMBA (25 micrograms for 2 weeks and 51 micrograms thereafter). Regardless of the tumor induction protocol, tumor formation was inhibited by vitamin A-deficiency, while RA (3 micrograms/g of diet) or retinyl palmitate (RP, 6 micrograms/g) supplementation permitted the appearance of tumors. In addition, in comparison to the purified diets and regardless of their RA levels, the non-purified Purina chow diet enhanced tumor yield especially in the two-stage tumorigenesis protocol. This effect was less striking in mice with tumors induced by the complete tumorigenesis protocol. In summary, dietary retinoids are essential for skin tumor formation induced either by the two-stage or the complete tumorigenesis protocol.     

Initiation of urinary bladder carcinogenesis in rats by freeze ulceration with sodium saccharin promotion

Sodium saccharin was previously shown to induce a significant incidence of transitional cell carcinoma of the bladder when administered to rats either immediately or beginning 2 weeks after ulceration of the bladder epithelium induced by freezing or cyclophosphamide injection. However, the marked regenerative hyperplasia following ulceration by either of these methods is not completely repaired until 3 to 4 weeks after ulceration. To determine whether initiation in this model was due to the ulceration and regenerative hyperplasia alone or if it was due to the administration of sodium saccharin acting on the hyperplastic epithelium, the effect of administering sodium saccharin at various times after ulceration was examined. Five-week-old F344 male rats were given sodium saccharin as 5% of the diet beginning either immediately (Group 1) or 2, 4, 6, or 18 weeks (Groups 2, 3, 4, or 5, respectively) after freezing of the bladder, and sacrificed 2 years after the start of the experiment. The incidences of rats with transitional cell carcinoma of the bladder were 11 of 36 rats (31%) in Group 1, 6 of 36 (17%) in Group 2, 12 of 40 (30%) in Group 3, 7 of 36 (19%) in Group 4, and 9 of 39 (23%) in Group 5. Sodium saccharin without prior ulceration induced a transitional cell papilloma in one rat, and freeze ulceration without subsequent sodium saccharin induced a transitional cell carcinoma in one rat. No bladder lesions were seen in the untreated control rats. Scanning electron microscopic examination of rats fed sodium saccharin after ulceration showed evidence of multifocal hyperplasia and significant surface changes either at Week 18 of the experiment (Groups 1 to 3) or 18 weeks after beginning sodium saccharin administration (Groups 4 and 5). These results indicate that freeze ulceration of the bladder induced irreversible changes in the epithelial cells related to bladder cancer initiation even though the regenerative hyperplasia is morphologically reversible, and that sodium saccharin promotes the tumorigenic expression of those freeze ulceration-induced cellular changes even after healing from the injury.     

An evolutionary model for initiation, promotion, and progression in carcinogenesis

Human carcinogenesis is a multistep process in which epithelial cells progress through a series of premalignant phenotypes until an invasive cancer emerges. Extensive experimental observations in carcinogenesis have demonstrated this process can be divided into three general eras: initiation, promotion, and progression. However, this empirically derived, tissue-level explanation of carcinogenesis has not been reconciled with the step-wise genotypic and phenotypic changes encompassed in evolutionary paradigms such as the Feoron-Vogelstein diagram. Here, we analyze an evolutionary model of cellular dynamics that defines mutual interactions of cellular and subcellular events and tissue level changes in tumor growth and morphology. Results are expressed using an adaptive landscape that illustrates the evolutionary potential of cells that allow them to adapt to specific microenvironmental selection forces. It is shown that normal epithelial cells have a novel adaptive landscape that permits coexistence of normal cellular populations but also allows invasion by mutant phenotypes. Subsequent cancer evolution is possible due to a relaxation of tissue growth constraints (as mediated by cell-cell and cell-extracellular matrix interactions) and adaptations in response to perturbations in microenvironmental substrate concentrations (due to separation of evolving tumor cells from their blood supply by an intact basement membrane). Simulations, based on the dynamic model, produce three distinct stages of carcinogenesis that are consistent with the initiation, promotion, and progression stages observed experimentally. The simulations provide insight into the underlying cellular and microenvironmental dynamics that govern these empirical observations and suggest novel prevention strategies that may be tested experimentally.     

Inhibition of mouse skin tumor promotion by adriamycin and daunomycin in combination with verapamil or palmitoylcarnitine

The anti-cancer drugs Adriamycin (ADR) and Daunomycin (DAU) alone were unable to inhibit the promotion of skin papillomas by repeated applications of 8.5 nmol of 12-O-tetradecanoylphorbol-13-acetate (TPA) in 7,12-dimethylbenz(a)anthracene (DMBA)-initiated mice. Pretreatments with 50 micrograms of ADR also failed to alter the tumor-promoting activities of smaller doses of TPA. Therefore, the effects of the anthracycline antibiotics on skin tumor promotion were evaluated in combination with the Ca2+ antagonist verapamil (VRP) and the protein kinase C (PKC) inhibitor palmitoylcarnitine (PC), compounds known to circumvent drug resistance. When applied simultaneously with each promotion treatment with 8.5 nmol of TPA, 2.5 mg of VRP inhibited the number of papillomas/mouse by 26%. But the combination of VRP + 50 micrograms of ADR or DAU inhibited the yields of papillomas by 50 or 47%, respectively, suggesting that VRP was required to reveal the antitumor-promoting activities of otherwise ineffective drugs. Similarly, the promotion of skin tumors by TPA was inhibited synergistically by the combinations of 2 mumol of PC + 50 micrograms of ADR or DAU. For instance, ADR and DAU had no effects alone but inhibited the incidence of skin papillomas by 78 and 86%, respectively, in the presence of PC, a compound which alone inhibited the tumor incidence by only 44%. The results indicate that ADR and DAU are effective against the promoting component of skin carcinogenesis only if they are applied in combination with Ca2+ antagonists or PKC inhibitors at a time when they can inhibit the early biochemical effects induced by TPA.     

Genetics of chemical carcinogenesis: analysis of bidirectional selective breeding inducing maximal resistance or maximal susceptibility to 2-stage skin tumorigenesis in the mouse

We report on bidirectional selective breeding, initiated from a genetically defined foundation population and carried out to selection limit, for producing lines of mice endowed with maximal resistance (Car-R) or maximal susceptibility (Car-S) to 2-stage skin tumorigenesis. The initial population resulted from a balanced intercrossing of 8 inbred strains of mice. The tumors, induced by a single application of DMBA (initiation) and twice weekly applications of TPA (promotion), were benign papillomas; their number at the end of the promotion period was the phenotype chosen for assortative mating. Afterward, the majority of them regressed while others progressed to malignant carcinomas. The Car-R line was selected through a strong challenge, while the Car-S line selection was based on responses to decreasing concentrations of DMBA and TPA. The selection limit was reached after 14 or 15 generations showing progressive interline divergence, which strongly suggests the interaction of several quantitative trait loci (QTL). The phenotypic difference was extremely large: the tumor response was 73 times higher in Car-S than in Car-R mice, though the applied concentrations of DMBA and TPA were 100 and 40 times lower, respectively. The mean heritability realized during the selective breeding was 0.20 in Car-R and 0.49 in Car-S. Our results are compatible with a minimal QTL estimate of 8 in the Car-R line and of 9 or 10 in the Car-S line. The Car-S line is also much more susceptible to carcinoma induction. An association of coat color with tumorigenesis was observed in interline F2 segregants. The Car-R and Car-S lines, obtained through a long-lasting breeding program, are a unique model for identifying the QTL involved in chemical tumorigenesis and will be provided to interested investigators.     

Frequent codon 12 Ki-ras mutations in mouse skin tumors initiated by N-methyl-N'-nitro-N-nitrosoguanidine and promoted by mezerein

The skin tumor initiators N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) and 7,12-dimethylbenz[a]anthracene (DMBA) differ in effectiveness when tumor formation is promoted by 12-O-tetradecanoylphorbol-13-acetate (TPA). Even at high doses, MNNG is less effective, producing fewer benign and malignant tumors with a longer latent period. In DMBA-initiated skin, 10 wk of TPA promotion produced a maximal tumor response. With MNNG, 20 wk of TPA promotion was required, producing nearly four times as many papillomas as 10 wk of promotion. Promotion of MNNG-initiated skin with mezerein induced the appearance of very rapidly-growing papillomas within 5 wk, 3 wk earlier than the first TPA-promoted papillomas. Thus, MNNG may induce a novel mutation resulting in a population of initiated cells that respond especially well to mezerein. Since ras mutations are common in experimental tumors in many tissues, we determined the frequency of activating mutations in the Ha-ras, Ki-ras, and N-ras oncogenes. Activating Ha-ras mutations were present in essentially all DMBA-initiated tumors and about 70% of MNNG-initiated tumors. No N-ras mutations were found in tumors lacking other ras mutations. Surprisingly, 41% of the papillomas arising in the first 11 wk in MNNG-initiated, mezerein-promoted mice bore mutations in codon 12 of the Ki-ras oncogene. Activating Ki-ras mutations were also found in more than 60% of squamous cell carcinomas and 40% of keratoacanthomas. Although mutations in Ha-ras are frequently detected in mouse skin tumors, mutations in Ki-ras are rare. This is the first report of mutated Ki-ras in skin tumors from mice initiated by MNNG.