Mechanisms of SEPA 0009-Induced Tumorigenesis in v- ras Ha Transgenic Tg.AC Mice

Abstract

Genetically engineered mouse models with altered oncogene or tumor suppressor gene activity have been utilized recently for carcinogen identification. The v-ras ^Ha transgenic Tg.AC mouse, with its enhanced susceptibility to skin tumorigenesis, is thought to be well suited for examining the carcinogenicity of topically applied agents. Tg.AC mice were used to examine the carcinogenicity of SEPA 0009, a rationally designed organic molecule designed to enhance drug penetration through the skin. Fifty mg SEPA 0009/kg body weight, 1500 mg SEPA 0009/kg body weight, or the vehicle alone was applied daily to the skin of Tg.AC mice. Nontransgenic FVB/N mice were also treated with the vehicle alone or 1500 mg SEPA 0009. Daily application of a high-dose of SEPA 0009 caused severe and chronic irritation by 1 week that was maintained throughout the experiment. The irritation was accompanied by increased proliferation, increased apoptosis, and expression of the wound-associated keratin 6. High-dose SEPA 0009 induced squamous papillomas in Tg.AC, but not in nontransgenic mice, by 6 weeks. In mice treated with the high dose SEPA 0009, transgene expression was detected in papillomas at week 9, well after the onset of skin irritation and hyperplasia. In contrast, low-dose SEPA 0009 was not irritating to the skin and did not induce papillomas. Thus, SEPA 0009-induced tumorigenesis was associated with chronic and severe irritation. We propose that SEPA 0009-induced tumorigenesis in Tg.AC mice proceeds through an indirect mechanism that is secondary to cutaneous irritation.

Introduction

Traditionally, the carcinogenic potential of chemicals has been tested using life-time tumor incidence assays in rodents where animals are repeatedly exposed to the test chemical for up to 2 years and the tumor incidence determined. The advent of transgenic animal models that constitutively overexpress 1 or more oncogenes held the promise of acquiring similar information within a shorter treatment period. One such model, the Tg.AC transgenic mouse, exhibits many of the characteristics of the 2-stage mouse skin carcinogenesis model that has been studied over the past 50 years.

The 2-stage mouse skin carcinogenesis model depends on an initiation step, commonly a single topically applied dose of a complete carcinogen in which a mutation is fixed in the c-ras ^Ha gene located in the skin (Yuspa, 1994). The second step requires repetitive topical application of a promoting agent at a frequency and dose that induces hyperplasia of the epidermis, inflammation, and other ill-defined events that foster the clonal expansion of tumor progenitor cells carrying a mutation, usually in codons 12, 13, or 61 of the c-ras ^Ha gene (Brown et al., 1990). Depending on the mouse strain, initiating dose and dosing frequency of the promoter, benign epidermal squamous papillomas can begin to appear at the site of application within weeks.

A variety of initiating agents have been described, including 7,12-dimethylbenz[a]anthracene (DMBA), 1-methyl-3-nitro-1-nitrosoguanidine (MNNG), and 3-methylcholanthrene (Brown et al., 1990). The most widely used initiator in these studies has been DMBA, which almost always results in tumors with mutations in codon 61 of the c-ras ^Ha gene. Over the years, many different promoting agents have been described. Early investigations were concerned with the promoting activity of surfactants and lipophilic agents that were used in a variety of cosmetic and other personal care preparations that were in everyday human use. A number of early reports identified a class of mouse skin promoters that include the alkanes and alkane alcohols or acids (Holsti, 1959; Sice et al., 1957; Sice, 1966), but the relevance of mouse skin promotion activity to human risk from these chemicals is not known. However, many tumor promoters in the 2-stage model have also caused tumors in lifetime rodent bioassays (Tennant, 1999).

The Tg.AC mouse line was created by pronucleus microinjection of the construct into the FVB/N mouse strain (Leder et al., 1990). The transgene confers on the animal the property of initiated skin in the context of the mutated c-ras ^Ha gene of the 2-stage mouse skin carcinogenesis model. Though the transgene is present in the cells of all tissues, the induction of transgene expression by exogenous agents is limited to only a few tissues and to the spontaneous tumors that characterize the Tg.AC mouse line (Hansen et al., 1996). Despite the fact that mutations in the c-ras ^Ha gene characterize chemically induced tumors in the liver of several different strains of mice, including the B6C3F1 mouse used in the standard National Toxicology Program (NTP) 2-year bioassay (Balmain and Brown, 1988), Tg.AC mice rarely develop liver tumors (Mahler et al., 1998). This pattern of tissue specific transgene expression is most likely influenced by the site of transgene integration, which has been located at chromosome 11-A2 (centromere proximal) (Humble et al., 2000). With the exception of the bone marrow, constitutive expression of the transgene is absent in most adult tissues (Leder et al., 1990). However transgene expression can be induced by topical application of chemical agents, e.g., genotoxic carcinogens or nongenotoxic carcinogens and promoters (Spalding et al., 1993), by UV radiation (Trempus et al., 1998), and by the wound repair process following full-thickness surgical incisions in the dorsal skin (Cannon et al., 1997).

SEPA 0009 (2-n-Nonyl-1,3-dioxolane) belongs to a series of rationally designed organic molecules aimed at enhancing drug penetration through skin. This compound will reversibly fluidize the lamellar lipids of stratum corneum, thereby allowing coformulated active ingredients to penetrate the skin barrier more readily. While a 2-year lifetime tumor incidence assay in rats showed that topical application of 50 mg SEPA 0009/kg body weight did not increase the incidence of tumors in the treated animals, very high doses of this enhancer (250 and 1500 mg/kg body weight/day) caused significant skin injury in Tg.AC skin and resulted in subsequent papilloma eruptions (unpublished observations). Doses of SEPA 0009 at or below 50 mg did not cause skin injury nor elicit papillomas in Tg.AC mice. The present study was undertaken to better characterize the mechanism of tumor development in Tg.AC mice treated with SEPA 0009 and to better characterize the temporal events leading to papilloma formation on the skin of these animals.

Materials and Methods

Materials

SEPA 0009 (2-n-nonyl-1,3-dioxolane) was provided by MacroChem, Inc. (Lexington, MA). 1500 mg/kg body weight SEPA 0009 dosing solution was prepared as a 25% solution in acetone (Sigma Aldrich, St. Louis, MO); 50 mg/kg body weight SEPA dosing solution was prepared as a 0.85% solution in acetone. Tetradecanoyl phorbol-13-acetate (TPA) and acetone were obtained from Sigma Aldrich (St. Louis, MO). SEPA 0009 solutions were stored at room temperature in a manner that prevented evaporation and were prepared fresh weekly.

Animals

Homozygous female Tg.AC mice (Taconic Farms, Germantown, NY) containing a fetal ζ-globin promoter fused to a v-ras ^Ha transgene were 13–14 weeks of age at the beginning of the experiment. Age-matched female nontransgenic FVB/N mice were obtained from Taconic Farms. All mice were maintained in compliance with NIH Guidelines for Humane Care and Use of Laboratory Animals and our Institutional Animal Care and Use Committee. Identification tags were fastened to the ear of each mouse. Mice were housed 5/cage and fed Purina Lab Diet Rodent Diet 5001 and tap water ad libitum while kept on a 12-hour light/dark schedule.

Tumorigenesis Experiment

One day prior to treatment, mice were weighed and approximately 8 cm² of the dorsal skin shaved with electric clippers. Mice were shaved throughout the experiment as necessary and weighed at intervals to adjust the dose. Then, 1500 mg SEPA 0009 or 50 mg SEPA 0009 in acetone, or acetone alone, was applied daily to 10 mice per group for 5 weeks and then 5 days per week for the duration of the experiment (Table 1). TPA (5 μg per mouse) in 200 μl acetone was applied twice weekly (Table 1). TPA, SEPA 0009, or acetone alone was applied in a total volume of 200 μl to the shaved surface of the skin, an area of approximately 8 mm². Skin-fold thickness was measured weekly using a sliding mechanical calipers by pinching a fold of skin at the center of the treated area between the sliding arms of the calipers. All of the papillomas present on the skin of each mouse were counted, up to 50 papillomas per mouse and recorded weekly. Gross observations of the treated skin, including skin color, hair growth or hair loss, desquamation, and roughness, were recorded weekly. Following euthanasia, portions of the dorsal skin or skin tumors were removed and fixed overnight in 10% neutral buffered formalin for histological analysis or flash-frozen. Final skin-fold thickness measurements, tumor counts, and photographs were also taken at the time of euthanasia.

Transgene Expression

RNA was isolated as described previously in Cannon et al. (1997). For reverse transcription (RT), RNA with oligo[dT]_12–18 (Invitrogen, Carlsbad, CA) and 10 mM dNTP mix (Invitrogen, Carlsbad, CA) was heated to 65°C. First-Strand Buffer (Invitrogen, Carlsbad, CA) and dithiothreitol (Invitrogen, Carlsbad, CA) were added, and the reaction mix was heated to 42°C in a water bath. After the addition of Superscript II Reverse Transcriptase (Invitrogen, Carlsbad, CA), the mix was incubated at 42°C and then heated to 70°C to stop the reaction. Polymerase chain reaction (PCR) was performed using a Thermo-Hybaid thermal cycler (Madison, WI) as described in Cannon et al. (1997) with primers selected from the SV40 polyadenylation splice region of the transgene construct. The primers, 5′-AATTCTGAAGGAAAGTCC-3′ (antisense) and 5′-TGGACAAACTACCTACAG-3′ (sense), encompass a 65 base pair intron and amplify 279 base pair (DNA or unnspliced RNA) and 214 base pair (spliced RNA) products. Thirty cycles of 1 minute denaturation at 94°C, 2 minutes annealing at 50°C, and 2 minutes extension at 72°C were run. The samples were analyzed by electrophoresis on a 2% agarose gel with ethidium bromide and photographed in ultraviolet light.

Histopathological and Immunofluorescence Analysis

Histopathological analysis was performed on hematoxylin and eosin stained sections. Immunofluorescence for Ki67 (Novacastra, Newcastle upon Tyne, UK), keratin 1 (Covance, Berkeley, CA), and keratin 6 (Covance, Berkeley, CA) was performed using Alexafluor 488 conjugated secondary antibodies (Molecular Probes, Carlsbad, CA) and DAPI-containing mounting medium (Vector Labs, Burlingame, CA). Terminal deoxynucleotidyl-transferase-mediated dUTP nick end labeling (TUNEL) was performed according to the manufacturer’s recommendations (Promega, Madison, WI). The number of Ki67-labeled cells or TUNEL-positive cells per 100 basal cells was determined by counting Ki67- or TUNEL-labeled epidermal keratinocytes and DAPI-positive basal keratinocytes. Counts were obtained in at least 5 randomly selected regions per slide for each of 3 mice per group with the investigator blinded as to the identity of the samples. Blood vessels were assessed qualitatively in at least 5 microscopic fields from sections of skin from 3 mice per group.

Results

SEPA 0009 Induces Dermal Irritation and Skin Tumors in Tg.AC Mice

In order to better understand the mechanisms of tumorigenesis in SEPA 0009-treated Tg.AC mice, the effects of SEPA 0009 on Tg.AC and nontransgenic FVB/N skin were examined over the course of 15 weeks of SEPA 0009 application to skin. Mice were treated daily with topical application of the vehicle acetone alone, 50 mg SEPA 0009/kg body weight, or 1500 mg SEPA 0009/kg body weight (Table 1). Dry, scaly, and thickened skin with patchy hair loss was observed 1 week after the start of high-dose SEPA 0009 in Tg.AC mice. Over the course of the next 2 weeks, this effect worsened to produce reddened, scaly, peeling, thickened, and crusted skin with desquamation and patchy hair loss. The skin of non-transgenic FVB/N mice treated with 1500 mg SEPA 0009 appeared similar to that of the transgenic mice. Tg.AC mice treated with acetone or low-dose SEPA 0009 was grossly normal. Skin-fold thickness, a measure of inflammation at early time points, was increased in high SEPA 0009 but not low SEPA 0009 treated mice between 1 and 9 weeks of treatment (Table 2). Nontransgenic FVB/N mice responded similarly to the Tg.AC mice except that skin-fold thickness was significantly less in high SEPA FVB/N compared to similarly treated Tg.AC mice at 9 weeks (Table 2). This difference may reflect the onset of tumor development in the Tg.AC mice or may be a result of a difference in inflammatory response between the 2 genotypes.

Skin tumors appeared on Tg.AC mice treated with 1500 mg SEPA 0009 beginning at 6 weeks. The number of tumors increased to a maximum of 50 per mouse by 15 weeks of treatment, at which time the experiment was terminated because of the heavy tumor burden of these mice (Figure 1A). All of the mice in the high-dose group developed tumors by 10 weeks of treatment (Figure 1B). Twice-weekly treatment of Tg.AC mice with 5 μg TPA in acetone, included as a positive control group, induced irritation. Irritation was indicated by the gross appearance of erythema, edema (indicated by an increase in skin-fold thickness), and desquamation. TPA induced skin tumors as well beginning at 8 weeks (data not shown). Nontransgenic FVB/N mice and Tg.AC mice treated with acetone or low-dose SEPA 0009 did not any develop skin tumors (Figure 1A,B).

SEPA 0009 Induces Cutaneous Inflammation, Irritation, and Epidermal Hyperplasia in Transgenic and Nontransgenic Mouse Skin

While the skin of Tg.AC mice treated with 50 mg SEPA 0009 or the vehicle acetone alone exhibited minor, if any, histopathological changes over the course of treatment, skin from mice treated with 1500 mg SEPA 0009 sustained significant damage. Epidermal hyperplasia was detected in mice treated with 1500 mg SEPA 0009 as early as 1 week after the start of SEPA 0009 treatment (Figure 2E). At this time point, both mild acanthosis and hyperkeratosis of the epidermis was noted (Figure 2E). Hyperplasia became more pronounced as the treatment continued. At 10 days from the start of treatment, the amount of acanthosis and hyperkeratosis had increased slightly and alternating ortho-and parakeratosis was evident (data not shown). When compared to acetone-treated controls, mice treated with 50 mg SEPA 0009 developed a qualitatively assessed, marginally detectable increase in nucleated epidermal thickness and the number of epidermal cell layers between 2 and 9 weeks of treatment (Figure 2C,D). In mice treated with the high dose of SEPA 0009, squamous papillomas developed with scattered dyskeratotic cells (keratinocytes with irregular shrunken nuclei next to dense eosinophilic cytoplasm) in the lower epidermis, consistent with our gross observations of tumor development. The papillomas were identified histologically as benign epithelial neoplasms with a lobular structure (Figure 2H). By 9 weeks, the epidermal lesions progressed further, with even more marked papillomas and epidermal hyperplasia (Figure 2F,H).

After 1 week of 1500 mg SEPA 0009, scattered inflammatory cells, including neutrophils and lymphocytes were present in the dermis (Figure 2E, inset). At 10 days, the degree of inflammation was similar to that of skin after 1 week of treatment, with the exception of increased perivascular mast cells in the papillary dermis. Two weeks after the start of high-dose SEPA 0009 application, the dermal inflammatory infiltrate increased, as did microvessel (blood vessel) density. Between 3 and 9 weeks of treatment, a moderate amount of acute and chronic inflammatory cells including neutrophils and mast cells were present in the papillary, or subepithelial, dermis at these time points. By 9 weeks, as the epidermal lesions progressed, the dermal infiltrate was increased (Figure 2F). By 9 weeks, the dermal inflammatory cell infiltrate intensified when compared with 7 weeks, with a particularly dense infiltrate of mast cells. Microvessel density was further increased within the dermis and vascular congestion was prominent.

In contrast to the changes in Tg.AC skin, nontransgenic FVB/N mouse skin exhibited a much less severe reaction to high-dose SEPA 0009. While very minor skin changes were seen in high-dose SEPA 0009-treated FVB/N skin at 3 weeks, by 9 weeks epidermal hyperplasia was present (Figure 2G). A moderate acute and chronic inflammatory infiltrate, which included mast cells, was present in the papillary dermis (Figure 2G and data not shown). Similarly to the results with the transgenic skin, no pathologic change was seen in the vehicle-treated FVB/N skin.

High-Dose SEPA 0009 Increases Keratinocyte Proliferation and Cell Death

The development of hyperplasia in high-dose SEPA 0009-treated mice was accompanied by increased epidermal cell proliferation. As measured by Ki67 immunofluorescence, cell proliferation was increased more than 13-fold after 1 week of high SEPA 0009 treatment, or from 6.2 ±1.1 in acetone-treated to 82.6 ± 14.5 Ki67 positive cells per 100 basal cells in high SEPA 0009-treated Tg.AC epidermis (Figure 3A,E). The elevation in proliferation was maintained out to 9 weeks of high-dose SEPA 0009 treatment (Figure 3F and Table 3). SEPA 0009-treated nontransgenic mice sustained similarly increased epidermal cell proliferation after 9 weeks of treatment (Table 3). Fifty mg SEPA 0009 did not significantly increase cell proliferation in the skin of Tg.AC mice after either 1 or 9 weeks of treatment (Figure 3C,D and Table 3).

In addition to its effects on cell proliferation and consistent with its irritating effect, high-dose SEPA 0009 also increased apoptotic cell death after 1–9 weeks, while the lower dose increased apoptosis only at the later time point, as shown by a Student’s t-test (Figure 3 and Table 3). Apoptosis was similarly high in nontransgenic FVB/N mice treated with high-dose SEPA 0009 (Table 3). However, TUNEL-positive cells in acetone-treated FVB/N mice were significantly higher than in acetone-treated Tg.AC skin (Table 3). Although this result seems anomalous, the analysis was repeated on additional slides with similar results.

The mechanisms of SEPA 0009-induced tumorigenesis were further investigated using immunohistochemical assays for keratin markers of differentiation. Keratin 1 is normally expressed in the differentiating stratum spinosum, as shown in the acetone-treated skin of Figure 4A,B. SEPA 0009 had no effect on keratin 1 localization, although the thickness of the stratum spinosum appeared increased in both low and high SEPA 0009-treated mice between 1 and 9 weeks of treatment (Figure 4E,F,I,J). This increase correlates with the increased epidermal thickness induced by SEPA 0009 (Figure 2D–F). Keratin 6 is a wound-associated marker of hyperproliferation in the epidermis that is not normally expressed in cutaneous keratinocytes outside of the hair follicles. As shown in Figure 4K, high-dose SEPA 0009 induced keratin 6 expression after only 1 week of treatment. Surprisingly, keratin 6 was also expressed in low-dose SEPA 0009-treated epidermis, only marginally detectable at 1 week, but with a much stronger signal by 9 weeks of treatment (Figure 4G,H). TPA-treated skin expressed keratin 6 by 9 weeks (earliest time point measured) and continued into 16 weeks (not shown).

The v-ras^Ha Transgene is Expressed Well After the Onset of Irritation and Hyperplasia

The v-ras ^Ha transgene is not constitutively expressed in Tg.AC mouse skin but rather is induced following treatments that induce papillomas. In order to determine whether SEPA 0009-induced Tg.AC skin tumorigenesis is an indirect result of its irritant properties rather than a more direct effect of SEPA 0009-induced transgene expression, transgene expression was determined in a time course after SEPA 0009 treatment. Transgene expression was assessed using RT-PCR as described in (Cannon et al., 1997). The bands that appear at 279 bp are due to amplification of transgenic genomic DNA, not dependent on reverse transcription (Figure 5). The bands at 214 bp are reverse transcriptase-dependent, indicating v-ras ^Ha transgene expression. As shown in Figure 5, no transgene expression was detected in skin from mice treated with either acetone alone or 50 mg SEPA 0009. In mice treated with the high dose of SEPA 0009, transgene expression was detected in papillomas at week 9 (Figure 5). Skin tumors from 1500 mg SEPA 0009 or TPA treated mice all expressed the transgene by week 9 (Figure 5).

Discussion

The purpose of this study was to examine the skin tumorigenicity of SEPA 0009 upon topical application in v-ras ^Ha transgenic Tg.AC mice. In earlier studies, Tg.AC mice that received topical daily applications of 250 and 1500 mg of SEPA 0009 in acetone developed multiple skin tumors (unpublished observations). Skin tumorigenesis was preceded by a chronic and dose-dependent dermal irritation, an effective tumor-promoting stimulus. We hypothesized that SEPA 0009-induced skin tumorigenesis in Tg.AC mice may be a consequence of SEPA 0009-induced irritation and skin injury rather than a result of SEPA 0009-induced transgene expression. In order to investigate this hypothesis, we examined the expression of the v-ras ^Ha transgene, dermal irritation, the histology of the skin, and several markers associated with tumor promotion in Tg.AC mice treated with SEPA 0009. Our hypothesis predicted that dermal irritation and skin injury would precede the onset of transgene expression and tumorigenesis in SEPA 0009-treated mouse skin.

As predicted, transgene expression was preceded by several weeks of dermal irritation and injury. Overt desquamation, increased cell proliferation, hyperplasia, hyperkeratosis, inflammation, and keratin 6 expression were observed very early in the time course of SEPA 0009 treatment using the high dose. All of these parameters were altered after only 1 week of daily 1500 mg SEPA 0009 application, many weeks earlier than the first appearance of squamous papillomas and transgene expression at 6 and 9 weeks, respectively. In TPA treated mice, transgene expression was detected in tumors, which first appeared at 8 weeks (earliest timepoint studied), but not in non-tumor-bearing skin. However, in other reports, transgene activation has been detected as early as 9 days from the start of TPA treatment, well before the appearance of tumors (Battalora et al., 2001). These differences may be due to both differences in the sensitivities of the assays used as well as to differences in the mechanisms of action of SEPA and TPA.

We propose that SEPA 0009 induces tumors by causing cutaneous irritation and injury. Consistent with this hypothesis, lower doses of SEPA 0009 that did not irritate the skin also did not produce tumors. The skin of mice treated with 50 mg SEPA 0009 was grossly normal, although a mild epidermal thickening and up-regulated expression of the wounding marker keratin 6 were identified following histopathological and immunohistochemical analyses. In addition, TUNEL analysis indicated increased apoptosis in the Low SEPA mouse skin, although morphological analysis would also be required to differentiate the TUNEL-positive apoptotic keratinocytes from necrotic cells. These data indicate that daily application of 50 mg SEPA, while largely innocuous, still altered the physiology of the skin. Application of acetone alone to murine skin causes barrier disruption and can result in increased DNA synthesis and epidermal hyperplasia (Proksch et al., 1991; Proksch, 1992). Thus, the effects of low-dose SEPA may result, in part, from the combined effects of SEPA and acetone. The findings suggest that severe and sustained nonneoplastic skin changes are related to tumorigenesis. However, in the absence of significant irritation, no tumors developed. As expected, nontransgenic FVB/N mice also did not develop tumors. In these mice, high-dose SEPA induced irritation, inflammation and hyperplasia, although with a slower time course than in the transgenic mice. The reason for the differences between the response of the Tg.AC mice and nontransgenic FVB/N mice to SEPA-induced irritation is not clear, but may be an effect of the transgene insertion.

These experiments were designed to determine the carcinogenic potential of SEPA following topical application. In the last few years, alternative approaches to the NTP 2-year bioassay, such as the Tg.AC mouse, have been proposed to improve upon methods of carcinogen identification. Genetically engineered mouse models with mutations in ras ^Ha and p53, which are appealing because these mutations are relevant for many human cancers and have been used most frequently for this purpose (Alden et al., 2002; Pritchard et al., 2003).

These models can provide mechanistic information, because of selectivity of response to genotoxic versus nongenotoxic carcinogens, and avoid false positives due to the predisposition of mice to liver and thyroid cancer in the NTP bioassay (Alden et al., 2002). The Tg.AC mouse has been proposed to be a good predictor of human carcinogens applied topically and responds with sensitivity to both mutagenic and nonmutagenic carcinogens (Spalding et al., 1993; Pritchard et al., 2003). However, the sensitivity of this model to wounding and chronic irritation may also be its greatest weakness. The sensitivity of Tg.AC mice to proliferative stimuli is evidenced by the development of tumors following a single full-thickness incision in the skin (Cannon et al., 1997).

Agents that cause severe, chronic and overt irritation and inflammation, are often tumorigenic in Tg.AC mice (Sistare et al., 2002). For example, resorcinol, which was negative in the rat and mouse NTP bioassay, induced papillomas in Tg.AC mice when given at a level that also caused inflammation in the skin (Sistare et al., 2002). However, investigations into the role of inflammation in tumorigenesis in Tg.AC mice have also produced contradictory results, necessitating further research (Albert et al., 1996; Murphy et al., 2003). Tumor development in SEPA 0009-treated Tg.AC mice, which is preceded by a severe and chronic skin irritation, is consistent with the response of Tg.AC mice to sustained irritation and inflammation. These results invite us to revisit questions about the influence of irritation on carcinogenesis and how irritants to rodent skin that are positive in Tg.AC tumorigenesis assays should be regulated. In contrast to our results, some agents produce tumors in Tg.AC mice without causing irritation or inflammation. Wyde et al. (2004) demonstrated a dose-response for cutaneous papillomas in Tg.AC mice following daily application of only nanogram quantities of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). At this dose, the effects of TCDD are clearly a receptor-mediated response that did not produce significant inflammation or irritation. We propose that tumorigenesis in Tg.AC mice induced by agents that also cause sustained irritation of the skin should be considered mechanistically distinct from assays in which significant irritation does not occur. This implies as well that irritant and nonirritant Tg.AC tumorigens might be considered differently for regulatory purposes.

In summary, we found that daily application of 1500 mg/kg SEPA 0009 rapidly induced dermal irritation and skin injury that were associated with increased inflammation, increased cell division and cell death, epidermal hyperplasia, and up-regulation of the hyperproliferation marker keratin 6 in genetically-initiated transgenic Tg.AC mice. Tumor development followed many weeks later, associated with expression of the v-ras ^Ha transgene. This pattern is consistent with an indirect mechanism of SEPA-induced tumorigenesis that is secondary to the cutaneous irritation and injury.