What is the Tg.AC transgenic mouse?

The Tg.AC transgenic mouse was developed in the Leder laboratory (Leder et al., 1990) to examine gene regulation in embryonic development. The line was one of several created on an FVB/N strain mouse background by pronuclear injection of a transgene construct containing a v-Ha-ras gene flanked 5' by a mouse -globin promoter and 3' by an SV-40 polyadenylation sequence (Figure 1). The Tg.AC mouse was created with two goals in mind: to define the putative promoter region of the mouse -globin gene, and to create an embryonic erythroid cell line from the -globin/v-Ha-ras-induced tumors (Leder et al., 1990). Four transgenic lines were created carrying 3–10 copies of the transgene (Leder et al., 1990). Three of the lines showed no transgene expression or apparent phenotype. A fourth, Tg.AC, expressed the transgene in day 12 embryonic blood, the hematopoietic fetal liver, and the placenta. In addition, low levels of transgene expression were detectable in the adult bone marrow (Leder et al., 1990). It was noted that the Tg.AC mice developed dorsal squamous cell papillomas at the site of bite wounds inflicted by cage mates. Transgene expression in normal Tg.AC skin was not detected (Leder et al., 1990).

Figure 1.

Transgene construct in v-Ha-ras Tg.AC transgenic mice. The Tg.AC transgene consists of a 1 kb region corresponding to the mouse -globin promoter, a 1.8 kb v-Ha-ras sequence, and a 1 kb SV-40 polyadenylation signal sequence. Restriction sites found at the junctions between each component and at the ends of the construct are indicated. The ras gene contains two activating mutations found at amino acids 12 and 59

Full figure and legend (11K)

Wound repair and chronic irritation are known tumor promoters (Pullinger, 1945; Hennings and Boutwell, 1970; Argyris, 1985). The observed promotional response to wound stimuli in Tg.AC was further explored though the use of the known chemical tumor promoter, 12-O-tetradecanoylphorbol-13-acetate (TPA). When treated topically with TPA, Tg.AC mice developed multiple papillomas, some of which progressed to malignancies (Leder et al., 1990; Spalding et al., 1993). Tg.AC mice also develop papillomas in response to full-thickness wounding (Leder et al., 1990; Hansen and Tennant, 1994a; Cannon et al., 1997). UVA/B radiation (Trempus et al., 1998a; Battalora et al., 2001) and to a lesser extent, UVA alone (Chignell et al., 2003b) can generate a tumorigenic response in Tg.AC, an effect that can be diminished by pretreatment with an inhibitor of the epidermal growth factor receptor (EGFR) (El-Abersari et al., 2005). Other wavelengths of radiation have been found to be tumor suppressive (Ohara et al., 2003). As the parent FVB/N strain does not develop tumors in response to these promotional stimuli in the absence of initiation, the tumorigenic response in Tg.AC mice must be due to the presence of the v-Ha-ras transgene.

Experimental skin carcinogenesis is typically a multistage process consisting of initiation, promotion, and progression. Initiation is thought to induce mutations in critical genes which confer a selective growth advantage to specific cells, which, during the proliferative process of tumor promotion, can acquire additional genetic or epigenetic changes resulting in tumor formation and a potential progression to malignancy (Deelman, 1927; Barry et al., 1935; Berenblum, 1941; Boutwell, 1964). Chemically induced mutations in the c-Ha-ras gene have been detected following initiation with the mutagen 7,12-dimethylbenz[a]anthracene (DMBA) in mouse skin carcinogenesis studies (Balmain and Pragnell, 1983). Because the tumorigenic response observed in Tg.AC mice occurred independent of an initiation step, and is due to the presence of the v-Ha-ras transgene, these mice have been characterized as a 'genetically initiated' model for mouse skin tumorigenesis. Tg.AC mice can develop papillomas within 20 weeks of topical applications of either mutagenic or nonmutagenic carcinogens (Leder et al., 1990; Spalding et al., 1993). Hemizygous and homozygous Tg.AC mice have been found to have similar tumorigenic responses (Hansen et al., 1998; Spalding et al., 1999).

In addition to the wound or chemically induced skin tumors observed on Tg.AC mice, a number of other tumor types have been reported. Spontaneous tumors develop at the sites of chronic abrasion, such as the mouth, nares, eyelids, and genitalia (Leder et al., 1990). Approximately 36% of homozygous Tg.AC males and females will develop odontogenic tumors within their first year of life, a phenomenon uncommon in mice (Wright et al., 1995), and likely due to local injuries caused by continuous gnawing. One type of odontogenic tumor appears to arise from cells in the periodontal ligament, which is located in the mid-root region of the incisor (Wright et al., 1995; Mahler et al., 1998). In a more recent investigation, it was found that genes displaying differential expression in the ameloblastoma-like tumors of Tg.AC mice were also the same genes thought to be important in the odontogenesis and odontogenic tumor formation in human ameloblatomas (Dodds et al., 2003). Other reported spontaneous tumors include squamous cell carcinomas (SCC) of the salivary gland, ovarian teratomas, a rare ovarian yolk sac carcinoma, squamous cell papillomas of the forestomach and lung alveolar/bronchiolar adenomas (Hansen et al., 1996; Mahler et al., 1998). The incidence of erythroleukemia in control hemizygous Tg.AC mice is 1–4% (Mahler et al., 1998). The erythroleukemia have an associated hepatosplenomegaly due to leukemic infiltration. No clear evidence of erythroleukemia has been associated with any particular chemical treatment (Trempus et al., 1998b), and erythroleukemias have not been observed in the parental FVB/N line. Tumors also form in the forestomach (Cannon et al., 2000).

Tg.AC transgene expression is detected in day 12 embryonic blood, the hematopoietic fetal liver, placenta, kidney, and at low levels in adult bone marrow (Leder et al., 1990; Hansen et al., 1996; Delker et al., 1999). Transgene expression was not detected in normal Tg.AC skin, but was detected in skin papillomas, carcinomas, malignancies, and all spontaneous tumors examined (Leder et al., 1990; Hansen and Tennant, 1994a). RT–PCR has failed to detect transgene expression in the liver (Hansen et al., 1996; Delker et al., 1999), although some adult Tg.AC spleens will test positive for transgene expression by RT–PCR or by an RNase protection assay (Cardiff et al., 1993; Hansen et al., 1996). It should be noted that not all Tg.AC mouse tissues respond with a tumorigenic outcome following wounding. Despite constitutive transgene expression in the kidneys, chloroform-induced kidney damage was not sufficient to generate a renal tumorigenic response (Delker et al., 1999). FVB/N mice are not susceptible to spontaneous kidney tumors (Mahler et al., 1996) nor were spontaneous kidney tumors seen in 33–34-week-old control hemizygous Tg.AC mice (Mahler et al., 1998).

Skin tumorigenesis in Tg.AC mice is strongly age dependent (Battalora et al., 2001). Limited (two to four) doses of TPA on mice 5 or 10 weeks of age develop a much lower tumor incidence and multiplicity when compared to mice 21 and 32 weeks of age treated in a similar manner. The age-dependent increase in response occurs following TPA exposure, wounding, or UV irradiation (Battalora et al., 2001). Young mice receiving a second treatment later in life will respond in a manner to similarly dosed older mice. It is interesting to note that FVB/N mice do not demonstrate an age-dependent response when treated with DMBA/TPA (Battalora et al., 2001).

Tg.AC mice can develop cutaneous malignancies, the sites of which have always been associated with spontaneous or induced papillomas (Hansen et al., 1996; Asano et al., 1998). Malignancies are primarily either squamous cell carcinomas or spindle cell carcinomas (Hansen et al., 1996), although malignant fibrous histiocytoma-like (MFH-like) spindle cell tumors have been reported (Cardiff et al., 1993; Asano et al., 1998). Based on subsequent immunohistochemical staining for keratins, as well as the presence of desmosomes, the spindle cell carcinomas have been shown to be of epithelial origin (Asano et al., 1998). Transgene expression, while usually found at the basal region of papillomas, may be found throughout spindle or squamous cell carcinomas (Hansen et al., 1996).

Tg.AC and chemical carcinogenesis

The Tg.AC mouse, along with additional transgenic mice, has been proposed as an alternative or adjunct to the rodent 2-year bioassay (Tennant et al., 1995; Tennant, 1997), a standardized scientific procedure developed by the National Toxicology Program (NTP) for identifying toxic and carcinogenic compounds that are hazardous to human health. The International Life Sciences Institute (ILSI) sponsored a collaborative evaluation program to examine alternative models for carcinogenicity testing (Cohen et al., 2001; Robinson and MacDonald, 2001; Tennant et al., 2001). With the ILSI initiative and individual lab testing, a wide variety of chemicals have been tested on Tg.AC mice in an effort to validate the model (Tennant et al., 1996; Spalding et al., 1999, 2000). An extensive analysis of those chemical tests has recently been performed (reviewed in Pritchard et al., 2003; Sistare et al., 2002) and a compilation of tests and references can be found on the Tg.AC website at http://dir.niehs.nih.gov/dirlecm/TgAC/TgAC_Home.htm. The data from these studies show that a variety of chemical carcinogens can generate a tumorigenic response in Tg.AC mice, including both genotoxic and nongenotoxic carcinogens. The Tg.AC data, and the ILSI initiative have demonstrated transgenic assays to be valuable adjuncts to the 2-year bioassay. There are concerns when the model is being utilized for human risk assessment of dermal formulations applied to human skin. Positive responses are likely to occur, if these formulations contain components, or their expected metabolites, that have been previously identified as promoters in the two-stage mouse skin carcinogenesis model. As noted, '... treatments that are tumor promoting in a two-stage skin carcinogenesis model would be expected to induce papillomas in this system, if the mutated ras gene products can be expressed and cooperates with events that follow the promotional stimulus (Sistare et al., 2002)'.

The induction of papillomas on topically treated Tg.AC skin serves as a reporter phenotype that defines the carcinogenic activity of a chemical compound (Figure 2). TPA, the most well-defined promoter in two-stage carcinogenesis protocols, is the most common promoter utilized to generate an experimental tumorigenic response, and its use constitutes most of the positive control group dosing in bioassays and experimental protocols. However, the stratified epithelium of the forestomach is also a potential target. In accordance with the NTP bioassay, dimethyl vinyl chloride (DMVC) was able to significantly increase the incidence of forestomach papillomas in Tg.AC mice in both a time and dose-related manner (Cannon et al., 2000). All forestomach tumors examined expressed the transgene, while transgene expression in normal, nontumorigenic forestomach was not detected (Cannon et al., 2000).

Figure 2.

Papilloma induction by TPA. TPA (2.50 g) was applied to the shaved dorsal surface, three times a week for up to 20 weeks or until a maximum tumor response of 30 papillomas is reached. Animals having 30 papillomas before the end of treatment were euthanized

Full figure and legend (65K)

The truncation of the tumorigenic timeline and the discriminate response to carcinogens highlights the use of Tg.AC mice as a short-term bioassay for identifying potential chemical carcinogens. The relevance to humans can be correlated to the finding that 30% of human tumors contain activating point mutations in the Ha-ras gene (Bos, 1987). While skin carcinogenesis in the mouse is not orthologous to that in humans, this correlation of ras-dependent pathways in the tumorigenic process strengthens the utility of the Tg.AC bioassay. Understanding the cellular targets and the molecular mechanisms by which the -globin promoter/v-Ha-ras transgene confers the tumorigenic phenotype in Tg.AC mice is paramount in its continued use as a model for carcinogenesis and as an adjunct to the 2-year bioassay.

Cellular aspects

A major focus of current research has been to identify the specific cells at risk for transgene-dependent proliferation. Untreated Tg.AC skin is grossly and histologically indistinguishable from that of FVB/N skin (Hansen and Tennant, 1994b). Also indistinguishable between the two strains is the hyperplasia induced by TPA treatment, as well as the early activation and subsequent downregulation of protein kinase C (PKC) (Hansen and Tennant, 1994a). However, TPA-promoted Tg.AC mice develop focal hyperplasias that eventually develop into squamous papillomas, while uninitiated FVB/N fail to do so. All tumors tested to date, whether spontaneous or induced, express the ras transgene, as determined by RT–PCR, Western blots, Northern analysis, and in situ hybridization, with the highest level of transgene expression associated with the proliferating cells of the basal epidermis (Figure 3) (Hansen and Tennant, 1994a; Wright et al., 1995; Hansen et al., 1996; Cannon et al., 1997, 2000).

Figure 3.

Expression of ras transgene message and PCNA. Tissue from treated Tg.AC skin was removed, fixed in 10% neutral-buffered formalin and assayed for proliferating cell nuclear antigen expression by immunohistochemistry (panel a) and ras transgene message by in situ hybridization (with a transgene-specific riboprobe) (panel b). Note that transgene expression is localized to the basal compartment of the lesion, coincident with proliferating cells

Full figure and legend (372K)

Transgene expression is not detectable in normal whole skin, but can be detected in focal hyperplasias and papilloma precursors that develop following a promotional stimulus. This observation can be accounted for in one of two ways. First, the ras transgene is not expressed in normal, untreated skin, but is induced in a small population of susceptible cells, which then clonally expand into ras transgene-expressing papillomas. Second, it is possible that transgene expression is constitutively present in a small population in the epidermis, but due to the small numbers, transgene expression is below the lower limit of detection using current methodologies. In the absence of additional signaling, neoplastic proliferation does not occur. In most cases, using whole skin, the earliest transgene expression detected is 18 days following the first of four TPA treatments (Hansen and Tennant, 1994a) or 16 days after a full-thickness wound (Cannon et al., 1997), concomitant with global hypomethylation (Cannon et al., 1998). However, using keratinocytes harvested from TPA-treated skin, it has been shown that the transgene can be detected by RT–PCR 14 days after treatment with TPA in older, more responsive mice, and as early as 9 days after TPA treatment using fractionated keratinocytes (Battalora et al., 2001). These results provide clues as to the temporal kinetics of transgene expression, but – given detection limits – it is likely that induction of the ras transgene occurs earlier in the neoplastic process than we have measured, and probably in specific cells 'at risk' in the epidermis and/or hair follicle.

The two-stage murine epidermal carcinogenesis model provides insights into the cellular origin of tumors. Initiated cells persist in the skin presumably for the lifetime of the animal despite renewal of the epidermis every 6 days (Potten, 1983). In order for latent neoplastic cells to persist in the skin, they must be slowly cycling to avoid removal to the suprabasal layers and loss by terminal differentiation, and they must reside within a protected niche or microenvironment within the skin (Lavker and Sun, 2000). Therefore, carcinogen target cells possess some qualities of a stem cell population (Morris, 2000). The most likely repository of keratinocyte stem cells (KSC) in the skin is the bulge region, which is located in the permanent portion of the hair follicle below the sebaceous gland at the point of attachment of the arrector pili muscle. Bulge cells are biochemically distinct in that they express keratin 15 (Lyle et al., 1998), and they are slowly cycling (label-retaining cells) (Cotsarelis et al., 1990). The bulge region itself is protected from cyclical changes in the hair follicle as well as from environmental damage (Cotsarelis et al., 1990).

The first evidence for a follicular origin of papillomas in Tg.AC mice was the observation of transgene expression in early focal hyperplasias, in a region consistent with the bulge (Hansen and Tennant, 1994b; Hansen et al., 1995). Transgene expression was also localized to the 20% highest expressing -1 integrin population in keratinocytes harvested from TPA-treated, nontumor-bearing skin (Battalora et al., 2001). It has been demonstrated that the 20% '-1 bright' keratinocytes possess a subset of cells that are slowly cycling and clonogenic in culture, thereby providing a selectable marker for KSCs (Jones and Watt, 1993). Data localizing expression to this population was evidence that the earliest transgene-expressing cells are in the pool of cells containing stem and progenitor cell characteristics (Figure 4).

Figure 4.

Follicular origin of papillomas in Tg.AC mice (adapted from (Hansen et al., 1995)). (a) Hair follicles of normal skin (E: epidermis; D: dermis; HF: hair follicle; DP: dermal papilla; S: sebaceous gland; APM: arrector pili muscle; B: bulge region). (b) Following a promotional stimulus, a small population of transgene-expressing cells in the bulge region of a follicle begin to proliferate. (c) A focal hyperplasia within the hair follicle is established. (d) The focal hyperplasia begins to push above the surface of the skin. (e) Formation of a pedunculated papilloma of follicular origin

Full figure and legend (40K)

Induction of the ras transgene in the skin via the hematopoietic -globin promoter was unexpected. In order to provide a biological rationale for the transcriptional activation of a blood-borne promoter in the skin, the hematopoietic system was examined for insights into the follicular microenvironment. As described below, members of the GATA family of transcription factors that are typically associated with the hematopoietic system have been found to coexpress with the transgene (Trempus et al., 1998b; Delker et al., 1999; Humble et al., submitted). It has also been reported that a reduction in GM-CSF reduces tumor multiplicity in Tg.AC mice (Germolec et al., 1997). The expression of CD34, a 105–120 kDa glycoprotein expressed in hematopoietic stem and progenitor cells, was examined in the skin and intense membrane staining was found in cells in the follicular bulge region that colocalized with both slowly cycling (label retaining) and keratin-15-expressing cells (Trempus et al., 2003). CD34 expression has been used in combination with -6 integrin, which is expressed on basal keratinocytes, and fluorescent-activated cell sorting (FACS) to isolate living bulge keratinocytes that are both slowly cycling and form large colonies in culture (Trempus et al., 2003) and unpublished observations). Preliminary evidence suggests that CD34⁺ keratinocytes express the ras transgene following TPA treatment, providing additional support for the hypothesis that KSCs are the carcinogen target cell population (Carol Trempus, unpublished observations).

The identification of TPA-inducible genes that contribute to tumor development is important with regard to understanding the biology of the neoplastic process. The identification of CD34 as a selectable determinant of follicular bulge cells allowed us to identify several important candidate genes that may contribute to tumor development. CD34⁺ keratinocytes were isolated from either TPA-treated or untreated Tg.AC skin, and a gene expression profile was developed using cDNA nylon arrays (Wei et al., 2003). A total of 11 genes were identified as either upregulated (nine genes, including galectin 7, nucleoside diphosphate kinase B, keratin 14, Dss1, and a double-strand break repair RAD21 homolog) or downregulated (two genes, including keratin 15 and apoliprotein E precursor) in CD34⁺ cells from TPA-treated mice. Further characterization of Dss1 has revealed that (1) Dss1 is overexpressed in TPA-induced hyperplastic skin, papillomas, and cutaneous malignancies; (2) constitutive expression of Dss1 promoted cell proliferation and soft agar growth in preneoplastic epidermal cell lines, demonstrating in vitro transformation potential and defining a possible role for Dss1 in early TPA-induced neoplastic development (Wei et al., 2003).

Molecular aspects

In addition to identifying the cell(s) responsible for transgene-dependent tumorigenesis, it is also important to gain an understanding of the molecular controls regulating transgene expression and tumor formation. Only one of the four founder lines in the creation of the -globin/v-Ha-ras transgenic mouse exhibited the unique skin tumorigenic response observed in the Tg.AC (Leder et al., 1990), leading to a hypothesis that the transgene-driven tumorigenic response is the result of a position effect. To explore the possibility that transgene expression is initiated not by the transgene's -globin promoter, but rather by the promoter of a nearby gene or possibly the promoter of a gene in which the transgene construct is integrated, RNA was isolated from Tg.AC erythroleukemic liver and spleen (leukemic infiltrates) as well as from Tg.AC carcinoma cell lines and examined for the start of transgene transcription by an S1 protection assay (Trempus et al., 1998b). Results indicated that the start of transcription initiated from the natural start site within the transgene mouse -globin promoter, and not from an adjacent gene in the flanking chromosome.

The role of the -globin promoter

The regulatory portion of the Tg.AC -globin proximal promoter has three regions with adjacent GATA and Sp1-binding sites, as well as an AP1-binding site, all of which are occupied by transcription factors according to DNase I footprinting assays (Humble et al., submitted). Functional assays utilizing deleted -globin promoter/secreted alkaline phosphatase (SEAP) reporter vectors demonstrated the importance of the most proximal and distal GATA–Sp1 sites for transgene promoter function, while the central GATA–Sp1 contributed little to promoter function (Humble et al., submitted).

The specific transcription factors interacting with the Sp1, GATA, and AP1 sequences within the regulatory region of the transgene -globin promoter may account for the tissue-specific gene expression observed in the Tg.AC mouse. Electrophoretic mobility shift assays have identified Sp1 binding to the three Sp1 sequences in the regulatory promoter, as well as a cooperativity between Sp1 and a GATA-binding factor (Humble et al., submitted). Expression of the GATA family of transcription factors is typically associated with hematopoietic tissues and cells, but has been observed in additional nonhematopoietic tissues such as kidney, developing jaw, brain (Orkin, 1992), and hair follicle (Kaufman et al., 2003). In Tg.AC mice, GATA-1 has been reported to coexpress with the transgene in erythroleukemias (Trempus et al., 1998b). GATA-3 expression has been found in the kidney of Tg.AC mice, a site where transgene expression has been shown to be constitutively expressed without tumor development (Delker et al., 1999). GATA-3 expression has also been identified in Tg.AC papillomas and carcinomas by RT–PCR (Humble et al., submitted). The tissue-specific expression of GATA-3 and the observed cooperative binding between Sp1 and a GATA-binding factor at sites within the transgene -globin promoter may help to account for the spatial and temporal expression of the Tg.AC transgene.

The palindrome

An important feature within the transgene array is a palindromic repeat formed from two transgene copies joined in a head-to-head inverted orientation ( ) (Thompson et al., 1998). A variety of 100 bp and larger symmetric or asymmetric deletions across the EcoR1 center of symmetry within the palindromic -globin promoters eliminates the tumorigenic response in hemizygous Tg.AC mice resulting in a nonresponding (NR) phenotype (Thompson et al., 1998). In homozygous Tg.AC mice, the presence of an intact transgene array on one allele and a deleted palindrome on the other allele creates a heterozygous nonresponder, with the deleted palindrome being recessive to the dominant intact palindrome. The effect of the NR allele in these mice is masked, as the remaining intact transgene palindrome enables transgene expression and a tumorigenic phenotype.

The NR phenotype was originally reported following poor TPA responses in positive control Tg.AC hemizygous mice (Weaver et al., 1998), and may have been an unknown confounding factor in other studies (Blanchard et al., 1998). Subsequent examination of the entire Tg.AC line demonstrated a germline transmission of the deleted palindrome. Careful monitoring of the breeding lines by DNA blotting for the presence of the NR allele (Kantz et al., 1999) has reduced the appearance of NR hemizygous Tg.AC mice to less than 2% per generation.

Palindromic sequences are known to be unstable and prone to deletion (Leach, 1994; Collick et al., 1996; Lewis et al., 1999). It has been reported that a Tg.AC carcinoma cell line continued to express the transgene despite loss of the promoter palindromes (Humble et al., 2000). To examine palindrome stability over time and its role in sustained transgene expression, DNA was isolated from normal and TPA-treated hemizygous Tg.AC skin, papillomas, carcinomas, and malignancies, and examined for the presence of intact or deleted palindromes. The results demonstrated a quantifiable but incomplete loss of the palindrome in carcinomas and malignancies (Moon and Cannon, 2001; Thompson et al., 2001).

An intact -globin promoter palindrome is required for the Tg.AC discriminate response to promotional stimuli, yet the function of the promoter palindrome is not known. It has been suggested that a stem-loop or hairpin structure may form as a result of the palindromic sequences, and that this structure in some manner enables transgene expression (Thompson et al., 1998). It is also possible that the palindromic transgene promoters act much the same as a traditional locus control region (LCR) by initiating an open chromatin domain for gene expression. LCRs may be many kilobases away from a gene, and may function by altering the topography of the chromatin domain, or by directly interacting with genes by a looping model (reviewed in Fraser and Grosveld, 1998; Grosveld, 1999). Deletion across the central palindrome may remove important transcription factor-binding sites as well as remove the palindrome's ability to initiate chromatin remodeling, rendering the transgene quiescent and noninducible and creating the nonresponder phenotype. Sequence analysis of the 5' end of the -globin promoter indicates transcription factor-binding sites, particularly GATA and AP1 sites, within the first 100 bp (Humble et al., submitted).

Flanking chromosomal sequences, such as LCRs, regulate the expression of a number of genes, including the - and -globin genes (Tuan and London, 1984; Forrester et al., 1987; Tuan et al., 1989; Higgs et al., 1990; Jarman et al., 1991). The 5' regulatory sequences, such as the -globin LCR, were not included with the -globin promoter in the Tg.AC transgene, which may account for transgene expression in nonhematopoietic tissues (Leder et al., 1990; Delker et al., 1999). To explore the possibility that cis-flanking regions near the integration site on chromosome 11 were involved in transgene expression, radial transformation-associated recombination (TAR) cloning (Kouprina et al., 1998) in yeast was utilized to create orientation-specific Yeast Artificial Chromosomes (YACs)/Bacterial Artificial Chromosomes (BACs) containing variable lengths of 5' or 3' flanking chromosome 11 DNA and the Tg.AC transgene (Humble et al., 2000). BACs were assayed for their ability to promote transcription of the transgene following stable transfection into an FVB/N carcinoma cell line. A transgene-specific RT–PCR assay demonstrated that all Tg.AC BACs expressed the transgene in this in vitro system regardless of their flanking DNA content (Humble et al., 2000). This suggests that transcriptional activity may not require cis elements flanking the transgene's integration site. It was noted, however, that the transgene promoter palindromes were deleted in the YAC/BACs at some point in the cloning or propagation process, likely due to the instability of palindromic sequences in Escherichia coli (Humble et al., 2000). Although necessary for in vivo induction of the transgene, the absence of intact promoter palindromes did not hinder transgene expression in vitro. It was postulated that conditions within the FVB/N carcinoma cell line may circumvent the need for the intact palindromes as these cells would have already achieved a carcinogenic potential and thus may not mimic the cellular environment or conditions of the nontransformed/promoted cell in vivo.

In addition to the presence of specific transcription factors, flanking control sequences, and chromatin packing, the control of gene expression is also affected by a gene's methylation status. Embryonic globin genes are hypomethylated while expressed, becoming methylated over time and silenced, while adult globin genes become hypomethylated as their expression begins ((Karlsson and Nienhuis, 1985; Yisraeli et al., 1988) and references therein). A variety of stimuli have been shown to affect gene expression by altering CpG methylation (reviewed in Cannon et al., 1998). The methylation status of the Tg.AC transgene was examined over the time course of wound-induced papillomagenesis (Cannon et al., 1997). A site-specific hypomethylation of the transgene is detectable by DNA blots 23 days after wounding (Cannon et al., 1998). This timing was in close correlation with additional time course parameters, primarily the RT–PCR detection of transgene mRNA 16 days post wounding, and blot detection of p21 v-Ha-ras protein 21 days post wounding (Cannon et al., 1998). Thus, the Tg.AC transgene exists in a hypomethylated state in expressing cells.

Initial induction of transgene expression

TPA is a multicyclic, phorbol ester containing a diacylglycerol-like moiety. Diacylglycerols (DAGs) are intracellular second messengers generated by the receptor-mediated hydrolysis of membrane phospholipids. sn-1,2-didecanoylglycerol (DIC₁₀), a model DAG, readily formed tumors in 100% of Tg.AC mice, while neither glycerol nor n-decanoic acid, potential degradation products of DAGs, were able to promote tumors (Mills et al., 1993; Owens et al., 1995). A similar result was obtained in Tg.AC mice treated with an initiating dose of DMBA followed by DIC₁₀ promotion (Owens et al., 1995). These results could implicate a role for phospholipid metabolism and the production of endogenous DAG signaling through protein kinase C (PKC) in Tg.AC tumorigenesis.

TPA and DAGs are known ligands for PKC. Upon activation by TPA, PKC sets in motion pathways leading to the upregulation and activation of transcription factors such as c-fos and c-jun (AP1), factors that may play an important role in initiating cell proliferation. AP1 may then bind sequences within the regulatory region of the -globin promoter as well as at the center of symmetry of the head-to-head -globin promoter palindromes. An AP1 sequence within the transgene -globin proximal promoter has been shown to be occupied in transgene-expressing tissues by DNase I footprinting (Humble et al., submitted). AP1 upregulation and its subsequent binding to the transgene palindrome and promoter may embody the first step towards transgene induction and represent a common link between seemingly disparate promotional stimuli. An upregulation of a GATA factor in the skin, its binding to the transgene palindrome, and its cooperative binding with Sp1 to key elements within the -globin promoter (Humble et al., submitted) may further account for the tissue specificity observed in the Tg.AC tumorigenic phenotype.

Post-translational modification/activation of ras

Post-translational controls also play a role in v-Ha-ras transgene expression. Post-translational addition of a C₁₅ or C₂₀ isoprenoid moiety to the C terminus of ras proteins enables ras function and anchoring to the cell membrane. Addition of the C₁₅ moiety is accomplished by the enzyme farnesyl protein transferase, a target for anti-ras drug therapies. Homozygous and hemizygous Tg.AC mice were treated with a farnesyl transferase inhibitor (FTI), SCH 56582, prior to exposure to TPA (Trempus et al., 2000). SCH 56582 was able to reduce TPA-induced papillomagenesis, indicating post-translational farnesylation is involved in ras-transgene function.

A common phenomenon observed early in the Tg.AC tumorigenic response is the induction of cell proliferation within the epidermis. Topical exposure to tumor promoters (Leder et al., 1990; Spalding et al., 1993), chemical carcinogens (Tennant and Spalding, 1996), UV light (Trempus et al., 1998a; Chignell et al., 2003a), full-thickness surgical wounding (Cannon et al., 1997), and depilation (Hansen and Tennant, 1994b) all result in epidermal hyperplasia. Within that hyperplasia, expression of the transgene in specific cells drives clonal expansion and tumor formation. Treatment of Tg.AC mice with FTI SCH 56582 did not affect TPA-induced epidermal hyperplasia (Trempus et al., 2000), implicating papillomagenesis to be a ras-dependent phenomenon, while TPA-induced hyperplasia is not. This is supported by the observation that a single low-dose treatment of TPA can induce hyperplasia, yet will not induce tumor formation or detectable transgene expression in Tg.AC mice. In addition, proliferating nontumorigenic cells adjacent to the tumors or in other nontumor tissues can be observed that do not express the transgene (Hansen et al., 1996).

A role for the endogenous c-Ha-ras gene in Tg.AC tumorigenesis has been explored. Sequence analysis of the endogenous c-Ha-ras gene in Tg.AC tumors revealed no activating mutations in codons 12, 59, or 61 (Hansen et al., 1996). In addition, no codon 12, 13, 59, nor 61 mutations were observed in c-Ha-ras following DMBA/TPA treatment of Tg.AC mice (Owens et al., 1995), although more tumors were generated in the DMBA-treated Tg.AC mice than in non-DMBA-treated mice. Additionally, no p53 mutations were detected.

Structural rearrangements within the Tg.AC or FVB/N genome were rarely observed (French et al., 1994). A lack of trisomies on c-Ha-ras bearing chromosome 7 has been reported (French et al., 1994), offering further support that endogenous Ha-ras is not likely playing a role in Tg.AC tumorigenesis. In addition, no report of trisomy 11 is seen, the chromosome on which the Tg.AC transgene is located. The absence of trisomy 7 in Tg.AC mice may be due to transgene ras expression since the multicopy transgene may compensate for a mutated c-Ha-ras (French et al., 1994).

Trisomy 6 or 15 are frequently occurring alterations in Tg.AC and FVB/N malignancies, regardless of treatment or tumor type (spindle cell carcinomas or squamous cell carcinomas) (French et al., 1994). Perhaps genes or a gene on chromosome 6 or 15 contributes to the development of malignant tumors. The fact that two chromosomes can be involved may indicated that multiple pathways exist in malignant conversions (French et al., 1994).

Pathway interactions

In addition to the expression of the transgene, it is a certainty that the Tg.AC tumorigenic response is dependent on the expression and involvement of additional genes. To understand the cellular mechanisms regulating (and regulated by) Tg.AC transgene, the Tg.AC mouse has been crossed with additional transgenic mice in an effort to elucidate ras-dependent pathways and signaling (Table 1).

Table 1 - Transgenic crosses with Tg.AC mice.

Full table

Ornithine decarboxylase

Tumor promoters such as TPA induce the enzyme ornithine decarboxylase (ODC). ODC is the first and rate-limiting enzyme in the biosynthesis of polyamines. Overexpression of ODC in the epidermis of a transgenic mouse resulted in squamous papilloma formation following a single subthreshold dose of a carcinogen (O'Brien et al., 1997).

The Tg.AC was crossed with a transgenic mouse that overexpressed ODC in the outer root sheath of the hair follicle (Smith et al., 1998). The resulting K6-ODC/ras bitransgenics generated spontaneous papillomas and carcinomas in the skin, giving the appearance of being both genetically initiated (from the Tg.AC background) and genetically promoted (from the ODC background) (Smith et al., 1998; Gilmour et al., 1999). Tumor formation in the bitransgenics could be prevented by administration (via drinking water) of -difluoromethylornithine (DFMO), a specific and irreversible inhibitor of ODC (Smith et al., 1998). The mechanisms of ODC and v-Ha-ras, although independent, appear to act in concert driving progression through the cell cycle.

In addition to preventing tumors in ODC/ras bitransgenics, DFMO also was able to prevent tumor formation in wounded Tg.AC mice, indicating the potential involvement of polyamine synthesis and ODC in wound-induced promotion (Smith et al., 1998). Furthermore, it was found that administration of DFMO after the onset of tumorigenesis resulted in tumor regression, despite continued v-Ha-ras transgene expression (Smith et al., 1998). Together, these data support a hypothesis that transgene expression, although a necessary precursor to papilloma formation, is in and of itself not sufficient for tumor formation.

Cyclin D1

Mutated ras expression is known to affect the expression of a number of genes. Mutated ras increases the level of cyclin D and the associated kinase activity of cyclin D/Cdk4. In Tg.AC mice, the upregulation of D1 following TPA exposure has been shown in papillomas and follicular hyperplasias, but does not occur in uninvolved TPA-exposed skin (Robles et al., 1998).

Tg.AC/cyclin D1-null bitransgenic mice were created to further explore the relationship between cyclin D1 and tumorigenesis (Robles et al., 1998). Following TPA promotion, tumor multiplicity in cyclin D1-null bitransgenics was sixfold lower than in the matched cyclin D1-hemizygotes. In addition, the onset of tumors was delayed 1 week in D1 nulls. Similar results were obtained in the D1 nulls treated with DMBA followed by multiple applications with TPA, indicating the result was not simply a Tg.AC phenomenon.

Hyperphoshorylation of pRb (a blocker of G1 progression) by cyclin D1/Cdk4 allows cells to move through, but not out of G1. Elevated levels of ODC/polyamines indirectly activates or increases the kinase activity of cyclin E/Cdk2 in late G1, allowing the cells to move from late G1 to S phase DNA synthesis. Once a cell passes the G1–S checkpoint, progression through the cell cycle is independent of mitogenic stimulation. Elevated ODC/polyamines have the same effect at the S to G2 phase by increasing the kinase activity of cyclin A/Cdk2. p53 protein levels are increased in the skin of K6/ODC mice, indicating an increase in genomic instability or damage as the cells are pushed through the cell cycle (Gilmour et al., 1999).

Transforming growth factor

It has been shown that v-Ha-ras-mediated expression of cyclin D1 is also dependent on autocrine stimulation through the EGFR (Robles et al., 1998). A member of the epidermal growth factor (EGF) family, transforming growth factor (TGF) binds to and activates the EGFR. A potent mitogen in many mesenchymal and epithelial cells, including keratinocytes, a body of evidence supports a role for TGF in the promotional phase of multistage skin carcinogenesis (see references in Humble et al., 1998). To determine if promotional stimuli (TPA and wounding) could induce papillomagenesis in Tg.AC mice deficient (null) in TGF, Tg.AC mice were crossed with TGF-null mice. F1 mice, hemizygous for v-Ha-ras and heterozygous or null for TGF, were dosed with TPA or given full-thickness dorsal wounds. Papillomas were able to develop following both wounding and TPA treatment in TGF-null mice, with some papillomas converting to squamous or spindle cell carcinomas. Although no statistical difference was evident between dosing groups, trends were apparent, with TGF-null mice developing fewer average papillomas with an increased latency time over sex-matched heterozygotes. Although not ruling out the importance of signaling though the EGFR in the tumorigenic process, these results may demonstrate that redundant members of the EGF family may be able to take the place of TGF in its absence (Humble et al., 1998), and that these EGF family members can stimulate cyclin D upregulation, permitting cells to move through G1 into S phase synthesis.

Tyrosine kinase receptor signaling in Tg.AC

Kinase suppressor of ras (KSR) is a positive modulator of the map kinase pathway and appears to function by modulating the ras/raf activity of the mitogen-activated protein kinase signaling cascade (MAPK). To investigate its potential to modulate the ras-dependant tumorigenic response in Tg.AC mice, the v-Ha-ras transgene was crossed onto a KSR-/- background (Lozano et al., 2003). Mice that contained a transcriptionally active v-Ha-ras transgene but lacked KSR activity exhibited an abrogated TPA-induced tumorigenic response, indicating that KSR function appears to be required for TPA-induced papillomas in Tg.AC mice. Moreover, an interesting phenotypic observation was also noted; mice lacking KSR activity exhibited an unusual disorganized hair follicle phenotype similar to that of EGF knockout mice. Defects in EGFR signaling were also found in the mouse embryonic fibroblast (MEF). EGF-induced MAPK phosphorylation was blocked in KSR knockout mice demonstrating that KSR1 functions to integrate signaling through the Ras/MAPK complex via the EGFR receptor. Taken together, these data suggest that the v-Ha-ras transgene of Tg.AC mice signals through the MAPK pathway via the EGFR receptor.

It is also worth mentioning that another tyrosine kinase (TK) receptor, Ron, was found to modulate tumorigenesis in Tg.AC mice. Ron tyrosine kinase receptors are a family of multifunctional TK receptors that include the c-Met and c-Sea proto-oncogenes. These TK regulate very diverse biological responses, including proliferation, motility, invasiveness, and cellular dissociation. When a kinase domain-deficient Ron receptor was crossed onto Tg.AC, TPA-induced papillomas were found to be smaller in volume and less likely to convert to malignancy (Chan et al., 2005).

Bcl-2

The bcl-2 gene product is a negative regulator of apoptosis, or programmed cell death. It has been shown that bcl-2 is expressed in normal hair follicles in the mouse, the basal epidermis, and to a lesser extent in the postmitotic spinous epidemal layer (Stenn et al., 1994). As an inhibitor of cell death, it was of interest to examine the role bcl-2 plays in Tg.AC tumorigenesis. Tg.AC mice were crossed with mice deficient in bcl-2. The resulting F1 mice, hemizygous for the Tg.AC transgene and either homozygous or deficient in bcl-2, were exposed to TPA promotion. At a low TPA dose (1.25 g twice weekly for 10 weeks), a statistically significant decrease in papilloma formation was seen for mice deficient in bcl-2. At a higher TPA dose (2.5 g), no effect was seen. The decrease in papilloma development at the lower dose of TPA could be the result of an increase in apoptotic cell death due to the lowered bcl-2 expression, or a possible enhancement of apoptosis through the upregulation of genes known to promote apoptosis and that are negatively controlled by bcl-2, such as c-myc, c-fos, and/or TGF. This possibility is supported by results obtained in a apoptosis-sensitive keratinocyte cell line (Marthinuss et al., 1995).

Vascular endothelial growth factor

The role of angiogenesis in the production of Tg.AC tumors has been explored though several means (Larcher et al., 1996, 1998; Tober et al., 1998). Vascular endothelial growth factor (VEGF), also known as vascular permeability factor (VGP), stimulates the formation of new capillaries and increases vascular permeability. A variety of splice variants exist for VEGF, designated by their amino-acid content as murine VEGF 120, 144, 164, 188, and 205. The smaller variants (120 and 144) are secreted and are mitogenic for endothelial cells, while the larger variants (188 and 205) are secreted yet remain adherent to the cell surface and likely regulate vascular permeability. VEGF derived from keratinocytes is the major angiogenic factor in the skin. VEGF expression has been shown to be mediated by oncogenic ras (Larcher et al., 1996). In addition, TPA is able to upregulate VEGF/VGP expression in mouse skin (Larcher et al., 1996). The four smaller splice variants of VEGF have been identified in Tg.AC and FVB/N skin, papillomas, and carcinomas (Tober et al., 1998). Of interest is a fifth, larger splice variant present during the latter stages of tumor promotion and progression and perhaps related to the presence of v-Ha-ras (Tober et al., 1998). Its presence in the latter stages may ensure that growth factors, oxygen, and nutrients are able to nourish the tumors by ensuring vascular permeability.

The overexpression of VEGF and its effect on Tg.AC tumorigenesis was examined though the use of a K6-VEGF transgenic mouse (Larcher et al., 1998). The keratin-6 promoter utilized in this transgenic mouse confers a selective expression to the suprabasal layers of the epidermis in response to hyperproliferative stimuli. TPA-induced premalignant papillomas appear earlier in bitransgenic K6-VEGF/Tg.AC mice than in matched Tg.AC mice. Tumor size was greater in the majority of bitransgenic tumors, with total tumor mass also being significantly higher. Taken together, these studies demonstrate a role for VEGF in the tumorigenic process in Tg.AC mice.

Future directions

Following proliferative stimuli and the change in regulation of additional genes such as ODC or VEGF, specific transgene-expressing cells multiply to detectable levels, ultimately leading to clonal expansion and tumor formation. The fact remains, however, that only one in four founder lines carrying the -globin/v-Ha-ras transgene resulted in the unique phenotype now associated with the Tg.AC mouse. The influence of transgene position within the genome continues to be queried.

Transgene expression may be due to the fortuitous induction or expression of a neighboring gene on chromosome 11. This consequentially opens the chromatin surrounding or adjacent to the transgene integration site, resulting in -globin promoter regions becoming available to transcription factors. The multicopy transgene array has been located proximal to the centromere on chromosome 11 (Humble et al., 2000), and recently the specific site of integration is identified (Scott Barros, personal communication and Leder et al., 2002). Examination of adjacent features within the flanking chromosome or the function of adjacent genes may offer insights into transgene induction. A continued examination of the nonresponder phenotype and the palindromic promoters may elucidate the role of this feature in transgene expression.

The advent and use of gene array technology has revealed TPA-inducible genes involved in the Tg.AC tumorigenic response (Wei et al., 2003). Ongoing studies continue to explore the role Dss-1, NDPK-B, and other genes involved in the early stage of neoplastic development and skin tumorigenesis.

The correlation of Tg.AC tumorigenesis to ras-dependent pathways in the tumorigenic process strengthens the utility of the Tg.AC mouse as a model for experimental skin carcinogenesis. The utility of the Tg.AC as an adjunct to the 2-year bioassay will continue to be explored and debated. As a prototype alternative model, it has generated a healthy discussion on the future of transgenic bioassays, and opened the doors for subsequent models for toxicity testing. The continued exploration and elucidation of the cellular and molecular controls of transgene expression will enhance the usefulness of this mouse and enable a better understanding of the discriminate response to chemical carcinogens exhibited by Tg.AC mice.

References

1. Argyris TS. (1985) Am. J. Ind. Med. 8: 329–337. | PubMed | ISI | ChemPort |
2. Asano S, Trempus CS, Spalding JW, Tennant RW & Battalora MS. (1998) Toxicol. Pathol. 26: 512–519. | PubMed | ISI | ChemPort |
3. Balmain A & Pragnell IB. (1983) Nature 303: 72–74. | Article | PubMed | ISI | ChemPort |
4. Barry G, Cook JW, Haslewood GAD, Hewett CL, Hieger I & Kennaway EL. (1935) Proc. Roy. Soc. Biol. 117: 318–351.
5. Battalora MS, Spalding JW, Szczesniak CJ, Cape JE, Morris RJ, Trempus CS, Bortner CD, Lee BM & Tennant RW. (2001) Carcinogenesis 22: 651–659. | Article | PubMed | ISI | ChemPort |
6. Berenblum I. (1941) Cancer Res. 1: 807–814. | ChemPort |
7. Blanchard KT, Ball DJ, Holden HE, Furst SM, Stoltz JH & Stoll RE. (1998) Toxicol. Pathol. 26: 541–547. | PubMed | ISI | ChemPort |
8. Bos JL. (1987) Cancer Res. 49: 4682–4689.
9. Boutwell RK. (1964) Progress in Experimental Tumor Research Vol. 4. Karger: Basel/New York pp. 207–250. | ChemPort |
10. Cannon RE, Graves S, Spalding JW, Trempus CS & Tennant RW. (2000) Mol. Carcinog. 29: 229–235. | Article | PubMed | ISI | ChemPort |
11. Cannon RE, Spalding JW, Trempus CS, Szczesniak CJ, Virgil KM, Humble MC & Tennant RW. (1997) Mol. Carcinog. 20: 108–114. | Article | PubMed | ISI | ChemPort |
12. Cannon RE, Spalding JW, Virgil KM, Faircloth RS, Humble MC, Lacks GD & Tennant RW. (1998) Mol. Carcinog. 21: 244–250. | Article | PubMed | ISI | ChemPort |
13. Cardiff RD, Leder A, Kuo A, Pattengale PK & Leder P. (1993) Am. J. Pathol. 142: 1199–1207. | PubMed | ISI | ChemPort |
14. Chan EL, Peace BE, Collins MH, Toney-Earley K & Waltz SE. (2005) Oncogene 24: 479–488. | Article | PubMed | ISI | ChemPort |
15. Chignell CF, Haseman JK, Sik RH, Tennant RW & Trempus CS. (2003a) Photochem. Photobiol. 77: 77–80. | Article | ISI | ChemPort |
16. Chignell CF, Haseman JK, Sik RH, Tennant RW & Trempus CS. (2003b) Photochem. Photobiol. 77: 77–80. | Article | ISI | ChemPort |
17. Cohen SM, Robinson D & MacDonald J. (2001) Toxicol. Sci. 64: 14–19. | PubMed | ISI | ChemPort |
18. Collick A, Drew J, Penberth J, Bois P, Luckett J, Scaerou F, Jeffreys A & Reik W. (1996) EMBO J. 15: 1163–1171. | PubMed | ISI | ChemPort |
19. Cotsarelis G, Sun TT & Lavker RM. (1990) Cell 61: 1329–1337. | Article | PubMed | ISI | ChemPort |
20. Deelman HT. (1927) Br. Med. J. 1: 872.
21. Delker DA, Yano BL & Gollapudi BB. (1999) Toxicol. Sci. 50: 90–97. | Article | PubMed | ISI | ChemPort |
22. Dodds AP, Cannon RE, Suggs CA & Wright JT. (2003) Arch. Oral Biol. 48: 843–850. | Article | PubMed | ISI | ChemPort |
23. El-Abersari TB, Furhmann J, Trempus C, Shendrik I, Tennant RW & Hansen LA. (2005) Cancer Res. 65: 3958–3965. | Article | PubMed |
24. Forrester WC, Takegawa S, Papayannopoulou T, Stamatoyannopoulos G & Groudine M. (1987) Nucleic Acids Res. 15: 10159–10177. | PubMed | ISI | ChemPort |
25. Fraser P & Grosveld F. (1998) Curr. Opin. Cell Biol. 10: 361–365. | Article | PubMed | ISI | ChemPort |
26. French JE, Libbus BL, Hansen L, Spalding J, Tice RR, Mahler J & Tennant RW. (1994) Mol. Carcinog. 11: 215–226. | PubMed | ISI | ChemPort |
27. Germolec DR, Spalding J, Boorman GA, Wilmer JL, Yoshida T, Simeonova PP, Bruccoleri A, Kayama F, Gaido K, Tennant R, Burleson F, Dong W, Lang RW & Luster MI. (1997) Mutat. Res. 386: 209–218. | Article | PubMed | ISI | ChemPort |
28. Gilmour SK, Birchler M, Smith MK, Rayca K & Mostochuk J. (1999) Cell Growth Differ. 10: 739–748. | PubMed | ISI | ChemPort |
29. Grosveld F. (1999) Curr. Opin. Genet Dev. 9: 152–157. | Article | PubMed | ISI | ChemPort |
30. Hansen LA, Malarkey DE, Wilkinson JE, Rosenberg M, Woychik RE & Tennant RW. (1998) Carcinogenesis 19: 1837–1845. | Article | PubMed | ISI | ChemPort |
31. Hansen LA, Spalding JW, French JE & Tennant RW. (1995) Prog. Clin. Biol. Res. 391: 223–235. | PubMed | ChemPort |
32. Hansen LA & Tennant R. (1994a) Mol. Carcinog. 9: 143–154. | PubMed | ISI | ChemPort |
33. Hansen LA & Tennant RW. (1994b) Proc. Natl. Acad. Sci. USA 91: 7822–7826. | PubMed | ChemPort |
34. Hansen LA, Trempus CS, Mahler JF & Tennant RW. (1996) Carcinogenesis 17: 1825–1833. | PubMed | ISI | ChemPort |
35. Hennings H & Boutwell RK. (1970) Cancer Res. 30: 312–320. | PubMed | ISI | ChemPort |
36. Higgs DR, Wood WG, Jarman AP, Sharpe J, Lida J, Pretorius IM & Ayyub H. (1990) Genes Dev. 4: 1588–1601. | PubMed | ISI | ChemPort |
37. Humble MC, Faircloth RS, Tennant RW & Cannon RE. (Submitted).
38. Humble MC, Kouprina N, Noskov VN, Graves J, Garner E, Tennant RW, Resnick MA, Larionov V & Cannon RE. (2000) Genomics 70: 292–299. | Article | PubMed | ISI | ChemPort |
39. Humble MC, Szczesniak CJ, Luetteke NC, Spalding JW, Cannon RE, Hansen LA, Lee DC & Tennant RW. (1998) Toxicol. Pathol. 26: 562–569. | PubMed | ISI | ChemPort |
40. Jarman AP, Wood WG, Sharpe JA, Gourdon G, Ayyub H & Higgs DR. (1991) Mol. Cell. Biol. 11: 4679–4689. | PubMed | ISI | ChemPort |
41. Jones PH & Watt FM. (1993) Cell 73: 713–724. | Article | PubMed | ISI | ChemPort |
42. Kantz DC, Lacks GD & Cannon RE. (1999) Biotechniques 27: 278–280. | PubMed | ISI | ChemPort |
43. Karlsson S & Nienhuis AW. (1985) Annu. Rev. Biochem. 54: 1071–1108. | Article | PubMed | ISI | ChemPort |
44. Kaufman CK, Zhou P, Pasolli HA, Rendl M, Bolotin D, Lim K-C, Dai X, Alegre M-L & Fuchs E. (2003) Genes Dev. 17: 2108–2122. | Article | PubMed | ISI | ChemPort |
45. Kouprina N, Annab L, Graves J, Afshari C, Barrett JC, Resnick MA & Larionov V. (1998) Proc. Natl. Acad. Sci. USA 95: 4469–4474. | Article | PubMed | ChemPort |
46. Larcher F, Murillas R, Bolontrade M, Conti CJ & Jorcano JL. (1998) Oncogene 17: 303–311. | Article | PubMed | ISI | ChemPort |
47. Larcher F, Robles AI, Duran H, Murillas R, Quintanilla M, Cano A, Conti CJ & Jorcano JL. (1996) Cancer Res. 56: 5391–5396. | PubMed | ISI | ChemPort |
48. Lavker RM & Sun TT. (2000) Proc. Natl. Acad. Sci. USA 97: 13473–13475. | Article | PubMed | ChemPort |
49. Leach DR. (1994) BioEssays 16: 893–900. | Article | PubMed | ISI | ChemPort |
50. Leder A, Kuo A, Cardiff RD, Sinn E & Leder P. (1990) Proc. Natl. Acad. Sci. USA 87: 9178–9182. | PubMed | ChemPort |
51. Leder A, Lebel M, Zhou F, Fontaine K, Bishop A & Leder P. (2002) Oncogene 21: 6657–6668. | Article | PubMed | ISI | ChemPort |
52. Lewis S, Akgun E & Jasin M. (1999) Ann. NY Acad. Sci. 870: 45–57. | PubMed | ChemPort |
53. Lozano J, Xing R, Cai Z, Jensen HL, Trempus C, Mark W, Cannon R & Kolesnick R. (2003) Cancer Res. 63: 4232–4238. | PubMed | ISI | ChemPort |
54. Lyle S, Christofidou-Solomidou M, Liu Y, Elder DE, Albelda S & Cotsarelis G. (1998) J. Cell Sci. 111: 3179–3188. | PubMed | ISI | ChemPort |
55. Mahler JF, Flagler ND, Malarkey DE, Mann PC, Haseman JK & Eastin WC. (1998) Toxicol. Pathol. 26: 501–511. | PubMed | ISI | ChemPort |
56. Mahler JF, Stokes W, Mann PC, Takaoka M & Maronpot RR. (1996) Toxicol. Pathol. 24: 710–716. | PubMed | ISI | ChemPort |
57. Marthinuss J, Lawrence L & Seiberg M. (1995) Cell Growth Differ. 6: 239–250. | PubMed | ISI | ChemPort |
58. Mills KJ, Reynolds SH & Smart RC. (1993) Carcinogenesis 14: 2645–2648. | PubMed | ISI | ChemPort |
59. Moon A & Cannon RE. (2001) Biochemistry North Carolina State University: Raleigh p. 46.
60. Morris RJ. (2000) J. Clin. Invest. 106: 3–8. | PubMed | ISI | ChemPort |
61. Murphy J, Morales RE, Scott J & Kupper TS. (2003) J. Immunol. 170: 5697–5703. | PubMed | ISI | ChemPort |
62. O'Brien TG, Megosh LC, Gilliard G & Soler AP. (1997) Cancer Res. 57: 2630–2637. | PubMed |
63. Ohara M, Kawashima Y, Kitajima S, Mitsuoka C & Watanabe H. (2003) Cancer Sci. 94: 205–209. | Article | PubMed | ISI | ChemPort |
64. Orkin SH. (1992) Blood 80: 575–581. | PubMed | ISI | ChemPort |
65. Owens DM, Spalding JW, Tennant RW & Smart RC. (1995) Cancer Res. 55: 3171–3178. | PubMed | ISI | ChemPort |
66. Potten CS. (1983) Stem Cells: Their Identification and Characterization Potten CS (ed). Churchill Livingstone: Edinburgh, London, Melbourne, and New York pp. 200–232.
67. Pritchard JB, French JE, Davis BJ & Haseman JK. (2003) Environ. Health Perspect 111: 444–454. | PubMed | ISI | ChemPort |
68. Pullinger BD. (1945) J. Pathol. Bacteriol. 57: 447–481.
69. Robinson DE & MacDonald JS. (2001) Toxicol. Pathol. 29: 13–19. | Article | PubMed | ISI | ChemPort |
70. Robles AI, Rodriguez-Puebla ML, Glick AB, Trempus C, Hansen L, Sicinski P, Tennant RW, Weinberg RA, Yuspa SH & Conti CJ. (1998) Genes Dev. 12: 2469–2474. | PubMed | ISI | ChemPort |
71. Sistare FD, Thompson KL, Honchel R & DeGeorge J. (2002) Int. J. Toxicol. 21: 65–79. | Article | PubMed | ISI | ChemPort |
72. Smith MK, Trempus CS & Gilmour SK. (1998) Carcinogenesis 19: 1409–1415. | Article | PubMed | ISI | ChemPort |
73. Spalding JW, French JE, Stasiewicz S, Furedi-Machacek M, Conner F, Tice RR & Tennant RW. (2000) Toxicol. Sci. 53: 213–223. | Article | PubMed | ISI | ChemPort |
74. Spalding JW, French JE, Tice RR, Furedi-Machacek M, Haseman JK & Tennant RW. (1999) Toxicol. Sci. 49: 241–254. | Article | PubMed | ISI | ChemPort |
75. Spalding JW, Momma J, Elwell MR & Tennant RW. (1993) Carcinogenesis 14: 1335–1341. | PubMed | ISI | ChemPort |
76. Stenn KS, Lawrence L, Veis D, Korsmeyer S & Seiberg M. (1994) J. Invest. Dermatol. 103: 107–111. | Article | PubMed | ISI | ChemPort |
77. Tennant R. (1997) Toxicol. Pathol. 25: 240–241. | PubMed | ISI | ChemPort |
78. Tennant RW, French JE & Spalding JW. (1995) Environ. Health Perspect. 103: 942–950. | PubMed | ISI | ChemPort |
79. Tennant RW & Spalding J. (1996) Environ. Health Perspect. 104 (Suppl. 5): 1095–1100. | PubMed | ISI | ChemPort |
80. Tennant RW, Spalding J & French JE. (1996) Mutat. Res. 365: 119–127. | PubMed | ISI | ChemPort |
81. Tennant RW, Stasiewicz S, Eastin WC, Mennear JH & Spalding JW. (2001) Toxicol. Pathol. 29: 51–59. | Article | PubMed | ISI | ChemPort |
82. Thompson KL, Rosenzweig BA, Honchel R, Cannon RE, Blanchard KT, Stoll RE & Sistare FD. (2001) Mol. Carcinog. 32: 176–186. | Article | PubMed | ISI | ChemPort |
83. Thompson KL, Rosenzweig BA & Sistare FD. (1998) Toxicol. Pathol. 26: 548–555. | PubMed | ISI | ChemPort |
84. Tober KL, Cannon RE, Spalding JW, Oberyszyn TM, Parrett ML, Rackoff AI, Oberyszyn AS, Tennant RW & Robertson FM. (1998) Biochem. Biophys. Res. Commun. 247: 644–653. | Article | PubMed | ISI | ChemPort |
85. Trempus CS, Bishop WR, Njoroge FG, Doll RJ, Battalora MS, Mahler JF, Haseman JK & Tennant RW. (2000) Mol. Carcinog. 27: 24–33. | Article | PubMed | ISI | ChemPort |
86. Trempus CS, Haseman JK & Tennant RW. (1997) Mol. Carcinog. 20: 68–77. | Article | PubMed | ISI | ChemPort |
87. Trempus CS, Mahler JF, Ananthaswamy HN, Loughlin SM, French JE & Tennant RW. (1998a) J. Invest. Dermatol. 111: 445–451. | Article | ISI | ChemPort |
88. Trempus CS, Morris RJ, Bortner CD, Cotsarelis G, Faircloth RS, Reece JM & Tennant RW. (2003) J. Invest. Dermatol. 120: 501–511. | Article | PubMed | ISI | ChemPort |
89. Trempus CS, Ward S, Farris G, Malarkey D, Faircloth RS, Cannon RE & Mahler JF. (1998b) Am. J. Pathol. 153: 247–254. | ISI | ChemPort |
90. Tuan D & London IM. (1984) Proc. Natl. Acad. Sci. USA 81: 2718–2722. | PubMed | ChemPort |
91. Tuan DY, Solomon WB, London IM & Lee DP. (1989) Proc. Natl. Acad. Sci. USA 86: 2554–2558. | PubMed | ChemPort |
92. Weaver JL, Contrera JF, Rosenzweig BA, Thompson KL, Faustino PJ, Strong JM, Ellison CD, Anderson LW, Prasanna HR, Long-Bradley PE, Lin KK, Zhang J & Sistare FD. (1998) Toxicol. Pathol. 26: 532–540. | PubMed | ISI | ChemPort |
93. Wei S-J, Trempus CS, Cannon RE, Bortner CD & Tennant RW. (2003) J. Biol. Chem. 278: 1758–1768. | Article | PubMed | ISI | ChemPort |
94. Wright JT, Hansen L, Mahler J, Szczesniak C & Spalding JW. (1995) Arch. Oral. Biol. 40: 631–638. | Article | PubMed | ISI | ChemPort |
95. Yisraeli J, Frank D, Razin A & Cedar H. (1988) Proc. Natl. Acad. Sci. USA 85: 4638–4642. | PubMed | ChemPort |

Acknowledgements

We dedicate this review to the memory of two valued friends and collaborators, Dr Ghanta N Rao and Dr Joel F Mahler, who have strongly contributed to the development and use of the Tg.AC mouse model. Dr Rao's dedication to the care and breeding of research animals, as well as his independent research and consultation, contributed to the high standard of quality that the Tg.AC mouse currently enjoys. Dr Mahler was instrumental in first describing, then exploring, the unique characteristics displayed not only by the Tg.AC mouse, but many other transgenic strains as well. His participation as pathologist and scientist considerably advanced our understanding of the biology and pathology of the Tg.AC mouse model. This work was partly funded by Grant NIEHS T32 ES07126 (MCH).