Mutational activation of the ras proto-oncogenes is frequently found in skin cancers. However, the nature of downstream signaling pathways from Ras involved in skin carcinogenesis remains poorly understood. Recently, we and others identified phospholipase C (PLC) ε as an effector of Ras. Here we have examined the role of PLCε in de novo skin chemical carcinogenesis by using mice whose PLCε is genetically inactivated. PLCε−/− mice exhibit delayed onset and markedly reduced incidence of skin squamous tumors induced by initiation with 7,12-dimethylbenz(a)anthracene followed by promotion with 12-O-tetradecanoylphorbol-13-acetate (TPA). Furthermore, the papillomas formed in PLCε−/− mice fail to undergo malignant progression into carcinomas, in contrast to a malignant conversion rate of approximately 20% observed with papillomas in PLCε+/+ mice. In all of the tumors analyzed, the Ha-ras gene is mutationally activated irrespective of the PLCε background. The skin of PLCε−/− mice fails to exhibit basal layer cell proliferation and epidermal hyperplasia in response to TPA treatment. These results indicate a crucial role of PLCε in ras oncogene-induced de novo carcinogenesis and downstream signaling from TPA, introducing PLCε as a candidate molecular target for the development of anticancer drugs.
The ras proto-oncogenes are mutationally activated in about 15% of human neoplasms (1). Their products, Ras small GTPases, control cell proliferation and differentiation through interaction with multiple effector proteins, among which Raf kinases have been implicated in carcinogenesis from studies on in vitro transformation of fibroblast cell lines (2) and on genomic mutations in malignant melanoma (3). However, downstream signaling pathways from Ras involved in epithelial cell carcinogenesis remain poorly understood, despite the fact that ras mutations are more frequently found in epithelial cell-derived neoplasms (1). Likewise, the role of phosphoinositide-specific phospholipase C (PLC) in carcinogenesis remains obscure (4). PLC produces two vital intracellular second messengers, diacylglycerol and inositol 1,4,5-trisphosphate, which induce activation of protein kinase C and mobilization of Ca2+ from intracellular stores, respectively. Among 12 mammalian PLC isoforms classified into 5 classes (β, γ, δ, ε, and ζ), PLCε is characterized by possession of the Ras-associating domains, which are responsible for PLCε activation through direct association with the GTP-bound active forms of the small GTPases Ras (5, 6), Rap1 (7), and Rap2 (8). PLCε was also reported to be regulated by α12, α13, and β1γ2 subunits of heterotrimeric G proteins and Rho small GTPase (9). Identification of PLCε as a Ras effector has prompted us to examine the role of PLCε in carcinogenesis. Here we show that PLCε-deficient mice are resistant to chemical carcinogen-induced skin tumor formation, suggesting a crucial role of PLCε in tumor development downstream of Ras signaling.
Materials and Methods
Targeted inactivation of the PLCε gene was performed by a standard embryonic stem cell-based method.4 The targeted allele (PLCε−) expresses a mutant PLCε with an in-frame deletion of amino acids 1333 to 1408 corresponding to the NH2-terminal part of the catalytic X domain. This mutant completely lost its PLC catalytic activity. PLCε−/− mice were maintained on a mixed 129/Sv × C57BL/6 background.
Reverse Transcription-Polymerase Chain Reaction Analysis.
Reverse transcription-polymerase chain reaction (RT-PCR) was performed as described previously (10). Primers used for amplification of PLCε were 5′-TCAGTGCCTGGAGCAGCAG-3′ and 5′-CTTGAAGGGGATCTTGGTTG-3′.
Skin Tumor Formation.
A dorsal area of skin of 8-week–old mice was shaved and treated with a single application of 7,12-dimethylbenz(a)anthracene [DMBA (25 μg in 100 μL of acetone; Sigma, St. Louis, MO] and subsequently treated with 12-O-tetradecanoyl-phorbor-13-acetate [TPA (0.2 mmol/L in 100 μL of acetone; Sigma] twice a week for 20 weeks (11). Tumors were assessed weekly for up to 30 weeks and defined as raised lesions with a minimum diameter of 1 mm. P values were determined by unpaired Student’s t test using GraphPad InStat software (GraphPad Software, Inc., San Diego, CA).
Paraffin-embedded sections were prepared and stained with hematoxylin and eosin or with a specific antibody against mouse PLCε (10), keratin 14 (PRB-155P; BAbCO, Berkeley, CA), or keratin 1 (PRB-165P; BAbCO). Detection of immunoreactive signals was performed with HistoMouse Plus kit (Zymed Laboratories, South San Francisco, CA) or with a fluorescein isothiocyanate-conjugated secondary antibody (AP182F; Chemicon, Temecula, CA).
12-O-Tetradecanoylphorbol-13-acetate–Induced Skin Hyperplasia.
A dorsal area of skin of 10-week–old mice was treated with TPA (0.2 mmol/L in 100 μL of acetone). The mouse skin was analyzed by staining with an anti-proliferating cell nuclear antigen (PCNA) antibody (M0879; Dako Cytomation, Copenhagen, Denmark) or hematoxylin and eosin. The thickness of the epidermis was measured at a minimum of five different points on the specimens and averaged.
Analysis of Ha-ras Gene Mutations.
Ha-ras gene mutations at the 61st codon of the tumors were analyzed as described previously (12).
Results and Discussion
RT-PCR analysis of skin RNA detected two amplified products whose sizes were identical to those predicted from the wild-type and mutant PLCε mRNAs (Fig. 1,A). Immunohistochemical analysis showed that PLCε is expressed in the epidermis (Fig. 1,B), including keratin 14-positive proliferative keratinocytes and keratin 1-positive differentiating keratinocytes, but not in the dermis, except for hair follicles (Fig. 1,C). To address the role of PLCε in de novo skin carcinogenesis, we applied the skin two-stage chemical carcinogenesis protocol (11) on PLCε−/− mice. Initiation was carried out with a single application of DMBA, which almost invariably introduced oncogenic mutations on the Ha-ras gene (11, 12). Subsequent promotion by repeated treatment with TPA for 20 weeks caused the selective clonal outgrowth of the initiated cells to produce benign squamous tumors (Fig. 2,A). PLCε−/− mice showed significant delay in the average time of tumor onset compared with PLCε+/+ mice [average ± SE: 12.63 ± 0.42 weeks (PLCε−/−; 21 mice analyzed) versus 10.14 ± 0.47 weeks (PLCε+/+; 14 mice); P < 0.001; Fig. 2,B]. PLCε+/− mice showed an intermediate phenotype (11.79 ± 0.31 weeks; 23 mice; P < 0.01), indicating the existence of an apparent gene-dosage effect. The time to develop the first tumor also showed a significant difference [PLCε+/+, 6.06 ± 0.36 weeks; PLCε+/−, 7.87 ± 0.30 weeks (P < 0.001); PLCε−/−, 9.86 ± 0.43 weeks (P < 0.0001)]. The number of tumors reached a maximum at 15 weeks. At this point, the average number of tumors per mouse was reduced by approximately 70% in PLCε−/− mice (4.14 ± 0.40; P < 0.0001) compared with PLCε+/+ mice (14.36 ± 1.25). Again, PLCε+/− mice showed an intermediate phenotype (10.22 ± 0.65; P < 0.0001; Fig. 2,B). In PLCε−/− mice, no tumor greater than 6 mm in diameter was observed at 20 weeks (Fig. 2,C). In the two-stage protocol, a population of papillomas undergo progression into squamous cell carcinoma (SCC) (11). At 30 weeks after initiation, tumors of at least 2 mm in diameter were isolated and subjected to histologic analysis (Table 1; Fig. 2 D). In PLCε+/+ mice, approximately 20% of the tumors were carcinomas. In contrast, essentially no carcinoma was found in PLCε−/− mice. PLCε+/− mice showed a partial resistance to malignant progression. Thus, PLCε deficiency strongly suppressed malignant progression. All of the tumors tested carried the activating mutations at the 61st codon of the Ha-ras gene, irrespective of the PLCε genetic background (data not shown).
We next investigated the effect of PLCε deficiency on TPA-induced proliferation of the skin epidermis. Before or after treatment with acetone, there was no apparent difference between PLCε+/+ and PLCε−/− mice in the skin architecture and the number of proliferating cells positive for PCNA (Fig. 3). On TPA treatment, PLCε+/+ mouse skin showed a marked increase in the number of PCNA-positive cells in the basal layer cells (Fig. 3,A). In striking contrast, PLCε−/− mouse skin showed only a moderate increase (Fig. 3,A). TPA-induced epidermal hyperplasia was also suppressed in PLCε−/− mice (Fig. 3 B). The average thickness of the epidermis after 48 hours of TPA treatment was 98.4, 66.3, and 31.3 μm in PLCε+/+, PLCε+/−, and PLCε−/− mice, respectively, whereas that after acetone treatment was 27.7, 25.4, and 24.6 μm, respectively.
We have shown here that PLCε plays a crucial role in skin papilloma formation and malignant progression, which are induced by ras activation followed by TPA treatment. Furthermore, PLCε is shown to function downstream of TPA to induce hyperproliferation of the basal layer cells and skin hyperplasia. Thus, it is likely that PLCε functions in TPA-induced tumor promotion of the initiated cells carrying the activated ras genes. There are two possible mechanisms linking TPA to PLCε activation. TPA may activate PLCε through Ras activation, which is mediated by RasGRP1, a TPA-regulated Ras-specific guanine nucleotide exchange factor (GEF) expressed in keratinocytes (13). Rap1, whose activation is mediated by TPA-responsive Rap GEFs including CalDAG-GEFI (14) and RasGRP2 (15), may also be responsible for PLCε activation. Alternatively, TPA may activate PLCε through secretion of tumor necrosis factor (TNF)-α from keratinocytes (16) and subsequent TNF-α–induced Ras activation (17). TNF-α has been implicated in both two-stage skin carcinogenesis and TPA-induced skin hyperplasia (16).
Because targeted inactivation of protein kinase C (PKC) η resulted in enhancement of both papilloma formation and TPA-induced skin hyperplasia, TPA-induced down-regulation of PKCη is thought to play a crucial role in induction of these phenomena (18). In the present study, TPA treatment failed to compensate for the deficiency in papilloma formation and skin hyperplasia of PLCε−/− mice, although TPA is known to mimic diacylglycerol, a product of PLCε, in regulating PKCη. The result indicates that the PLCε pathway has an intrinsic role in skin hyperplasia and carcinogenesis, which is independent of the PKCη pathway. This intrinsic function may be mediated by another of its products, inositol 1,4,5-trisphosphate. On the other hand, activation of PLCε in DMBA-initiated cells, which must be induced by constitutively active Ras and produce diacylglycerol, could not substitute for TPA treatment in promoting papilloma formation. This suggests that TPA possesses another target that is also required for tumor promotion. In addition, papillomas developed in PLCε−/− mice failed to undergo malignant conversion. It was reported that prostaglandins are involved in skin tumor progression in addition to promotion (19) and play a key role in intestinal polyposis (20). Considering that arachidonic acid, a precursor of prostaglandins, can be produced from diacylglycerol, it is possible that the role of PLCε may be mediated through prostaglandin signaling.
Our present results have shown that PLCε plays a crucial role in ras oncogene-induced de novo carcinogenesis of skin epithelial cells. They also provide the first concrete evidence for the importance of the PLC signaling in carcinogenesis. This leads to the idea that specific inhibitors of PLCε may be useful for treatment and prevention of certain types of cancer.
Grant support: Grant-in-Aid for Priority Areas 12215098 (T. Kataoka); Grants-in-Aid for Scientific Research 15390093 (T. Kataoka), 16790187 (H. Edamatsu), and 15570117 (T. Satoh); and 21st Century COE Programs (T. Kataoka and T. Satoh) from the Ministry of Education, Science, Sports and Culture of Japan.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Note: Y. Bai and H. Edamatsu contributed equally to this work.
Requests for reprints: Tohru Kataoka, Division of Molecular Biology, Department of Molecular and Cellular Biology, Kobe University Graduate School of Medicine, 7-5-1 Kusunoki-cho, Chuo-ku, Kobe 650-0017, Japan. E-mail: firstname.lastname@example.org
M. Tadano, H. Edamatsu, S. Minamisawa, U. Yokoyama, Y. Ishikawa, N. Suzuki, H. Saito, D. Wu, M. Masago-Toda, Y. Yamawaki-Kataoka, T. Setsu, T. Terashima, S. Maeda, T. Satoh, and T. Kataoka. Congenital semilunar valvulogenesis defect in mice deficient in phospholipase Cε, submitted for publication.
|PLCε genotypes .||+/+ (n = 6) .||+/− (n = 14) .||−/− (n = 9) .|
|Total no. of tumors analyzed||48||70||26|
|PLCε genotypes .||+/+ (n = 6) .||+/− (n = 14) .||−/− (n = 9) .|
|Total no. of tumors analyzed||48||70||26|
NOTE. n represents the number of mice analyzed.
We thank Dr. Atsu Aiba, Dr. Ushio Kikkawa, Dr. Makoto Tadano, Shuzo Ikuta, and Tadashi Murase for helpful discussion and Shuichi Matsuda for excellent technical assistance.