Experimental Basosquamous Carcinoma Model in Rats

Previous studies have established that 7,12-dimethylbenz(a)anthracene (DMBA) can initiate skin tumorigenesis in conventional furred mouse models by acting on hair follicle stem cells. In this work, we have developed a simple and convenient rat model of basosquamous carcinoma (BSC) using DMBA-induced carcinogenesis in male specific pathogen free (SPF) Wistar rats with no additional tumor promoter agents. The results showed that topical application of 0.1% solution of DMBA in benzene in a volume of 40 μL twice a week produced skin tumors after 8–9 months. As a result, during the 11–12th months, we obtained 15 animals with BSCs, 22 with basal cell carcinomas (BCCs), and 3 with squamous cell carcinomas (SCCs). This chemically driven skin cancer model in Wistar rats can serve as a suitable alternative to the mouse skin cancer model and can be reliably replicated as demonstrated by the experiment.


Introduction
Topical administration of chemical carcinogens can induce skin carcinogenesis in laboratory animals and thus provides an easy and convenient model for studying factors influencing tumor susceptibility, growth, and progression [1]. Despite a long history of use in dermato-oncology and skin toxicology, the chemically induced carcinogenesis remains a costeffective model for the identification of molecular mechanisms implicated in skin tumorigenesis as well as for topical antitumor drug testing.
In particular, previous studies have established that the topical administration of 7,12-dimethylbenz[a]anthracene (DMBA) can initiate skin tumorigenesis in conventional rodent models. Typically, experimental protocol involves a twostage application of DMBA to the skin for the initiation and promotion of tumors. For instance, DMBA as the initiator and 12-O-tetradecanoylphorbol-13-acetate (TPA) as the promoter can produce skin tumors as early as 6 weeks with malignant transformation occurring at 18 weeks in furred SENCAR mouse strains [2]. However, there are some limitations of this approach. Thus, the efficiency of the DMBA-TPA carcinogenesis varies highly between the animal strains [3,4]. Another limitation is that this protocol tends to produce predominantly benign papillomas and hyperplastic lesions that regress upon TPA discontinuation [5]. Several studies have shown that DMBA can serve as both the tumor initiator and promoter, i.e., a complete carcinogen [6][7][8]. However, the tumorigenic response to DMBA may differ, and there is a need in further investigation of factors, such as timeline and the dosage regime, influencing skin tumorigenesis and progression.
In this work, we have developed a simple and convenient rat model of basosquamous carcinoma (BSC) using DMBAinduced carcinogenesis. Basosquamous carcinoma is a rare but an aggressive type of basal cell carcinoma (BCC) with an increased risk of recurrence and metastases [9]. The reported incidence of basosquamous carcinoma ranges from 1.2 to 2.7% of all cases of BCC, with the incidence of metastasis at least 5% and the published recurrence rates between 12 and 51% for surgical excision and 4% for Mohs micrographic surgery. The aggressive biological behavior and clinical course distinguish BSC from other forms of BCC.
O. Braun-Falco et al. characterize BSC as a unique rare tumor that is histologically difficult to differentiate from low-grade basal cell carcinoma and squamous cell skin cancer with high invasive potential and ability to metastasize [10].
J. Darier and Ferrand M. drew attention to the existence of two types of basosquamous epithelium [11]. The first type they called mixed was characterized by a local cell keratinization with the formation of pearls with colloidal or parakeratotic center. The second type is intermediate, characterized by an outer number of small, dark-colored basaloid cells and an inner zone consisting of larger cells with light eosinophilic cytoplasm. The mixed type is supposed to be keratotic basal cell carcinoma and intermediate-basal cell carcinoma with differentiation into two cell types. It is also possible that the mixed type can be a tumor collision when squamous cell carcinoma is in contact with the basal cell carcinoma. Most likely, in these cases, squamous cell carcinoma develops secondary to the basal cell carcinoma, which, as burns and ulcers, can stimulate its development; in this case, it is necessary to exclude the possibility pseudocarcinomatous hyperplasia in the background of basal cell carcinoma [10][11][12].
Among the histological variants of BSC, a solid, adenoid, morphea-like, and mixed histological types of BSC are allocated. In the background of a solid, adenoid, or morphea-like variant of the tumor, as a rule, there are keratotic areas resembling Bhorn pearls^in case of squamous skin cancer (SCC). BSC is also characterized by cellular polymorphism of tumor proliferate, infiltrative nature of tumor growth, and the presence of surrounding tissue manifestations a local immune response in the form of lymphoid-plasmacytic infiltration of the dermis and subcutaneous tissue. All researchers emphasize the difficulties of morphological diagnosis of BSC and its distinction from ulcerative basal cell carcinoma and SCC [13,14].
There is a need in reliable, convenient, and costeffective animal model of BSC for the identification of biological, immunological, and molecular pathways implicated in this form of skin tumor and for testing of antitumor agents. This is the first study, to our knowledge, to demonstrate that DMBA with no additional promoter agent can induce BSC. Our aim was to study the timeline of skin tumorigenesis and malignant progression in Wistar rats in response to treatment with DMBA without additional promoter agents in order to establish a useful experimental BSC model.

Materials and Methods
This experimental animal protocol was approved by the Animal Research Committee of the Sechenov First Moscow State Medicinal University (FMSMU). A total of 50 male Wistar rats (weight 160-180 g) aged 2 months were purchased from a colony that was certified as specific pathogen free (SPF). These rats were housed under SPF conditions at the animal house of FMSMU in metal cages (five animals per cage). The rats were given an acclimatization period of 2 weeks before the experiment.
The cages were lined with Lignocel bedding (Tecnilab-BMI BV, The Netherlands) with a thickness layer of 1.5 cm. Animal cages and beddings were changed every week according to the standard protocols. The rats were provided with a 12-h light/dark cycle (08:00/20:00 h in the summer and 07:00/ 19:00 h in the winter). Temperature (23 ± 1°C) and humidity (55 ± 10%) were monitored daily. Standard food pellets and pure water were provided ad libitum during the entire study period (12 months). All procedures were carried out with care to avoid and minimize animal discomfort.
In our developed methodology, we used the monotonous (in the same concentration and dose) application of 0.1% solution of DMBA in benzene as carcinogen in a volume of 40 μL using graduated pipette with an interval not exceeding 48-72 h, twice a week on the pre-shaved back skin of the rat with an area 5 × 5 cm during 10-12 months. We studied the clinical rat skin changes dynamics and histological image of biopsied skin using serial histological samples of material fixed in 10% formalin with staining it with hematoxylin and eosin (H&E).
The animals were sacrificed by exsanguination under ether anesthesia. For histological study, the tumor was completely excised and then the procedure of preparation of the sections was carried out with the aim of studying the structure of the tumor. Each animal was performed a single collection of material in the form of complete excision of the lesion. Consecutive biopsies were collected from various lesions. The presence/absence of metastases was assessed using autopsy and internal examination of organs. A total of 40 biopsies were made during the experiment.
The study was carried out in accordance with the Directive for the Protection of Vertebrate Animals used for Experimental and other Scientific Purposes of the European Union (86/609/ EEC), in compliance with the Helsinki Declaration. The local ethical committee for experimental animals approved the protocols of the study.

Results
The first skin changes such as erythema, small areas of keratosis 0.2-0.3 cm in diameter, papillomas and round-shaped pink papules with dotted crust were detected in 5 of 50 rats on the 8th month of observation. Forty-five rats did not show tumor formation, only small erythema areas were observed.
Further, skin changes were recorded weekly and summarized monthly (Fig. 1). On the 9th month of the observation, 38 rats showed similar skin changes such as papillomas, small keratosis, and papules. In five rats, skin changes developed with the formation of a vast number of micronodular basal cell carcinoma 5-8 mm in diameter. Histological image showed the initial stages of BCC formation in the form of trabecular arrangement of basaloid cells accompanied by moderate differentiation of hair follicles with the cellular and nuclear monomorphism (Fig. 2). At the same time, the skin of the seven rats was not affected by tumor process. One rat died from pneumonia.
On the 10th month, initial symptoms of neoplastic process (keratosis, papillomas, small papules) were detected in 3 rats; in 20 rats, micronodular basal cell carcinomas were observed. In 21 rats, skin changes developed; tumor nodules >0.5-1.0 cm in diameter were formed, with the formation of dense thick brown color crusted areas of irregular shape in central part, covering ulceration surfaces surrounded by a zone of hyperemia. The histological analysis of biopsies from such lesions indicated formation of ulcerated solid BCC with a large number of lagoons, expressed inflammatory mononuclear cell infiltrate, and sharply dilated capillaries. Tumors were surrounded by abundant cellular fibrous stroma, and an expressed deep invasive infiltrating tumor growth pattern was observed (Fig. 3). During the 10th month, five rats did not show any skin changes and three rats died.
On the 11th month, the initial neoplastic processes were observed in two rats, and four rats showed the formation of micronodular carcinomas and keratoacanthomas 5-8 mm in diameter. In the major part of rats, process progressed toward the formation of tumor nodules with ulceration (nodulo-ulcerative process). Herewith, in 12 rats, they were 0.5-1 mm in diameter; in 26 rats, the tumor nodules were characterized by nodulo-ulcerative lesions >2 cm in diameter. Thus, for example, in rat No. 15, the nodule size was 3.1/2.2 cm; in rat No. 3, 2.5/2.8 cm; in rat No. 22, 3.3/2.7 cm. Ulceration in the central part of the tumors had reached 3 mm in depth; the edges of the ulcer were uneven, sometimes steep or completely covered with a thick layered crust, with rejection of which (scraping by claws) bleeding of uneven bottom of the ulcerative defect was noted. It was observed that the rats became restless and aggressive. Tumors were dense, movable on palpation, without metastases. The clinical picture was suspicious for BSC, SCC and also for the infiltrative nodular BCC with ulceration. Histological studies, which were performed using layer-by-layer histological analysis of morphology of the tumor structures, identified the image of skin BSC on the 11-12th month from such nodules in 15 rats.
In all specimens, the tumor structure was represented by substantial, solid-adenoid epithelial complexes, characterized by destructive growth pattern. Elongated trabecular arrangements of basaloid cells had irregular shape; in some areas, the loss of structure of Bpalisade^, typical for BCC, was noted. In some tumor areas, monomorphic cells were formed, while other areas comprised polymorphic cells and nuclei. Cells taken from the central areas of tumor complexes were typically larger than basaloid cells with eosinophilic cytoplasm. Their mitotic activity was increased, and multipolar and monocentric mitoses as well as three-group metaphases were observed.
We also noticed that basaloid cells usually differentiated to spinous elements. Formation of concentrically located flattened cells and horn pearls was also detected. Local immune response occurred in the form of stromal infiltration by plasma cells and lymphocytes. A primitive angiogenesis was also significant (Fig. 4).
A total of 40 biopsies were made during the experiment. As a result, during the 11-12th months, we obtained 15 BSCs, 22 BCCs, and 3 SCCs (Fig. 5).

Discussion
In this study, we developed a simple and convenient rat model of basosquamous carcinoma (BSC) with high yield of malignant tumors (40%). During the experiment, we used a simple protocol of DMBA application as complete carcinogen with no additional promoter agent instead of a two-stage model of induction, when after a single application of the initiator mutagen repeated applications of a pro-inflammatory agent should be carried out.
In contrast to the previous studies of mouse model of skin cancer [15,16], for the development of basosquamous carcinoma model, we used specific pathogen-free Wistar rats. Rat    [17][18][19] but very rarely for skin cancer modeling.
The first visible tumors occurred on the 9th month of the experiment; the first BSCs were obtained on the 10th month. On the 11th month of the experiment, BSCs were obtained in 12 rats. At equal doses of DMBA, the process of tumor formation in rats required more time than in mouse models; at the same time, this technique allowed to investigate the process of tumorigenesis in more detail and with low mortality rate of animals (only eight rats have died during the experiment).
The stages of BSC formation observed in the described experiment allow us to conclude that BSC is formed from basaloid tumor complex as a result of changes in metaplasiatype differentiation of a part of basaloid cells. These results are contradictory to the previous assumption about BSC formation from two initial tumors, BCC and SCC (a tumor-collision model) [20]. The basaloid tumor complex was formed on the 10th month of the experiment, while BSC was detected only to the 11-12th months, i.e., after prolonged application of the carcinogen on the skin with expressed indicators of oncogenesis, such as ulcerative BCC with the signs of primitive angiogenesis, infiltrative growth, and active stroma formation. Apparently, these events are related to activation of a number of oncogenes and immune depression.

Conclusion
In this study, we have developed a convenient BSC model in Wistar rats. It is based on a simple carcinogenesis protocol that only utilized twice a week low-dose DMBA applications. The investigation of tumor progression dynamics suggests that BSC is formed from basaloid tumor complex as a result of changes in metaplasia-type differentiation of basaloid cells. However, additional studies, such as proteome analysis of cells taken from the tumor tissue and experiments with increasing doses of carcinogen, are required to clarify the mechanisms of carcinogenesis. The developed skin cancer model provides an easy and convenient research tool for studying factors influencing BSC growth and progression, identification of molecular mechanisms implicated in BSC tumorigenesis, and for topical antitumor drug testing.