The following procedure describes a method for spatial and temporal control of melanocytic tumor initiation in murine dorsal skin, using a genetically engineered mouse model. This protocol describes macroscopic as well as microscopic cutaneous melanoma initiation.
Cutaneous melanoma is well known as the most aggressive skin cancer. Although the risk factors and major genetic alterations continue to be documented with increasing depth, the incidence rate of cutaneous melanoma has shown a rapid and continuous increase during recent decades. In order to find effective preventative methods, it is important to understand the early steps of melanoma initiation in the skin. Previous data has demonstrated that follicular melanocyte stem cells (MCSCs) in the adult skin tissues can act as melanoma cells of origin when expressing oncogenic mutations and genetic alterations. Tumorigenesis arising from melanoma-prone MCSCs can be induced when MCSCs transition from a quiescent to active state. This transition in melanoma-prone MCSCs can be promoted by the modulation of either hair follicle stem cells' activity state or through extrinsic environmental factors such as ultraviolet-B (UV-B). These factors can be artificially manipulated in the laboratory by chemical depilation, which causes transition of hair follicle stem cells and MCSCs from a quiescent to active state, and by UV-B exposure using a benchtop light. These methods provide successful spatial and temporal control of cutaneous melanoma initiation in the murine dorsal skin. Therefore, these in vivo model systems will be valuable to define the early steps of cutaneous melanoma initiation and could be used to test potential methods for tumor prevention.
Melanoma, the malignant form of melanocytic tumors, is the most aggressive cancer in the skin with cutaneous melanoma responsible for the majority of skin cancer deaths1. In the United States, melanoma is commonly diagnosed; it is projected to be the 5th to 6th most common cancer type among the estimated new cancer cases in 20182. Furthermore, while overall cancer incidences have shown the trend of gradual reduction in recent decades, cutaneous melanoma incidence rates during the last few decades demonstrate a continuous and rapid increase in both genders2.
To combat cutaneous melanoma more efficiently, it is important to clearly understand the early events of melanoma development from their cells of origin to better identify clinically effective preventative methods. It is now known that adult stem cells can significantly contribute to tumor formation as the cellular origins of cancers in many different types of organs including skin tissues3,4,5. Similarly, adult melanocyte stem cells (MCSCs) in the murine dorsal skin can work as melanoma cells of origin upon aberrant activation of the Ras/Raf and Akt pathways6. However, these oncogenic mutations alone are not able to efficiently induce tumor formation from quiescent MCSCs. Melanocytic tumors eventually develop when MCSCs become active during natural hair cycling. However, in this protocol, we will describe methods to artificially induce cellular activation of tumor-prone MCSCs, thus facilitating precise control of melanoma initiation6,7.
Through the methods described here, spatial and temporal control of cutaneous melanoma initiation from tumor-prone MCSCs expressing oncogenic BrafV600E together with loss of Pten expression can be successfully performed, as we have previously reported6. This method incorporates previous findings that demonstrate hair follicle stem cell activation through depilation as well as how exposure of murine skin to UV-B can facilitate activation of MCSCs resident to the hair follicle and translocation of this cellular subpopulation to the interfollicular epidermis6,8,9,10. These in vivo model systems can provide valuable information regarding how physiological and environmental changes can alter the cellular status of melanoma-prone MCSCs, which in turn can induce significant initiation of melanoma in the skin.
All animal procedures are performed in accordance with Cornell University Institutional Animal Care and Use Committee (IACUC).
1. Preparation
2. Tamoxifen Treatment for Cre-Lox genetic Recombination
3. Chemical Depilation
4. UV-B Irradiation
5. Tissue Processing
Cutaneous melanoma initiation induced by chemical depilation
The procedure of chemical depilation is depicted in Figure 2. When mice are 7-weeks postnatal, their dorsal skin is in telogen. During telogen, hair follicle stem cells and MCSCs are known to be in a quiescent, resting state. The skin should show no significant hair growth after shaving. On the other hand, chemical depilation can induce hair follicle stem cell activation, which in turn shows significant hair growth from the region of interest (Figure 2). Similarly, tumor-prone MCSCs expressing oncogenic mutations also need to be active for accelerated tumor formation. Chemical depilation can significantly induce the activation of both hair follicle stem cell and MCSCs. Melanoma-prone MCSCs in an active state can form tumors, and microscopic melanoma initiation can be observed within 2 weeks after chemical depilation using the lineage tracing allele LSL-tdTomato (Figure 3).
Cutaneous melanoma initiation induced by UV-B irradiation
Similar to chemical depilation, UV-B can induce tumor initiation from quiescent tumor-prone MCSCs (Figure 4). While the murine skin containing tumor-prone MCSCs in a quiescent state shows no significant tumor initiation and lack of significant pigmentation, the dorsal skin exposed to UV-B (twice, 180 mJ/cm2) shows black pigmented early melanocytic tumors which are macroscopically evident (Figure 4A, B). While UV-B can significantly induce macroscopically evident tumor initiation within 2 weeks, the application of sunscreen can protect UV-B-mediated melanoma initiation in the skin, similar to a UV-resistant cloth (Figure 4C, D).
Importantly, UV-B induces the direct translocation of MCSCs from their follicular stem cell niche to the interfollicular epidermis. Furthermore, UV-B can induce MCSC-originating cutaneous melanoma formation throughout the interfollicular epidermis, and the microscopic phenotype can be observed by the lineage tracing marker, tdTomato (Figure 5). As these tumor cells are malignant, they grow rapidly and invasively (Figure 5). Similar to the dorsal skin, UV-B-induced melanoma initiation can be notably observed in other skin areas, such as the ear skin (Figure 6). Compared to the control skin covered by UV-resistant cloth, the ear skin exposed to UV-B demonstrates higher pigmentation due to a higher burden of melanoma initiation (Figure 6).
Figure 1: Skin tissue isolation and cryo-embedding. Following euthanasia, shave the mouse dorsal skin region of interest and embed skin in cryomold containing OCT. The dotted line on the mouse represents the midline. Typically, 3 or 4 pieces of skin can be isolated with line segment 3-4 along the midline; one such piece is indicated by rectangle 1/2/3/4. These skin pieces should be embedded in the OCT as shown in the figure. Please click here to view a larger version of this figure.
Figure 2: Macroscopic phenotypes after chemical depilation. (A) Diagram of chemical depilation. (B) After shaving, the hair cycle was confirmed to be in telogen. (C) Chemical depilation was performed to remove hair shafts from the hair follicles, which can activate both hair follicle and melanocyte stem cells. (D) Around a week later, early hair growth will be easily noticeable. (E) Then, significant hair growth can be observed over time. Please click here to view a larger version of this figure.
Figure 3: Microscopic observation of melanoma initiation controlled by chemical depilation. Chemical depilation can significantly induce the activation of quiescent tumor-prone MCSCs. Then, microscopic tumor initiation demonstrated by the lineage tracing marker tdTomato can be observed within 2 weeks. This figure demonstrates the microscopic phenotype 16 days post 3 days tamoxifen by IP injection followed by chemical depilation using different visualizations of the same field of view. (A) tdTomato+ tumor cells originating from the hair follicle merged with Dapi nuclear counter stain. (B) Hair follicles and interfollicular epidermis are outlined in white. (C) A schematic showing how tdTomato+ tumor cells invade into the dermis from the hair follicle, modified from Moon, H. et al.6. Please click here to view a larger version of this figure.
Figure 4: Macroscopic phenotype of cutaneous melanoma induced by UV-B irradiation. Mice were treated by tamoxifen for 3 days (day 1 to 3) followed by 2 UV-B exposures (day 4 and 6). (A) Diagram of UV-B irradiation on the dorsal skin containing mutant MCSCs. (B) Macroscopic phenotype between control and UV-B exposed dorsal skin. Significant black pigmentation related to melanocytic tumor formation can be observed within 2 weeks after the second UV-B irradiation. (C) Diagram of sunscreen application. (D) While UV-B can significantly induce tumor initiation from mutant MCSCs, application of sunscreen showed significantly suppressed UV-B-mediated tumor initiation (16 days post the second UV-B irradiation). Please click here to view a larger version of this figure.
Figure 5: Microscopic observation of UV-B-induced cutaneous melanoma. Mice were treated with tamoxifen by IP injection for 3 days (day 1 to 3) followed by 2 UV-B exposures (day 4 and 5). (A-C) Early tumor initiation throughout the interfollicular epidermis demonstrated by lineage tracing tdTomato can be shown at day 17. (A) tdTomato+ tumor cells are shown migrating from the hair follicle into the interfollicular epidermis. Dapi is used as a nuclear counter stain. (B) Dapi staining is removed and dotted line indicates location of the hair follicles and interfollicular epidermis. (C) Schematic indicating tdTomato+ cell migration from the hair germ to the interfollicular epidermis following UV-B exposure. Modified from Moon et al.6. (D-F) Then, tumor cells grow rapidly and invade the dermal tissues below (day 25). (D) tdTomato+ tumor cells have continued to proliferate and invade into the dermal tissues. Dapi is used as a nuclear counter stain. (E) Dapi staining is removed and dotted line indicates the location of the hair follicles and interfollicular epidermis. (F) Schematic indicating tdTomato+ cell invasion into the dermis and melanocytic tumor growth. Modified from Moon et al.6. Please click here to view a larger version of this figure.
Figure 6: Macroscopic and microscopic observation of UV-B-induced melanoma in ear skin. (A) Experimental scheme. Mice were treated by tamoxifen for 3 days (day 1 to 3) followed by 3 exposures to UV-B irradiation (day 4, 6 and 8). (B) Significantly increased melanoma initiation by UV-B irradiation was observed macroscopically as well as microscopically at day 28. Please click here to view a larger version of this figure.
Hallmark genetic alterations frequently found in cutaneous melanoma tumors have been well described13. The most dominant driver mutation is BrafV600E, and a genetically engineered mouse model for BrafV600E-mediated melanoma was generated by the Bosenberg group14 and deposited in the Jackson Laboratory. Using this mouse model, our recent study demonstrated the requirement of cellular activation of tumor-prone MCSCs for the significant initiation of BrafV600E-mediated melanoma in the dorsal skin6.
Depilation is known to be an effective method to induce murine hair follicle stem cell activation on the dorsal skin6,10. Murine MCSCs are located within the hair follicles, and furthermore, the cellular status of follicular MCSCs is synchronized with the state of hair follicle stem cells8. Artificial modulation of hair follicle stem cells can alter MCSC status, thus depilation can directly activate both melanocyte and hair follicle stem cells. This activation method for MCSCs can be applied for tumor-prone MCSCs to examine early steps of cutaneous melanoma initiation in genetically engineered melanoma mouse models. Through a chemical depilation method, melanoma formation in the region of interest in the murine dorsal skin can be directly initiated. Depilation can be performed by various methods such as waxing or plucking, both of which are types of mechanical hair removal where hair shafts are extracted from the hair follicle. However, these mechanical depilation methods can be imprecise with respect to a specific region of interest. On the other hand, a chemical depilation method using hair removal cream is simple to perform and allows for precise application with reliable activation of the hair cycle, and may be less harsh on the skin than mechanical removal.
Among various risk factors, UV-B is the most well-known risk factor in cutaneous melanoma15. UV-B irradiation is known to induce the direct migration of follicular MCSCs into the interfollicular epidermis9. Furthermore, UV-B is a strong environmental stressor which can significantly induce melanoma initiation from tumor-prone MCSCs6. In the experimental protocol described above, a small region of interest on the murine dorsal skin can be exposed to UV-B light while the use of a UV resistant cloth easily protects the rest of the mouse from unwanted exposure. This treatment can be performed while the mouse is under anesthesia. Importantly, the migration of MCSCs and formation of MCSC-originating melanoma following UV-B exposure are UV-B-dose dependent6. Therefore, to have successful results, it is important to regularly check the intensity of light bulb using a UV-B meter and adjust the exposure time and distance between the skin and light bulbs as needed. As is generally known, the skin can burn in response to a high dose of UV-B light, but weak UV-B exposure is also detrimental to the experiment as it will be insufficient to induce MCSC-originating melanoma.
The protocol described here primarily focuses on the environmental risk factor, UV-B, in the dorsal skin. However, the experimental protocol above can be applied to different aims. For instance, the protocol can be modified with light bulbs of differing spectrums to determine the effect of alternative light sources or wavelengths of UV, such as UV-A. This protocol primarily targeted the dorsal skin as it is accessible and the status of MCSCs is well defined. However, it can also be altered to target UV light exposure to appendages such as the ear, tail or palm skin, including foot pads. As an example, in Figure 6, it is possible to examine possible difference in early melanocytic tumor formation at alternative tissue sites. However, unlike the dorsal skin, the status of MCSCs in these sites are less well defined; hence, the potential spontaneous activation of MCSCs leading to random occurrences of melanoma formation should be considered. Therefore, compared to the protocol in the dorsal skin, alternative approaches will require an increased number of experiments to obtain accurate experimental results.
Furthermore, while the current protocol primarily demonstrates BrafV600E-mediated melanoma initiation, our previous study also reported similar early melanoma formation with an oncogenic Ras/Pten combination, using the LSL-KrasG12D genetic allele6. However, other driver mutations such as mutant Nras genes can also be considered with the experimental system described here to understand early melanoma initiation driven by other oncogenic combinations.
Taken together, described here is an in vivo model system which allows for the examination of early cutaneous melanoma initiation in genetically engineered mice. These murine models provide the methods for temporal and spatial control of melanoma initiation from mutant MCSCs in the skin. These protocols provide a novel paradigm to study the mechanisms of early melanoma initiation from tumor-prone MCSCs as well as a platform to determine the physiological and environmental stressors that can contribute to melanoma initiation. This level of spatiotemporal control can also provide a tool for more precise melanoma induction to test drug efficacy in preventing melanoma formation, as well as tumor growth.
The authors have nothing to disclose.
This work was supported by the Office of the Assistant Secretary of Defense for Health Affairs, through the Peer Reviewed Cancer Research Program under award W81XWH-16-1-0272. Opinions, interpretations, conclusions, and recommendations are those of the authors and are not necessarily endorsed by the Department of Defense. This work was also supported by a seed grant from the Cornell Stem Cell Program to A.C. White. H. Moon was supported by the Cornell Center for Vertebrate Genomics Scholar Program.
Tamoxifen | Sigma | T5648-1G | For systemic injection |
Tamoxifen | Cayman Chemical | 13258 | For systemic injection |
Corn oil | Sigma | 45-C8267-2.5L-EA | |
4OH-tamoxifen | Sigma | H7904-25MG | For topical treatment |
26g 1/2" needles | various | Veterinary grade | |
1 mL syringe | various | Veterinary grade | |
Pet hair trimmer | Wahl | 09990-502 | |
Hair removal cream | Nair | n/a | Available at most drug stores |
Cotton swabs | various | ||
Ultraviolet light bulb | UVP | 95-0042-08 | model XX-15M midrange UV lamp |
200 proof ethanol | various | pure ethanol | |
Histoplast PE | Fisher Scientific | 22900700 | paraffin pellets |
Neutral Buffered Formalin, 10% | Sigma | HT501128-4L | |
Clear-Rite 3 | Thermo Scientific | 6901 | xylene substitute |
O.C.T. Compound | Thermo | 23730571 | |
Tissue Cassette | Sakura | 89199-430 | for FFPE processing |
Cryomolds | Sakura | 4557 | 25 x 20 mm |
FFPE metal mold | Leica | 3803082 | 24 x 24 mm |
Isoflurane | various | Veterinary grade | |
Anesthesia inhalation system | various | Veterinary grade | |
Fine scissor | FST | 14085-09 | Straight, sharp/sharp |
Fine scissor | FST | 14558-09 | Straight, sharp/sharp |
Metzenbaum | FST | 14018-13 | Straight, blunt/blunt |
Forcep | FST | 11252-00 | Dumont #5 |
Forcep | FST | 11018-12 | Micro-Adson |
Tyr-CreER; LSL-BrafV600E; Pten-f/f | Jackson Labs | 13590 | |
LSL-tdTomato | Jackson Labs | 007914 | ai14 |
Cre-1 | n/a | GCATTACCGGTCGATGCAACGAGTGATGAG | |
Cre-2 | n/a | GAGTGAACGAACCTGGTCGAAATCAGTGCG | |
Braf-V600E-1 | n/a | TGAGTATTTTTGTGGCAACTGC | |
Braf-V600E-2 | n/a | CTCTGCTGGGAAAGCGGC | |
Kras-G12D-1 | n/a | AGCTAGCCACCATGGCTTGAGTAAGTCTGCA | |
Kras-G12D-2 | n/a | CCTTTACAAGCGCACGCAGACTGTAGA | |
Pten-1 | n/a | ACTCAAGGCAGGGATGAGC | |
Pten-2 | n/a | AATCTAGGGCCTCTTGTGCC | |
Pten-3 | n/a | GCTTGATATCGAATTCCTGCAGC | |
tdTomato-1 | n/a | AAGGGAGCTGCAGTGGAGTA | |
tdTomato-2 | n/a | CCGAAAATCTGTGGGAAGTC | |
tdTomato-3 | n/a | GGCATTAAAGCAGCGTATCC | |
tdTomato-4 | n/a | CTGTTCCTGTACGGCATGG |