Mouse melanoma models and cell lines


  • William E. Damsky Jr,

  • Marcus Bosenberg

Contact: William E. Damsky Jr and Marcus Bosenberg, Department of Dermatology, Yale School of Medicine, 15 York Street, New Haven, CT 06520, USA e-mail:

Mouse models of melanoma have roots in the early 1900′s when melanocytic tumors arose spontaneously in inbred mouse strains (Cloudman, 1941; Green, 1962; Harding and Passey, 1930). These spontaneous melanomas (probably the most well known of which is B16) were transplantable to congenic mice and could also be cultured, studied, and manipulated in vitro. Due to the utility of these models, they have been a central means by which to address basic questions in melanoma biology. In the latter part of the 20th century, researchers began to employ mutagens like UV irradiation and 7,12-Dimethylbenz(a)anthracene (DMBA), as well as tumor promoting agents such as croton oil/12-O-tetradecanoylphorbol-13-acetate (TPA) in order to generate melanomas in mice. These pioneering studies (summarized in Table 1), gave rise to melanomas and cutaneous neoplasms.

Table 1.  Mouse models of melanoma
Author(year)PMIDGeneticmodificationBackgroundSpontaneousMelanomaInducedMelanomaMetastasisCell lines?Notes
  1. Abbreviations - PMID: PubMed identifier and LN: lymph node, others are standard.

Nogueira et al. (2010)20711233Tyr::H-RasV12GInk4a/Arf −/−Pten ± 75%, medlatency18.9 weeks rare LN, lungyes 
Ferguson et al. (2010)20718941Tyr::N-RasQ61KCdk4R24C/R24C100% penetranceby ∼300 daysUV accelerates/incpenetrancenot reportednot reportednevusprecursors?
Tyr::N-RasQ61KArf −/−∼25% penetranceUV accelerates/incpenetrancenot reportednot reported
Tyr::N-RasQ61KTyrCreERT2/p53F/F 100% penetranceby 200 daysUV accelerates/incpenetrancenot reportednot reported
Yang et al.(2010)20530876Tyr::H-RasV12GTyr::rtTA tetO-CreIkkbF/FInk4a/Arf −/−4.6% pentrance,latency 79 days not reportednot reportedIKKB lossinhibitstumorigenesis
Monahan et al.(2010)20697345Tyr::CreERT2/LSLK-rasG12D/p53F/F 45%, 31 weeksmed latency noneyes 
Tyr::CreERT2/LSLK-rasG12D/p16F/F73%, 24 weeksmed latencynoneyes
Tyr::CreERT2/LSLK-rasG12D, p53F/F,p16F/F100%, 9 weeksmed lantencynoneyes
Woods and Bishop(2010)20577802Dct::rtTAtetO-cmyc  UVR, 10–17 monthlatencynot reportednot reportedworks withlacZ/H2B-eGFPreporter
Milagre et al.(2010)20516123Tyr::CreERT2LSLK-rasG12D 100%, 4 monthsmedian latency noneyesmetastaticpotential inxenografs
VanBrocklin et al. (2010)20444198Dct::TVA Ink4a/ArfF/F(RCAS-NrasQ61R-IRES-Cre) 63% pentrance,med survival 47 days noneyesmetastaticpotential inxenografs
Kumasaka et al. (2010)20048069MT::RetEdnrb ± about 40% byabout 70 weeks lung ∼40%not reportedinc mets relto MT::RET
Heidorn et al. (2010)20141835Tyr::CreERT2LSL-KrasG12DLSL-BrafD594A 100% by6 months not reportedyesno nevusprecursors
Held et al.(2010)20048081Tyr::CreERT2 PtenF/FCdkn2aF/F 100% by40 weeks not reportedyes (1118/1111)tumorinitiatingsub-populations
Tyr::CreERT2 PtenF/FCdkn2aF/FB-cateninloxex3/wt100% by40 weeksnot reportedyes (1445)tumor initiatingsub-populations
Chawla et al.(2010)20703083Tyr::HrasG12VCdk4R24C/R24C33% by>15 monthsenhanced byDMBA/TPAnonot reported 
Goel et al.(2009)19398955Tyr::BrafV600E <10% 295–595med survival(2 founders) LNyesmelanocytic nevi
Tyr::BrafV600ECdkn2a ± 11–38% pentrance,185–485 daysmed latanceyLNyesmelanocytic nevi
Tyr::BrafV600Ep53 ± 6–53% penetrance,107–457 daysmed latencyLNyesmelanocytic nevi
Dhomen et al.(2009)19345328Tyr::CreERT2LSL-BrafV600E 64%, medlatency 12 months noyesmelanocytic nevi
Tyr::CreERT2LSL-BrafV600Ep16−/−80%, medlatency 7 mosnoyesmelanocytic nevi
Dankort et al.(2009)19282848Tyr::CreERT2BrafCA/wt PtenF/F 100% within10 weeks 100% lung/LNyes (2697)Rapamycin/MEK-Isensitive
Inoue-Narita et al.(2008)18632629Dct::Cre PtenF/F  DMBA/TPA: 50%by 25 weeks20% lungnot reporteddec hair greying,neuorlogicphenotypes
Delmas et al.(2007)18006687Tyr::B-cateninstaTyr::NrasN61K 85%, 27.6med latency not reportedyescongenic toC57BL/6
Ha et al. (2007)17576930MT::HGF/SFp16−/− neonatal UVR,100% by<450 days melanocyte lines 
MT::HGF/SFp19−/− neonatalUVR, ∼70% by200 daysnot reportedmelanocytelinesp53 independentfunction of Arf
Yang et al.(2007)17575131Xpc −/− Ink4a/Arf −/− ∼70% by∼350 daysneonatal UVRnot reportednot reportedxpc−/− alsohas increasedRas mutants
Hujibers et al.(2006)16540681Tyr::CreER-iRasP1A(HRasG12V) p16F/Fp19F/F 33%, median73–253 days(pig/non-pig) noyesP1A antigenexpression
Tormo et al.(2006)16877364MT::HGF/SF  DMBA/TPA,∼50% by30 weeksLNnot reportedcongenic toC57BL/6
MT::HGF/SFCdk4R24C/R24C DMBA/TPA,100% by12 weeksLNnot reportedcongenic toC57BL/6
Hacker et al.(2006)16540642Tyr::HrasG12VCdk4R24C/R24C58% in 1 yrUVR increasesto 80%in 1 yrLNyes 
Yamazaki et al.(2005)16117793K14::SCF XPA −/−  UVR, 55%by week 70LNnot reported 
Ackermann et al.(2005)15899789Tyr::NrasQ61K 94% by 6 months    
Tyr::NrasQ61Kp16−/−29%, median12.5 monthsLN, 36%lung/liveryesmetastatic inNOD/SCID
Von Felbert et al.(2005)15743795MT::RetIL6−/−47% by 65 weeks cerebralyesIL6−/− delaysmelanomaformation
Hacker et al.(2005)16297212Tyr::HRasG12V  UVR, 57%by 12 monthsnot reportedyes 
Sharpless et al.(2003)12902988Tyr::HRasG12Vp19−/−52% by 81 weeks not reportednot reported 
Tyr::HRasG12Vp16−/−35% by 89 weeksnot reportednot reported
Pollock et al.(2003)12704387Dct::Grm1 100%, up to20 months LNnot reportedhyperpigmentedprecursors
Kannan et al.(2003)12538879Tyr::HRasG12Vp19−/−∼50% by 50 weeks>75% by50 weeksnot reportednot reportedUV induces Cdk6amplification
Tyr::HRasG12Vp16−/−∼50% by 50 weeks∼50% by50 weeksnot reportednot reported
Recio et al.(2002)12438273MT::HGF/SFp16−/− p19−/− 100% by 50 daysLN, livernot reported 
Yang et al.(2001)11719444Tyr::MIP-2p16 ± p19 ±18.5% through40 weeksDMBA, 12%198 daysnoneyesmetastatic innude mice
Kligman and Elenitsas (2001)11479419Skh-hr-2 8% through30–47 weeksDMBALN, lungnot reportedpigmentedmacules inother mice
Bardeesy et al. (2001)11238948Tyr::HrasG12Vp53−/−26%, medlatency 17 weeks noyes 
Krimpenfort et al. (2001)11544530p16−/− p19 ±   DMBA, 50%between 3 and9 monthsLN, lung, liver, spleennot reported 
Sotillo et al. (2001)11606789Cdk4R24C/R24C  DMBA/TPA,70% by25 weeksnonot reported 
Noonan et al. (2001)11565020MT::HGF/SF  UVR, ∼20–40%by 450 daysnot reportednot reportedneonatal UVRimportant fortumorigenesis
Strickland et al. (2000)10989613   aloe emodin/UVR,50–70% by30 weeksnot reportednot reportedB3H congenic
Noonan et al. (2000)10919643MT::HGF/SF  UVR, >50%by 21 monthsnot reportednot reported 
Kato et al. (2000)11121157MT::Ret  UVR, 80% by28 weekslung up to 50%not reportedUV super-activates Ret
Chin et al. (1999)10440378Tyr::rtTAtetO::HRasG12VCdkn2a −/−25%, medianlatency 60 days not reportedyesKrasG12V addiction
Broome Powell et al. (1999)10469620Tyr::HrasG12V  UVR, DMBA, TPA,combinations,variablesome lung, LNyesalso melanocyticnevus development
Kunisada et al., 19989584135K14::SCF nonononoepidermal retentionof melanocytes
Otsuka et al. (1998)9823327MT::HGF/SF 22%, meanlatency 15.6months LN, liver,spleenyes 
Kato et al. (1998)9778055MT::Ret 65% malignant,mean latency130 days LN, lung, brain,othersyes100% benignmelanocyticproliferation
Zhu et al. (1998)9506443TG3 100% penetrance LN, lung,brain, othersnot reportedunknown transgeneinsertion site
Kelsall and Mintz (1998)9751610Tyr::SV40E  UVR 12.5%,mean latency77 weeksLN, lung,kidneynot reported 
Chin et al. (1997)9353252Tyr::HrasG12VCdkn2a −/−60% penetranceby 6 months noyes 
Chen et al. (1996)8618055TG3 100% penetrance not reportednot reportedunknowntransgeneinsertion site
Powell et al. (1995)7662120Tyr::HrasG12V nonononot reportedmelanoccytichyperplasia
Klein-Szanto et al. (1994)8062242Tyr::SV40E  UVR, earlylesions transplantedat 20 weeksinirectly(LN, lung)not reported 
Mintz and Silvers (1993)8415613Tyr::SV40E 25–100%,average 46–51weeks inirectly(LN, lung)not reportedearly eyemelanomas, musttransplant
Husain et al. (1991)1909931Skh-hr-2  DMBA/UVR,25–33% by20–30 weeksnonot reported 
Bradl et al. (1991)1846036Tyr::SV40E <10% cutaneousmelanoma not reportednot reportedearly ocularmelanomas
Klein-Szanto et al. (1991)1846037Tyr::SV40E 17% cutaneousmelanoma nonot reportedmelanosis,ocular lesions
Iwamoto et al. (1991)1915289MT::Ret variable dependingon founder not reportedyesmelanosis,ocular lesions
Romerdahl et al. (1989)2517915   UVR/DMBA/crotonoil, > 31%penetrance??C3H congenic
Takizawa et al. (1985)3924435   DMBA/croton oil,0–80% dependingon strainnot reportednot reportedDBA resistantto macule/melanoma
Holman et al. (1983)6578359?      
Berkelhammer et al. (1982)7093959   DMBA/croton oil,10% penetranceaftertransplantationyes 
Epstein et al. (1967)6016644  ?DMBA/UVR?? 
Green (1962)n/a  spontaneous  yesB16 melanoma
Cloudman (1941)n/a  spontaneous  yesCloudmanmelanoma
Harding and Passey (1930)n/a  spontaneous  yesHarding-Passeymelanoma

After the advent of transgenic mouse technology, two early approaches were successful in producing melanocytic neoplasms in mice (Iwamoto et al., 1991; Klein-Szanto et al., 1991). One used the methallothionein promoter to drive Ret expression, while the latter used the tyrosinase promoter to drive SV40 T cell antigen expression. In addition to generalized melanosis, these transgenic mice developed ocular melanocytic neoplasms that often precluded development of cutaneous melanomas. Through the 1990s and early 2000s, several more transgenic models were developed (summarized in Table 1), which mostly employed melanocyte-specific expression of mutated Ras or activation of the c-Met-HGF/SF signaling axis. Changes to cell-cycle control elements also aided this effort by using either melanocyte-specific activation of Cdk4 and/or Cdkn2a-deficient backgrounds (p16 and/or p19). Although these models were important in confirming the tumor initiating role of these genetic changes, they often had long latencies and were incompletely penetrant.

More recently, an increased understanding of the genetic changes that occur in human melanoma coupled with technical advances in mouse modeling have led to the development of novel and particularly useful models. Braf activating mutations have been described to occur in 50% of melanomas (Davies et al., 2002). Additionally, near universal activation of the PI3K/Akt/mTOR signaling pathway (often through inactivation of Pten) has also been documented in human melanomas. Concurrent to this increased understanding of genetic changes influencing melanoma formation, inducible lox-Cre based recombination technology in mice has allowed for more precise control over genetic recombination in mouse models (Figure 1). In 2009, three novel mouse models based on Braf activation were described (Dankort et al., 2009; Dhomen et al., 2009; Goel et al., 2009). In particular, Dankort et al. (2009) describe a model based on melanocyte specific Braf (V600E) activation and Pten inactivation. In this model, metastatic melanoma forms with 100% penetrance and virtually no latency. Additionally, tumor formation can be controlled in both a temporal and anatomically-restricted fashion, which occurs only after topical application of 4-hydroxytamoxifen.

Figure 1.

 Schematic of inducible genetically engineered mouse models of melanoma. (A). In the inducible Cre models, the Cre-recombinase::estrogen receptor (CreER) fusion protein is constitutively expressed melanocytes by Tyr or Dct transgenic promoter elements. Upon exposure to 4-hydroxytamoxifen (either topical or systemic application), CreER is released from Hsp90 in the cytoplasm, translocates to the nucleus and recombines chromosomal sites, resulting in removal of DNA sequences located between paired loxP sites in a gene of interest (GOI). This strategy can result in either genetic inactivation or activation, depending on the design of the allele and generally cannot be reversed. (B). In the doxycycline inducible system, administration of doxycycline alters of tet-transactivator (tTA) or reverse tet-transactivator(rtTA) function, resulting in activation of transcription (rtTA) or inactivation of transcription (tTA) at target promoters the containing tetO promoter sequence element. This system allows for reversible induction or inactivation of transgene expression.

Due to the reproducibility and ease-of-use of the inducible ‘Pten/Braf’ model, Jackson Laboratories will soon distribute the three alleles required for the model in a congenic C57/BL6 background. This model is advantageous for many reasons, some of which include: short latency and known kinetics of tumor formation, attractiveness for tumor-immunological studies and pre-clinical testing of novel therapeutics, transplantability of congenic tumor cells, and ability to incorporate additional genetic changes with different floxed alleles. This model and derivatives of it will likely be central to further advancing our understanding of melanoma biology as well as providing a platform for which to test novel therapeutics based.

Several melanoma tumor lines from this and other Cre/lox mouse melanoma models have been generated (Dankort et al., 2009; Held et al., 2010). Analogous mouse melanoma cell lines derived from congenic mice are currently being developed with the intent of making them available for distribution shortly after derivation (within the next 12–18 months). These lines will be very useful for studying genotype-specific signaling, chemosensitivity, and tumor immunology in vitro and upon transplantation into immune competent C57Bl/6 mice.