COURTESY OF DOMINIK PAQUET, THE ROCKEFELLER UNIVERSITY, NEW YORK, USA
From frogs to dogs and people, cancer wreaks havoc across the animal kingdom—and fish are no exception. Coral trout, for example, develop melanoma from overexposure to sun, just as humans do. Rainbow trout develop liver cancer in response to environmental toxins. And zebrafish—small, striped fish indigenous to the rivers of India and a widely used model organism—are susceptible to both malignant and benign tumors of the brain, nervous system, blood, liver, pancreas, skin, muscle, and intestine.
Importantly, tumors that arise in the same organs...
Zebrafish are an increasingly popular choice among cancer biologists. Between 1995 and 2012, there was a 10-fold increase in the number of yearly PubMed citations of cancer studies in the species, with more than 200 research papers published last year. Although dwarfed by cancer studies using human tissue and mouse models, the optical transparency of zebrafish embryos and larvae—and now, adult fish of a recently created strain—allows researchers to track tumors in a way that is not possible in other vertebrate models. Furthermore, their small size—embryos are small enough to be reared in 96-well plates—make them a more practical laboratory system than other cancer models. Indeed, researchers are now using these fish to identify druggable oncogenic drivers of specific tumor types, to tease apart the complex network of cancer genes that cooperate in tumor formation and progression, to probe the interplay between the genes that govern embryonic development and those that cause cancer, and to uncover how tumors metastasize and kill their host. The zebrafish model offers a major opportunity to discover important pathways underlying cancer and to identify novel therapies in high-throughput drug screens in a way that mice never could.
The zebrafish toolbox
Zebrafish (Danio rerio) have fast made their way from pet stores and home aquaria into research laboratories worldwide. Their weekly matings produce 100 to 200 embryos that rapidly and synchronously march through embryonic development, so that within 5 days of fertilization, they are mature, feeding larvae. Zebrafish are small and inexpensive to maintain in high numbers, facilitating large-scale experimentation and cheap in vivo drug screens. Famously, the fish are transparent during early larval stages, allowing investigators to directly observe internal development and making the fish a favorite of developmental biologists since the 1960s. But in recent years, the utility of zebrafish has been proven beyond developmental fields, and they are now being found in more and more laboratories studying behavior, diabetes, heart disease, regeneration, stem cell biology—and cancer.
Critically, zebrafish can be used to identify the important pathways and processes that cause cancer in people. Common organ systems and cell types are shared between human and zebrafish, and whether induced by transgenesis or carcinogens, cancers arising from the blood (leukemia and lymphoma), pigmented cells of the skin (melanoma), and the cells that line the bile ducts (cholangiocarcinoma) have microscopic features that are essentially indistinguishable between humans and zebrafish.
The zebrafish model offers a major opportunity to discover important pathways underlying cancer and to identify novel therapies in high-throughput drug screens, in a way that mice never could.
Comparing gene-expression profiles of tumors across various species provides a powerful mechanism for identifying genes that likely represent core functions of cancer. For example, microarray gene-expression analyses have compared the gene signatures of fish hepatocellular carcinoma to that of human liver, gastric, prostate, and lung tumors. Remarkably, this analysis revealed that fish and human liver tumors are more similar to each other than either tumor type is to human tumors derived from different tissues. Moreover, comparative studies can often be used to pinpoint pathways that are active in human disease. This is illustrated by work on a zebrafish model of rhabdomyosarcoma (RMS), a cancer of skeletal muscle, which revealed a gene signature that is also commonly found in human RMS, highlighting the importance of the RAS signaling pathway in the genesis of human RMS.1
A window into cancer
In 2003, the laboratory of A. Thomas Look at Children’s Hospital Boston was the first to realize the long-held dream of following the behavior of cancer cells as they initiate tumor growth and invade structures within live animals. Specifically, the researchers engineered leukemia-afflicted T cells to express green fluorescent protein (GFP) and visualized cancer onset within the zebrafish thymus.2 Moreover, these GFP-positive tumor cells were transplantable into recipient fish, a hallmark of the malignant cell type. Following up on this work, several researchers have now begun transplanting fluorescently labeled human cancer cells into zebrafish larvae to visualize tumor growth and spread in a manner not achievable in more common mouse xenograft models.
Capitalizing on the lack of an acquired immune system during larval stages and the ability to rear zebrafish at temperatures that mimic the human core temperature, Stefania Nicoli of the University of Brescia in Italy and colleagues implanted human cancer cells expressing high levels of vascular endothelial growth factor (VEGF) into zebrafish larvae with GFP-labeled blood vessels.3 VEGF is a factor commonly produced by growing cancers and is responsible for coaxing blood vessels to invade the developing tumor. Nicoli’s work allowed the direct visualization of vasculature remodeling and new vessel formation—and showed that it could be blocked by the addition of VEGF-inhibitory drugs to the water in which the larvae lived. Using similar approaches, many laboratories have successfully engrafted human cancer cells from a range of tumor types into zebrafish embryos.
Researchers have also used zebrafish to visualize the role of tumor heterogeneity within cancer over time. Work from one of our labs (David Langenau’s) has utilized a model of RAS-induced RMS to fluorescently label tumor cells based on differentiation status. This allowed the team to watch the never-before-visualized birth of cancer—the acquisition of invasive properties by normal muscle stem cells and the breakdown of normal muscle architecture, clearing the way for continued tumor expansion. The researchers also characterized two molecularly distinct cell populations, one that is responsible for tumor growth and another that drives cancer spread or metastasis. (See photos below—Imaging Blood Cancers; Solid Tumor Development.)
CLOCKWISE FROM TOP: © MARTIN LOBER; COURTESY OF RICHARD WHITE; COURTESY OF GRAHAM SCOTT; COURTESY OF KIRSTEN EDEPLIT-cell leukemia and RMS are pediatric diseases, favoring cancer development in early larval stages of these zebrafish models. However, cancer is predominantly a disease that affects adults, and zebrafish lose their transparency at around 3 weeks of age. To visualize tumor formation in older zebrafish, one of us (Richard White) and Len Zon of Children’s Hospital Boston have developed a strain of zebrafish called casper, which lacks pigment and is optically clear into adulthood.4 (See photo here.) Using these animals, investigators have implanted pigmented melanomas and witnessed local spread and metastasis over time. First described in 2008, casper is now the zebrafish strain of choice for imaging studies in the field.
Zebrafish have truly proven to be ideal organisms for visualizing cancer; there is no other animal system that allows researchers to literally watch tumors grow and spread. Because we can follow cells as they escape from the primary tumor, migrate, and form metastases in a variety of organs, zebrafish provide an unsurpassed model to describe the distinct steps of cancer progression, and researchers using this model are already contributing much to our understanding of cancer.
Screening for drugs
In terms of drug discovery, zebrafish have emerged as the only vertebrate organism amenable to high-throughput and high-content chemical screening in vivo. The small size of freshly hatched zebrafish embryos means that up to 20 embryos can be dispensed into the individual wells of a 96-well plate, thereby providing a platform for the robotic delivery and testing of hundreds or thousands of compounds in living animals. Because of the parallels between embryonic development and cancer, compounds producing changes in the growth or proliferation of developing organs may also be relevant to cancer. However, to more directly search for small molecules capable of putting the brakes on cancerous growth, researchers are turning to several established tumor-prone zebrafish lines.
One prominent success from such endeavors is the identification of a drug to treat melanoma. Like human melanoma skin lesions, zebrafish melanomas exhibit a gene-expression signature characteristic of the embryonic neural crest, a multipotent group of stem cells that give rise to dozens of cell types, including pigmented skin cells called melanocytes. The genes in this signature, including sox10 and mitf, were hypothesized to be important for melanoma growth, so in 2011 White and Zon performed an in vivo screen to identify small molecules that suppressed the expression of these neural crest genes in developing embryos.5 After screening 2,000 molecules, they identified leflunomide, an approved treatment for rheumatoid arthritis. Importantly, leflunomide was then found to inhibit the growth of both zebrafish and human melanoma xenografts in vivo, and the drug moved from discovery to Phase 1/2 trials in only 4 years, demonstrating just how quickly discoveries in zebrafish can have clinical impact. Similar efforts to discover novel modulators of leukemia growth have recently been reported to work in both fish and humans, suggesting that this approach will be broadly applicable to a wide range of solid and liquid tumors.
Probing the cancer genome
The genesis of cancer generally depends on the inactivation of one or more tumor suppressor genes in conjunction with signaling from oncogenes. Indeed, rapid advances in sequencing technologies and efforts such as The Cancer Genome Atlas (TCGA) have revealed surprisingly few “driver” mutations capable of causing cancer alone. Instead, TCGA and other sequencing studies have identified vast genetic heterogeneity both across and within tumor types, with mutations extending well beyond genes likely to represent classical oncogenes or tumor suppressor genes. How these mutations influence tumor growth remains a major unanswered question in cancer biology.
COURTESY OF DAVID LANGENAUEnter zebrafish, and a range of high-throughput reverse-genetic techniques for cancer gene discovery. By transiently overexpressing each of 30 candidate genes in zebrafish larvae, for example, Craig Ceol and Yariv Houvras in Zon’s group identified a single cooperating oncogene, SETDB1, as a new player in melanoma.6 In this study, the researchers created and analyzed more than 3,000 transgenic animals. Because they used a transposon-based transgenic approach that leads to high-level, uniform expression, they could directly assess how the injected genes affected tumor onset without going through the lengthy process of germline transgenics, a major bottleneck in mouse genetics. This rapid screening approach is a prime example of how the mountains of data generated by TCGA can be quickly assessed for biological function, and zebrafish are the only in vivo whole-animal vertebrate system that enables researchers to rapidly sift through these data to understand which mutations drive cancer.
Other new approaches are advancing loss-of-function analyses. Until recently, such studies relied on TiLLING (targeting induced local lesions in genomes), in which a chemical carcinogen, ethylnitrosourea (ENU), introduces point mutations throughout the genome and high-throughput methods then look for mutations in genes of interest. This method yielded several valuable strains of tumor-prone zebrafish harboring clinically relevant mutations in the well-known tumor-suppressor genes p53, apc, and pten, and these have been pivotal to the development of multiple zebrafish cancer models. However, the unbiased nature of ENU mutagenesis makes TiLLING a labor-intensive and impractical business in most laboratory settings. Instead, precision editing of the genome has emerged as the method of choice for the systematic creation of knockout and mutant animals. Specifically, homology-based editing, using TALENs (transcription activator–like effector nucleases) and, more recently, CRISPR (clustered regularly interspaced short palindromic repeats)/Cas systems, has revolutionized the field.7,8 Using relatively simple procedures, virtually any gene can now be mutated in zebrafish, allowing for very large-scale, in vivo assessments of novel cancer genes and the analysis of interacting mutations—one of the greatest challenges facing the cancer field in the coming decade.
COURTESY OF DAVID LANGENAUTo complement approaches that directly inactivate genes within the genome, strategies to achieve interference RNA-mediated gene silencing in zebrafish have come of age as well. Expression of short hairpin RNAs, for example, have produced stable and tissue-specific knockdown in cancer-related genes such as chordin and wnt5b.9,10 Because of the ease of manipulating the genome as well as the large number of well-characterized zebrafish gene promoters, such strategies immediately afford the opportunity to knockdown known gene functions in a tissue-specific fashion, and it is likely that temporal control will be readily achievable as well.
In all, combining the descriptive data from TCGA with zebrafish transgenesis, high-throughput overexpression and knockout techniques, and unbiased genetic screens offers an unprecedented opportunity to functionally probe the cancer genome.
The future of the field
Given its power for imaging, transplantation, small-molecule screens, and high-throughput transgenesis, the zebrafish model should become a major platform for deeply interrogating cancer biology in vivo over the next decade. One major area where zebrafish are particularly valuable is in teasing apart the extreme complexity of cancer. Because combinations of genetic pathways can be assessed simultaneously, potentially dozens of genomic alterations found in human cancer could be tested for their effects in the fish, allowing us to sort biologically meaningful alterations from neutral ones. These techniques will also allow us to understand how numerous small changes, which on their own have little phenotypic effect, can combine to cause cancer.
The success of the field will depend upon improved funding for zebrafish cancer research, however. Currently, only a small fraction of National Institutes of Health RO1 grants for cancer research are awarded for zebrafish studies, with the vast majority going to work in mice, humans, and human cells. Consortium efforts analogous to the Mouse Model Consortium will be necessary to develop more faithful zebrafish models of human cancer, which can then be used as the basis for further screens. Whereas mouse models of cancer have delivered great insights into the biological mechanisms underlying human malignancy, we view zebrafish models as a springboard for the rapid launch of unbiased genetic and chemical screens.
With any cancer model, bridging the gap between the animals and human patients is the ultimate proof of its utility. For the zebrafish, this can occur not only through bringing drugs to the clinic, but also in the development of novel biomarkers and early detection methods. The next 10 years will be an exciting time, and we have great confidence that the zebrafish will contribute major discoveries to the treatment of human cancers.
Joan K. Heath is an associate professor in the ACRF Chemical Biology Division at the Walter and Eliza Hall Institute of Medical Research and the Department of Medical Biology at the University of Melbourne, Australia, where her laboratory is studying the genetic regulation of intestinal organogenesis and colorectal cancer.
David Langenau, an assistant professor of pathology at Harvard Medical School, studies the mechanisms that drive pediatric cancer relapse within the Molecular Pathology Unit and the Cancer Center at Massachusetts General Hospital.
Kirsten C. Sadler is an assistant professor in the Division of Liver Diseases/Department of Medicine and in Developmental and Regenerative Biology at the Icahn School of Medicine at Mount Sinai in New York City, where she studies the mechanisms of liver development, regeneration, and cancer.
Richard White is an assistant professor at the Memorial Sloan-Kettering Cancer Center and Weill Cornell Medical College in New York City. His laboratory studies the evolutionary mechanisms by which tumors develop the capacity for metastasis.
- D.M. Langenau et al., “Effects of RAS on the genesis of embryonal rhabdomyosarcoma,” Genes Dev, 21:1382-95, 2007.
- D.M. Langenau et al., “Myc-induced T cell leukemia in transgenic zebrafish,” Science, 299:887-90, 2003.
- S. Nicoli et al., “Mammalian tumor xenografts induce neovascularization in zebrafish embryos,” Cancer Res, 67:2927-31, 2007.
- R.M. White et al., “Transparent adult zebrafish as a tool for in vivo transplantation analysis,” Cell Stem Cell, 2:183-89, 2008.
- R.M. White et al., “DHODH modulates transcriptional elongation in the neural crest and melanoma,” Nature, 471:518-22, 2011.
- C.J. Ceol et al., “The histone methyltransferase SETDB1 is recurrently amplified in melanoma and accelerates its onset,” Nature, 471:513-17, 2011.
- V.M. Bedell et al., “In vivo genome editing using a high-efficiency TALEN system,” Nature, 491:114-18, 2012.
- W.Y. Hwang et al., “Efficient genome editing in zebrafish using a CRISPR-Cas system,” Nat Biotechnol, doi: 10.1038/nbt.2501, 2013.
- M. Dong et al., “Heritable and lineage-specific gene knockdown in zebrafish embryo,” PLOS ONE, 4:e6125, 2009.
- G. De Rienzo et al., “Efficient shRNA-mediated inhibition of gene expression in zebrafish,” Zebrafish, 9:97-107, 2012.