Date Published: July 8, 2012
Publisher: Hindawi Publishing Corporation
Author(s): Finola E. Moore, David M. Langenau.
Zebrafish have emerged as a powerful model of development and cancer. Human, mouse, and zebrafish malignancies exhibit striking histopathologic and molecular similarities, underscoring the remarkable conservation of genetic pathways required to induce cancer. Zebrafish are uniquely suited for large-scale studies in which hundreds of animals can be used to investigate cancer processes. Moreover, zebrafish are small in size, optically clear during development, and amenable to genetic manipulation. Facile transgenic approaches and new technologies in gene inactivation have provided much needed genomic resources to interrogate the function of specific oncogenic and tumor suppressor pathways in cancer. This manuscript focuses on the unique attribute of labeling leukemia cells with fluorescent proteins and directly visualizing cancer processes in vivo including tumor growth, dissemination, and intravasation into the vasculature. We will also discuss the use of fluorescent transgenic approaches and cell transplantation to assess leukemia-propagating cell frequency and response to chemotherapy.
Zebrafish models of hematological malignancies exhibit striking similarities with human and mouse disease [1–7], yet afford unique avenues of study due to imaging modalities that permit direct visualization of fluorescently labeled blood cells within live animals. As with mouse and human disease, zebrafish leukemias are distinguished from lymphomas by the infiltration of leukemic cells into the marrow. Lymphomas are predominantly located as masses throughout the body, including lymph nodes in mouse and human, and have no or little infiltration into the marrow . Leukemias are also classified as acute or chronic. Acute leukemias are arrested at early stages of maturation, are highly proliferative, and advance quickly in patients . By contrast, chronic leukemias are arrested at later stages of maturation and resemble functional, yet abnormal, blood cell counterparts. Although characterized by increased circulating white blood counts, chronic leukemias are often much slower growing and take months or years to progress. Leukemias can be further subdivided based on the blood lineage in which cells have become transformed . To date, zebrafish models of Acute Lymphoblastic Leukemias (ALL), Acute Myeloid Leukemia (AML), and Myeloproliferative Neoplasms (MPN) have been described.
Many studies have employed the use of stable transgenic zebrafish to drive oncogenic transgene expression in a tissue-specific manner including pancreatic adenocarcinoma , hepatocellular carcinoma , melanoma [18–20], embryonal rhabdomyosarcoma , and leukemia. By and large, investigators have used oncogene fusions with GFP to create tumors that are fluorescently labeled. For example, we and others have generated EGFP-Myc, NOTCH1-GFP, EGFP-TEL-AML1, and MYST3/NCOA2-EGFP fusions to drive leukemogenesis while also fluorescently labeling leukemic cells [2, 6, 7, 12]. Although these approaches have been largely successful in generating fluorescently labeled leukemias, it is worth noting that fluorescent protein expression is linked with oncogene localization within the cell and protein stability. For example, MYC is a nuclear transcription factor with a half-life of ~30 minutes in non-transformed cells. Thus, the EGFP-MYC fusion protein is rapidly turned over in normal thymocytes prior to GFP maturation into a functional fluorescent molecule, precluding the use of fluorescence to identify stable transgenic Tg(rag2:EGFP-Myc) animals at 5 days of life. However, the EGFP-Myc transgene is stabilized following transformation leading to weak, nuclear fluorescent protein expression in T-ALL. Fluorescent protein fusions can also exhibit reduced transforming activity depending on cellular context. For example, we have developed a zebrafish model of kRASG12D-induced embryonal rhabdomyosarcoma but have been unable to model this disease using the same transgene promoter to drive expression of a GFP fusion with kRASG12D. By contrast, others have used similar RAS fusion constructs to generate fluorescently labeled hepatocellular carcinoma, pancreatic adenocarcinoma, and melanoma [16, 17, 19, 20]. To obviate issues surrounding the function of fluorescent protein-oncogene fusions, it is possible to utilize dual transgenic approaches to drive both the oncogene and fluorescent protein within the same cell types. For example, Tg(rag2:Myc) lines could be bred to Tg(rag2:GFP) fish. The resulting progeny would develop T-ALL that expresses high fluorescent protein expression.
Investigators have utilized cell transplantation of fluorescently labeled cancer cells into sublethally irradiated adult zebrafish to assess tumorigenicity . For example, Traver et al. optimized cell transplantation of both blood and leukemic cells into gamma-irradiated animals [7, 25]. Specifically, recipient fish were irradiated with 20–25 Gy two days prior to cell transplantation and then injected with fluorescently labeled donor cells into the peritoneal cavity or sinus venosis. For T-ALL, animals can be injected with 1 × 106 cells and assessed for fluorescently labeled leukemia engraftment at 10 days posttransplantation [7, 25]. Imaging of engraftment can be further facilitated by transplantation into optically clear strains of zebrafish that lack iridiphores and melanocytes—aptly named casper . Casper fish were created by breeding together roy and nacre mutants and must be maintained as double homozygous mutant animals. These fish are transparent as adults, facilitating detailed imaging of cell migration, metastasis, and kinetics of tumor growth. For example, recent work has shown that blood cells can be tracked and counted within the circulation of live adult fish using an integrated optical system that combines a laser scanning confocal microscope and an in vivo flow cytometer .
Blood cells and their dynamic cell movements can be easily visualized in live fluorescent transgenic zebrafish. For example, researchers have tracked the migration of various blood lineages including erythroid and macrophage progenitors [25, 32–34]. Importantly, hematopoietic stem cell (HSC) movement can also be followed in Tg(CD41:eGFP), Tg(cmyb:GFP), Tg(runx1:GFP), and Tg(lmo2:GFP) transgenic zebrafish larvae [35–40]. Moreover, fluorescently labeled blood cells can also be tracked in adult fish [27, 41]. Capitalizing on cell transplantation approaches, investigators have also utilized fluorescence imaging to visualize normal hematopoietic cell homing in live animals. For example, Bertrand et al. visualized HSC homing to the caudal hematopoietic tissue by transplanting Tg(CD41:eGFP; gata1:dsRed) cells into irradiated recipients . We have also described the homing of Tg(lck:GFP)+ T cells back to the thymus following transplantation of cells into larval wildtype fish . While malignant GFP+ T-ALL lymphoblasts also migrate to the thymus, they exhibit robust and specific homing to the olfactory bulb [6, 7]. These studies demonstrate the ease of visualizing cell migration and homing to specific anatomically defined sites within live animals using fluorescently labeled normal hematopoietic and leukemic cells.
Fluorescence imaging of transplanted cancer cells can also be used to visualize response to chemotherapy and radiation. For example, the Revskoy group recently showed that GFP-labeled T-ALL cells could be serially transplanted into syngeneic strain larvae . Treatment of transplant recipients with vincristine or cyclophosphamide reduced tumor burden (Figure 3) and extended lifespan significantly . These experiments established that high-throughput cell transplantation assays can generate large cohorts of animals for drug screens and showed that zebrafish T-ALL responds to the same drugs that are used to treat human T-ALL patients . In addition, fluorescently labeled cells can be assessed for response to radiation. For example, we have shown that engrafted GFP-labeled T-ALLs that coexpress EGFP-bcl2 and the Myc transgene failed to undergo apoptosis following 20 Gy of gamma-irradiation ; however, T-ALLs that express only Myc were ablated by 4 days postirradiation, suggesting that Myc-induced T-ALL have an intact p53 DNA damage pathway.
Leukemia-propagating cells (LPCs) have the capacity to produce all the other cell types contained within the leukemia, are responsible for continued tumor growth, and ultimately drive relapse. Investigators have used fluorescence-activated cell sorting (FACS) to identify unique cell populations and limiting dilution cell transplantation to assess if molecularly defined leukemia cells retain LPC activity in human disease. For example, in AML a rare CD34+, CD38− cell enriches for leukemia-propagating potential [45, 46]. In T-ALL, it has been suggested that CD34+ CD7+ cell populations are enriched in LPCs . Despite enormous efforts aimed at defining if and what cell surface markers define LPC activity, relatively little is known about the molecular mechanisms that drive leukemia propagating activity. For example, elegant work from Jean Soulier’s group has shown xenograft transplantation of primary human T-ALL into immune-compromised mice selected for a small subset of clones found within the diagnosis leukemia . These clones contained specific genomic lesions that likely increase leukemia aggression and increase the frequency of LPCs within the bulk of the leukemia mass . Yet, despite the identification of recurrent genomic changes associated within continued clonal evolution, the mechanisms driving these relapse-associated processes are largely unknown.
Zebrafish has fast emerged as a powerful model of leukemia. When coupled with fluorescent transgenic approaches and powerful imaging techniques, these models are uniquely positioned to uncover mechanisms driving tumor dissemination, progression, and relapse. Moreover, the use of multifluorescent transgenic animals will allow for labeling of tumor cell compartments similar to those defined in RAS-induced rhabdomyosarcoma models [21, 50] and for the visualizing of leukemia growth in relation to supportive cell types including vasculature, fibroblasts, and macrophages. Moreover, though not the focus of this paper, cell transplantation approaches that utilize fluorescently labeled, human leukemia cells into either zebrafish embryos or adults will likely provide novel experimental models to assess tumor growth and response to therapy [51–60], capitalizing on the numbers of disease animals that can be created by microinjection and direct visualization of tumor growth in vivo.