The aim of this method is to generate an in vivo model of tumor angiogenesis by xenografting mammalian tumor cells into a zebrafish embryo that has fluorescently-labelled blood vessels. By imaging the xenograft and associated vessels, a quantitative measurement of the angiogenic response can be obtained.
Tumor angiogenesis is a key target of anti-cancer therapy and this method has been developed to provide a new model to study this process in vivo. A zebrafish xenograft is created by implanting mammalian tumor cells into the perivitelline space of two days-post-fertilization zebrafish embryos, followed by measuring the extent of the angiogenic response observed at an experimental endpoint up to two days post-implantation. The key advantage to this method is the ability to accurately quantitate the zebrafish host angiogenic response to the graft. This enables detailed examination of the molecular mechanisms as well as the host vs tumor contribution to the angiogenic response. The xenografted embryos can be subjected to a variety of treatments, such as incubation with potential anti-angiogenesis drugs, in order to investigate strategies to inhibit tumor angiogenesis. The angiogenic response can also be live-imaged in order to examine more dynamic cellular processes. The relatively undemanding experimental technique, cheap maintenance costs of zebrafish and short experimental timeline make this model especially useful for the development of strategies to manipulate tumor angiogenesis.
Angiogenesis is one of the classic hallmarks of cancer and represents a target of anti-cancer therapy1,2. To study this process, xenograft models of cancer have been created by implanting mammalian tumor cells into animals such as mice3. A zebrafish xenograft model has also been developed, which involves the implantation of tumor cells into 2 days post fertilization (dpi) zebrafish which results in rapid growth of zebrafish blood vessels into the xenograft4.
This protocol describes an in vivo zebrafish embryo tumor xenograft model in which the angiogenic response can be accurately quantitated across the entire xenograft. This method allows the investigator to examine, in vivo, the molecular mechanisms that underpin the tumor angiogenic response. The genetic tractability of the zebrafish allows the host contribution to be interrogated, while selection of different tumor cell lines allows the tumor contribution to angiogenesis to also be examined5,6,7. In addition, as zebrafish larvae are permeable to small molecules, specific pathway inhibitors can be used or drug libraries can be screened to identify novel inhibitors of tumor angiogenesis8,9,10,11.
The zebrafish embryo xenograft model presents unique advantages compared with other mammalian xenograft models. Zebrafish xenografts are cheaper and easier to perform, large numbers of animals can be examined and live cell imaging allows detailed examination of cell behaviour4. Unlike other in vivo models, which require up to several weeks to observe significant vessel growth, angiogenesis in zebrafish xenografts can be observed within 24 h following implantation3,4. However, the lack of an adaptive immune system in embryonic zebrafish, while beneficial to maintaining the xenograft, means that the adaptive immune response and its contribution towards tumor angiogenesis cannot be examined. In addition, the lack of tumor stromal cells, the inability to orthotopically implant the tumor and the difference in maintenance temperature between zebrafish and mammalian cells are potential weaknesses of this method. Nonetheless, the amenability of this model for live imaging and the ability to accurately quantitate the angiogenic response makes it uniquely beneficial for studying the cellular processes that regulate tumor angiogenesis in vivo.
1. Preparation of Microinjection Needles
2. Cell Culture for Implantation
NOTE: When using this protocol, any mammalian cancer cell line can be used for implantation into zebrafish embryos as xenografts. However, there is much variation in the angiogenic response induced among different cell lines5,11,12. B16-F1 murine melanoma cells have been shown to induce a strong angiogenic response in zebrafish embryos11 and are therefore appropriate for use in this protocol.
3. Labelling B16-F1 Cells with Fluorescent Dye
NOTE: In order to differentiate between the implanted tumor cells and other cells in the embryo, the tumor cells must be labelled with an appropriate fluorescent dye before implantation. This step can be skipped if the cells already express fluorescent reporters.
4. Preparation of Embryos for Implantation
NOTE: Choose a transgenic zebrafish line that has fluorescently labelled blood vessels (e.g. kdrl:RFP, fli1a:EGFP, etc.)13,14.
5. Perivitelline Injection of Mammalian Cancer Cells into 2 dpf Embryos
NOTE: To ensure the cells clump together as a graft when implanted, the cells must be mixed together with an extracellular matrix mixture (ECM). We have described the steps of making such a cell/ECM mixture when using an ECM mixture referenced in the Table of Materials. If an alternative matrix is used, the steps should be adjusted accordingly.
6. Live Imaging
7. Time Lapse Imaging
NOTE: This model is highly suited to imaging dynamic cellular processes due to its transparency and the availability of zebrafish transgenics that fluorescently label different cell types. This makes time-lapse imaging a key application of this model.
8. Quantitation of the Angiogenic Response to the Zebrafish Xenograft
NOTE: The following steps use 3D image analysis software. Specific steps will vary depending on the software used.
By imaging an individual xenograft at 6, 24 and 48 hpi, the angiogenic response at different timepoints can be calculated as shown in Figure 1A-C. The largest angiogenic response is observed between 24-48 h post implantation, with the maximum levels of graft vascularization seen around 2 dpi (Figure 1A-C). A time-lapse movie of a typical angiogenic response to a B16-F1 xenograft from 20.75 hpi until 46 hpi is seen in Supplementary Movie 1. This movie was generated by imaging a xenograft implanted following this protocol into a 2 dpf kdrl:EGFP embryo. The vessels that grow through the xenograft form a tortuous network with vessels of irregular size and morphology, which is typical of the abnormal vascular network seen in mammalian tumors20. The angiogenic response in our model is a result of vessels that sprout from the common cardinal vein into the xenograft as opposed to previous studies, which have reported the sub-intestinal veins as the source of the vessels that grow into the xenograft4. The difference in vessel origin is due to the fact we have implanted our xenografts in a more ventral location in the perivitelline space, as this makes it easier to visualize the shape and size of the xenograft while injecting and imaging20.
Figure 1D-F shows representative results of B16-F1 xenografts and their associated vasculature following treatment with either DMSO (control) or 50 nM of a VEGFR inhibitor (Tivozanib)21,22. Immediately after implantation, xenografted embryos were split into two groups and either 50 nM Tivozanib or DMSO (0.5% v/v) was applied to each group. They were maintained in these drugs before imaging at 2 dpi. These results show that the vessels associated with xenografts incubated in the VEGFR inhibitor (Figure 1E) are far less numerous than the vessels associated with xenografts incubated in DMSO (Figure 1D). In control embryos, there is an expansive vascular network that stretches across the entire xenograft region (Figure 1D), which is the typical angiogenic response observed following implantation of B16-F1 cells. In Figure 1F, the angiogenic response has been quantitated by following this protocol and it shows a clear reduction in graft vascularization in the VEGFR inhibitor-treated xenografts when compared to controls. Despite normalizing the levels of graft vascularization against graft volume, there remains a large variation in the levels of graft vascularization at 48 hpi (coefficient of variation of 54.2%). This is due to the variation in the total volume of the tumor vessels (coefficient of variation 73.3%); we observed that even similarly-sized grafts had large variation in the level of vascularization. Because of this, it is recommended that around 20 xenografts are used per treatment group.
The steps taken to quantitate the vessels associated with the xenograft from Figure 1D are shown in Figure 2. The tumor channel is shown alone in Figure 2A, where a mass of B16-F1 cells fluorescently-labelled (false-colored green) can be seen below the yolk sac, which displays autofluorescence. An ROI must carefully be drawn around the xenograft (Figure 2A'), taking care not to include any of the autofluorescence from the yolk sac as the Tumor Volume protocol may then identify this autofluorescence as part of the tumor. Performing the Tumor Volume protocol identifies all the B16-F1 cells in the ROI, measures their volume and clips the ROI to their volume (Figure 2A''). Performing the Tumor Vessel Volume protocol in this new clipped ROI allows the identification of all the vessels associated with the mass of B16-F1 cells and measures their volume (Figure 2A'''). The values from the Tumor Volume and Tumor Vessel Volume protocols can be used to give a measure of graft vascularization which is then plotted in a graph as seen in Figure 1F.
Figure 1: Zebrafish tumor xenograft vascularization is inhibited by a VEGFR signaling inhibitor. (A-C) Confocal images of a live embryo at 6 hpi (A), 24 hpi (B) and 48 hpi (C), with GFP-labelled blood vessels (magenta). The % graft vascularization was calculated and included in the corresponding panel. (D-E) Confocal images of live zebrafish larvae shown at 48 hpi with fluorescently-labelled B16-F1 xenografts (green) and GFP-labelled blood vessels (magenta) which were treated immediately after implantation with either 0.5% (v/v) DMSO in E3 (D) or 50 nm of a VEGFR inhibitor (Tivozanib) in E3 (E). The vessel channels for D and E are shown in isolation in D' and E', respectively. (F) Graph displaying the data from quantifying the % graft vascularization at 48 hpi in these xenografts, n = 42 (DMSO), n = 20 (VEGFR inhibitor). **p>0.01 by Mann-Whitney test, error bars represent s.d. Scale bar = 50 μm. Data contained in Figure 1F is reproduced11.
Figure 2: Quantifying graft vascularization. Confocal image of live 2 dpi zebrafish larvae with fluorescently-labelled B16-F1 xenografts (green) and GFP-labelled blood vessels (magenta). (A) Tumor channel prior to drawing of a ROI. (A') An ROI is drawn around the xenograft region, ensuring that the autofluorescence in the yolk is not included. (A'') The “Tumor Volume” protocol is performed to identify the volume of xenograft inside the ROI and also creates a new ROI. (A''') The “Tumor Vessel Volume” protocol is used to identify the volume of vessels inside the new ROI. Scale bar = 50 μm.
Supplementary Movie 1: Tumor xenograft angiogenesis. Confocal time-lapse imaging of a B16-F1 xenograft implanted into a 2 dpf zebrafish embryo with GFP-labelled blood vessels (kdrl:EGFP). Movie taken from 20.75 hpi to 46 hpi. Time-lapse image z-stacks were collected 10 min apart; movie was made at 7 frames per second. Scale bar = 50 μm. Please click here to download this file.
The first critical step in the protocol is the implantation of tumor cells. It is essential that cells are injected into a location that will allow the xenograft to implant successfully in the embryo without making the embryo edematous. An injection that is too anterior can allow the cells to move towards the heart, blocking the bloodstream and leading to edema, while an injection that is too posterior will result in a poorly implanted xenograft. An anterior injection is best avoided by inserting the needle through the yolk sac in an anterior to posterior direction so that it is pointing away from the heart as it injects. A posterior injection is best avoided by injecting the cells at a location in the anterior half of the round part of the yolk sac. In addition, cells must be injected ventral to the yolk sac and not lateral to it, as the lateral location is more difficult to view and subsequently image. Once the researcher is reasonably adept at this process, approximately 200-300 embryos can be implanted with xenografts each day. Imaging of the angiogenic response will take longer due to the time taken to mount and image the xenografts; it takes approximately one hour to image 15-20 xenografts using a standard laser-scanning confocal microscope.
The second critical step is the measurement of graft vascularization using 3D image analysis software. Quantitation by tumor volume is required as it is difficult to obtain consistently-sized xenografts through microinjection. It is necessary to calibrate the settings of the software protocols carefully and use them consistently in order to obtain accurate and unbiased quantitative measurements for all experiments. Setting the minimum threshold for intensity (for both Tumor Volume and Tumor Vessel Volume) is an important part of this step as it allows the protocol to correctly discriminate between fluorescence that is part of the tumor/vessels and background fluorescence. Drawing the ROI around the tumor xenograft to include only the tumor xenograft and not any bright autofluorescing parts of the embryo (such as the yolk sac) is also important; the accidental inclusion of any non-tumor autofluorescing regions or non-tumor blood vessels into the ROI will result in these parts being measured as part of "Tumor volume" or "Tumor Vessel volume", creating significant inaccuracies in the final calculation. There is always a large variation in the angiogenic response (see Figure 1F), however, rather than being a limitation of the model, this may merely reflect the natural variance in tumor vessel density observed in both human patients and murine xenografts23,24. The quantitation of graft vascularization in 60 xenografts will take approximately 1 h.
Careful selection of tumor cell line is also important as there is a wide variation in the angiogenic response induced between different mammalian cell lines, which may relate to the differing levels of pro-angiogenic molecules produced by these cells4,11. In our hands, B16-F1 mouse melanoma cells give a robust angiogenic response, possibly due to the high level of VEGFA secretion from these cells11. Possible modifications to this protocol would be the use of zebrafish melanoma cells, which would allow the assay temperature to be lowered to 28° and therefore be optimal for both the host and tumor25 or the use of cancer cells that overexpress pro-angiogenic growth factors such as FGF4,7,20.
Some of the advantages of this model lie in its short time span, low cost, and the relative ease with which the xenotransplantation procedure can be conducted compared with other in vivo models such as murine xenografts, all of which make it highly amenable to be used for large scale studies (i.e., chemical screens for anti-angiogenic compounds26). The ability to live image the model and the availability of transgenic fish with fluorescently labelled blood vessels and other cell types allows researchers to study the details of the angiogenesis process (such as vessel sprouting) as well as the cell-cell interactions (such as macrophage-endothelial interactions) that take place during angiogenesis in this model11,20,27. As vessel normalization is also a key objective of antiangiogenic therapy, high-resolution live-imaging of the vascular network in the xenografts means that this model has the potential to identify treatments that are directed at this objective. Network tortuosity could be examined by counting the number of vessel branchpoints, while vessel morphology could be examined by measuring variation in vessel width, allowing pro-normalization treatments to be tested and evaluated in this model28. In addition, fluorescence microangiography can be used to determine the patency and integrity of tumor vessels29. The genetic tractability of the zebrafish also means that it can be genetically modified to examine the role of various host signaling pathways in tumor angiogenesis6. This model can therefore be used to identify novel anti-angiogenic compounds that are active in a tumor setting as well as studying the cellular and molecular interactions that regulate tumor angiogenesis.
The authors have nothing to disclose.
We thank Mr. Alhad Mahagaonkar for managing the University of Auckland zebrafish facility and the Biomedical Imaging Research Unit, School of Medical Sciences, University of Auckland, for assistance in time-lapse confocal microscopy. This work was supported by a Health Research Council of New Zealand project grant (14/105), a Royal Society of New Zealand Marsden Fund Project Grant (UOA1602) and an Auckland Medical Research Foundation Project Grant (1116012) awarded to J.W.A.
Air cylinder | BOC | 011G | Xenotransplantation |
B16-F1 cells | ATCC | Cell culture | |
BD Matrigel LDEV-free (extracellular matrix mixture) | Corning | 356235 | Xenotransplantation |
Borosillicate glass capillaries | Warner Instruments | G100T-4 | OD=1.00 mm, ID=0.78 mm, Length =10 cm Cell injection |
Cell culture dish -35 mm diameter | Thermofisher NZ | NUN153066 | Fish husbandry |
Cell culture dish -100 mm diameter | Sigma-Aldrich | CLS430167-500EA | Fish husbandry |
Cell culture flask 75 cm2 | In Vitro Technologies | COR430641 | Cell culture |
CellTracker Green | Invitrogen | C2925 | Cell labelling, Stock concentration (10 mM in DMSO), working concentration (0.2 μM in serum-free media) |
Dimethyl sulfoxide | Sigma-Aldrich | D8418 | Drug treatment, Cell labelling |
E3 Media (60x in 2 L of water) 34.8 g NaCl 1.6 g KCl 5.8 g CaCl2·2H2O 9.78 g MgCl2·6H2O adjust to pH 7.2 with NaOH |
In house [1] | Fish husbandry | |
Ethyl-3-aminobenzoate methanesulfonate (Tricaine) | Sigma-Aldrich | E10521 | Xenotransplantation, Imaging |
Filter tip 1000 μL | VWR | 732-1491 | Used during multiple steps |
Filter tip 200 μL | VWR | 732-1489 | Used during multiple steps |
Filter tip 10 μL | VWR | 732-1487 | Used during multiple steps |
Fluorescence microscope | Leica | MZ16FA | Preparation of embryos |
FBS (NZ origin) | Thermofisher Scientific | 10091148 | Cell culture |
Gloves | Any commercial brand | Used during multiple steps | |
Haemocytometer cell counting chamber Improved Neubauer | HawksleyVet | AC1000 | Xenotransplantation |
Heraeus Multifuge X3R Centrifuge | Thermofisher Scientific | 75004500 | Cell culture, Cell labelling |
Hoechst 33342 | Thermofisher Scientific | 62249 | Cell labelling, Stock concentration (1 mg/ml in DMSO), working concentration (6 μg/ml in serum-free media) |
Low Melting Point, UltraPure Agarose | Thermofisher Scientific | 16520050 | Imaging |
Methycellulose | Sigma-Aldrich | 9004 67 5 | Xenotransplantation |
Methylene blue | sigma-Aldrich | M9140 | Fish husbandry |
Microloader 0.5-20 μL pipette tip for loading microcapillaries | Eppendorf | 5242956003 | Xenotransplantation |
Micropipettes | Any commercial brand | Used during multiple steps | |
Micropipette puller P 87 | Sutter Instruments | Xenotransplantation | |
Microscope cage incubator | Okolab | Time-lapse imaging | |
Microwave | Any commercial brand | Imaging | |
Mineral oil | Sigma-Aldrich | M3516 | Xenotransplantation |
Minimal Essential Media (MEM) – alpha | Thermofisher Scientfic | 12561056 | Cell Culture |
MPPI-2 Pressure Injector | Applied Scientific Instrumentation | Xenotransplantation | |
Narishige micromanipulator | Narishige Group | Xenotransplantation | |
Nikon D Eclipse C1 Confocal Microscope | Nikon | Imaging | |
N-Phenylthiourea (PTU) | Sigma-Aldrich | P7629 | Fish husbandry |
PBS | Gibco | 10010023 | Cell culture |
Penicillin Streptomycin | Life Technologies | 15140122 | Cell culture |
S1 pipet filler | Thermoscientific | 9501 | Cell culture |
Serological stripette 10 mL | Corning | 4488 | Cell culture |
Serological stripette 25 mL | Corning | 4489 | Cell culture |
Serological stripette 5 mL | Corning | 4485 | Cell culture |
Serological stripette 2 mL | Corning | 4486 | Cell culture |
Terumo Needle 22 gauge | Amtech | SH 182 | Fish husbandry |
Tissue culture incubator | Thermofisher Scientfic | HeraCell 150i | Cell culture |
Tivozanib (AV951) | AVEO Pharmaceuticals | Drug treatment | |
Transfer pipette 3 mL | Mediray | RL200C | Fish husbandry |
Trypsin/EDTA (0.25% ) | Life Technologies | T4049 | Cell culture |
Tweezers | Fine Science Tools | 11295-10 | Fish husbandry |
Volocity Software (v6.3) | Improvision/Perkin Elmer | Image analysis |