The zebrafish/tumor xenograft angiogenesis assay

PROTOCOL

The zebrafish/tumor xenograft angiogenesis assay Stefania Nicoli & Marco Presta General Pathology and Immunology, Department of Biomedical Sciences and Biotechnology, University of Brescia Medical School, Viale Europa 11, 25123 Brescia, Italy. Correspondence should be addressed to M.P. ([email protected]).

© 2007 Nature Publishing Group http://www.nature.com/natureprotocols

Published online 8 November 2007; doi:10.1038/nprot.2007.412

Here we describe a method to study tumor angiogenesis in zebrafish (Danio rerio) based on the injection of proangiogenic mammalian tumor cells into the perivitelline space of zebrafish embryos at 48 h post-fertilization. Within 24–48 h, proangiogenic tumor grafts induce a neovascular response originating from the developing subintestinal vessels. This can be observed at macroscopic and microscopic levels after whole-mount alkaline phosphatase staining of wild-type zebrafish embryos, or by fluorescence microscopy in transgenic VEGFR2:G-RCFP embryos in which endothelial cells express the green fluorescent protein under the control of the VEGFR2/KDR promoter. Angiogenesis inhibitors added to the injected cell suspension or to the fish water prevent tumor-induced neovascularization. The assay is rapid and inexpensive, representing a novel tool for investigating tumor angiogenesis and for antiangiogenic drug discovery. Also, gene inactivation by antisense morpholino oligonucleotides injection in zebrafish embryos may allow the identification of genes involved in tumor angiogenesis.

INTRODUCTION Angiogenesis plays a key role in tumor growth and metastasis1. Angiogenic growth factors released by tumor cells, including members of the vascular endothelial growth factor (VEGF)2 and fibroblast growth factor (FGF)3 families, are responsible for tumor neovascularization. In its absence, the tumor remains in a state of dormancy1. Thus, the identification of antiangiogenic drugs and of angiogenesis-related targets may have significant implications for the development of antineoplastic therapies, as shown by the positive outcomes in the treatment of cancer patients with the monoclonal anti-VEGF antibody bevacizumab2. The use of tumor cell syngrafts or xenografts in animal models may allow continuous delivery of angiogenic factors produced by a limited number of tumor cells, thus mimicking the initial stages of tumor angiogenesis and metastasis. Various animal models have been developed in rodents and in the chick embryo to investigate the angiogenesis process and for screening pro and antiangiogenic compounds, each with its own unique characteristics and disadvantages4. The teleost zebrafish (Danio rerio) represents a promising alternative model in cancer research5. When compared to other vertebrate model systems, zebrafish offers many advantages, including ease of experimentation, drug administration, amenability to in vivo manipulation and feasibility of reverse and forward genetic approaches6. Also, zebrafish possesses a complex circulatory system similar to that of mammals, and the optical transparency and ability to survive for 3–4 d without functioning circulation make the zebrafish embryo amenable for vascular biology studies7. Recent studies have shown the feasibility of injecting human melanoma cells in zebrafish embryos to follow their fate and study their impact on zebrafish development8. In these studies, tumor cells were injected at the blastula stage to explore potential bidirectional interactions between cancer cells and embryonic stem cells. The results indicate that developing zebrafish can be used as a biosensor for tumor-derived signals. However, grafting of tumor cells at this stage, well before vascular development, results in their reprogramming toward a non-tumorigenic phenotype, thus hampering any attempt to investigate tumor-driven vascularization. At variance, injection of melanoma cells into the hindbrain 2918 | VOL.2 NO.11 | 2007 | NATURE PROTOCOLS

ventricle or yolk sac of 48 h post-fertilization (hpf) embryos results in the formation of tumor masses within 4 d9. Immunostaining analysis of the grafts reveals the presence of blood vessels within the brain and abdominal lesions, even though the high vascularity of the invaded regions may not allow easy discrimination between developmental and tumor-induced angiogenesis9. Here, we describe a novel experimental procedure to study tumor angiogenesis triggered by mammalian tumor cells grafted in 48 hpf zebrafish embryos. Cells are injected in the perivitelline space of the embryo, a region that remains avascular throughout the experimental period in the absence of an exogenous stimulus, and the experiment is concluded 24–48 h later. This allows a simple and unambiguous identification of newly formed ectopic vessels driven by the tumor graft in live embryos and their characterization by immunohistochemistry and whole-mount mRNA in situ hybrydization. This method is suitable for assessing the effect of antiangiogenic chemicals and for the identification of non-redundant gene products involved in tumor angiogenesis by antisense morpholino oligonucleotide strategies. The basic vascular plan of the developing zebrafish embryo shows strong similarity to that of other vertebrates10. At the 13somite stage, endothelial cell precursors migrating from the lateral mesoderm originate the zebrafish vasculature, and a single blood circulatory loop is present at 24 hpf. Blood vessel development continues during the subsequent days by angiogenic processes. In particular, at 48 hpf, subintestinal veins (SIVs) originate from the duct of Cuvier. During the next 24 h, SIVs will form a vascular plexus across most of the dorsal–lateral aspect of the yolk ball (Fig. 1) and will provide, together with the supraintestinal artery, blood supply to the digestive system. The zebrafish/tumor xenograft angiogenesis assay described here is based on the grafting of mammalian tumor cells in the proximity of the developing SIV plexus at 48 hpf. Proangiogenic factors released locally by the tumor graft will affect the normal developmental pattern of the SIVs by stimulating the migration and growth of sprouting vessels toward the implant. Because of the expected variability frequently observed in any in vivo assay, at least 20 embryos should be injected per experimental

PROTOCOL a SIV development 2d

2.5 d

3d


b

c

Figure 1 | Normal SIV development in zebrafish embryo. (a) Schematic representation of SIVs (in green) at different stages of zebrafish embryo development. Lateral (b) and dorsal (c) view of a 3-d-old zebrafish embryo (head on the left) after whole-mount alkaline phosphatase staining. The arrowheads point toward the normally developed SIV plexus; original magnification: 11.5.

point. The timing of analysis of the angiogenic response is also of importance, best results being obtained 24–48 h after tumor cell grafting, depending on the cell line used. At these times, macroscopic evaluation of the angiogenic response can be performed, even though histological evaluation of vessel density represents an unbiased measurement of vascularity. The use of transgenic VEGFR2:G-RCFP zebrafish embryos, in which endothelial cells express the green fluorescent protein (GFP) under the control of the VEGFR2/KDR promoter11, may represent an improvement of the method, allowing the observation and time-lapse recording of newly formed blood vessels in live embryos by epifluorescence microscopy as well as by in vivo confocal microscopy. Advantages and disadvantages When compared to other in vivo tumor angiogenesis assays, the zebrafish/tumor xenograft model presents both advantages and disadvantages that should be considered: The zebrafish/tumor xenograft model allows the delivery of a very limited number of cells (as low as 1,000–2,000 cells per embryo), thus mimicking the initial stages of tumor angiogenesis and metastasis. Labeled tumor cells (e.g., GFP-transduced or fluorescent dyeloaded cells) can be easily visualized within the embryo. Thus, analysis of the spatial/temporal relationship among tumor cells and newly formed blood vessels may represent an important feature of this model.

Several techniques can be applied within the constraints of paraffin or gelatin embedding, including histochemistry and immunohistochemistry. Electron microscopy can also be used in combination with light microscopy. Whole-mount mRNA in situ hybridization can be used to study the expression of selected genes in newly formed blood vessels. Moreover, reverse transcriptase-polymerase chain reaction analysis with speciesspecific primers allows the study of gene expression by grafted tumor cells and by the host under different experimental conditions12. Because of the immaturity of the immune system in zebrafish embryos at 48–72 hpf, no graft rejection occurs at this stage. However, 5–6 d after injection, xenografts interfere with zebrafish development and the larvae eventually die, thus hampering the possibility to perform long-term studies on tumor vascularization in this model. Also, it should be pointed out that zebrafish embryos are maintained at 28 1C. This may not represent an optimal temperature for mammalian cell growth and metabolism, even though we have observed mitotic figures with no sign of apoptosis in grafted tumors throughout the experimental period12. In this respect, the possibility to raise the incubation temperature up to 35 1C with no apparent gross effects on zebrafish development has been reported9, even though the activation of stress response(s) cannot be ruled out. When compared to the rabbit cornea and chick embryo chorioallantoic membrane assays, the zebrafish/tumor xenograft model has shown a similar capacity to discriminate between highly angiogenic and poorly angiogenic tumor cell lines (Table 1 and see ref. 12). However, only parallel screening of a variety of cell lines will establish whether the zebrafish-based assay may provide results fully over-imposable with those obtained with the classic rodent and chick embryo angiogenesis assays. Because of the permeability of its embryos to small molecules, zebrafish allows disease-driven drug target identification and in vivo validation, thus representing an interesting bioassay tool for small-molecule testing and dissection of biological pathways, alternative to other vertebrate models13. Accordingly, systemic exposure of live zebrafish embryos to antiangiogenic compounds dissolved in fish water results in a significant inhibition of developmental neovascularization14 and angiogenesis triggered by the tumor graft12. Thus, the zebrafish/tumor xenograft model may represent a short-term assay suitable for the identification of novel tumor angiogenesis inhibitors. Clearly, any in vivo animal assay based on the grafting of tumor cells, including the zebrafish/tumor xenograft assay, is labor-intensive and timeconsuming. On the other hand, physiological and tumor angiogenesis differ and simpler models based on the study of the effect

TABLE 1 | Comparison between different tumor angiogenesis assays. Animal model Zebrafish embryo12 Rabbit cornea18 Chick embryo CAM21

Vehicle 0/15 0/12 0/10

MAE cells 2/40 5/12 0/10

FGF2-T-MAE cells 60/60 10/12 10/10

Parental, non-tumorigenic murine aortic endothelial (MAE) cells and highly tumorigenic, FGF2overexpressing FGF2-T-MAE cells18 were grafted in zebrafish embryos, avascular rabbit cornea and chick embryo chorioallantoic membrane (CAM). The table summarizes the findings of the zebrafish/tumor xenograft angiogenesis assay compared with those obtained with the other animal models12,18,21. Data are shown as the number of positive implants/total number of implants.

NATURE PROTOCOLS | VOL.2 NO.11 | 2007 | 2919


PROTOCOL

of chemicals on developmental angiogenesis may provide erroneous information. The zebrafish/tumor xenograft model is characterized by a rapid response to angiogenesis inhibitors (24–48 h) when compared to other tumor graft/angiogenesis assays, including the chick embryo chorioallantoic membrane assay (3–4 d), the s.c. murine Matrigel plug assay (5–7 d), the murine (1 week) and rabbit (2–3 weeks) cornea assays and the s.c. mouse syngraft and xenograft assays (several weeks)4. A trained laboratory technician can inject a large number of zebrafish embryos (200–250 embryos per day) that are maintained in 96-well plates, thus allowing systemic in vivo treatment of the animals with minimal amounts of compound. Therefore, dose–response experiments can be easily performed and numerous compounds can be tested in an effective manner. This is usually not feasible in most of the laboratories when using mammalian models (mice, rats and rabbits). Last but not least, a zebrafish facility is much cheaper and its logistics is much simpler than a mammalian facility. Thus, despite some challenging technical aspects described below, the zebrafish/tumor xenograft assay appears to be suitable for a medium-scale throughput screening of putative angiogenesis inhibitors. A possible drawback of the zebrafish/tumor xenograft model may be represented by the possibility that the metabolic fate of the drug (either in terms of its activation or inactivation) may differ in zebrafish embryos compared to mammalian species. At variance with low-molecular-weight compounds, highmolecular-weight antiangiogenic molecules (e.g., neutralizing antibodies and protein inhibitors) cannot be delivered in the fish water. However, they can be dissolved in the tumor cell suspension before injection. The availability of inbred, transgenic, gene knockout/knockin animals, of a wide array of antibodies, as well as of bioinformatic, genomic, transcriptomic and proteomic information represent important tools for tumor angiogenesis studies performed in murine models. Several of these tools have now become available for zebrafish as well. The identification of genes essential for blood vessel formation is of pivotal importance for the understanding of the angiogenesis process and for the discovery of novel therapeutic targets. In zebrafish, antisense morpholino oligonucleotides induce a translational block in gene function15. Gene inactivation by this approach is easy and fast (3–4 d for a single gene-inactivation experiment) when compared to the generation of knockout mice (several months). Also, the simultaneous injection of different morpholinos may allow the inactivation of more than one gene at the same time. The morpholino strategy has been used successfully to phenocopy various null mutations and to identify the role of numerous genes in zebrafish16. Nevertheless, caution

MATERIALS

a b

c

Yolk syncytial layer

nc

Periderm Yolk Tumor cells

Figure 2 | Visualization of the site of injection of the tumor cell graft in the zebrafish embryo. (a) A 48 hpf embryo was grafted with GFP-transduced tumor cells to highlight the site of injection (arrowhead). The vertical bar indicates the site of the transverse section shown in b. (b) DAPI-stained transverse section of an injected embryo at 72 hpf showing the cell graft (*) (nc, notochord) (original magnification, 200). (c) Schematic representation of grafted tumor cells injected into the perivitelline space between the periderm and the yolk syncytial layer.

should be used to avoid misinterpretation of the experimental data. Dose–response experiments and proper controls (represented by the use of two morpholinos targeting the same gene, control mismatch morpholinos and RNA rescue experiments) should be included in the experimental protocols to rule out possible toxic or mistargeting effects (see ref. 16 for a detailed discussion on the use of morpholinos in zebrafish). When proper controls are used, the morpholino strategy applied to the zebrafish/tumor xenograft model may provide significant advantages when compared to any other mammalian assay available, and it can be exploited for the identification of novel gene(s) involved in tumor neovascularization. As a proof of concept, we have demonstrated that morpholino-induced inactivation of the VE-cadherin zebrafish gene ortholog results in a significant inhibition of the angiogenic process triggered by the tumor graft in zebrafish embryos12. The morpho-functional differences between normal and tumor blood vessels are well known1. The zebrafish/tumor xenograft model allows the study of the impact of angiogenesis inhibitors or of morpholino-induced gene inactivation on tumor-driven neovascularization as well as on physiological angiogenesis in the same embryo (e.g., angiogenesis that occurs in developing intersegmental vessels of the trunk). Indeed, we have reported the capacity of VE-cadherin morpholinos to hamper tumor vascularization, without affecting the development of the normal embryo vasculature12.

REAGENTS

. Tumor cell suspension (from 2 106 to 1 107 cells resuspended on ice in

(provided by A. Rubinstein, Zygogen) . E3 embryo medium (5.0 mM NaCl, 0.17 mM KCl, 0.33 mM MgSO4) . Agarose-modified microinjection plates prepared as described17 . Matrigel (Cultrex Basement Membrane Extract, R&D Systems) . Tricaine (Sigma-Aldrich, A-5040) . 1-Phenyl-2-thiourea (PTU; Sigma-Aldrich, P7629) . Nitroblue tetrazolium and X-phosphate (Roche)

20–30 ml of 12.0 mg ml 1 Matrigel solution). If required, angiogenesis inhibitors can be dissolved in Matrigel before cell resuspension and injection . Standard solutions for paraffin embedding, including ethanol series, toluene and embedding paraffin (Sigma-Aldrich) . Standard solutions for gelatin embedding, including sucrose series and embedding gelatin (Sigma-Aldrich) . 4¢,6-diamidino-2-phenylindole (DAPI; Sigma-Aldrich). Use a 0.25 mg ml 1 aqueous solution

. Wild-type AB zebrafish strain and the VEGFR2:G-RCFP zebrafish line11

2920 | VOL.2 NO.11 | 2007 | NATURE PROTOCOLS

PROTOCOL EQUIPMENT

. Air incubator set at 28 1C (International PBI S.p.A) . Curved-tip forceps . Straight-tip forceps . Borosilicate needles for microinjection (C100F-10, Harvard Apparatus) . Magnetic glass microelectrode horizontal puller (PN-30, Tritech Research)

. Microloader tips (0.5–10 ml, Eppendorf)

. FemtoJet 5247 and Inject Man NI2 microinjection apparatus (Eppendorf)

. Stereomicroscope (MZ75, Leica) . Epifluorescence stereomicroscope (MZ16F, Leica) equipped with digital camera (DFC480, Leica)

. Epifluorescence Axiovert 200M microscope (Zeiss) . Software for computerized image analysis (Image-Pro Plus; MediaCybernetics)


PROCEDURE The zebrafish/tumor xenograft assay 1| Incubate fertilized wild-type AB zebrafish or VEGFR2:G-RCFP transgenic eggs at 28 1C in fish water for 24 h. 2| At 24 hpf, collect the embryos, remove the chorion by forceps, soak the dechorionated embryos in E3 embryo medium containing 0.2 mM PTU and incubate them for further 24 h at 28 1C. 3| Anesthetize 48 hpf zebrafish embryos in E3 embryo medium with 0.02 mg ml Pasteur pipette and place it on an agarose-modified Petri dish.

1

tricaine added, collect each embryo with a

4| Remove excess water from the dish using a syringe and orient the embryos with the yolk on a flank. m CRITICAL STEP Proper orientation of the embryos is of pivotal importance for subsequent correct cell injection, as it is difficult to inject into the perivitelline space when embryos are oriented with the yolk on the top. To maintain the correct orientation of the embryo, remove most of the water until a light shaking of the Petri dish will not cause any movement of the embryo. 5| Load pre-cooled borosilicate needles with 5 ml of tumor cell suspension in Matrigel (a solubilized basement membrane extract) by using a 20-ml Eppendorf pipette. ? TROUBLESHOOTING 6| Put the Petri dish with the correctly orientated embryos under a stereomicroscope at 50 magnification and point the cellloaded needle toward the region of the embryo body between the yolk and the duct of Cuvier area, close to the sinuous venous. By using the microinjector, insert the tip of the needle into the yolk and pull it back slightly to create an artificial space between the periderm and the yolk syncytial layer where the cells will be grafted (Fig. 2). Then, inject a drop of the cell suspension (approximately 4–10 nl) by setting the proper pressure (500–1,000 hPa) and time (0.5–1.0 s) parameters of the apparatus. Injection of the same volumes of the Matrigel solution or of a non-tumorigenic cell suspension may be used as mock and negative controls, respectively. ! CAUTION Warming up of the tip of the needle under the light of the stereomicroscope may cause gelling of the Matrigel solution, thus hampering the injection of the cell suspension. ? TROUBLESHOOTING 7| Collect the injected embryos and incubate them for 24–48 h at 28 1C in fresh E3 embryo medium. If desired, this medium can contain angiogenesis inhibitors. m CRITICAL STEP Small organic molecules, including synthetic angiogenesis inhibitors, may require organic solvents. Dimethylsulfoxide may be used at concentrations up to 1.0% in E3 embryo medium without adverse effects on embryo development and survival. 8| At the end of the incubation, fix the embryos with 4% paraformaldehyde in PBS (pH 7.0) for 2 h at 23–25 1C, rinse them in PBS and store them at 4 1C in PBS until further use. ? TROUBLESHOOTING 9| Perform whole-mount staining of the fixed embryos for endogenous alkaline phosphatase activity according to standard procedures14. Briefly, embryos are dehydrated in 25%, 50%, 75% and 100% methanol in PBS and rehydrated stepwise to 100% PBS with 1.0% Tween 20 added. Then, embryos are equilibrated in 0.1 M Tris-HCl (pH 9.5), 50 mM MgCl, 0.1 M NaCl and 0.1% Tween 20 for 30 min and stained with

a

c

b

d

Figure 3 | Angiogenic responses triggered by tumor cell grafts in the zebrafish embryo. VEGFR2:G-RCFP transgenic embryos (a,b) and wild-type AB strain embryos (c,d) were injected with fluorescent dye-loaded tumor cells (in red in a and b; * in c and d). After 24 h, embryos were photographed under an epifluorescence stereomicroscope (a,b) or stained for alkaline phosphatase activity (c,d). Note the different morphological features of the angiogenic response when the graft is located 50–100 mm apart from (a,c) or proximal to (b,d) the SIV plexus. In both cases, a positive angiogenesis score is assigned to the embryo. See text for further details. Lateral view, head on the left; magnification: 11.5. NATURE PROTOCOLS | VOL.2 NO.11 | 2007 | 2921

PROTOCOL 0.34 mg ml 1 nitroblue tetrazolium and 0.15 mg ml dure at 23–25 1C.

1

X-phosphate in the same buffer for 10–20 min. Perform the whole proce-

10| Dehydrate fixed embryos in an ethanol series, clear them in toluene and immerse them in embedding paraffin for 2 h. For cryosections, incubate fixed embryos in 5% sucrose/PBS for 1 h and in 15% sucrose/PBS for 3 h at 23–25 1C. Then, embryos are embedded in 7.5% gelatin/15% sucrose in PBS for 4 h at 37 1C and snap-frozen12.


11| Cut 8 mm serial transversal sections of paraffin-embedded embryos in the region of the tumor graft and observe them under a light microscope. Stain gelatin-embedded sections with DAPI for 5 min at room temperature (23–25 1C) and observe them under an epifluorescence microscope. Macroscopic evaluation of the angiogenic response 12| Perform macroscopic evaluation of the vasoproliferative response triggered by the mammalian tumor xenografts in zebrafish embryos by whole-mount analysis of the modifications of SIV development as evidenced by alkaline phosphatase staining in wildtype AB strain embryos and/or by fluorescence microscopy in VEGFR2:G-RCFP transgenic embryos. These modifications are characterized by the convergence of SIVs toward the graft with different angiogenic morphological features depending on the site of injection of the xenograft. In details, (a) when the xenograft is located approximately 50–100 mm apart from the developing SIV plexus, new blood vessels will sprout and migrate toward the graft. These vessels will eventually reach, surround and penetrate the graft within 48 h after injection (Fig. 3); (b) when the graft is closer to the plexus, it becomes rapidly invaded by SIVs with a consequent local increase in blood vessel branching and density when compared to mock and control embryos (Fig. 3). In both cases, a positive angiogenic score is assigned to the embryo. Routinely, each experimental point consists of 20 embryos and each experiment is repeated thrice. Data are expressed as the ratio between positive and total grafted embryos. Microscopic evaluation of the angiogenic response 13| Apply a computerized image analysis on transversal embryo sections to obtain a quantitative evaluation of the angiogenic response, expressed as microvessel density. This can be performed on digitized images acquired at 600 magnification under a light microscope for alkaline phosphatase-stained embryos or under an epifluorescence microscope for VEGFR2:G-RCFP transgenic embryos. However, we recommend the usage of the VEGFR2:G-RCFP line owing to the high specificity and higher sensitivity of the GFP fluorescence signal in zebrafish endothelial cells that can be merged with the DAPI nuclear staining of the whole graft section (Fig. 4). For each section, vascular density is calculated as the ratio between the blood vessel area and the total area of the graft. Routinely, five sections are analyzed for each xenograft. Then, mean values ± 1 s.d. are determined for each zebrafish specimen.

TIMING Preparation for injection of 50 zebrafish embryos, Step 2: 30 min Injection of tumor cell suspension in 50 zebrafish embryos, Steps 3–7: 2 h Histological processing and staining of the samples, Steps 8–11: 24 h Macroscopic evaluation of the angiogenic response in 50 embryos, Step 12: 1 h Microscopic evaluation of microvessel density in one embryo, Step 13: 1 h ? TROUBLESHOOTING During Steps 5 and 6, maintain a low temperature throughout the whole injection procedure by using ice bath and pre-cooled equipment to avoid gelling of the tumor cell suspension in the Matrigel solution. During Step 6, the injection of the cell suspension in the correct region of the embryo body is of pivotal importance for obtaining a positive angiogenic response. For this FGF2-T-MAE FGF2-T-MAE + PTX3 purpose, take care that the injected cell pellet causes a a c protrusion of the periderm when the embryo is observed from

Figure 4 | Effect of long pentraxin 3 (PTX3), an FGF2 antagonist20, on the angiogenic response induced by proangiogenic FGF2-T-MAE cells. Zebrafish embryos were injected with tumor cells resuspended in a Matrigel solution in the absence (a,b) or in the presence (c,d) of 0.22 mM PTX3. Note the reduction in the number of macroscopic alkaline phosphatase-positive SIVs (arrowheads) converging versus the graft (*) in PTX3-treated (c) versus control (a) graft. Consequently, a remarkable reduction of GFP-positive blood vessel density is observed within the PTX3-treated tumor graft (defined by the dotted line) in DAPI-stained transversal sections of VEGFR2:G-RCFP transgenic embryos. Original magnifications: (a,c) 11.5; (b,d) 600. 2922 | VOL.2 NO.11 | 2007 | NATURE PROTOCOLS

b

d

PROTOCOL a dorsal view. Discard all the embryos in which cells were injected into the yolk sac. During Step 8, paraformaldehyde fixation will cause autofluorescence of the yolk with consequent masking of the SIV-associated GFP fluorescence signal in the VEGFR2:G-RCFP zebrafish embryos. Thus, assessment of the macroscopic angiogenic response in these embryos (see below) must be performed and eventually recorded under an epifluorescence stereomicroscope before fixation.

TABLE 2 | Angiogenic activity of different tumor cell lines in the zebrafish/tumor xenograft angiogenesis assay. Tumor cell line Positive embryos (%) 18 Murine aortic endothelial FGF2-T-MAE cells 100 Human endometrial adenocarcinoma 75 Tet-FGF2 cells22 Human ovarian carcinoma A2780 cells 80 Human breast carcinoma MDA-MB-435 cells 60 Murine melanoma B16-BL16 cells 80


Cells were grafted in zebrafish embryos at 48 hpf (20 embryos per group). The macroscopic angiogenic response was scored 24–48 h thereafter on whole-mount alkaline phosphatase-stained embryos. ANTICIPATED RESULTS At 24–48 h after injection of proangiogenic tumor cells, macroscopic observation of zebrafish embryos shows the growth of SIV neovessels converging toward the tumor graft. In a typical experiment, between 60% and 100% of injected embryos showed a positive response characterized by the migration of new blood vessels from the SIV plexus toward the implant or by an increased blood vessel density within the plexus (see Step 12 and Fig. 3). The percentage of positive response depends on the grafted tumor cell lines, as illustrated in Table 2. No positive responses are observed in embryos injected with Matrigel alone, whereas the injection of poorly angiogenic cells results in a percentage of positive embryos lower than 5%. Histological analysis of tumor grafts shows the presence of numerous blood vessels infiltrating the implant, with a consequent significant increase in vascular density. In a typical experiment, tumor grafts originating by the injection of highly angiogenic, FGF2-overexpressing murine aortic endothelial cells (FGF2-T-MAE cells)18 show a vascular area equal to 17 ± 3% of the total tumor area12. Addition of an angiogenesis inhibitor to the cell suspension (Fig. 4) or in the fish water12 will result in impairment of the angiogenic response triggered by the tumor graft. This can be observed both at the macroscopic level, as a decrease in the percentage of positive embryos, and at the microscopic level, as a decrease in vascular density of the graft. Similarly, knock-down of genes relevant to the angiogenesis process following injection of the corresponding antisense morpholino oligonucleotides will cause a significant decrease in tumor angiogenesis, as demonstrated after inactivation of the VE-cadherin zebrafish gene ortholog12. Recent observations have shown the possibility to inject mammalian tumor cells in the peritoneal cavity of 1 month-old zebrafish19. Also in this case, VEGF-dependent tumor-induced angiogenesis and cancer cell intravasation is observed. These data underlie the possibility to develop zebrafish/tumor xenograft models to investigate the mechanisms of cancer neovascularization and for the development of antiangiogeneic drugs.

ACKNOWLEDGMENTS This work was supported by grants from Istituto Superiore di Sanita` (Oncotechnological Program), Ministero dell’Istruzione, Universita` e Ricerca (Centro di Eccellenza per l’Innovazione Diagnostica e Terapeutica), Associazione Italiana per la Ricerca sul Cancro, Fondazione Berlucchi and NOBEL Project Cariplo. Published online at http://www.natureprotocols.com Reprints and permissions information is available online at http://npg.nature.com/ reprintsandpermissions 1. Carmeliet, P. & Jain, R.K. Angiogenesis in cancer and other diseases. Nature 407, 249–257 (2000). 2. Ferrara, N. Vascular endothelial growth factor: basic science and clinical progress. Endocr. Rev. 25, 581–611 (2004). 3. Presta, M. et al. Fibroblast growth factor/fibroblast growth factor receptor system in angiogenesis. Cytokine Growth Factor Rev. 16, 159–178 (2005). 4. Hasan, J. et al. Quantitative angiogenesis assays in vivo—a review. Angiogenesis 7, 1–16 (2004). 5. Lam, S.H. et al. Conservation of gene expression signatures between zebrafish and human liver tumors and tumor progression. Nat. Biotechnol. 24, 73–75 (2006). 6. Thisse, C. & Zon, L.I. Organogenesis—heart and blood formation from the zebrafish point of view. Science 295, 457–462 (2002). 7. Weinstein, B. Vascular cell biology in vivo: a new piscine paradigm? Trends Cell Biol. 12, 439–445 (2002). 8. Topczewska, J.M. et al. Embryonic and tumorigenic pathways converge via Nodal signaling: role in melanoma aggressiveness. Nat. Med. 12, 925–932 (2006). 9. Haldi, M., Ton, C., Seng, W.L. & McGrath, P. Human melanoma cells transplanted into zebrafish proliferate, migrate, produce melanin, form masses and stimulate angiogenesis in zebrafish. Angiogenesis 9, 139–151 (2006). 10. Isogai, S., Horiguchi, M. & Weinstein, B.M. The vascular anatomy of the developing zebrafish: an atlas of embryonic and early larval development. Dev. Biol. 230, 278–301 (2001).

11. Cross, L.M., Cook, M.A., Lin, S., Chen, J.N. & Rubinstein, A.L. Rapid analysis of angiogenesis drugs in a live fluorescent zebrafish assay. Arterioscler. Thromb. Vasc. Biol. 23, 911–912 (2003). 12. Nicoli, S., Ribatti, D., Cotelli, F. & Presta, M. Mammalian tumor xenografts induce neovascularization in zebrafish embryos. Cancer Res. 67, 2927–2931 (2007). 13. Pichler, F.B. et al. Chemical discovery and global gene expression analysis in zebrafish. Nat. Biotechnol. 21, 879–883 (2003). 14. Serbedzija, G.N., Flynn, E. & Willett, C.E. Zebrafish angiogenesis: a new model for drug screening. Angiogenesis 3, 353–359 (1999). 15. Nasevicius, A. & Ekker, S.C. Effective targeted gene ‘knockdown’ in zebrafish. Nat. Genet. 26, 216–220 (2000). 16. Sumanas, S. & Larson, J.D. Morpholino phosphorodiamidate oligonucleotides in zebrafish: a recipe for functional genomics? Brief. Funct. Genomic. Proteomic. 1, 239–256 (2002). 17. Gilmour, D.T., Jessen, J.R. & Lin, S. in Zebrafish (eds. Nusslein-Volhard, C. & Dahm, R.) 121–143 (Oxford University Press, Oxford, 2002). 18. Gualandris, A. et al. Basic fibroblast growth factor overexpression in endothelial cells: an autocrine mechanism for angiogenesis and angioproliferative diseases. Cell Growth Differ. 7, 147–160 (1996). 19. Stoletov, K. et al. Nigh-resolution imaging of the dynamic tumor cell-vascular interface in transparent zebrafish. Proc. Natl. Acad. Sci. USA 104, 17406–17411 (2007). 20. Presta, M., Camozzi, M., Salvatori, G. & Rusnati, M. Role of the soluble pattern recognition receptor PTX3 in vascular biology. J. Cell. Mol. Med. 11, 723–738 (2007). 21. Ribatti, D. et al. Alterations of blood vessel development by endothelial cells overexpressing fibroblast growth factor-2. J. Pathol. 189, 590–599 (1999). 22. Giavazzi, R. et al. Distinct role of fibroblast growth factor-2 and vascular endothelial growth factor on tumor growth and angiogenesis. Am. J. Pathol. 162, 1913–1926 (2003).

NATURE PROTOCOLS | VOL.2 NO.11 | 2007 | 2923