**9. Embryonic zebrafish metastasis and tumor model**

Zebrafish embryos in the first few weeks of development lack an adaptive immune system, which makes them unable to reject non-zebrafish grafts and they are therefore perfect recipients for implantation of mammalian tumor cells. Due to the transparent nature of zebrafish embryos, fluorescently labeled tumor cells may furthermore be traced in a completely non-invasive fashion, leading to continuous monitoring of the behavior of the tumor cells, including their dissemination from the primary tumor mass *in vivo* (Rouhi et al, 2010).

The induction of tumor angiogenesis is a hallmark of advanced tumors, and therefore also of established tumor cell lines. Thus, mammalian tumor cells, implanted into a space around the yolk sac called the perivitelline space of transgenic *fli1:EGFP* zebrafish embryos, is a powerful way of studying early events of tumor-induced angiogenesis and especially of tumor cell dissemination and metastasis via the vasculature (Lee et al, 2009; Rouhi et al, 2010). Depending on their invasive capabilities, implanted tumor cells may start disseminating from the primary site very shortly after implantation (Fig. 15). Angiogenic factors produced by the tumor cells can affect the host's existing and developing vasculature which furthermore is a good way of studying the function and morphology of vessels induced by the tumor-derived factors in question. Different kinds of manipulation e.g. knock-down or over-expression of a specific gene in the tumor cells can affect the dissemination and invasion pattern of a specific cell line implanted in the embryo. Genetic manipulation of tumor cells can be performed with help of different tools e.g. small interfering RNA (siRNA), small hairpin RNA (shRNA) or by retroviral integration of expression vectors containing angiogenic factors.

Fig. 15. Metastasis model in zebrafish embryos. DiI labeled tumor cells (red) are injected in the perivitelline space of approximately 2 day old zebrafish embryos (left panels). 3 days after injection many disseminated cells are observed (arrowheads) proximal to the tumor (middle panels). After 6 days tumor cells have spread throughout the embryo, via the vasculature (right panels). The top row shows brightfield images combined with the redfluorescent signal from the tumor cells. The bottom row shows the EGFP-positive vasculature (green) in transgenic F*li1:EGFP* zebrafish embryos combined with the redfluorescent signal from the tumor cells.

It is also possible to study the effects of host gene expression on tumor cell dissemination and metastasis by modifying expression of host genes by injecting either morpholinos or mature mRNA immediately after fertilization, and thus prior to tumor implantation, as described above (Lee et al, 2009). This model is thus a simple system to parse out the contributions of factors derived from the tumor relative to the host on tumor angiogenesis, dissemination and metastasis.

Since disseminated tumor cells in this model will be detected by fluorescent microscopy, they should either permanently express a fluorescent protein or they can – prior to injection – be labeled with fluorescent dyes such as DiI, which can be easily distinguished from the green fluorescent color emitted from the embryo vasculature. Tumor cells are preferably implanted in the non-vascularized area in the perivitelline cavity as it can otherwise be difficult to distinguish between existing vessels, those formed by normal developmental processes and those induced by the tumor mass.

Prior to implantation, embryos will be dechorionated, anesthetized and placed on a modified agarose gel. Transplantation of tumor cells is easily carried out using a micromanipulator which is connected to a microinjector. Approximately 100 cells will be implanted into each embryo.

After microinjection, based on the purpose of the study, embryos will be immediately transferred into appropriate embryo water such as E3 water or Danieu's buffer (Fig. 15). As mentioned above, drugs may be added to the water and their effects on tumor angiogenesis, dissemination and metastasis can thus easily be evaluated.

Each tumor bearing embryo may be placed in a separate well of a multi-well plate and examined individually to monitor tumor angiogenesis, dissemination and metastasis in as high temporal resolution as required in the experiment. This model furthermore allows generation of time-lapse video sequences of the tumor cells invading into and out from the vasculature at metastatic sites, and thus examine the tumor-endothelium communications involved in tumor dissemination and metastasis in detail.

Fig. 15. Metastasis model in zebrafish embryos. DiI labeled tumor cells (red) are injected in the perivitelline space of approximately 2 day old zebrafish embryos (left panels). 3 days after injection many disseminated cells are observed (arrowheads) proximal to the tumor (middle panels). After 6 days tumor cells have spread throughout the embryo, via the vasculature (right panels). The top row shows brightfield images combined with the redfluorescent signal from the tumor cells. The bottom row shows the EGFP-positive vasculature (green) in transgenic F*li1:EGFP* zebrafish embryos combined with the red-

It is also possible to study the effects of host gene expression on tumor cell dissemination and metastasis by modifying expression of host genes by injecting either morpholinos or mature mRNA immediately after fertilization, and thus prior to tumor implantation, as described above (Lee et al, 2009). This model is thus a simple system to parse out the contributions of factors derived from the tumor relative to the host on tumor angiogenesis,

Since disseminated tumor cells in this model will be detected by fluorescent microscopy, they should either permanently express a fluorescent protein or they can – prior to injection – be labeled with fluorescent dyes such as DiI, which can be easily distinguished from the green fluorescent color emitted from the embryo vasculature. Tumor cells are preferably implanted in the non-vascularized area in the perivitelline cavity as it can otherwise be difficult to distinguish between existing vessels, those formed by normal developmental

Prior to implantation, embryos will be dechorionated, anesthetized and placed on a modified agarose gel. Transplantation of tumor cells is easily carried out using a micromanipulator which is connected to a microinjector. Approximately 100 cells will be

After microinjection, based on the purpose of the study, embryos will be immediately transferred into appropriate embryo water such as E3 water or Danieu's buffer (Fig. 15). As mentioned above, drugs may be added to the water and their effects on tumor angiogenesis,

Each tumor bearing embryo may be placed in a separate well of a multi-well plate and examined individually to monitor tumor angiogenesis, dissemination and metastasis in as high temporal resolution as required in the experiment. This model furthermore allows generation of time-lapse video sequences of the tumor cells invading into and out from the vasculature at metastatic sites, and thus examine the tumor-endothelium communications

fluorescent signal from the tumor cells.

processes and those induced by the tumor mass.

dissemination and metastasis can thus easily be evaluated.

involved in tumor dissemination and metastasis in detail.

dissemination and metastasis.

implanted into each embryo.

As described in more detail below, zebrafish are – compared to mice – highly amenable to studies on the physiological and pathological effects of hypoxia. In order to investigate the effects of hypoxia on tumor angiogenesis, dissemination and metastasis, tumor celltransplanted embryos can be placed in hypoxic water in a special aquarium (Lee et al, 2009). As tumor hypoxia in mammalian models is very difficult to control and monitor, this system allows studies of the effects of highly defined oxygen concentrations and periods of hypoxia exposure. In all conditions above fish embryos will be kept at 28.5ºC, which is the standard temperature for rearing zebrafish embryos and larvae. Death rate is relatively high in hypoxia experiments and in order to have enough embryos at the end to make statistically correct conclusions, a relatively high number of embryos should be implanted with tumor cells and placed inside the hypoxia chamber.
