Live Imaging and Gene Expression Analysis in Zebrafish Identifies a Link between Neutrophils and Epithelial to Mesenchymal Transition

Chronic inflammation is associated with epithelial to mesenchymal transition (EMT) and cancer progression however the relationship between inflammation and EMT remains unclear. Here, we have exploited zebrafish to visualize and quantify the earliest events during epithelial cell transformation induced by oncogenic HRasV12. Live imaging revealed that expression of HRasV12 in the epidermis results in EMT and chronic neutrophil and macrophage infiltration. We have developed an in vivo system to probe and quantify gene expression changes specifically in transformed cells from chimeric zebrafish expressing oncogenic HRasV12 using translating ribosomal affinity purification (TRAP). We found that the expression of genes associated with EMT, including slug, vimentin and mmp9, are enriched in HRasV12 transformed epithelial cells and that this enrichment requires the presence of neutrophils. An early signal induced by HRasV12 in epithelial cells is the expression of il-8 (cxcl8) and we found that the chemokine receptor, Cxcr2, mediates neutrophil but not macrophage recruitment to the transformed cells. Surprisingly, we also found a cell autonomous role for Cxcr2 signaling in transformed cells for both neutrophil recruitment and EMT related gene expression associated with Ras transformation. Taken together, these findings implicate both autocrine and paracrine signaling through Cxcr2 in the regulation of inflammation and gene expression in transformed epithelial cells.


Introduction
Epithelial to mesenchymal transition (EMT) is essential for normal embryonic development, but also occurs during wound healing and the invasion of transformed cells during cancer progression [1,2]. One main difference between the activation of programmed EMT during early development and EMT associated with pathology is the presence of inflammation [3]. Chronic inflammation associated with diseases like Rheumatoid Arthritis, Crohns disease, Chronic obstructive pulmonary disease (COPD) and pancreatitis is known to increase cancer risk [4,5] and neutrophil depletion has been shown to reduce tumor burden in a mouse model of lung cancer [6]. The connection between inflammation and cancer risk is further supported by studies that suggest that treatment with anti-inflammatory agents can decrease the incidence of some cancers [7][8][9]. Although chronic inflammation has been linked to EMT and cancer invasion [10][11][12], it is not clear how innate immune inflammation is associated with EMT and participates in cancer progression.
Understanding how transformed cells induce chronic innate immune inflammation in tissues and if this inflammation plays a role in EMT progression has been limited due to the difficulty visualizing the early tissue microenvironment around small clusters of transformed epithelial cells in live animals. Zebrafish is therefore an ideal model system to study these questions since zebrafish larvae are transparent allowing for real time imaging of EMT in live animals. Indeed, zebrafish have emerged as a powerful model system to study the pathogenesis of cancer since many of the signaling pathways and mechanisms are conserved [13,14]. In fact, recent progress in zebrafish has identified signaling pathways that mediate leukocyte recruitment to wounds [15][16][17], infection [18] and transformed melanocytes [19]. Additionally, chronic inflammation of the epidermis in zebrafish has been associated with the activation of EMT programs and developmental defects in epidermal homeostasis, further supporting the use of zebrafish to study the relationship between inflammation and EMT [10,20].
Chemokine-dependent signaling in immune cells is an important mechanism that mediates leukocyte recruitment to bacterial infections and wounds. A large body of work has explored the role of the chemokine (C-X-C motif) receptor 2 (CXCR2) during leukocyte recruitment to sites of inflammation [6,[15][16][17][18][19][21][22][23]. Previous studies, utilizing the zebrafish model system, have shown that Cxcr2 receptor signaling axis is involved in both long-range systemic neutrophil mobilization from hematopoietic tissue as well as local recruitment to infection or purified Cxcl8 (IL-8, a potent Cxcr2 ligand) [17,18]. Previous studies have also implicated CXCR2 signaling in tumor progression [21,22,[24][25][26][27][28][29]. CXCR2 expression is increased in some tumor types [26,[30][31][32] and pharmacological inhibition of CXCR1 and CXCR2 inhibits neutrophil recruitment into A547 lung tumor spheroids resulting in slower tumor growth [33]. Mouse models have also been instrumental in demonstrating that CXCR2 signaling is involved in the recruitment of myeloid-derived suppressor cells into the tumor microenvironment. For instance, expression of human CXCL8 in mice results in increased mobilization of immature myeloid cells, which exacerbates inflammation and accelerates colon carcinogenesis [34]. Tumor trafficking of myeloid-derived suppressor cells is inhibited by CXCR2 deficiency in a mouse model of rhabdomyosarcoma [35] and data from a mouse model of colitis-associated cancer suggests that CXCR2 is required for recruitment of myeloid-derived suppressor cells [24,36]. Moreover, activation of CXCR2 on Ras-transformed keratinocytes contributes to tumor progression in a mouse model of skin cancer [23]. These findings support a role for CXCR2 signaling in inflammation and cancer progression; however the connection between CXCR2-mediated neutrophil recruitment and EMT remains unclear.
In this study we have exploited advances in real time imaging and analysis of tissue-specific gene expression in zebrafish to interrogate the role of Cxcr2 and neutrophil recruitment in HRas V12 -induced EMT. We found that signaling through Cxcr2 is required for neutrophil, but not macrophage, recruitment to transformed epithelial cells. We further show that both Cxcr2 in transformed cells and the presence of neutrophils, but not macrophages, in the microenvironment are required for expression of EMT-related genes. These findings establish an essential role for Cxcr2 in regulating chronic neutrophil inflammation and EMT-related gene expression via both autocrine and paracrine mechanisms.

HRas V12 expression in epithelial cells induces EMT and inflammation
Tissue damage or oncogenic signals induce EMT [37]. To visualize EMT in zebrafish, we used the krt4 promoter to drive gene expression in epidermal cells (krt4; previously termed krt8) [38]. We found that the krt4 promoter drives GFP expression in the developing epidermis during early gastrulation and by 12 hours post fertilization (hpf) uniform expression is observed throughout the epidermis ( Figure 1A). To express HRas V12 in the zebrafish epidermis, RFP-HRas V12 was cloned into a Tol2 containing plasmid and co-injected with transposase RNA into one-cell stage embryos ( Figure 1B). Active Ras signaling promotes cell transformation [39][40][41][42][43] and has been shown to drive chemokine and cytokine expression [44][45][46]. Early expression of HRas V12 using the krt4 promoter induced cell extrusion in developing embryos ( Figure 1C), as has previously been reported for apoptotic epithelial cells in zebrafish [47] or in response to activating Src [48]. To reduce early transgene expression, we designed an antisense morpholino oligonucleotide (MO) targeting the 39 end of the krt4 promoter 25 bases directly 59 to the AUG translational start site of RFP-HRas V12 ( Figure 1B). Microinjection of the krt4 MO inhibited GFP expression in krt4:GFP transgenic larvae ( Figure 1D) and reduced transgene expression in 24 hpf embryos injected with Tol2 flanked:krt4: RFP-HRas V12 ( Figure 1E). By reducing early HRas V12 transgene expression, we were able to significantly reduce the early apical extrusion of HRas V12 expressing cells ( Figure 1F).
High resolution confocal imaging revealed that chimeric embryos expressing wild type HRas WT had membrane localization of the transgene and displayed a cuboidal morphology typical of epithelial cells at 3.5 days post fertilization (dpf) ( Figure 2B). Cells expressing constitutively active HRas V12 also had membrane localization of the transgene but displayed altered cell morphology ( Figure 2C). Live imaging, of chimeric 3.5 dpf embryos, revealed that HRas WT cells maintained their shape over a four hour time period ( Figure 1D, Movie S1) while, HRas V12 cells displayed an abnormal morphology with dynamic protrusions ( Figure 2E, Movie S2), quantified by reduced 2D cell area and roundness ( Figure 2F-G).
To determine if HRas V12 expression in epithelial cells resulted in changes in gene expression consistent with EMT, we investigated the expression of a transcriptional activator of EMT, Slug (also known as Snail2) which has been identified as a driving factor of EMT in keratinocytes during wound healing [37] and is increased during cancer progression [49]. We also tested expression of a matrix metalloproteinase (Mmp9) that has been linked to EMT, and a type III intermediate filament protein (Vimentin) that is expressed in mesenchymal cells and has been previously shown to be a reliable marker of EMT [50][51][52]. Double label Whole Mount In Situ Hybridization (WMISH) revealed that the EMT associated genes mmp9, slug and vimentin were enriched in HRas V12 transformed epithelial cells, compared to control HRas WT expressing cells ( Figure 2H-I). To better quantify these changes in gene expression we used translating ribosome affinity purification (TRAP) [53] to isolate RNA specifically from the transformed epithelial cells followed by Quantitative Reverse Transcriptase Polymerase Chain Reaction (qRT-PCR). We found that transforming HRas V12 induced EMT gene expression specifically in the transformed epithelial cells, including increased slug and mmp9, compared to HRas WT control at 3.5 dpf ( Figure 2J).
To characterize the host innate immune responses to transformed epithelial cells, we used the transgenic zebrafish lines Tg(mpx:GFP) and Tg(mpeg:Dendra) to visualize neutrophils and macrophages respectively in vivo [20,[54][55][56]. In zebrafish larvae, neutrophils are generally localized to the caudal hematopoietic tissue (CHT) [57], which has a hematopoietic function similar to the fetal liver in mammals [58]. Live high resolution spinning disk confocal microscopy revealed that neutrophils ( Figure 3B, Movie S4) and macrophages ( Figure 3E, Movie S6) are recruited to HRas V12 but not control HRas WT expressing epithelial cells ( Figure 3A and D, Movie S3 and S5). Quantification of leukocytes as a ratio of transformed cell number revealed a significant increase in neutrophils ( Figure 3C) and macrophages ( Figure 3F) per HRas V12 transformed cell compared to control HRas WT cells. These findings indicate that chimeric HRas V12 expression in the zebrafish epidermis is sufficient to stimulate neutrophil and macrophage recruitment, similar to the effects reported with oncogene-transformed melanoblasts in zebrafish [19].
Neutrophils, but not macrophages, are necessary for mmp9 and slug expression in transformed epithelial cells To determine if there is a cell autonomous role for neutrophils in regulating EMT, we characterized EMT related gene expression in larvae with impaired neutrophil function. We took advantage of a zebrafish model of primary immunodeficiency (zf307Tg, Tg(mpx:mCherry,rac2 D57N )), where neutrophil recruitment to tissue damage is impaired. In this model, expression of the human inhibitory Rac2 D57N mutation in neutrophils results in reduced neutrophil migration and recruitment to wounds and infection [57]. We found a significant decrease in neutrophil recruitment to HRas V12 expressing cells in Rac2 D57N larvae compared to control ( Figure 4B and D). To ensure that macrophages were still recruited to transformed cells in the absence of neutrophil recruitment we quantified macrophage numbers at transformed cells in Rac2 D57N larvae and found that macrophage recruitment was not affected ( Figure 4G and I). Surprisingly, we found that EMT associated gene expression is impaired in neutrophil-deficient larvae ( Figure 4K) indicating that neutrophils are necessary for the expression of EMT associated genes in transformed epithelial cells.
To determine if there is a cell autonomous role for macrophages in regulating EMT associated gene expression, we utilized a previously published MO targeting interferon regulatory factor 8 (Irf8), which is essential for directing macrophage but not neutrophil differentiation [59]. We found a significant decrease in macrophage recruitment to HRas V12 expressing cells in irf8 morphants compared to control ( Figure 4 H and J). To determine if neutrophils were still recruited to transformed cells in irf8 morphants we quantified neutrophil numbers at transformed cells and found that neutrophil recruitment was not affected ( Figure 4C and E). Moreover, mmp9 and slug transcripts were not reduced in irf8 morphants compared to control ( Figure 4K), suggesting that macrophages do not induce EMT gene expression in transformed cells. Interestingly, mmp9 expression was increased in irf8 morphants, likely due to the increase in total numbers of neutrophils in irf8 morphants. Taken together, these findings suggest that neutrophils but not macrophages influence EMT associated gene expression in transformed epithelial cells in vivo.

Cxcr2 is required for neutrophil recruitment to transformed cells
Previous work has predominantly focused on the role of macrophages in cancer progression and few studies have characterized the role of neutrophils within the tumor microenvironment [60][61][62][63][64][65]. Our findings suggest that neutrophils play a critical role in influencing gene expression in transformed epithelial cells. To identify the signals that mediate neutrophil recruitment to transformed epithelial cells, we targeted specific pathways that have been implicated in neutrophil wound attraction, including Cxcl8 signaling [66]. Indeed, human liver epithelial cells from hepatocellular carcinoma have been shown to produce the CXC chemokine CXCL8 and promote neutrophil infiltration [67]. To determine if Cxcl8 is up-regulated by epithelial cell transformation we performed double label WMISH and found an increase in Cxcl8 in regions with HRas V12 expressing cells ( Figure 4M), compared to HRas WT control ( Figure 4L).
We recently reported that Cxcr2, but not Cxcr1, is necessary for neutrophil recruitment to exogenous purified Cxcl8, suggesting that Cxcr2 may mediate neutrophil recruitment to transformed epithelial cells that produce Cxcl8 [18]. To determine if Cxcr2 mediates neutrophil recruitment we depleted cxcr2 using a previously published MO [18]. We found that depletion of cxcr2 resulted in a significant decrease in neutrophil recruitment to HRas V12 expressing cells, quantified as a ratio of neutrophils per transformed cell ( Figure 5B and E). To determine if Cxcr2 also mediates macrophage recruitment we quantified macrophage infiltration, and found that, cxcr2 depletion did not have a significant impact on macrophage recruitment ( Figure 5D and F), suggesting that Cxcr2 mediates neutrophil but not macrophage recruitment to transformed epithelial cells.
To determine if Cxcr2 mediates the invasive progression of transformed cells we characterized the effect of cxcr2 depletion on the HRas V12 -induced expression of mmp9 and slug. We found that depletion of cxcr2 blocked the HRAS V12 -induced expression of EMT related genes ( Figure 5G). It is important to note that, although neutrophil recruitment required Cxcr2, the early morphological changes induced by HRas V12 were not affected by cxcr2 depletion. These findings indicate the Cxcr2 is necessary for HRAS V12 -induced expression of EMT associated genes.
Cxcr2 signaling in transformed epithelial cells is necessary for neutrophil recruitment and HRas V12 induced expression of vimentin, mmp9 and slug in epithelial cells Previous studies have shown that Cxcr2 expression is necessary for neutrophil recruitment to exogenous Cxcl8, indicating that neutrophil Cxcr2 mediates chemotaxis to Cxcl8 in vivo. It is possible that Cxcr2 mediates EMT gene expression in transformed epithelial cells indirectly through its effects on neutrophil recruitment. However, Cxcr2 is also expressed in tumor cells, suggesting that Cxcr2 may play cell autonomous roles in epithelial cells. Indeed, we found that zebrafish epithelial cells express cxcr2 ( Figure 6A), suggesting that Cxcr2 in epithelial cells may affect gene expression changes induced by cell transformation independent of the effects of Cxcr2 in neutrophils. To determine if there is a cell autonomous role for Cxcr2 in transformed epithelial cells we utilized a cell transplantation strategy to deplete Cxcr2 specifically in transformed cells without altering Cxcr2 signaling in neutrophils ( Figure 6B). Cells from embryos expressing RFP-HRas V12 in cxcr2 morphants or control morphants were transplanted into Tg(mpx:GFP) embryos. Interestingly, we found a significant decrease in neutrophil recruitment toward Cxcr2-deficient HRas V12 expressing cells even though the neutrophils expressed Cxcr2 ( Figure 6D and E), suggesting that Cxcr2 signaling in transformed cells is necessary for neutrophil recruitment. To determine if Cxcr2-deficient epithelial cells that express HRas V12 induce EMT associated genes, we tested the expression of vimentin, mmp9 and slug in the Cxcr2-deficient and control transformed epithelial cells. Surprisingly, expression of vimentin, slug and mmp9 were reduced in Cxcr2-deficient transformed epithelial cells compared to control ( Figure 6F); indicating a cell autonomous role for Cxcr2 signaling in epithelial cells in the induction of EMT related gene expression.

Discussion
Here, we uncovered roles for Cxcr2 and neutrophils in oncogene induced EMT gene expression in epithelial cells in zebrafish. By exploiting the transparency of zebrafish we showed that expression of oncogenic HRas V12 induced EMT and invasive growth. Interestingly, early expression of HRas V12 induced cell extrusion, suggesting that depending on the developmental stage of the epithelium the fate of the transformed cell may be either extrusion or EMT. This provides a powerful model system to understand how early EMT is regulated within a live host, and Previous studies have identified roles for both macrophages [60][61][62][63][64][65] and neutrophils [6,33,68] during cancer progression. For example, neutrophil activation is associated with the progression of head and neck cancers [68] supporting the idea that neutrophils influence tumor progression. CXCR2 has a known role in the recruitment of neutrophils to cancer cells. Of interest CXCR2 expressing ovarian cancers are aggressive with poor outcomes [69]. However, the mechanisms of these effects remain poorly understood.
Here we show directly that depletion of Cxcr2 in transformed cells results in decreased neutrophil recruitment and impairs the early expression of EMT genes in transformed epithelial cells. Our findings reveal a pro-inflammatory role for Cxcr2 in Ras transformed keratinocytes and provides direct evidence that neutrophils impact the progression of transformed cells in a live animal. It is likely that chronic neutrophil infiltration results in increased amounts of neutrophil derived elastase which has been previously shown to support the progression of pancreatic tumor cells [70] and lung cancer in mice [6]. To test this possibility we exposed zebrafish larvae transiently expressing HRas V12 in the epidermis to a specific inhibitor of human neutrophil elastase (Sivelestat) from 1-3.5 dpf, however we did not observe any difference in either neutrophil recruitment or the transcription of EMT related genes. It is possible that Sivelestat may not be effective in inhibiting zebrafish neutrophil elastase and future experiments are needed to determine the role of neutrophil derived elastase in the transcription of EMT related genes. Additionally, it is also possible that a positive feedback loop between neutrophil derived Cxcl8 and Cxcr2 signaling in the transformed cells may contribute to these changes. This is especially interesting since recent evidence suggests that CXCL8 is associated with progression of some cancers [45].
The zebrafish model is particularly powerful since the transparency allows for the real time visualization of the earliest events that occur after an oncogene turns on. We found that early after HRas V12 was first expressed in epithelial cells both neutrophils and macrophages are recruited to small foci of transformed cells. A previous report showed that transformed melanoblasts in zebrafish also induces early leukocyte recruitment [19]. The previous study showed that hydrogen peroxide generated by the transformed cells mediates leukocyte recruitment. We found that reactive oxygen species inhibition with DPI did not impair neutrophil recruitment to the transformed epithelial cells, but this may have been due to the short time of inhibitor treatment as longer treatments resulted in lethality. It is also possible that different signals mediate neutrophil recruitment to transformed melanocytes versus epithelial cells. Our findings support an essential role for Cxcr2 signaling in chronic tumor associated inflammation.
In summary, here we report a new in vivo model of epithelial cell transformation that is amenable to live imaging and probing the microenvironment. We have identified a pathway that specifically mediates neutrophil but not macrophage recruitment to transformed epithelial cells. We have also provided evidence that Cxcr2 and neutrophils are both necessary for HRas V12induced changes in gene expression in epithelial cells ( Figure 6G). This study highlights a novel cell autonomous role for Cxcr2 signaling in transformed cells in both influencing neutrophil infiltration and EMT-related gene expression, suggesting that both autocrine and paracrine signaling contribute to Cxcr2 effects on EMT. Future studies will be necessary to identify the specific neutrophil factors that influence gene expression induced by oncogenic HRas. It is possible that neutrophils regulate gene expression in transformed epithelial cells at least in part through the release of Cxcl8 that signals to epithelial cell Cxcr2 in a positive feedback loop. The zebrafish model is poised to contribute to a better understanding of the trophic factors involved in the reactivation of EMT programs in vivo, which will likely aid in identifying novel therapeutic targets that modulate EMT with implications to both wound and cancer biology.

Zebrafish maintenance and general procedures
University of Wisconsin -Madison (A3368-01) Institutional review board specifically approved this study. Adult AB fish, including Transgenic (Tg) zebrafish lines Tg(mpx:GFP), Tg(mpeg:- Dendra), Tg(mpx: EGFP-L10a [53]), Tg(krt4: EGFP-L10a [53]) and Tg(mpx:mCherry,rac2 D57N , zf307Tg [57]) were maintained at 28uC in a 14-hour light/10-hour dark cycle in the Research Animal facilities at the University of Wisconsin, which are fully accredited by the American Association for the Accreditation of Laboratory Animal Care. Embryos were staged by both hours post fertilization (hpf) at 28.5uC and by using morphological criteria [71]. To prevent pigment formation, larvae were maintained in E3 containing 0.2 mM N-phenylthiourea (PTU, Sigma Aldrich). For live imaging, 1-3.5 dpf larvae were anesthetized in E3 containing 0.2 mg/mL Tricaine (ethyl 3-aminobenzoate, Sigma Aldrich) and mounted in in 1% low melting point agarose and/or corresponding culture medium. Zebrafish that needed to be euthanized were placed into.05% Tricaine (diluted in E3 water) for 10-15 minutes. After this time the fish were checked to ensure they are not moving or breathing and are then placed into a latex glove for disposal. Alternative way to euthanize: Larval forms between 4 and 10 dpf were maintained in an ice water bath for a minimum of 20 minutes. Death was confirmed when heart and gill movements have ceased.

Tol2 plasmid injections
Zebrafish embryos were microinjected with a pressure injector with approximately 3 nano-liter volumes at the 1-cell stage. All DNA expression vectors contained the krt4 promoter for epithelial cell expression [72,73]. All expression vectors contain minimal Tol2 elements flanking the promoter and gene of interest for efficient integration [74] and an SV40 polyadenylation sequence (Clontech Laboratories, Inc). The following constructs were generated: krt4-RFP-HRAS WT , krt4-RFP-HRAS V12 , krt4-EGFP-L10a and krt4-GFP-H2B. Mosaic expression of transgenes was obtained by injecting 3 nano-liter of solution containing 12.5 ng/mL of Tol2 DNA plasmid and 17.5 ng/mL in vitro transcribed (Ambion) transposase mRNA into the cytoplasm of one-cell stage embryo.

Live imaging
For figure 1A, C, D and E, fluorescence images were acquired using Nikon SMZ-1500 zoom microscope equipped with epifluorescence and a CoolSnap ES camera (Roper Scientific, Duluth, GA). For all confocal imaging 1-3.5 dpf larvae were  Figure 6C-D fluorescence images were acquired using a spinning disk confocal microscope (Yokogawa CSU-X) with confocal scanhead on a Zeiss Observer Z.1 inverted microscope (NA1.3/636 water immersion objective). Maximum intensity projection images were made using the Zen 2012 (blue edition) software (Carl Zeiss). Neutrophils and macrophages were tracked and analyzed by using plugins MTrackj (3D tracking) and Chemotaxis and Migration tool (ibidi) for ImageJ (NIH, Bethesda, MD). Transformed cell 2D area and cell roundness was measured using the analyze function of ImageJ. Data of time-lapse images represent at least three separate movies. For Figure 2H-I and Figure 4L-M WMISH images were acquired using Nikon SMZ-1500 zoom microscope equipped with a color camera.

MO mediated Gene knockdown
Morpholino oligonucleotides (Gene Tools) were suspended in distilled water, and stored at room temperature at a concentration of 1 mM. Zebrafish embryos were microinjected with a pressure injector with approximately 3 nano-liter volumes at the 1-cell stage. For cxcr2 knockdown, a cxcr2 MO targeting the ATG region (59-ACTCTGTAGTAGCAGTTTCCATGTT-39) [18], was used at 100 mM.
For irf8 knockdown, a previously published splice blocking irf8 MO [59] (59-AATGTTTCGCTTACTTTGAAAATGG-39) [18], was used at 100 mM. For knockdown of exogenous Krt4 driven genes, a krt4 MO targeting the region directly upstream of the ATG of targets cloned downstream of Krt4 using the BamHI cloning site (krt4 MO: 59-GCTGCTGAGAGACACGCAGAGG-GAT-39) was used at 20 mM (knockdown through 15 hpf). As a control, Gene Tools standard control morpholino (59-CCTCTTACCTCAGTTACAATTTATA-39) was used at 100 mM. Gene knockdown was obtained by injecting 3 nano-liter of solution into the cytoplasm of one-cell stage embryo.

Whole Mount In situ Hybridization
For in whole mount in situ hybridization, Larvae were fixed in 4% paraformaldehyde in PBS and mRNA was labelled by in situ hybridization as previously described [75]. In short, both Dig and fluorescein-labeled antisense probes were hybridized using a 55u hybridization temperature. Purple color was developed with APconjugated anti-Dig and BM purple (Roche Applied Science), and red color was developed with AP-conjugated anti-fluorescence and fast red (Roche Applied Science). Reactions were stopped in PBS. Imaging was performed with a Nikon SMZ-1500 stereomicroscope.
The T7 promoter was attached 39 of the coding sequence of primers to make the DNA templates for the probes. After sequence confirmation of the DNA templates, labelled RNA probes were transcribed with the use of T7 RNA polymerase (Ambion).

Quantitative PCR
All quantitative PCR was performed using 2 ml purified TRAP RNA as template. Following RNA isolation, one step qPCR was performed using superscript 3 one step qPCR kit (Invitrogen). All experiments had at least two identical samples and were done in three biological replicates with the reference gene ef1a [77] and gene specific primers, which were checked to have produced clean melt curves with one sharp peak indicating specific amplification of the target genes. Fold change was determined using the efficiency-corrected delta comparative quantification method and students t-test (unpaired, two tailed) were performed to determine significance.
Oligo sequences used for qPCR were as follows:

Cell transplants
One cell stage donor embryos were microinjected with approximately 3 nano-liter volumes of injection mixes containing 45 ng/mL in vitro transcribed (Ambion) RFP-HRas V12 mRNA mixed with either 100 mM cxcr2 or control MO. During the blastula period, 10-20 cells were transplanted from donor embryos into transgenic Mpx:GFP hosts. Cell transplants resulted in ,45% success rate of incorporation of RFP labeled donor cells into host embryos. To generate chimeric embryos for TRAP experiments and qPCR analysis one cell stage transgenic krt4-EGFP-L10a donor embryos were injected with 3 nano-liter volumes of injection mixes containing 45 ng/mL in vitro transcribed RFP-HRas V12 mRNA mixed with either 100 mM cxcr2 MO or control MO and during the blastula period 10-20 cells were transplanted into AB host embryos. Of the chimeric embryos generated ,60% had RFP expression in epithelial cells in the trunk region and not in the surrounding tissues. These chimeric zebrafish were used for analysis.

Statistical Analysis
Each in vivo experiment was done at least three times. Dot plots contain data from one representative experiment from at least three biological replicates. Each dot is from one foci of HRas labeled cells in the trunk region of an individual embryo and each embryo is represented by one dot. Assuming Gaussian distribution of overall population of values, P values were derived by the following analyses. One-way ANOVA with Dunnett post-test: Figure 3C and F, Figure 4 D-E and I-J, Figure 5E-F, and Figure 6E. Students T-test (unpaired, two tailed) was used in: Figure 1G, Figure

Supporting Information
Movie S1 Live imaging of HRas WT expressing epithelial cells. Live imaging of epidermal cells in the trunk region of a 3.5 dpf chimeric zebrafish expressing GFP-H2B (green-nucleus) and HRas WT (magenta) in the epidermis. Images were collected at 10 minute intervals for 4 hours. Movie shows the stationary phenotype of HRas WT expressing cells.

(AVI)
Movie S2 Live imaging of HRas V12 expressing epithelial cells. Live imaging of epidermal cells in the trunk region of a 3.5 dpf chimeric zebrafish expressing GFP-H2B (green-nucleus) and HRas V12 (magenta) in the epidermis. Images were collected in 10 minute intervals for 4 hours. Movie shows the dynamic phenotype of HRas V12 expressing cells characterized by cell division and active protrusions. (AVI) Movie S3 Neutrophil recruitment to HRas WT expressing cells. Live imaging of epidermal cells in the trunk region of a 3.5 dpf transgenic mpx:GFP (green-neutrophils) chimeric zebrafish expressing HRas WT (magenta). Images were collected for 30 minutes with 2 minute intervals. Movie shows that neutrophils are not recruited to HRas WT expressing cells. (AVI) Movie S4 Neutrophil recruitment to HRas V12 expressing cells. Live imaging of epidermal cells in the trunk region of a 3.5 dpf transgenic mpx:GFP (green-neutrophils) chimeric zebrafish expressing HRas V12 (magenta). Images were collected for 30 minutes with 2 minute intervals. Movie shows that neutrophils are recruited to HRas V12 expressing cells.

(AVI)
Movie S5 Macrophage recruitment to HRas WT expressing cells. Live imaging of epidermal cells in the trunk region of a 3.5 dpf transgenic mpeg:Dendra (green-macrophages) chimeric zebrafish expressing HRas WT (magenta). Images were collected for 30 minutes with 2 minute intervals. Movie shows that macrophages are not recruited to HRas WT expressing cells.

(AVI)
Movie S6 Macrophage recruitment to HRas V12 expressing cells. Live imaging of epidermal cells in the trunk region of a 3.5 dpf transgenic mpeg:Dendra (green-macrophages) chimeric zebrafish expressing HRas V12 (magenta). Images were collected for 30 minutes with 2 minute intervals. Movie shows that macrophages are recruited to HRas V12 expressing cells. (AVI)