BRIEF COMMUNICATIONS DsRed* erythroid cells circulating in both embryos. This indicates colonizing the CHT and thymus of mib mutants( data not shown) that their vascular systems had anastomosed in the fused region. whereas in parabiotes in which the mib embryo had no blood circu By 4 d p f, 22 of the 24 parabiotes were still alive, and in 18 of lation, we observed only thymus colonization(n=3, Fig. 21). These them DsRed+ cells were still circulating in both embryos results show that Notch signaling deficiency does not affect the We applied this technique to follow the behavior of primitive ability of the Cht and thymus to attract and host wild-type HSPCs hematopoietic cells in parabiotes. We fused two embryos harbor- They also confirm and extend previous results showing that blood different reporter transgenes: Pul: gfp2, which labels primitive circulation is required for CHT but not thymus colonization,6. myeloid cells(macrophages and granulocytes), and gatala: dsred 3 We here demonstrate the usefulness of the zebrafish blastula Fig. 2a-h; note: pul is also known as spilb) At 18.5 h post fertiliza- fusion technique for studying hematopoietic cells and their inter tion(h p f )in parabiotic embryos displaying separate trunks and actions with stromal niches. Beyond hematopoietic cells, this tails, GFP primitive myeloid cells and DsRed* primitive erythroid technique can be used for studying other migratory cells such as and myeloid cells were detected in only the embryo of origin, at neural crest cells or circulating signals and their interaction with their normal location for this developmental stage(Supplementary their target tissue or processes such as innervation or vascular bud Fig 3). By 23 h p f, primitive myeloid cells born in the yolk sac of formation. It can be a powerful alternative to cell transplantation either embryo had started to invade the other embryo(Fig 2a and experiments for investigating cell-autonomous versus non-cell Supplementary Video 2). By 25 h.p. f, blood circulation started, autonomous gene function notably as both situations to be tested and the DsRed+ erythroid cells born in the trunk of the gatala: (mutant-cell migration into wild-type tissue and vice versa)occur dsred embryo were then seen circulating within both embryos, within every parabiotic pair. Finally, reverse genetic tools such as the shared circulation e parabiotes(Fig 2a-h, antisense morpholinos and mRNA injection can be applied to either 9 Supplementary Fig 3 and Supplementary Video 2). By 48 h.p.f., partner before fusion, thereby further extending the range of poten- a GFP+ myeloid cells and DsRed* myeloid and erythroid cells were tially valuable applications of this simple and powerful technique g found in the CHT of both parabiotes(100%, n=35; Fig 2c-h) Next we investigated the migration and homing of defini- METHODS tive HSPCs in parabiotes In the kdr: gfp transgenic line, GFP is Methods and any associated references are available in the online expressed in all vascular endothelial cells and their HSPC progeny version of the pape upon fusion of a kdr: gfp with a nontransgenic wild-type embryo, Note: Supplementary information is available in the online version of the paper. g colonization of the wild-type ChT by GFP+ HSPCs was detected 7 of 7 parabiotic pairs with separate trunks and tails examined, ACKNOWLEDGMENTS 是mxd31A3dpP6p山 expanded over the油可的乙la时t的 starting at 48h. p f. (Fig. 2i), and this HSPCs were found to colo- nize the thymic rudiment of the wild-type parabiote(Fig. 2k and La Recherche sur le Cancer. Z R was supported by a short-term fellowship from 9 a pattern remarkably similar to what has been previously observed the mia itere re e enseignement Sup. ieua et d e a rea chehe and thAe oin dation s the pronephric kidney of the wild-type parabiote(Fig. 2k). Thus, from la Region Languedoc-Roussillon o HSPCs born from the aorta of the karl: fp partner were able to AUTHOR CONTRIBUTIONS each and settle in the successive hematopoietic niches of both K.K. designed the experiments. D.L. D, Z.R. J -M G, M.G. and K.K. performed the parabiotes with the same timeline as in normal development. experiments. K.K. wrote the manuscript with input from D L D. and PH Another useful aspect of this approach is the opportunity to track COMPETING FINANCIAL INTERESTS circulating or migrating cells marked by widely expressed reporter The authors declare no competing financial interests transgenes In pul: gfpllgatala: dsred parabiotes-that is, parabiotes witheachembryoharboringaseparatetransgene-gfp+myeloidPublishedonlineathtpi://www.nature.com/doifinder/10.1038/nmeth.2362. cells could be clearly observed in the atala: dsred tail(Fig. 2f) com/reprints/index. html. without the strong ectopic GFP expression observed in the muscles of the pul gfp embryo(Fig. 2e); conversely, observation of Ds Red+ 1. HerbomeL P, Thisse, B.& Thisse, C. Dew. Biol. 238, 274-288(2001) myeloid and erythroid cells was easier in the pul;gfP partner tail 2. Le Guyader, D. et al. Blood 111, 132-141(2008) 3. Godin, I.& Cumano, A. Nat. Rev. Immunol. 2, 593-604(2002) (Fig. 2c), where DsRed was not expressed in epidermal mucous 4. Murayama, E. et aL Immunity 25, 963-975(2006) cells(Fig 2d). In kdr: gfp//wild-type parabiotes, tracking GFP+ 5. Jin, H, Xu, J& Wen, Z Blood 109, 5208-5214(2007) 6. Gering, M.& Patient, R Dev. Cell 8, 389-400(2005 HSPCs in the CHT, thymus and kidney was also easier in the 7. Kissa, K& Herbomel, P Nature 464, 112-115(2010) nontransgenic parabiote(Fig 2i-k and Supplementary Video 3) 8. Kissa, K et al. Blood 111, 1147-1156(2008) owing to the lack of vascular GFP expression. 9. Goldman, D.C. et al Blood 114, 4393-4401(2009) Last, we applied the blastula fusion technique to study the role of 10. Wright, D.E. Wagers. A.d., Gulati, A P, Johnson, F.L.& Weissman, L.L. in which Notch signaling is din ing the mind bomb(mib)mutant, 11. Dieterlen-Lievre, F. Martin, C. Beaupain, D. Foti biol(Praho)25 genes affecting hematopoiesis upted and definitive HSPCs do not 93-295(1979) form. We fused mib mutant blastulae with cd41: gfp transgenic 12. Hsu, K et al Blood 104, 1291-1297(2004) blastulae, in which HSPCs weakly express GFP6(note: cd41 is 13. Traver, D. et al Nat. Immunol. 4, 1238-1246(2003) Miller, C T, Schilling, T F, Lee, K, Parker, J.& KimmeL, C.B. also known as itga2b) In parabiotes with separate trunks and tails Development127.3815-3828(200 that shared a common bloodstream, we observed GFPlow HSPCs 15. Carmany-Rampey, A.& Moens, C.B. Methods 39, 228-238(2006). NATURE METHODS I ADVANCE ONLINE PUBLICATION3© 2013 Nature America, Inc. All rights reserved. nature methods |  ADVANCE ONLINE PUBLICATION  | brief communications DsRed+ erythroid cells circulating in both embryos. This indicates that their vascular systems had anastomosed in the fused region. By 4 d.p.f., 22 of the 24 parabiotes were still alive, and in 18 of them DsRed+ cells were still circulating in both embryos. We applied this technique to follow the behavior of primitive hematopoietic cells in parabiotes. We fused two embryos harbor￾ing different reporter transgenes: pu1:gfp12, which labels primitive myeloid cells (macrophages and granulocytes), and gata1a:dsred13 (Fig. 2a–h; note: pu1 is also known as spi1b). At 18.5 h post fertiliza￾tion (h.p.f.) in parabiotic embryos displaying separate trunks and tails, GFP+ primitive myeloid cells and DsRed+ primitive erythroid and myeloid cells were detected in only the embryo of origin, at their normal location for this developmental stage (Supplementary Fig. 3). By 23 h.p.f., primitive myeloid cells born in the yolk sac of either embryo had started to invade the other embryo (Fig. 2a and Supplementary Video 2). By 25 h.p.f., blood circulation started, and the DsRed+ erythroid cells born in the trunk of the gata1a: dsred embryo were then seen circulating within both embryos, demonstrating the shared circulation of the parabiotes (Fig. 2a–h, Supplementary Fig. 3 and Supplementary Video 2). By 48 h.p.f., GFP+ myeloid cells and DsRed+ myeloid and erythroid cells were found in the CHT of both parabiotes (100%, n = 35; Fig. 2c–h). Next we investigated the migration and homing of defini￾tive HSPCs in parabiotes. In the kdrl:gfp transgenic line, GFP is expressed in all vascular endothelial cells and their HSPC progeny5. Upon fusion of a kdrl:gfp with a nontransgenic wild-type embryo, colonization of the wild-type CHT by GFP+ HSPCs was detected in 7 of 7 parabiotic pairs with separate trunks and tails examined, starting at 48 h.p.f. (Fig. 2i), and this population expanded over the next days (Fig. 2j). At 3 d.p.f., GFP+ HSPCs were found to colo￾nize the thymic rudiment of the wild-type parabiote (Fig. 2k and Supplementary Video 3) by migrating through the mesenchyme in a pattern remarkably similar to what has been previously observed in unmanipulated larvae6. By 4.75 d.p.f., HSPCs were detected in the pronephric kidney of the wild-type parabiote (Fig. 2k). Thus, HSPCs born from the aorta of the kdrl:gfp partner were able to reach and settle in the successive hematopoietic niches of both parabiotes with the same timeline as in normal development. Another useful aspect of this approach is the opportunity to track circulating or migrating cells marked by widely expressed reporter transgenes. In pu1:gfp//gata1a:dsred parabiotes—that is, parabiotes with each embryo harboring a separate transgene—GFP+ myeloid cells could be clearly observed in the gata1a:dsred tail (Fig. 2f) without the strong ectopic GFP expression observed in the muscles of the pu1:gfp embryo (Fig. 2e); conversely, observation of DsRed+ myeloid and erythroid cells was easier in the pu1:gfp partner tail (Fig. 2c), where DsRed was not expressed in epidermal mucous cells (Fig. 2d). In kdrl:gfp//wild-type parabiotes, tracking GFP+ HSPCs in the CHT, thymus and kidney was also easier in the nontransgenic parabiote (Fig. 2i–k and Supplementary Video 3) owing to the lack of vascular GFP expression. Last, we applied the blastula fusion technique to study the role of genes affecting hematopoiesis using the mind bomb (mib) mutant, in which Notch signaling is disrupted and definitive HSPCs do not form4. We fused mib mutant blastulae with cd41:gfp transgenic blastulae, in which HSPCs weakly express GFP6 (note: cd41 is also known as itga2b). In parabiotes with separate trunks and tails that shared a common bloodstream, we observed GFPlow HSPCs colonizing the CHT and thymus of mib mutants (data not shown), whereas in parabiotes in which the mib embryo had no blood circu￾lation, we observed only thymus colonization (n = 3, Fig. 2l). These results show that Notch signaling deficiency does not affect the ability of the CHT and thymus to attract and host wild-type HSPCs. They also confirm and extend previous results showing that blood circulation is required for CHT but not thymus colonization2,6. We here demonstrate the usefulness of the zebrafish blastula fusion technique for studying hematopoietic cells and their inter￾actions with stromal niches. Beyond hematopoietic cells, this technique can be used for studying other migratory cells such as neural crest cells or circulating signals and their interaction with their target tissue or processes such as innervation or vascular bud formation. It can be a powerful alternative to cell transplantation experiments14 for investigating cell-autonomous versus non–cell autonomous gene function15, notably as both situations to be tested (mutant-cell migration into wild-type tissue and vice versa) occur within every parabiotic pair. Finally, reverse genetic tools such as antisense morpholinos and mRNA injection can be applied to either partner before fusion, thereby further extending the range of poten￾tially valuable applications of this simple and powerful technique. Methods Methods and any associated references are available in the online version of the paper. Note: Supplementary information is available in the online version of the paper. Acknowledgments We thank B. Robert, Y. Lallemand, D. Montarras and T. Schilling for their critical reading of the manuscript. This work was partially supported by the Caisse Autonome Nationale de Sécurité Sociale dans les Mines and the Association pour la Recherche sur le Cancer. Z.R. was supported by a short-term fellowship from the Weissman International Internship Program; D.L.D. by PhD fellowships from the Ministère de l’Enseignement Supérieur et de la Recherche and the Fondation pour la Recherche Médicale; and J.-M.G. and M.G. by a Chercheur d’Avenir grant from la Région Languedoc-Roussillon. AUTHOR CONTRIBUTIONS K.K. designed the experiments. D.L.D., Z.R., J.-M.G., M.G. and K.K. performed the experiments. K.K. wrote the manuscript with input from D.L.D. and P.H. COMPETING FINANCIAL INTERESTS The authors declare no competing financial interests. Published online at http://www.nature.com/doifinder/10.1038/nmeth.2362. Reprints and permissions information is available online at http://www.nature. com/reprints/index.html. 1. Herbomel, P., Thisse, B. & Thisse, C. Dev. Biol. 238, 274–288 (2001). 2. Le Guyader, D. et al. Blood 111, 132–141 (2008). 3. Godin, I. & Cumano, A. Nat. Rev. Immunol. 2, 593–604 (2002). 4. Murayama, E. et al. Immunity 25, 963–975 (2006). 5. Jin, H., Xu, J. & Wen, Z. Blood 109, 5208–5214 (2007). 6. Gering, M. & Patient, R. Dev. Cell 8, 389–400 (2005). 7. Kissa, K. & Herbomel, P. Nature 464, 112–115 (2010). 8. Kissa, K. et al. Blood 111, 1147–1156 (2008). 9. Goldman, D.C. et al. Blood 114, 4393–4401 (2009). 10. Wright, D.E., Wagers, A.J., Gulati, A.P., Johnson, F.L. & Weissman, I.L. Science 294, 1933–1936 (2001). 11. Dieterlen-Lièvre, F., Martin, C. & Beaupain, D. Folia biol. (Praha) 25, 293–295 (1979). 12. Hsu, K. et al. Blood 104, 1291–1297 (2004). 13. Traver, D. et al. Nat. Immunol. 4, 1238–1246 (2003). 14. Miller, C.T., Schilling, T.F., Lee, K., Parker, J. & Kimmel, C.B. Development 127, 3815–3828 (2000). 15. Carmany-Rampey, A. & Moens, C.B. Methods 39, 228–238 (2006)