Role of the Bone Morphogenetic Protein Pathway in Definitive Endoderm Patterning

Role of the Bone Morphogenetic Protein Pathway in Definitive Endoderm Patterning Carla Gonçalves Biologia Celular e Molecular Departamento de Biologia 2015 Orientadores Professora Anne Grapin-Botton, DanStem, Universidade de Copenhaga Doutora Laurence Lemaire, DanStem, Universidade de Copenhaga Coorientador Professor José Pissarra, Faculdade de Biologia, Universidade do Porto Todas as correções determinadas pelo júri, e só essas, foram efetuadas. O Presidente do Júri, Porto, ____/____/_______ FCUP Role of the BMP Pathway in Definitive Endoderm Patterning ii Acknowledgements As my first year at DanStem comes to an end, it seems awfully hard to address all the people that helped me make this work a reality. First and foremost, I have to thank Anne, because she believed in me from the start and has always been inspiring as a scientist and as a person. I feel so lucky to work with her. Next, I have to thank Laurence for all her invaluable help in and out of the bench. She was always invested in my project, even if her work was already so much to take. I want to thank all the members of the Grapin-Botton group for discussions, advice, for making me grow so much overall as a scientist. And to everyone in DanStem, thank you for making this place so wonderful to work in, thank you for making me feel welcome and thank you for your science! Agora em português, para os portugueses. Primeiro de tudo, agradeço aos meus pais por me terem ensinado a ser inquisidora e confiante, capacidades tão importantes para um cientista. Os meus pais que tão bem souberam dar-me asas para voar e mostrar-me um mundo a conquistar. Agradeço a todos os meus queridos amigos, quase irmãos, que me acompanharam através de lágrimas, sorrisos e suor. Não há palavras suficientes para explicar a saudade que senti durante este ano. E por último, não porque menos importante, quero agradecer ao Ricardo. Ele que sempre soube fazer-me sorrir quando não parecia possível, mesmo a milhares de quilómetros de distância. FCUP Role of the BMP Pathway in Definitive Endoderm Patterning iii Table of Contents Resumo .......................................................................................................................... 1 Abstract ........................................................................................................................... 2 Figure Index .................................................................................................................... 3 Table Index ..................................................................................................................... 4 Glossary .......................................................................................................................... 5 Chapter I. Introduction ................................................................................................. 7 1.1. Formation of the Primitive Gut Tube .................................................................... 7 1.2. Patterning of the Primitive Gut Tube .................................................................... 8 1.3. BMP Signalling ..................................................................................................... 9 1.4. Bmp receptors: Alk2, Alk3 and Alk6 ................................................................... 11 1.5. BMP signalling in patterning of the gut tube ....................................................... 11 1.6. Sox17 Expression .............................................................................................. 13 1.7. The Cre Recombinase System .......................................................................... 14 1.8. General Aims and Strategy ................................................................................ 16 Chapter II. Materials and Methods ............................................................................ 19 2.1. Mouse breeding and genotyping ........................................................................ 19 2.2. Immunofluorescence .......................................................................................... 20 2.2.1. Immunofluorescence on sections - General protocol .................................. 20 2.2.2. Antigen Retrieval ......................................................................................... 21 2.2.3. Antibody Stripping ....................................................................................... 21 1.2.4. Phosphatase Assay ..................................................................................... 22 1.2.5. Wholemount Immunofluorescence .............................................................. 22 Chapter III. Results and discussion .......................................................................... 25 3.1. Characterization of the new Sox17CreERT2 line ............................................... 25 3.2. Expression of molecular markers in the gut ....................................................... 33 3.2.1. Nkx2.1 and Sox2 ......................................................................................... 33 FCUP Role of the BMP Pathway in Definitive Endoderm Patterning iv 3.2.2. Gcm2 and Foxn1 ......................................................................................... 34 3.2.3. Prox1 ........................................................................................................... 35 3.2.4. Hlxb9 ........................................................................................................... 36 3.2.5. Pitx2 ............................................................................................................. 36 3.2.6. Effectors downstream of BMP: pSMAD1/5/8 ............................................... 36 3.2.7. Sequential immunofluorescence ................................................................. 37 3.3. Inactivation of the BMP pathway ........................................................................ 40 3.3.1. Inactivation of both receptors is lethal before E10.5 .................................... 40 3.3.2. Hz embryos have several organ development defects ................................ 41 Chapter IV. Final Remarks ......................................................................................... 50 Cited Literature ............................................................................................................. 51 FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 1 Resumo A Endoderme Definitiva (DE) é uma das três camadas germinativas formadas durante a gastrulação e origina o tubo digestivo primitivo e os orgãos a este associados. Durante o desenvolvimento, cada região do tudo digestivo primitivo é caracterizada por motivos de expressão distintos que determinam a localização exacta onde os diferentes orgãos deverão surgir ao longo do eixo antero-posterior. Ao mesmo tempo, o tubo digestivo primitivo também é especificado ao longo to eixo dorso-ventral, e a maioria dos orgãos surgem ventralmente. O mecanismo que leva à formação do eixo dorso-ventral no tubo digestivo primitivo é ainda mal compreendido. No tubo neural este processo foi extensivamente estudado e é descrito como um equilibro entre a expressão de Bone Morphogenetic Protein (BMP)4 - dorsalmente - e de Sonic Hedgehog (SHH) - ventralmente. Para além disso, diversos estudos indicam que sinalização pela BMP é necessária para a formação de alguns orgãos ventrais associados ao tubo digestivo primitivo. Tendo isto em conta, propomos que a via de sinalização da BMP poderá ser responsável por conferir identidade ventral ao tubo digestivo. Neste projecto, caracterizamos um ratinho mutante previamente gerado pelo nosso grupo, em que CreERT2 é expressa sobre o controlo do promotor de Sox17. Este fator de transcrição é expresso na DE antes da formação do tubo digestivo primitivo. Após a caracterização, o ratinho foi utilizado para inativar condicionalmente a via de sinalização da BMP no tubo digestivo, através de inativação dos receptores principais da BMP. Os resultados preliminares estão geralmente de acordo com outros estudos que se focaram no papel da BMP na formação do tubo digestivo e orgão adjacentes. Para além disso, os resultados sugerem que a via de sinalização da BMP é essencial para a formação de orgãos ventrais como os pulmões e o primórdio pancreático ventral, algo que não foi previamente descrito. Palavras chave: Endoderme definitiva, eixo dorso-ventral, BMP, Sox17 FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 2 Abstract The definitive endoderm (DE) is one of three germ layers segregated during gastrulation and it gives rise to the primitive gut tube and all associated organs. During development, each region of the primitive gut tube is characterized by unique expression patterns that determine the precise locations where organs emerge along the anterior-posterior axis. At the same time, the primitive gut is also patterned along the dorsal-ventral axis, with most organs emerging on the ventral side of the gut. Dorsal-ventral patterning is still poorly understood in the gut tube. In the neural tube, dorsal-ventral patterning is well described as a balance between expression of Bone Morphogenetic Protein (BMP)4 - dorsally - and Sonic Hedgehog (SHH) - ventrally. Furthermore, several distinct studies indicate that BMP is required in the formation of some of the ventral gut organs. Taking this into consideration, we theorized that BMP signalling could be a general cue providing ventral identity to the gut tube. In this study, a mouse line previously generated by our group is characterized in which CreERT2 is expressed under the control of the promoter of Sox17. This transcription factor is expressed in definitive endoderm cells at a very early stage, previous to primitive gut formation. Thereafter, this mouse line is used as a tool in order to conditionally inactivate the BMP pathway in the DE, by knocking out its main receptors. Preliminary data are generally in accordance with previous observations made for the role of BMP in gut patterning and organ formation. Furthermore, we find BMP signalling to be essential for the formation of ventral organs like the lungs and the ventral pancreas, which had not been reported previously. Key words: Definitive endoderm, Patterning, BMP, Sox17 FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 3 Figure Index Figure 1.1. Gastrulating mouse embryo. (Adapted from Grapin-Botton2) Figure 1.2. Gut tube closure and turning. (Adapted from Grapin-Botton and Melton3) Figure 1.3. BMP signalling cascade (Balemans and Van Hul4) Figure 1.4. Primitive gut tube and associated organs (Adapted from Zorn and Wells5) Figure 1.5. Mirror DV patterning in the neural tube and the gut tube Figure 1.6. Diagram of the Sox17 locus, targeting vector, Sox17LCA allele, CreERT2 exchange cassette, Sox17GFPCre(þHygroR), and Sox17CreERT2 allele. (Adapted from Choi, et al1) Figure 3.1. Recombination rates in E12.5 embryos after administration of out of date tamoxifen at E7.5 Figure 3.2. Recombination rates in E9.5 embryos after administration of tamoxifen at E7.5 Figure 3.3. Recombination rates in E9.5 and E12.5 embryos after administration of tamoxifen twice, at E6.5 and E7.5 Figure 3.4. Recombination rates in E14.5 embryos after administration of tamoxifen at E10.5 Figure 3.5. Recombinantion rates in E14.5 pancreas after administration of tamoxifen at E10.5 Figure 3.6. NKX2.1 in WT E10.5 embryos Figure 3.7. SOX2 in WT E10.5 embryos Figure 3.8. GCM2 in WT E10.5 embryos Figure 3.9. PROX1 in WT E10.5 embryos Figure 3.10. HLXB9 in WT E10.5 embryos Figure 3.11. pSMAD1/5/8 in WT embryos and phosphatase treatment Figure 3.12. pSMAD1/5/8 in a E10.5 embryo after stripping primary antibodies Figure 3.13. Percentages of genotypes obtained overall compared to the theoretical percentages at E10.5. Figure 3.14. Three-dimensional projection of WT (A) and Hz (B) littermates, at E10.5 Figure 3.15. Comparison of lung and liver in the WT and Hz littermates at E10.5 Figure 3.16. Comparison of foregut in the WT and Hz littermates at E10.5 Figure 3.17. Surface rendering of prox1 expression domain Figure 3.18. Total Prox1 volumes in the Hz embryos (% of WT littermate) Figure 3.19. Comparison of dorsal pancreas in the WT and Hz littermates at E10.5. FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 4 Table Index Table 2.1 Breeding scheme for BMP mutants Table 2.2. Genotyping primers Table 2.3. Primary Antibodies Table 2.4. Secondary Antibodies Table 3.1. Summary of characterization of the Sox17CreERT2 line Table 3.2. Molecular markers expressed in the primitive gut FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 11 1.4. Bmp receptors: Alk2, Alk3 and Alk6 The abundance of ligands in the BMP family is not matched by similar numbers of receptors. Thus the BMP pathway has highly promiscuous ligand-receptor interactions. Two Type I receptors - BMPR1A;ALK3 and BMPR1B;ALK6 - are known to translate signals from BMP2/421. BMP4 has also been reported to bind to Type I A activin receptor ActR-I;ALK2 in the visceral endoderm (VE), although this pathway leads only to the activation of SMAD1/5 and not SMAD825. This difference in the activation pattern of R-Smads by ALK3 and ALK6 or ALK2 is translated into distinct biological responses26. In contrast to the localized expression patterns of BMP ligands, expression of BMP receptors is widespread during early embryonic mouse development. Alk3 and BmpRII are expressed in most tissues throughout development. Alk6 expression starts later during development, after the gastrula stage. It is first observed at E7.5 in the AIP and is later expressed along the endoderm and in the liver primordium at E9.0. This expression pattern is maintained until E10.5, albeit at lower levels22,27. Alk2 is expressed in pre-gastrulating and gastrulating embryos, mostly in extra-embryonic tissues (VE, chorion, amnion), and later, at E10.5, it is expressed in the head mesoderm as well as in the endocardium28. Mice lacking Alk6 show only mild skeletal defects in the adult29. However, mouse embryos lacking Bmp4, Alk3, or BmprII are arrested at gastrulation, and mesoderm does not form23,30,31. 1.5. BMP signalling in patterning of the gut tube Very little is known about how the DE-derived organs acquire their positions along the DV axis in mammals. During organ specification when the DE still forms a sheet, the morphogens giving this information, if any, should first be positioned along the medio-lateral axis and after, the gut tub is closed, along the DV axis. Contrary to the endoderm, this process is well characterized in the ectoderm and seems to be a conserved system across vertebrates and insects, where antagonistic secreted factors determine first medial or lateral identity and later dorsal or ventral identity. For example, BMP4 orthologs found in fly and in vertebrates share similar functions and mechanisms for medio-lateral and later DV patterning in the ectoderm32. In mice, opposing gradients of sonic hedgehog (SHH) and bone morphogenetic protein (BMP) signalling are involved. BMP signalling is necessary to form laterally the non-neural ectoderm or surface ectoderm which becomes thereafter located dorsally in the neural tube. After FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 12 the neural tube is closed, the roof plate situated dorsally becomes a new organizing centre that produces BMPs which induce a dorsal fate in the interneurons present in the dorsal side of the neural tube (reviewed in Liu and Niswander 33). In the endoderm, several studies describe a role of BMP in the formation of some ventral organs all along the anterior-posterior axis (Figure 1.4). The thymus, which emerges on the ventral domain of the thymus-parathyroid primordium, is severely reduced when BMP signalling is inhibited by Noggin, an antagonist of BMP. Besides, it does not reach its final destination in the mediastinum34. Conversely, the presence of BMP4 in the dorsal domain where the parathyroid originates, reduces the expression of the parathyroid specific marker Gcm235. BMP4 is necessary for the formation of the trachea in the ventral foregut 36. The oesophagus does not form in absence of Noggin, a BMP antagonist. If BMP signalling is disrupted after lung specification, lung development is delayed and less branches are formed 37In the absence of BMP4, the liver development is also delayed. In E9.5 mutant embryos, the hepatic epithelium did not yet bud contrary to their WT littermates 38. Moreover, in vitro specification of hepatocytes from ESC-derived DE cells requires the presence of BMP439. More globally, Xenopus ectodermal explants adopt a ventrolateral endodermal fate when Bmp4 is overexpressed 15. These observations prompt the hypothesis that in mouse BMP signalling is a global cue that initially induces lateral identity to the DE and after gut tub closure, ventral identity forming a mirror-image of ectodermal and neural tube patterning. For the DE, BMPs secreted by the mesoderm would induce cell fate in the lateral regions of the DE which progressively join at the midline to form a tube. After gut closure, BMP Figure 1.4. Primitive gut and associated organs. At E10.5 all organs associated to the gut tube have started to develop with a strict organization along the AP and DV axis. FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 13 Figure 1.5. Mirror DV patterning in the neural tube and the gut tube. A. BMP is expressed in the lateral mesoderm (red) and secreted and patterns lateral ectoderm, while SHH secreted by the notochord (red dot) patterns medial ectoderm (blue). B. During the dorsal folding of the neural tube, the lateral ectoderm becomes the dorsal non-neural ectoderm (skin) and medial ectoderm forms the neural tube. In the neural tube, BMP starts being secreted in the roof plate (yellow) and patterns the dorsal region of the neural tube, opposed by SHH in the ventral region. We hypothesize that concurrently, BMP secreted in the mesoderm patterns lateral DE (orange) while SHH secreted in the notochord may pattern medial DE. C. The gut tube folds in a mirror image to the neural tube and thus lateral tissues end-up ventrally in the tube. After ventral closure, SHH secreted by the notochord may pattern the dorsal region of the primitive gut tube and BMP secreted in the mesoderm may pattern the ventral region. BMP secretion may form a medio-lateral and later dorsal-ventral gradient. Black arrows - BMP signalling; Blue arrows - neural tube closure; Orange arrows - gut tube closure. D - Dorsal; V - Ventral; M - Medial; L - Lateral. would act on the ventral side of the tube. Therefore, BMP signalling would be required for the ventral identity of the gut (Figure1.5). However, the requirement of BMP signalling during gastrulation precludes further study during endoderm patterning. In order to explore the role of BMP in DV patterning of the gut tube, a time and tissue specific inactivation of the pathway is necessary. For example, in the respiratory tract, conditional inactivation of the BMP pathway was previously achieved by inactivating its two well characterized receptors - Alk3 and Alk6 in the future lung epithelium after organ specification37. Our aim was to inactivate it in the endoderm. 1.6. Sox17 Expression The SRY (sex determing region Y)-box 17 (SOX17) is a transcription factor belonging to the Sox protein family. Sox proteins share similar DNA binding properties, FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 14 however individual Sox proteins appear to regulate specific sets of target genes in vivo due to restricted patterns of expression and in combination with specific cofactors interactions. Sox proteins are key players in the regulation of embryonic development and determination of cell fate (reviewed in Lefebvre, et al. 40). SOX17 is first expressed at the blastocyst stage between E3.25 and E4.5, in a salt and pepper pattern, where it promotes the primitive endoderm cell fate over the epiblast fate41. Subsequently, it is also expressed in the visceral endoderm around E6.0-E6.5 42. SOX17 is dynamically expressed in the DE, forming a temporal and spatial wave of expression from the anterior to the posterior region of this tissue. At first, expression is detected in the anterior end of the primitive streak at E7.0, which coincides with the time of ingression of future anterior DE cells. By E8.0 SOX17 is observed in the prospective posterior gut, while its expression in the foregut is already reduced. Its expression in the hindgut shuts down around E9.0. SOX17 is a key player in the definitive endoderm development as it is necessary for its specification42,43. After E9.0, SOX17 expression is observed in the hemogenic endothelial cells (ECs), which are of mesodermal origin44. SOX17 is then necessary for definitive hematopoiesis and the maintenance of the hematopoietic stem cell pool, both at the fetal and neonatal stages1,45. Notably, at around E9.5, SOX17 is again expressed in the ventrolateral region of the most posterior foregut, where the bile duct and gall bladder originate, and persists until at least E15.546. Based on this expression pattern, it appears that Sox17 may be used as a driver to target gene inactivation widely in endoderm. 1.7. The Cre Recombinase System Cre (cyclization recombination) gene encodes a site-specific DNA recombinase of the bacteriophage P1 which is required for the circularisation of the phage DNA - a critical step in the bacteriophage life-cycle. The enzyme recognizes a specific sequence of 34-bp, termed loxP, and catalyses both intra and intermolecular recombination between two loxP sites. Cre–loxP mediated recombination between two directly repeated loxP sites excises all DNA sequences located within the two sites as a covalently bound circular molecule47. The conditional deletion of a gene in mice (conditional knock-out) is achieved by excising with a Cre the gene flanked by two LoxP sites, also called floxed gene. The gene promoter driving Cre expression determines tissue or stage specificity. Temporal control of the onset of the mutation can be further achieved by using a Cre fused to a modified Estrogen Receptor (ERT2). CreERT2 is sequestered in the cytoplasm unless FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 15 tamoxifen, an estrogen analogue, is present and has been metabolically activated in the liver48. Sox17 promoter has already been used to drive the expression of Cre recombinase in the endoderm and in the hemogenic endothelial cells1,49. As it is specifically expressed in the DE cells at a particular stage of development, it is possible to target these cells using a CreERT2 system. This strategy has already been confirmed by generating a Sox17CreERT2 mouse line50. Another Sox17CreERT2 line has contemporarily been developed in our laboratory. The Cre recombinase fused to an estrogen receptor has been targeted to the Sox17 locus disrupting the gene after the second exon, in contrast to the previously published line (Figure 1.6) (Marine Rentler-Courdier-Kraus, unpublished data). Figure 1.6. Diagram of the Sox17 locus, targeting vector, Sox17LCA allele, CreERT2 exchange cassette, Sox17GFPCre(þHygroR), and Sox17CreERT2 allele. A targeting vector for the mouse Sox17 gene was constructed where the sequence including exons 3–5, which contains the coding region of Sox17, was replaced with a puromycin resistance-D-thymidine kinase fusion gene (puDTK) and an EM7-driven kanamycin resistance gene (KanR) flanked by lox66 (open triangle) and lox2272 (black triangle) sites. The GFPCre exchange cassette was flanked by lox71 (gray triangle) and lox2272 sites and contained a phosphoglycerol kinase-driven hygromycin resistance gene (HygroR) flanked by flippase recognition target sites (open circles). This prepares the locus to easy replacement by any insertion and was previously used to insert a CreGFP fusion1. Following exchange into Sox17LCA-containing mouse embryonic stem cells by recombinase-mediated cassette exchange (RMCE), mice containing the Sox17CreERT2-(þHygroR) allele, were bred with FLPe-expressing transgenic mice, thereby generating the final Sox17CreERT2 allele. Abbreviations: DT-A, Diphtheria toxin A; LA, long arm; LCA, loxed cassette acceptor; SA, short arm. FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 16 1.8. General Aims and Strategy Prior evidence suggests a role of the BMP pathway in the formation of several ventral endodermal organs. We hypothesised that BMP signalling acts as global ventralizing factor in the gut, mirroring its action in the neural tube where it is necessary for dorsal identity (Figure 1.3). In the present study, we characterized the Sox17CreERT2 mouse line that was previously generated in our lab in order to determine the most reliable way to induce recombination in the DE without affecting the other tissues. The newly characterized mouse line was then used to inactivate the BMP signalling pathway in the DE by deleting Alk3 using the Sox17CreERT2, in an Alk6 null background. The outcome of the inactivation has been thoroughly analysed by whole mount imaging. For this purpose, efficient labelling of dorsal and ventral endoderm organ primordia was established. FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 17 FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 18 Chapter II Materials and Methods FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 19 Materials and Methods 2.1. Mouse breeding and genotyping The Sox17CreERT2 allele was maintained within an ICR background for experiments. Mice with the Rosa26YFP, Alk3flox or Alk6allele were previously described29,30,51 . The mating scheme used in order to obtain BMP pathway mutants is described in Table 2.1, along with the ratios of each genotype obtained in the progeny and the designations attributed for simplicity. Mice were housed at the University of Copenhagen. The Dyreforsøgstilsynet approved the mouse housing and experiments. Midnight before a vaginal plug was observed was considered as the time of fertilization (E0). Pregnant females received an intraperitoneal injection of warmed tamoxifen (Sigma) dissolved in corn oil at a concentration of 10mg/mL, with a 25 gauge needle. Each female was weighed before injection and the volume of injected tamoxifen was calculated accordingly. To collect the embryos at the different time points (E9.5, E10.5, E12.5), the females were euthanized by cervical dislocation and the embryos were dissected out of the uterus in PBS. A part of the yolk sac was removed and used for genotyping. The embryos were fixed in PFA 4% (Sigma) for 2 hours on ice with shaking. Table 2.1. Breeding scheme for BMP mutants Parent 2 Alleles Parent 1 Alleles Alk3fl Alk6 + Sox17 + Alk3fl Alk6 - Sox17 + Alk3fl Alk6+ Sox17CreERT2 Alk3fl/fl Alk6+/+ Sox17CreERT2/+ Alk3fl/fl Alk6+/- Sox17CreERT2/+ 1/8 1/4 Alk3 KO Hz Alk3fl Alk6Sox17CreERT2 Alk3fl/fl Alk6+/- Sox17CreERT2/+ Alk3fl/fl Alk6-/- Sox17CreERT2/+ 1/4 1/8 Hz dKO Alk3fl Alk6+ Sox17+ Alk3fl/fl Alk6+/+ Sox17+/+ Alk3fl/fl Alk6+/- Sox17+/+ 1/8 1/4 WT Alk6 +/- Alk3fl Alk6Sox17+ Alk3fl/fl Alk6+/- Sox17+/+ Alk3fl/fl Alk6-/- Sox17+/+ 1/4 1/8 Alk6 +/- Alk6 KO FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 20 PCR genotyping was performed on tail tip genomic DNA and embryonic tissue after lysis in 100µL PCR direct tail buffer (Viagen) containing 2,5µL 20,6 mg/mL Proteinase K (Roche) overnight at 55ºC, followed by Proteinase K heat inactivation for 45 min at 85ºC. Each PCR reaction contained 2 to 4 µL of genomic DNA digestion solution, 5 µL of 5x Green GoTaq buffer (Promega), 1 µL of 5mM dNTPs (Thermo Fisher Scientific), 1 µL of 10 µM primer stock (Table 2.1), 0,2 µL of 5u/µL GoTaq enzyme (Promega) and miliQ to a final volume of 25 µL. All primers were ordered from Integrated DNA Technologies. Table 2.2 Genotyping Primers Locus Primers Annealing temperature Cº product size WT mutant Sox17 5'-TGCCAC GACCAAGTGACAGC-3' 58 no product 700 5'-CCAGGTTACGATAT AGTTCATG-3' Rosa26 5'- AAAGTCGCTCTGAGTTGTTAT-3 58 600 300 5'-GCGAAGAGTTTGTCC TCAACC-3' 5'-GGAGCGGG AGAAATGGATATG-3' Alk3 5'-GCAGCTG CTGCTGCAGCCTCC -3' 50 350 600 5'-TGGCTACAATTTGTCT CATGC-3' Alk6 5'-CCCAAGATCCTACGT TGTAA-3' 62 150 230 5'-GAGTGGTTACAACAAGATC AGC A-3' 5'-GCCCTGAATG AACTGCA GG-3' Electrophoresis gels were prepared with 2% ultra pure agarose (Thermo Fisher Sci.) in TAE 1x (in house) containing 0,003% ethidium bromide (Thermo Fisher Sci.). In each well, 6 µL of the PCR reaction was loaded as well as a 1kb DNA ladder (Thermofisher Scientific). The gels were submitted to 80mV voltage on a standard power pack p25 (Biometra) for roughly 20 min and imaged with a molecular imager (GelDoc XR+, BioRad). 2.2. Immunofluorescence 2.2.1. Immunofluorescence on sections - General protocol Fixed embryos were thoroughly washed with PBS (1.8 mM KH2PO4, 10 mM Na2HPO4, 137 mM NaCl, 2.7 mM KCl pH7.4)a and incubated in a solution of 0.12M a Steps where the temperature is not mentioned were performed at room temperature. FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 27 activation and by the dynamic expression of Sox17. But also, considering the rapid shift in Sox17 expression, it is likely that even small discrepancies in the time of tamoxifen injection will result in the recombination of different groups of cells. Sox17 expression is first found in the prospective foregut (E7.0) and gradually shifted until it is only found in the prospective hindgut (E9.0)42. In line with the dynamic expression pattern, it was observed that often more cells were recombined in the posterior region of the gut than in the foregut (Figure 3.2 A-C), which indicates that the DE cells of the foregut were no longer Sox17+ at the time of Cre activation. According to the previous experiment, an earlier injection than E7.5 would be needed in order to induce recombination in the prospective foregut cells. Preliminary experiments in the lab have shown that injection of 0.7mg/10g at E6.5 leads to high recombination rate (around 90%) in the whole gut at E9.5 but results in embryonic lethality since no embryo survived after E10.5 (unpublished data). It is noteworthy that the tamoxifen solution used for this experiment was not fresh. Therefore, a strategy Figure 3.2. Recombination rates in E9.5 embryos after administration of tamoxifen at E7.5. (A-C) Immunofluorescence for E-Cadherin and GFP on sections of Sox17CreERT2/+ RosaYFP/+ E9.5 embryos shows the recombination when tamoxifen is injected at E7.5 with a dose of 0,5 mg/10g or (D) a dose of 0,6 mg/10g. Few cells of the gut endoderm highlighted by E Cadherin (red) were recombined and expressed YFP detected by the GFP antibody. Often, more cells were recombined in the posterior region gut as exemplified in the hindgut compared to the foregut (A and C). Nuclei are counterstained with DAPI (blue). Abbreviations: es - esophagus; tr - trachea; vp - ventral pancreas; mg - midgut. Scale bars - 100 µm. FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 28 combining lower doses of tamoxifen (0.4mg/10g) injected at two different time points was undertaken in order to improve the recombination in Sox17+ DE cells and the viability of the embryos. Using this approach, we observed that between 70 to 90% of the cells in the gut endoderm were recombined both at E9.5 (n=2) and E12.5 (n=4). For example, most of the cells in the liver primordium were recombined at E9.5 (Figure 3.3 - B) and high recombination rates in the pancreas were observed at E12.5 (Figure 3.3 - E). The recombination in the vasculature might be caused by the presence of tamoxifen metabolites long after injection. The precise length of time that tamoxifen continues to induce recombination is highly variable depending on the dose and mode of administration54,55. Nevertheless, the low rates of recombination observed in the vasculature (~1%) in this experiment would be unlikely to interfere when analysing DE Figure 3.3. Recombination rates in E9.5 and E12.5 embryos after administration of tamoxifen twice, at E6.5 and E7.5. (A-C) Immunofluorescence for E-Cadherin and GFP on sections of Sox17CreERT2/+ RosaYFP/+ E9.5 embryos shows the recombination when tamoxifen is injected at E6.5 and E7.5 with a dose of 0,4mg/10g/day. Around 70 to 90% of the gut endoderm cells highlighted by E Cadherin (red) were recombined and expressed YFP detected by the GFP antibody as exemplified in the foregut (A), the liver primordium (B) and the hindgut (C). (D-F) Immunofluorescence for E-Cadherin and GFP on sections of Sox17CreERT2/+ RosaYFP/+ E12.5 embryos shows the recombination when tamoxifen is injected at E6.5 and E7.5 with a dose of 0,4mg/10g/day. Around 70 to 90% of the gut endoderm highlighted by E Cadherin (red) were recombined and expressed YFP detected by the GFP antibody as exemplified in the esophagus and main bronchi (D), dorsal pancreas (E) and the midgut (F). Nuclei are counterstained with DAPI (blue). Abbreviations: br - bronchi; es - esophagus; dp - dorsal pancreas; mg - midgut. Scale bars - 100 µm. FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 29 lineage-restricted conditional mutants. This double injection strategy was found to be the most effective and reliable way to induce DE specific Cre activity with this new Sox17CreERT2 line. After E9.0, Sox17 expression is found on a subset of endothelial cells of the blood vessels, including the dorsal aorta45. When Sox17 is no longer expressed in the DE, it is re-expressed in the ventrolateral region of the most posterior foregut, where the bile duct and gall bladder originate46. In order to verify the ability of the Sox17CreERT2 line to recombine cells in these tissues, 0.7mg/10g of tamoxifen were administered at E10.5 and the embryos harvested at E14.5 (n=2), when the organs and vasculature are mainly formed. We observed that many endothelial cells in the vasculature were recombined, e.g in the dorsal aorta (Figure 3.4 - A-C). Furthermore, recombined cells were found in the bile duct, in accordance with Sox17 expression (Figure 3.4 - D-F). FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 30 It has been shown that Sox17 is still expressed at E9.0 in the ventral pancreas50. However, none of the pancreatic cells expressed YFP indicating that by E10.5, Sox17 is no longer expressed in this organ (Figure 3.5). After birth, SOX17 is essential for the regulation of insulin secretion in betacells. Mice lacking Sox17 during pancreas organogenesis are more susceptible to develop diabetes56. Although no mature beta cells are present at E10.557, there are some cells co-expressing insulin and glucagon. They were not GFP positive at E14.5 indicating that neither the progenitors of beta cells nor the glucagon/insulin doublepositive cells express Sox17 at E10.5 (Figure3.5). Figure 3.4. Recombination rates in E14.5 embryos after administration of tamoxifen at E10.5. (A-C) Immunofluorescence for GFP (green, single channel image (A)) and CD31 (gray, single channel image (B)) on sections of Sox17CreERT2/+ RosaYFP/+ E14.5 embryos shows the recombination when tamoxifen is injected at E10.5. Very high numbers of recombined cells were observed in the vasculature. (D-F) Immunofluorescence for GFP (green, single channel image (D)) and E-Cadherin (gray, single channel image (E)) on sections of Sox17CreERT2/+ RosaYFP/+ E14.5 embryos shows the recombination when tamoxifen is injected at E10.5. Very high numbers of recombined cells were observed in the bile duct.. Nuclei are counterstained with DAPI (blue). Abbreviations: da - dorsal aorta; bd - dile duct. Scale bars - 100 µm. FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 31 Overall, these experiments confirm that the new Sox17CreERT2 mouse line is inducible upon tamoxifen injection and expresses Cre in the expected cell types (Table 3.1). Furthermore, we showed that recombination in the Rosa26 locus is very efficient in the endoderm when tamoxifen is administered at E6.5 and E7.5, as in these conditions most cells in the endoderm expressed YFP. Recombination can also be induced in precursors of the vascular endothelial lineage as well as in the bile duct. The spatiotemporal induction of Cre in this line allows genetic lineage tracing of distinct Sox17+ populations, as well as tissue-specific gene edition. As stated previously, the Sox17 gene is disrupted in this Sox17CreERT2 line. It has been shown that this haploinsufficiency may be an issue depending of the genetic background of the animals. In a C57BL/6 background, 90% of the Sox17+/- mice suffer perinatal lethality due to aberrant development of the liver, gallbladder and bile duct network. The same study reported that in the ICR background a mild phenotype of gallblader hypoplasia is observed only in adults58. This new Sox17CreERT2 line is bred on an ICR background. We did not observe obvious defects; the mice could reach adulthood and were fertile. However, when performing additional mutations this must be taken into account, as it may have an unpredictable effect. On the other hand, this same characteristic raises the possibility to perform interesting experiments regarding the fate of Sox17 deficient cells, which has been only briefly explored to date59. Figure 3.4. Recombinantion rates in E14.5 pancreas after administration of tamoxifen at E10.5. (A-E) Immunofluorescence for GFP (green, single channel image (A)), insulin (red, single channel image (B)) and E-Cadherin (gray, single channel image (C)) on sections of Sox17CreERT2/+ RosaYFP/+ E14.5 embryos shows the recombination when tamoxifen is injected at E10.5.Insulin secreting cells (yellow arrows , magnification (E)) in the E14.5 pancreas were not recombined. Pannel (E) represents a magnification of the dashed area in (D). Nuclei are counterstained with DAPI (blue). Abbreviations: dp - dorsal pancreas. Scale bars - 100 µm. FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 32 Table 3.1. Summary of characterization of the Sox17CreERT2 line Time of Injection Cells recombined VE DE Vasculature E7.5 minimal variable, usually higher on posterior minimal E6.5-7.5 minimal yes (80-90%) minimal E10.5 none bile duct yes (~80%) FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 33 3.2. Expression of molecular markers in the gut To determine the effect of the absence of BMP signalling on the dorso-ventral patterning of the endoderm, the organ domain will be assessed at E10.5 using molecular markers. At this time point, morphological features are present only for some organs e.g. lungs, the pancreas and the liver. A series of organ-specific markers were selected based on their early expression in the different endoderm-derived organ domains (Table 3.1). The immunohistofluorescence protocol for every marker was optimized on E10.5 wild-type (WT) embryos both for whole-mount staining and on sections. The subsequent figures show the results of the optimization on sections. Table 3.2. Molecular markers expressed in the primitive gut and assessed in this study Molecular maker Region Onset of expression NKX2.1 thyroid, lungs, trachea 8.5 60 SOX2 esophagus 9.561 FOXN1 thymus 1162 GCM2 parathyroid 9.562 PROX1 liver 8.5 63 HLXB9 dorsal gut, pancreas 8.064 PITX2 caecum 1165 3.2.1. Nkx2.1 and Sox2 Nkx2.1 and Sox2 encode transcription factors that inhibit each other's expression. Therefore, their expression is mutually exclusive in different domains of the foregut. They are necessary for the proper formation of the trachea and the esophagus, respectively37,61. NKX2.1 is expressed in the ventral foregut endoderm as well as in the lungs and the thyroid, as was observed in E10.5 WT embryos (Figure 3.6). Wholemount stainings of the thyroid and lungs were also successful (Supplementary video 1 and 2). Complementary to the ventral expression of NKX2.1, high levels of SOX2 marked the dorsal foregut endoderm (Figure 3.7). SOX2 expression was also detected at lower levels in the main bronchi (Figure 3.7 - B, C). Indeed, this transcription factor has been shown to inhibit lung branching and its overexpression in the respiratory epithelium causes a severe reduction in the number of airways66. FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 34 3.2.2. Gcm2 and Foxn1 The thymus and parathyroid glands originate from the same endodermal primordium and develop bilaterally, in the third branchial pouch. In the adult, the thymus is situated above the heart and is responsible for T cell production, whereas the parathyroids are found near the thyroid and regulate calcium homeostasis. By E9.5 the thymus and parathyroid start to be specified, but they cannot be morphologically Figure 3.7. SOX2 in WT E10.5 embryos. (A-C) Immunofluorescence for SOX2 (green) was performed on WT E10.5 embryonic sections. SOX2 is found in the esophagus. Trachea can be morphologically distinguished (white lines) (A) as well as the main bronchi (B, C). SOX2 is found at lower levels in the main bronchi (C). Nuclei are counterstained with DAPI (blue). Abbreviations: es - esophagus; tr - trachea; br - bronchi. Dorsal is towards the top and ventral towards the bottom. Scale bar - 100 µm. Figure 3.6. NKX2.1 in WT E10.5 embryos. (A-D) Immunofluorescence for NKX2.1 (green) was performed on WT E10.5 embryonic sections. NKX2.1 is found in the thyroid (A) in the trachea (B,C) and in the main bronchi (D). The esophagus can be morphologically distinguished and do not expressed NKX2.1 (B, C, D - yellow lines). Nuclei are counterstained with DAPI (blue). Abbreviations: th - thyroid; es - esophagus; tr - trachea; br - bronchi. Dorsal is towards the top and ventral towards the bottom. Scale bar - 100 µm. FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 35 distinguished at this point. GCM2 marks the dorsal region in the common primordium which later becomes the parathyroid (Figure 3.8), while Foxn1 is expressed in the ventral domain that originates the thymus. However its expression only starts at E11 making it less suitable for this study62. Wholemount stainings of the parathyroid was also successful (Supplementary video 3). 3.2.3. Prox1 PROX1 is found in the liver, in both pancreatic buds and in the bile duct at E10.5 (Figure 3.9). Wholemount stainings of the liver, pancreas and bile duct with Prox1 were also successful (Supplementary video 4). Expression in the liver domain starts at E8.5 and is first observed in the budding dorsal pancreas at E9.563. Figure 3.7. PROX1 in WT E10.5 embryos. (A, B) Immunofluorescence for PROX1 (green) was performed on WT E10.5 embryonic sections. PROX1 is found in the liver (A, B), the bile duct (A) and in the pancreatic buds, e.g. dorsal pancreatic bud (B). Nuclei are counterstained with DAPI (blue). Abbreviations: bd - bile duct; dp - dorsal pancreas; li - liver. Scale bar - 100 µm. Figure 3.7. GCM2 in WT E10.5 embryos. Immunofluorescence for GCM2 (green) was performed on WT E10.5 embryonic sections. GCM2 is found in a small domain located dorsally on the third branchial pouch, where the parathyroid will develop. Nuclei are counterstained with DAPI (blue). Abbreviations: pt - parathyroid. Scale bar - 100 µm. FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 36 3.2.4. Hlxb9 HLXB9 is found all along the dorsal wall of the gut epithelium as well as in the dorsal pancreas (Figure 3.10). Hlxb9 is also expressed transiently in the ventral pancreas. After E10.5, only the differentiating beta-cells and the beta cells expressed Hlxb9 in the pancreas64. Wholemount stainings of with hlxb9 were also successful (Supplementary video 5). 3.2.5. Pitx2 Pitx2 is expressed in the epithelium and the mesenchyme of the caecum primordium from E11.0. It is required for the formation of the caecum67. Commercial antibodies against PITX2 were tested on E10.5 embryos. However, the different tests did not give any conclusive results. This may result from PITX2 not yet being expressed at this stage or defective antibodies, hypotheses that were not yet tested. 3.2.6. Effectors downstream of BMP: pSMAD1/5/8 In order to monitor BMP pathway activity, we evaluated the presence of a downstream effector, the phosphorylated form of SMAD1/5/8 (pSMAD1/5/8. Figure 3.10. HLXB9 in WT E10.5 embryos. (A, B) Immunofluorescence for HLXB9 (green) was performed on WT E10.5 embryonic sections. HLXB9 is found dorsally all along the gut endoderm. (A, B, C). HLXB9 was also found in the dorsal pancreatic bud (D). Nuclei are counterstained with DAPI (blue). Dorsal is towards the top and ventral towards the bottom. Abbreviations: bd - bile duct; dp - dorsal pancreas; li - liver. Scale bars - 100 µm. FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 43 At E10.5, PROX1 was expressed in the liver, in both pancreatic buds and in the gallbladder in WT embryos. Moreover, each structure was distinguishable morphologically and the liver already had its characteristic sinusoidal shape (Figure 3.14 A). In the Hz mutant, none of the aforementioned structures was distinguishable (Figure 3.14 B). The position and the arrangement of the cells suggested that the majority of PROX1+ cells were hepatocytes, while the rest might have a dorsal pancreatic identity. Nevertheless, more markers should be used in order to assess the identity of the PROX1+ cells. Figure 3.15. . Comparison of foregut in the WT and Hz littermates at E10.5. (D, H) Immunofluorescence for HLXB9 (green, single channel (A, E)), Prox1 (white, single channel (B, F)) and NKX2.1 (red, single channel (C, G)) on E10.5 embryonic sections, after wholemount immunofluorescence. HLXB9 is found in the dorsal region of the gut in both WT and Hz littermates. Expression of HLXB99 is also observable in the neural tube and the notochord (A, E). NKX2.1 is found in the ventral region of the gut in the WT (G) but not in the Hz littermate (C). The overexposed signal in the gut lumen in H is an artefact of wholemount immunofluorescence. Nuclei are counterstained with DAPI (blue). The gut epithelium is outlined. Dorsal is towards the top and ventral towards the bottom. Abbreviations: fg - foregut; nc - notochord; nt - neural tube. Scale bars - 100 µm FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 44 PROX1+ domain was also drastically reduced. Its total volume was estimated in the Hz mutants and in their WT littermates (n=3) using the Imaris 8.1 software (Figure 3.17). However, slight differences of the embryonic stage at the time of collection cause a high variation in the organ domain size, as organogenesis is fast evolving at these stages. Combined with the small number of cells (Figure 3.14 ) forming PROX1+ domain, it could explain why its volume might be 6 times higher in the largest compared to the smallest WT of different litters. For this reason the volume of PROX1 in the Hz mutant was normalized to the volume of its WT littermate (Figure 3.18). This analysis revealed that the total PROX1 expression domain is significantly reduced in the Hz mutant. Figure 3.16. Comparison of lung and liver in the WT and Hz littermates at E10.5. (D, H) Immunofluorescence for HLXB9 (green, single channel (A, E)), PROX1 (white, single channel (B, F)) and NKX2.1 (red, single channel (C, G)) on E10.5 embryonic sections, after wholemount immunofluorescence. HLXB9 is found in the dorsal region of the gut in both WT an Hz littermates (A, E). NKX2.1 is found in the ventral region of the gut where the lungs are budding in the WT (G) but not in the Hz littermate (C). PROX1 is expressed both in the WT (F) and in the Hz littermates (B). In the WT different organs can be distinguished by morphological aspects (H - liver/gallbladder). In the Hz however the PROX1 domain is reduced and no evident morphological structures were observed. The overexposed signal in the gut lumen is an artefact of wholemount immunoflourescence. Nuclei are counterstained with DAPI (blue). The gut epithelium is outlined. Dorsal is towards the top and ventral towards the bottom. Abbreviations: gb - gallbladder; li - liver, lu - lungs. Scale bar - 100 µm FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 45 Figure 3.18. Total Prox1 volumes in the Hz embryos (% of WT littermate). The volumes obtained for the Hz mutants through surface rendering in the Imaris software were plotted normalized to the WT volumes (n=3). PROX1 domains are significantly smaller in the Hz mutants than in the WT littermates. Orange - Litter 1; Green - Litter 2; Blue - Litter 3. *One sample t test p=0,01 Figure 3.17. Surface rende ring of prox1 expression domain. The three dimensional PROX1 domains were isolated from the remaining three dimensional image (white, A, B). Using the Imaris 8.1 software surfaces were rendered based on PROX1 signal (green, C, D). The same settings were replicated for each surface simulation. Before ventral closure, the liver starts developing bilaterally in the lateral endoderm. Both primordia meet and form a single organ when the lateral endoderm migrates ventrally70. Interestingly, the putative liver observed in the Hz mutant was located on the right side of the embryo (Figure 3.14 B), suggesting that the specification process of the right and left liver primordia are differently affected by BMP signalling. Furthermore, the pre-cardiac mesoderm which is necessary for the specification of the hepatogenic endoderm also receives reciprocal signals from this tissue to further develop71. In the Hz mutant, the heart also appears smaller when visualized with PROX1 staining (Figure 3.14 B). It might be a consequence of lack of reciprocal signalling between the liver and the pre-cardiac mesoderm. FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 46 Remarkably, no ventral pancreatic bud was formed in any of the Hz embryos while the outcome of the alteration of BMP signalling on the dorsal pancreatic bud was variable. In one instance, the dorsal pancreas seemed completely absent based on PROX1 and HXLB9 staining. In another litter, there was no apparent budding but PROX1 and HLXB9 expression overlapped dorsally, in the prospective region of the dorsal pancreatic bud (Figure 3.14 B). Finally, in the third litter, a bud was observed, although it was underdeveloped compared to its WT counterpart (Figure 3.19 C, F). The variable dorsal pancreatic phenotype may be linked to a slight difference in the developmental stage of the embryos or to differences in recombination rates between litters. These data indicate that BMP signalling is probably required for the specification of the ventral pancreatic bud while the effect of BMP signalling on dorsal pancreatic bud is less clear but suggest it is required for pancreatic growth, as previously suggested in chick72. Figure 3.19. Comparison of dorsal pancreas in the WT and Hz littermates at E10.5. (C, F) Immunofluorescence for HLXB9 (green, single channel (A, D)) and PROX1 (white, single channel (B, E)) on E10.5 embryonic sections, after wholemount immunofluorescence. Overlapping domains of HLXB99 and PROX1 as well as the morphology, indicate that the dorsal pancreas is present both in the WT (F) and the Hz mutant (C). Nuclei are counterstained with DAPI (blue). Abbreviations: dp - dorsal pancreas. Scale bar - 100 µm FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 47 The global dorsal molecular marker HLXB9 was expressed in the dorsal endoderm of both WT and Hz embryos. Although the size of the domain has not been quantified, it implies that the dorsal identity of the primitive gut tube remains unaffected. In summary, the resulting phenotype shows that the inactivation of BMP signalling through the ALK3 receptor and partially through the ALK6 receptor disturbs the ventral patterning of the anterior foregut as well as the formation of several organs, including the dorsal pancreas. Globally, the gut epithelium of Hz mutants appears to be thinner and less shapely while the surrounding mesoderm did not seem affected by the alteration of BMP signalling, indicating that conditional deletion of Alk3 in the endoderm was successful. The heart developmental defect is likely associated to the lack of reciprocal signalling from the liver endoderm, since this organ is significantly reduced. In this study we provide evidence that the disruption of BMP signalling in the endoderm after onset of gastrulation causes embryonic dead before E10.5. Uncovering the developmental defects that cause death in these embryos will provide further information on the role of BMP signalling in the endoderm during these stages. Furthermore, the preliminary results on characterization of the Hz phenotype reveal that disturbance of BMP signalling in the endoderm results in global defects in the endoderm. To achieve a deeper understanding of the phenotype of the Hz mutants, several further experiments could be conceived. The extent of BMP signalling inactivation in the Hz mutant should be assessed by evaluating pSMAD1/5/8 levels in the gut, as described in section 3.2.6. This experiment would reveal to which extent endodermal cells have undergone recombination and the level of BMP signalling in other cells. Alternatively, a recombination reporter could be included in the breeding scheme. The analysis of other endoderm-derived organs is also required. Indeed the global ventralizing activity of BMP is unclear since the dorsal pancreatic bud is either hypoplastic or absent and HLXB9 domain does not appear to be extended ventrally. Therefore, in order to assess our hypothesis, it will be important to evaluate the presence of the other dorsal organ, the parathyroid, by analysing the expression of GCM2. In addition, a ventral posterior organ, the caecum, could be assessed by expression of Pitx2 in the Hz embryo, possibly by in situ hybridization. It will also be interesting to examine pancreatic specific markers, such as PTF1A or PDX1, in order to differentiate the liver, the pancreas and the gallbladder. Additional global markers like SOX2 (dorsal foregut) and Islet1 (global ventral in chick, Palle Serup unpublished data) might offer further insight into which targets are affected by BMP signalling. As FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 48 our analyses cannot rule out a defect in the maintenance of the identity of the endoderm-derived organs, the Hz mutants should also be studied at earlier stages. A part of the observed phenotype might be associated with a defect in the number of endodermal cells. Thus, the levels of endoderm proliferation and cell death in the Hz mutant should also be evaluated at an early stage such as E8.5. FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 49 Chapter IV Final Remarks FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 50 Final Remarks BMP signalling is essential for many aspects of development, including mesoderm formation and ectoderm patterning. In the DE, several studies have identified roles of BMP signalling in the formation of ventral gut organs 34-38. We proposed that BMP signalling acts as a global cue in ventral patterning of the DE (Figure 1.3). The findings presented in this study are in accordance with the hypothesis. However, many questions remain before it can be confirmed. For instance, BMP signalling may be essential for the initial invagination of the AIP as suggested previously73, or it may also be required for ventral migration of the lateral DE and consequent ventral closure of the gut tube. The role of BMP in organogenesis is also still unclear. Is it required for specification, proliferation, maintenance of identity? Perhaps it is required for distinct processes depending on the organ in question. Challenges in the future are to uncover new targets of BMP during endoderm development. Human Embryonic Stem cell derived endoderm could be used in order to identify which genes are downstream targets of BMP. A global understanding of how BMP signalling acts during dorsal-ventral patterning of the DE will facilitate the development of more efficient protocols that better emulate the development of organs like the lungs, liver or the pancreas, bringing us one step forward into the exciting field of in vitro organogenesis74. FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 51 Cited Literature 1. Choi, E., et al. Dual lineage-specific expression of Sox17 during mouse embryogenesis. Stem cells (Dayton, Ohio) 30, 2297-2308 (2012). 2. Grapin-Botton, A. Endoderm Specification. in StemBook (ed. Community, T.S.C.R.) (2008). 3. Grapin-Botton, A. & Melton, D.A. Endoderm development: from patterning to organogenesis. Trends in genetics : TIG 16, 124-130 (2000). 4. Balemans, W. & Van Hul, W. Extracellular regulation of BMP signaling in vertebrates: a cocktail of modulators. Dev Biol 250, 231-250 (2002). 5. Zorn, A.M. & Wells, J.M. Vertebrate endoderm development and organ formation. Annu Rev Cell Dev Biol 25, 221-251 (2009). 6. Desrochers, T.M., Palma, E. & Kaplan, D.L. Tissue-engineered kidney disease models. Adv Drug Deliv Rev 69-70, 67-80 (2014). 7. Wang, L., et al. Engineering Three-Dimensional Cardiac Microtissues for Potential Drug Screening Applications. Curr Med Chem 21, 2497-2509 (2014). 8. Patthey, C. & Gunhaga, L. Signaling pathways regulating ectodermal cell fate choices. Experimental cell research 321, 11-16 (2014). 9. Kimelman, D. Mesoderm induction: from caps to chips. Nat Rev Genet 7, 360-372 (2006). 10. Nowotschin, S. & Hadjantonakis, A.-K.K. Cellular dynamics in the early mouse embryo: from axis formation to gastrulation. Current opinion in genetics & development 20, 420427 (2010). 11. Kwon, G.S., Viotti, M. & Hadjantonakis, A.-K.K. The endoderm of the mouse embryo arises by dynamic widespread intercalation of embryonic and extraembryonic lineages. Developmental cell 15, 509-520 (2008). 12. Viotti, M., Nowotschin, S. & Hadjantonakis, A.-K.K. Afp::mCherry, a red fluorescent transgenic reporter of the mouse visceral endoderm. Genesis (New York, N.Y. : 2000) 49, 124-133 (2011). 13. Spooner, B.S., Walther, B.T. & Rutter, W.J. The development of the dorsal and ventral mammalian pancreas in vivo and in vitro. J Cell Biol 47, 235-246 (1970). 14. Cordier, A.C. & Haumont, S.M. Development of thymus, parathyroids, and ultimobranchial bodies in NMRI and nude mice. The American journal of anatomy 157, 227263 (1980). 15. Wills, A., Dickinson, K., Khokha, M. & Baker, J.C. Bmp signaling is necessary and sufficient for ventrolateral endoderm specification in Xenopus. Dev Dyn 237, 2177-2186 (2008). 16. Urist, M.R. Bone: formation by autoinduction. Science 150, 893-899 (1965). 17. Wagner, D.O., et al. BMPs: from bone to body morphogenetic proteins. Sci Signal 3, mr1 (2010). 18. Cui, Y., Jean, F., Thomas, G. & Christian, J.L. BMP-4 is proteolytically activated by furin and/or PC6 during vertebrate embryonic development. EMBO J 17, 4735-4743 (1998). 19. Harrison, C.A., Al-Musawi, S.L. & Walton, K.L. Prodomains regulate the synthesis, extracellular localisation and activity of TGF-beta superfamily ligands. Growth Factors 29, 174-186 (2011). FCUP Role of the BMP Pathway in Definitive Endoderm Patterning 52 20. Heldin, C.H. & Moustakas, A. Role of Smads in TGFbeta signaling. Cell Tissue Res 347, 21-36 (2012). 21. Mueller, T.D. & Nickel, J. Promiscuity and specificity in BMP receptor activation. FEBS letters 586, 1846-1859 (2012). 22. Danesh, S.M., Villasenor, A., Chong, D., Soukup, C. & Cleaver, O. BMP and BMP receptor expression during murine organogenesis. Gene expression patterns : GEP 9, 255-265 (2009). 23. Winnier, G., Blessing, M., Labosky, P.A. & Hogan, B.L.M. Bone Morphogenetic Protein4 Is Required for Mesoderm Formation and Patterning in the Mouse. Genes & Development 9, 2105-2116 (1995). 24. Zhang, H.B. & Bradley, A. Mice deficient for BMP2 are nonviable and have defects in amnion chorion and cardiac development. Development 122, 2977-2986 (1996). 25. de Sousa Lopes, S.M., et al. BMP signaling mediated by ALK2 in the visceral endoderm is necessary for the generation of primordial germ cells in the mouse embryo. Genes & development 18, 1838-1849 (2004). 26. Miyazono, K., Kusanagi, K. & Inoue, H. Divergence and convergence of TGF /BMP signaling. Journal of cellular physiology 187, 265-276 (2001). 27. Dewulf, N., et al. Distinct spatial and temporal expression patterns of two type I receptors for bone morphogenetic proteins during mouse embryogenesis. Endocrinology 136, 2652-2663 (1995). 28. Gu, Z., et al. The type I serine/threonine kinase receptor ActRIA (ALK2) is required for gastrulation of the mouse embryo. Development (Cambridge, England) 126, 2551-2561 (1999). 29. Yi, S.E., Daluiski, A., Pederson, R., Rosen, V. & Lyons, K.M. The type I BMP receptor BMPRIB is required for chondrogenesis in the mouse limb. Development 127, 621-630 (2000). 30. Mishina, Y., Hanks, M.C., Miura, S., Tallquist, M.D. & Behringer, R.R. Generation of Bmpr/Alk3 conditional knockout mice. Genesis (New York, N.Y. : 2000) 32, 69-72 (2002). 31. Beppu, H., et al. BMP type II receptor is required for gastrulation and early development of mouse embryos. Developmental biology 221, 249-258 (2000). 32. Holley, S.A. & Ferguson, E.L. Fish are like flies are like frogs: conservation of dorsalventral patterning mechanisms. BioEssays : news and reviews in molecular, cellular and developmental biology 19, 281-284 (1997). 33. Liu, A. & Niswander, L. Bone morphogenetic protein signalling and vertebrate nervous system development. Nature Reviews Neuroscience 6, 945-954 (2005). 34. Bleul, C.C. & Boehm, T. BMP signaling is required for normal thymus development. Journal of immunology (Baltimore, Md. : 1950) 175, 5213-5221 (2005). 35. Frisbie, T., et al. Determining the effect of BMP signaling on thymus fate specification and differentiation of the third pharyngeal pouch in the mouse. (HEM7P.235). The Journal of Immunology 194, 188.115 (2015). 36. Li, Y., Gordon, J., Manley, N.R., Litingtung, Y. & Chiang, C. Bmp4 is required for tracheal formation: a novel mouse model for tracheal agenesis. Developmental biology 322, 145-155 (2008). 37. Domyan, E.T., et al. Signaling through BMP receptors promotes respiratory identity in the foregut via repression of Sox2. Development (Cambridge, England) 138, 971-981 (2011).