Herr P et al. (2008), Regulation of TGF-(beta) signalling by N-acetyl...

XB-ART-37584

Development 2008 May 01;13510:1813-22. doi: 10.1242/dev.019323.

Show Gene links Show Anatomy links

Regulation of TGF-(beta) signalling by N-acetylgalactosaminyltransferase-like 1.

Herr P , Korniychuk G , Yamamoto Y , Grubisic K , Oelgeschläger M .

???displayArticle.abstract???
The TGF-beta superfamily of secreted signalling molecules plays a pivotal role in the regulation of early embryogenesis, organogenesis and adult tissue homeostasis. Here we report the identification of Xenopus N-acetylgalactosaminyltransferase-like 1 (xGalntl-1) as a novel important regulator of TGF-beta signalling. N-acetylgalactosaminyltransferases mediate the first step of mucin-type glycosylation, adding N-acetylgalactose to serine or threonine side chains. xGalntl-1 is expressed in the anterior mesoderm and neural crest territory at neurula stage, and in the anterior neural crest, notochord and the mediolateral spinal cord at tailbud stage. Inhibition of endogenous xGalntl-1 protein synthesis, using specific morpholino oligomers, interfered with the formation of anterior neural crest, anterior notochord and the spinal cord. Xenopus and mammalian Galntl-1 inhibited Activin as well as BMP signalling in the early Xenopus embryo and in human HEK 293T cells. Gain- and loss-of-function experiments showed that xGalntl-1 interferes with the activity of the common TGF-beta type II receptor ActR-IIB in vivo. In addition, our biochemical data demonstrated that xGalntl-1 specifically interferes with the binding of ActR-IIB to Activin- and BMP-specific type I receptors. This inhibitory activity of xGalntl-1 was dependent on mucin-type glycosylation, as it was sensitive to the chemical inhibitor benzyl-GalNAc. These studies reveal an important role of a N-acetylgalactosaminyltransferase in the regulation of TGF-beta signalling. This novel regulatory mechanism is evolutionarily conserved and, thus, might provide a new paradigm for the regulation of TGF-beta signalling in vertebrates.

???displayArticle.pubmedLink??? 18417620
???displayArticle.link??? Development

Species referenced: Xenopus
Genes referenced: actl6a ag1 alk bmp4 bmpr1a bmpr1b chrd dlx5 fgf4 galnt14 galnt16 galnt2 gata1 msx1 myc myod1 ncam1 ncoa3 odc1 otx2 rax sia1 six3 smad1 smad2 snai1 snai2 tbxt ventx1.2 ventx2.2 wnt8a
???displayArticle.morpholinos??? galnt16 MO1 galnt16 MO2

???attribute.lit??? ???displayArticles.show???

	Fig. 1. Sequence and expression of xGalntl-1. (A) Comparison of Xenopus Galntl-1, human GALNTL1 (hGalntl-1), zebrafish Galnt14 (zGalnt-14) and Drosophila Galnt2 (dGalnt-2) protein sequences. The transmembrane domain (TM, blue), the catalytic domain (red), the Ricin domain (green) and amino acid identities/similarities for the different protein domains are indicated. (B) Detection of mucin-type glycosylation at the cell surface and in the Golgi compartment in xGalntl-1-transfected Cos-7 cells using Helix pomatia agglutinin (HPA) conjugated to Alexa Fluor 488. (C) RT-PCR analysis of xGalntl-1, xGalnt-1, xGalnt-6 and xGalnt-7 expression, using RNA from whole embryos (left) or dorsal and ventral marginal zone explants, isolated at gastrula stages and analysed at stage 20 (right). The blood markerα -globin, neural marker β-tubulin, dorsal mesodermal marker chordin and the nieuwkoop centre-specific gene siamois were used as references for the indicated developmental stages. Dorsal (ncam, myoD) and ventral (sizzled, msx1) marker genes were used as controls for the marginal zone explants. (D) Whole-mount in situ hybridisation analysis of xGalntl-1 expression at stage 13. Numbered lines indicate the positions of vibratome sections that reveal xGalntl-1 expression in the anterior mesoderm and in the deep layer of the lateral neural plate. (E) In situ analysis of xGalntl-1 expression at stage 27. The paraffin sections below show specific expression of xGalntl-1 in the anterior brain (1), neural crest (2), mediolateral spinal cord (2-4) and notochord (3,4).
	Fig. 2. Inhibition of mesoderm formation and Activin signalling by xGalntl-1. (A) Stage 40 uninjected control Xenopus embryo and (B) sibling embryos microinjected radially with 400 pg xGalntl-1 mRNA at 2- to 4-cell stages. (C) In situ hybridisation for the pan-mesodermal marker xbra in gastrula stage embryo microinjected either radially with xGalntl-1 mRNA alone or into single blastomeres together with lacZ mRNA at the 4-cell stage. (D) RT-PCR analysis of dorsal (DMZ) and ventral marginal zone (VMZ) explants at stage 20. The expression of α-globin (blood), gata1 and xvent1 (ventral mesoderm), msx1 (neural crest, epidermis) was reduced in VMZ from embryos microinjected with 200 pg xGalntl-1 mRNA, whereas the expression levels of xag (cement gland), otx2, rx2a (forebrain) and ncam (neural) were increased. The expression of α-actin (muscle) was reduced in DMZ and slightly increased in VMZ. (E) RT-PCR analysis for chordin (dorsal mesoderm), sox17β (endoderm) and xbra (mesoderm) expression using stage 20 animal cap explants from embryos microinjected with mRNA encoding eFGF (100 pg) or Activin (2 pg), in the presence or absence of 200 pg xGalntl-1 mRNA. (F) Western blot analysis of Smad and phospho-Smad proteins from HEK 293T cells that were transiently transfected with Smad2-FLAG and xGalntl-1 expression vectors and cultured for 2 hours in the absence or presence of 2ng/ml recombinant Activin protein. (G) In vitro translation of xGalntl-1 mRNA containing (+5′-UTR) or lacking (-5′-UTR) morpholino target sequences. Two different morpholinos (Galntl-1-MO1 and Galntl-1-MO2, 2.5 μm) as well as a mixture of both (1.25 μm each) specifically interfered with translation of RNA containing the morpholino target sequences. (H) RT-PCR analysis of animal cap explants at stage 20 from embryos microinjected with 0.5 pg Activin mRNA, xGalntl-1 morpholinos (1 ng each) and 20 pg xGalntl-1 mRNA. The morpholino stimulated the induction of myoD (muscle) and xbra expression by Activin. In all RT-PCR experiments, odc served as a loading control.
	Fig. 4. Xenopus Galntl-1 loss-of-function phenotype. (A) Stage 40 uninjected control embryo and embryo injected radially with the xGalntl-1-specific morpholinos at the 2- to 4-cell stage. Numbered lines indicate the positions of the paraffin sections shown on the right that reveal a reduction of neural tissue in the morpholino-injected embryos. (B) Uninjected control embryo and embryos microinjected with xGalntl-1 morpholinos stained with the notochord-specific antibody MZ15 at stage 40 (left) and analysed by in situ hybridisation for slug expression at stage 30. The morpholino-injected embryos display defects in the anterior notochord and reduced anterior expression of slug. (C) Whole-mount in situ hybridisation for slug of stage 15 embryos after radial microinjection of the xGalntl-1 morpholinos at the 2- to 4-cell stage, and stage 13 embryo stained for msx1 after single injections of the xGalntl-1 morpholinos into a dorsal-animal blastomere at the 8-cell stage together with lacZ mRNA. (D) RT-PCR analysis of stage 20 dorsal marginal zone explants from uninjected embryos or embryos microinjected with xGalntl-1 morpholinos alone or together with xGalntl-1 mRNA (20 pg). The expression of otx2 and six3 (forebrain) as well as of snail (neural crest) was inhibited by the xGalntl-1 morpholinos, whereas the expression of dlx5 (epidermis), xvent2 (ventral mesoderm) and xbra (mesoderm) was unaffected.
	Fig. 3. Inhibition of BMP signalling by Galntl-1. (A-D) Stage 36 embryos that were injected animally at the 8-cell stage with a total of 2 ng of mRNA encoding xGalntl-1, human GALNTL1 (hGalntl-1) or secreted Galntl-1 (sGalntl-1), with the N-terminal transmembrane domain replaced by the Chordin leader peptide. (E) RT-PCR analysis of animal cap explants from embryos injected with 2 ng xGalntl-1 or 40 pg chordin mRNA at stage 20 for neural (pax6, ncam, otx2), anterior (xag), muscle (myf5) and epidermal (cytokeratin) marker genes. (F) Western blot of the supernatant (SN) and cell pellet of dissociated ectodermal cells expressing the xGalntl-1 or sGalntl-1 with a C-terminal HA-TAG. (G) RT-PCR analysis of animal cap explants from embryos microinjected with mRNA encoding human, mouse (m) or secreted forms of Galntl-1. (H) RT-PCR analysis of animal cap explants from embryos microinjected with mRNA encoding xGalntl-1 (1 ng) alone or in combination with mRNA encoding BMP7/OP-1 (400 pg), dnFGFR-4 (400 pg), dnAlk-4 (400 pg) or ΔN-TCF-3 (400 pg). (I) Western blot for Smad1 or phospho-Smad1 using animal cap lysates from uninjected embryos or embryos microinjected with 200 pg Smad1-FLAG mRNA alone or together with 200 pg xGalntl-1 mRNA. (J) Western blot analysis for Smad1 and phospho-Smad1 from transiently transfected Hek293T cells treated for 2 hours with 40 ng/ml recombinant BMP4 protein. (K) Western blot for phospho-Smad1 or Smad1 using lysates from HEK 293T cells transiently transfected with the indicated expression plasmids. In the RT-PCR experiments α-actin or myf-5 served as controls for mesoderm contamination; odc as a loading control.
	Fig. 5. Inhibition of ActR-IIB by Galntl-1. (A) RT-PCR analysis of animal cap explants analysed at stage 20 from embryos microinjected with 400 pg xGalntl-1 mRNA alone or together with 200 pg mRNA encoding human ALK3, mouse Alk6, human BMPR-II or Xenopus ActR-IIB. (B) RT-PCR analysis of stage 20 animal caps microinjected with 400 pg ActR-IIB mRNA either alone or together with 1 ng xGalntl-1 mRNA. The mesodermal marker genes xbra and wnt8 indicated ActR-IIB activity, whereas otx2 and xag indicated xGalntl-1 activity, and xvent2 served as a marker for ventral mesoderm and odc as a loading control. (C) RT-PCR analysis of stage 20 animal caps microinjected with 40 pg ActR-IIB mRNA alone, together with xGalntl-1 morpholinos, and xGalntl-1 morpholinos and 100 pg xGalntl-1 mRNA. (D) Western blot analysis of Smad and phospho-Smad levels from HEK 293T cells transiently transfected with Smad1-FLAG, BMPR-II and xGalntl-1 (left) or Smad2-FLAG, ActR-IIB and xGalntl-1 (right). (E) Western blot analysis of phospho-Smad2 and Smad2 levels from HEK 293T cells transiently transfected with Smad2-FLAG, ActR-IIB and xGalntl-1 and incubated overnight in the presence or absence of 2 mM benzyl-GalNAc. (F) Whole-mount in situ hybridisation for msx1 using stage 13 embryos microinjected with xGalntl-1 morpholinos, 100 pg xActR-IIB mRNA and 100 pg xGalntl-1 mRNA into dorsal-animal blastomeres at the 4- to 8-cell stage.
	Fig. 6. Galntl-1 inhibits the binding of ActR-IIB to type I receptors. (A) Immunohistochemical staining of transiently transfected Cos-7 cells for ActR-IIB containing a C-terminal HA-TAG and xGalntl-1 containing a C-terminal FLAG-TAG. (B) Western blot analyses of stage 11 Xenopus animal cap cells expressing HA-tagged ActR-IIB, Alk-3, Alk-4 and Alk-6 and xGalntl-1. (C) Western blot analyses of transiently transfected HEK 293T cells expressing HA-tagged xActR-IIB and xGalntl-1 and incubated overnight in the presence or absence of 2 mM benzyl-GalNAc. (D) Co-immunoprecipitation of HA-tagged Alk-3, Alk-4 and Alk-6 with MYC-tagged ActR-IIB in the presence or absence of FLAG-tagged xGalntl-1 from lysates of transiently transfected HEK 293T cells. (E) Co-immunprecipitation of Alk-3 with Myc-tagged ActR-IIB or FLAG-tagged BMPR-II in the presence or absence of untagged xGalntl-1. The binding of BMPR-II to Alk-3 does not seem to be inhibited by xGAlntl-1. (F) Co-immunprecipitation of Alk-4-HA with ActR-IIB-Myc in the presence of xGalntl-1, sGalntl-1 or xGalnt-6. Only xGalntl-1 interfered with the formation of the heteromeric complex. In all experiments, 10% of the lysates used for co-immunprecipitation was subjected to direct immunprecipitation and the quantity of protein analysed by western blotting for the different epitopes.
	Fig. S2. Effect of OP-1, dnFGFR4, dnTCF3 or dnALK-4 on the xGalntl-1 phenotype. (A-E) Stage 40 embryos microinjected with 200 pg xGalntl-1 mRNA alone (A) or together with 200 pg OP-1 (B), dnFGFR-4 (C), dnTCF-3 (D) or dnAlk-4 (E). (F) RT-PCR analysis of ectodermal explants obtained from sibling embryos at stage 9 and analyzed at stage 25.
	Fig. S3. Xenopus Galtnl-1 completely abolishes the ventralizing activity of Alk-3, Alk-6 and BMPR-II. (A) Stage 33 uninjected control embryo. (B) RT-PCR analyses of ectodermal explants from embryos microinjected with 200 pg mRNA encoding the indicated BMP type I and type receptors. (C-H) Stage 33 embryos microinjected with 200 pg mRNA encoding Alk-3 (C,D), Alk-6 (E,F) or BMPR-II (G,H) alone (C,E,G), or together with 200 pg xGalntl-1 mRNA (D,F,G).
	Fig 1. D. Whole-mount in situ hybridisation analysis of xGalntl-1 expression at stage 13. Numbered lines indicate the positions of vibratome sections that reveal xGalntl-1 expression in the anterior mesoderm and in the deep layer of the lateral neural plate.
	Figure 1, (E) In situ analysis of galnt16 (polypeptide N-acetylgalactosaminyltransferase 16; xGalntl-1) expression at stage 27. The paraffin sections below show specific expression of xGalntl-1 in the anterior brain (1), neural crest (2), mediolateral spinal cord (2-4) and notochord (3,4).