Bardine N et al. (2014), Vertical signalling involves transmission of Ho...

XB-ART-50429

PLoS One 2014 Jan 01;912:e115208. doi: 10.1371/journal.pone.0115208.

Show Gene links Show Anatomy links

Vertical signalling involves transmission of Hox information from gastrula mesoderm to neurectoderm.

Bardine N , Lamers G , Wacker S , Donow C , Knoechel W , Durston A .

???displayArticle.abstract???
Development and patterning of neural tissue in the vertebrate embryo involves a set of molecules and processes whose relationships are not fully understood. Classical embryology revealed a remarkable phenomenon known as vertical signalling, a gastrulation stage mechanism that copies anterior-posterior positional information from mesoderm to prospective neural tissue. Vertical signalling mediates unambiguous copying of complex information from one tissue layer to another. In this study, we report an investigation of this process in recombinates of mesoderm and ectoderm from gastrulae of Xenopus laevis. Our results show that copying of positional information involves non cell autonomous autoregulation of particular Hox genes whose expression is copied from mesoderm to neurectoderm in the gastrula. Furthermore, this information sharing mechanism involves unconventional translocation of the homeoproteins themselves. This conserved primitive mechanism has been known for three decades but has only recently been put into any developmental context. It provides a simple, robust way to pattern the neurectoderm using the Hox pattern already present in the mesoderm during gastrulation. We suggest that this mechanism was selected during evolution to enable unambiguous copying of rather complex information from cell to cell and that it is a key part of the original ancestral mechanism mediating axial patterning by the highly conserved Hox genes.

???displayArticle.pubmedLink??? 25514127
???displayArticle.pmcLink??? PMC4267835
???displayArticle.link??? PLoS One

Species referenced: Xenopus laevis
Genes referenced: chrd hoxb4 hoxb9 hoxc6 hoxd1 mmut myc nrp1 tbxt uqcc6

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

	Figure 2. Mesodermal ectopic expression of a single Hox gene copies expression of the same Hox gene from mesoderm to neurectoderm.A– A’’’, E, F Localisation of mesoderm and neurectoderm in the wrap assay shown by expression of the mesodermal markers Chordin (Chd, A) and Brachyury (Bra, A’), and the neural marker Nrp1 (A”). In A’’’, lineage labelling by ectopic expression of GFP in the NOM only. These recombinants were analyzed 6 to 8 h after tissue healing. These data show that there is no tissue intermingling during wrap culture. Mesodermal Chd (expressed in SO, A) and GFP (within NOM, A’’’) do not mix with each other within the time course of wrap culture. Consistently, Bra, a pan mesodermal marker, is expressed in both types of mesoderm in accordance with its expression domain in the embryo (A’).The neural marker Nrp-1 is expressed in the space between mesoderm and the outermost ectodermal layer of the wrap, consistent with its known pattern of expression within the embryo (A’’). B–B’: Induction of Hoxd1 B: A wrap containing SO and NOM and ectoderm [AC(SO/NOM)AC] shows induction of Hoxd1 in the neurectoderm as well as mesoderm. B’, A wrap containing normal SO and SO ectopically expressing Hoxd1 also shows the induction of endogenous Hoxd1 in the neurectoderm as well as in the mesoderm. Endogenous Hoxd1 expression was detected using a 3′UTR probe that recognizes only the endogenous messenger. C, ectopic Hoxb4 in SO induces its own expression within the neurectoderm and in the mesoderm as in B’. D, wrap as in B’ and C but with ectopic Hoxc6 expression. This shows induction of Hoxc6 in neurectoderm and in the mesoderm. We used 3′UTR probes to detect expression of the endogenous mRNA’s in each of these experiments. E, F sections showing expression of Nrp1 (neural) and Bra (mesodermal) in control or standard [AC(SO/NOM)AC] recombinant. E Nrp1 expression is internal in the recombinant but excluded from an internal cell mass that is clearly the mesoderm. It is particularly strong around one end of the cell mass which is the neural inducing SO. Expression is also absent from the very outer layer of the recombinant, which represents the outer non neural layer of the neurectoderm. F: Bra expression is in an internal cell mass (the mesoderm). Please note that the germ layer markers Bra, Ch, and the mesodermal lineage label GFP are confined to an internal cell mass excluding tissue intermingling and that Hox expression is detected in neurectoderm as well as the mesodermal cell mass. Each photo in this figure represents at least 20 recombinants and embryos with consistently the same results.
	Figure 3. Hoxd1 homeoprotein containing a penetratin sequence is transferred from mesoderm to neurectoderm and its homeodomain plays a cargo function for gfp.A: Penetratin sequence is shown. Above, Hoxd1 homeodomain (HD) contains a penetratin like sequence (in red). This conserved sequence is a feature of all homeodoproteins. Below, two amino-acids WF within the penetratin were mutated into SR (highlighted by stars) to create a mutated Hoxd1 HD, mut HD. This mutation abrogates the transfer function of the HD. B–D”: Localisation of different fluorescent chimeric GFP proteins in recombinants after 6–8 hrs of culture. These proteins were introduced into wrap recombinates in SO. B: wild type GFP. C: wild type GFP coupled to wild type Hoxd1 homeodomain (d1-HD-gfp). D: the mutated homeodomain version coupled to wild type GFP (mut d1-HD-gfp).The signal has spread in d1-HD-gfp but not in wild type GFP or mut d1-HD-gfp. The GFP fluorescence is combined with phalloidin staining to increase its visibility. B, B’, B”: GFP expressed in the SO stays confined within the SO explant and fails to spread into surrounding neurectoderm. C, C’, C”: d1-HD-gfp spreads outside the mesodermal explant to the neurectoderm (spreading indicated by arrowheads). D, D’, D”: mut d1-HD-gfp protein shows a SO localisation pattern as shown by wild type GFP. Each photo in this figure is representative of 23 recombinates, each of which gave the same result. Homeoproteins contain a separate HD sequence regulating homeoprotein secretion as well as the penetratin sequence. The mutations we made were in the ‘penetratin’ uptake regulating sequence. Please note that d1-HD-GFP, mut d1-HD-GFP and GFP in Fig. 3 evidently diffuse less than Myc tagged Hoxd1 in S1 Figure. This is expected, due to the large size of GFP.
	Figure 4. Uptake of d1-HD-GFP and GFP by Drosophila imaginal wing discs.Drosphila imaginal wing discs were incubated with Hoxd1-HD-GFP (d1-HD-gfp) recombinant protein (B) or wild type GFP protein (A). or mutated mut-d1-HD-GFP. Recombinant d1-HD-gfp was taken up by the discs while wild type GFP and mut-d1-HD-GFP were not. Each photo in this figure represents 10 imaginal discs giving the same result. These data clearly show the cargo function of Hoxd1 homeodomain and suggest that this uptake is by a species independent mechanism.
	Figure 5. Craniofacial structures of Xenopus laevis tadpoles upon injection of Hoxd1 mRNA or protein.a: Schematic ventral view of an uninjected untreated control embryo. b: uninjected embryo. c: embryo injected with wild type GFP into the blastoecel at blastula stage. d: Embryo injected with Hoxd1 mRNA at 4 cell stage. e: Embryo injected with recombinant HOXD1 protein into 4 cell stage embryo. f: injection of recombinant HOXD1 protein into the blastoecel. Please note that standard injection of Hoxd1 mRNA or its protein counterpart injection in the cytoplasm or in the extracellular matrix leads to similar phenotypes in d, e and f. The embryos are strongly posteriorised as shown by the reduction (or deletion) of anterior structures. These data strongly suggest that the Hoxd1 protein successfully crossed the cellular membranes and retained its function as it leads to a severe truncation of anterior cartilage structures. Infrarostrale (in), Meckel’s cartilage (me), palatoquadrate (pa), ceratohyale (ce), basibranchiale (ba), branchial arches (br), eye (ey), intestine (in). Each photo represents 10 identically treated embryos giving the same result.
	Figure 1. Mesodermal Hox loss of function of a single Hox gene prevents neurectodermal Hox expression of the same Hox gene. Aa, Wrap assay consists of a piece of non-organiser mesoderm (NOM) and a piece of Spemann organiser mesoderm (SO) combined between two ectodermal animal caps. All tissues are excised from early gastrulae (st. 10a) Ab, Ba, Ca: External views of late gastrula stage Xenopus laevis expressing Hoxd1, Hoxb4 and Hoxb9 respectively. Note that the midline of the embryo, overlying the SO, does not express any Hox gene. Ac, Bb, Cb: wraps containing only SO explants [AC(SO/SO)AC]. Ad, Bc, Cc: wraps containing SO and NOM treated with control morpholino (ctMO) [AC(SO/NOM+ctMO)AC]. Ae, Bd, Cd, wrap with NOM treated with Hoxd1-, Hoxb4 and Hoxb9 MO’s respectively. Please note that in each case, only the wraps containing NOM and SO show Hox expression in the neurectoderm (Ad, Bc, Cc) and those containing only SO do not show any expression in accordance with the embryo’s lack of Hox expression in SO (Ab, Ba, Ca, and Ac, Bb, Cb). In each case, Hox MO treatment of NOM mesoderm also prevents the expression of the homologous Hox gene in neurectoderm (Ae, Bd, Cd). These wraps were fixed and analysed 6–8 hrs after they were made. Each photo of two recombinates or an embryo in this figure is representative of at least 20 recombinates or embryos, all showing the same result.

References [+] :

Bardine, Two Hoxc6 transcripts are differentially expressed and regulate primary neurogenesis in Xenopus laevis. 2009, Pubmed, Xenbase