December 1, 1994;
Expression patterns of Hoxb genes in the Xenopus embryo suggest roles in anteroposterior specification of the hindbrain and in dorsoventral patterning of the mesoderm.
Hox genes are thought to participate in patterning the anteroposterior (a-p) axis during vertebrate embryogenesis. In this investigation, the spatial expression of six Hoxb genes was analyzed in early embryos of Xenopus laevis by in situ hybridization. Hoxb gene expression was first detected in late gastrulae/early neurulae, by which stage, the characteristic spatially colinear Hoxb gene expression sequence was already apparent. Dissection experiments indicated that the establishment of these localized expression patterns coincides with the acquisition of anteroposterior positional information along the main body axis. The Hoxb genes continued to be expressed in similar domains along the anteroposterior axis at all developmental stages examined, although there were some changes in expression at the cellular level. Interestingly, the 3'' genes, Hoxb-1
, and Hoxb-4 were expressed in very restricted domains in the future hindbrain
, while Hoxb-5, Hoxb-7, and Hoxb-9 transcripts were present along the entire presumptive spinal cord
. It was thus notable that the 5'' Hoxb genes exhibited different types of expression domain than the 3'' Hoxb genes. These observations suggest that there may be different mechanisms regulating the expression of the 3'' and 5'' Hoxb genes. Expression of all of the Hoxb genes analyzed, except Hoxb-4, was predominantly detectable in the central nervous system
or in neural crest
-derived structures. Hoxb-4 mRNA was detected in the central nervous system
, but interestingly, the major expression site for this gene was the somites
. The other Hoxb genes tested failed to show significant expression in the somitic mesoderm
, although transcripts from genes 5'' from Hoxb-4 were detected in other mesodermal tissues. In the vertebrate trunk
, anteroposterior patterning of the CNS
is thought to be regulated by the somites
. The results obtained here for Xenopus embryos did not explicitly support the idea of a Hoxb code for the somites
, although we cannot rule this out. Instead, interestingly, the data were consistent with a role for Hoxb genes in dorsoventral patterning of the mesoderm
[+] show captions
FIG. 1. In situ hybridization ofstage 28 (L B) of stage 31/2 (C-F) tailbud stage embryos using digoxigenin-labeled probes. Hybridization was visualized using an alkaline phosphatase reaction with nitroblue tetrazolium as the substrate. All embryos are shown on their sides with anterior to the left. (A) Hoxb-1. Staining is in the hindbrain adjacent to the otic vesicle. (B) Hoxb-3. The probe hybridized to the CNS immediately posterior to the otic vesicle and to posterior branchial arches. (C) Hoxb-4. Transcripts were detected in the hindbrain, in the fourth branchial arch, in somites, and in pronephric ducts. (D) Hoxb-5. (E) Hoxb-7. In both cases, the posterior-most part of the hindbrain, the spinal cord, the pronephric duets, and the tail bud were stained. Weak staining in the somites was also seen in controls. (F) Hoxb-9. Staining was detected in the spinal cord. The tail bud may also be weakly stained.
FIG. 2. Development of Hoxb gene expression patterns. In situ hybridization of Hoxb gene probes to early neurula (A, D, G, J, M), midneurula (13, E, H, K, N), and early tailbud stage embryos (C, F, I, L, O). Whole embryos were hybridized with probes from the following genes; (A-C) Hoxb-1, (D-F) Hoxb-3, (G-I) Hoxb-4, (J) Hoxb-5, (K, L) Hoxb-7, (M-O) Hoxb-t9. All embryos are shown with anterior to the left. All neurula stage embryos were photographed in methanol without clearing and with the exception of (K) all are shown as dorsal views. The later stage embryos were cleared in Murray and were photographed from the side. Note the Haxb staining at the lateral edges of the neural plate at stage 12.5- 13 in (A). This staining appears to be relatively posterior compared to the stage 17 embryos hybridized to the Hoxb-i probe. This is explained by the disproportionate elongation of the trunk region during neurulation. The Haxb-3-stained embryo in (D) is slightly later (stage 14/15) than the other early neurulae shown, so the neural folds are already raised and the staining appears relatively anterior because of the changing shape of the embryo. The embryo is tilted slightly to one side. The stage 13 embryo in (J) was unstained by the Hoxb-5 probe. In (F) the unfocused staining anterior to the main hindbrain and third branchial arch staining is on the other side of the cleared embryo.
Fig. 3. Localization of Hoxb gene expression on sections. Transverse (A-E, G) and sagittal (F) sections of whole embryos hybridized with Hoxh gene probes. Only the dorsal structures are shown. The hybridization signal is blue and the red color is a nuclear counterstain. (A) Strong Hoxb1 expression in the neural crest and weaker expression in the forming neural tube at stage 17. (B) Hoxb-1 expression in rhoxnbomere 4 at stage 24. Note the unstained floorplate of the neural tube. (C) Hoxb-3 expression in the neural crest and future neural tube at stage 17/18. (D) Hoxb-3 expression in the hindbrain at stage 27. The epidermal staining is nonspecific. This section was not counterstained with neutral red. (E) Hoxb-4 expression in the future neural tube of a stage 17 embryo The somites are also weakly stained (F) Sagittal section of a stage 32 embryo which had been hybridized with the Hoxb-4 probe, showing somite staining. (G) Hoxb-5 staining in the spinal cord of a stage 28 embryo. Abbreviations: nc, neural crest; n, notochord; ar, archenteron cavity; s, somites; e, ear anlage.
FiG. 4. Autonomous differentiation of embryo fragments. Stage 12.5 embryos were cut into three equal-sized fragments along the a-p axis and cultured until control embryos reached tadpole stages. (A) anterior one-third pieces all developed eyes, cement glands, and hearts. (B) middle one-third pieces all gave rise to somites and dorsal fin. Three of the explants developed rudimentary eyes indicating that specification of the a axis is Incomplete at this stage. (C) posterior one-third pieces all developed as tails, although some were small.
FIG. 5. Hoxb genes show two types of expression pattern in the ONS and dorsal ventral colinearity in the mesodenn. (A) Hozb-1, (B) Hcxrb3, (C) Hoxb-4, (D) Hoxb-5, (E) Ioxb-7, (F) I-Iocth-9. The hybridization patterns in tailbud embryos demonstrated in Fig. 1 are shown by solid shading. The positions of the branchial arches are approximate. The hatched area in (F) represents additiona] staining in the lateral plate mesoderm detected by other methods (Dekker et aL, 1992b; Wright et aL, 1990). Note the a-p colinearity of expression in the CNS and the restricted expression domains of the 3genes, Hoxb-l, Hoxb-3, and Hoxb-4 compared with the more extensive expression domains of the 51 genes, Hoxb-5, Hoxb-7, and Hoxb-9. The possible dorsoventra] colinearity of expression in the mesoderm is also illustrated.
hoxb5 ( homeobox B5 ) gene expression in Xenopus laevis embryos, NF stage 31, as assayed by in situ hybridization, lateral view, anterior left, dorsal up.
hoxb7. In both cases, the posterior-most part of the hindbrain, the spinal cord, the pronephric duets, and the tail bud were stained.