XB-ART-1812Development 2005 Jun 01;13212:2861-71. doi: 10.1242/dev.01872.
Show Gene links Show Anatomy links
Knockdown of the complete Hox paralogous group 1 leads to dramatic hindbrain and neural crest defects.
The Hox paralogous group 1 (PG1) genes are the first and initially most anterior Hox genes expressed in the embryo. In Xenopus, the three PG1 genes, Hoxa1, Hoxb1 and Hoxd1, are expressed in a widely overlapping domain, which includes the region of the future hindbrain and its associated neural crest. We used morpholinos to achieve a complete knockdown of PG1 function. When Hoxa1, Hoxb1 and Hoxd1 are knocked down in combination, the hindbrain patterning phenotype is more severe than in the single or double knockdowns, indicating a degree of redundancy for these genes. In the triple PG1 knockdown embryos the hindbrain is reduced and lacks segmentation. The patterning of rhombomeres 2 to 7 is lost, with a concurrent posterior expansion of the rhombomere 1 marker, Gbx2. This effect could be via the downregulation of other Hox genes, as we show that PG1 function is necessary for the hindbrain expression of Hox genes from paralogous groups 2 to 4. Furthermore, in the absence of PG1 function, the cranial neural crest is correctly specified but does not migrate into the pharyngeal arches. Embryos with no active PG1 genes have defects in derivatives of the pharyngeal arches and, most strikingly, the gill cartilages are completely missing. These results show that the complete abrogation of PG1 function in Xenopus has a much wider scope of effect than would be predicted from the single and double PG1 knockouts in other organisms.
PubMed ID: 15930115
Article link: Development
Species referenced: Xenopus laevis
Genes referenced: dll4 dlx2 egr2 en2 gbx2.1 gbx2.2 hoxa1 hoxa2 hoxa3 hoxa5 hoxb1 hoxb2 hoxb9 hoxc6 hoxc9-like hoxd1 hoxd3 hoxd4 lsamp myod1 nrp1 otx2 snai1 snai2
Antibodies: Lsamp Ab1
Morpholinos: hoxa1 MO1 hoxa1 MO2 hoxb1 MO1 hoxb1 MO2 hoxd1 MO1 hoxd1 MO2
Article Images: [+] show captions
|Fig. 1. Hox PG1 genes have overlapping expression domains during Xenopus laevis development. Hoxa1 (A-H), Hoxb1 (I-P) and Hoxd1 (Q-X) have overlapping expression patterns. Gastrula stage embryos (A,B,I,J,Q,R) are shown from the vegetal side, dorsal up. When compared with Otx-2 expression (* in A,I,Q), it can be seen that the Hox expression does not reach the presumptive midbrain region. Neurula embryos are shown from the dorsal side, anterior to the right (C,E,K,M,S,U) or from the anterior side, dorsal up (D,F,L,N,T,V). Later embryos are shown from the lateral side, anterior to the right (G,H,O,P,W,X). Note pharyngeal arch expression of all three PG1 genes (arrows in G,H,O,P,W,X).|
|Fig. 7. Hoxd1 mRNA rescues Hox expression in PG1 knockdown embryos. Embryos injected on the left-hand side (lhs) with Hoxd1 MO insensitive mRNA (A,D,G,J), all three PG1 morpholinos (B,E,H,K) or a combination of both (C,F,I,L) were analysed for expression of Hoxb2, Hoxa3, Hoxd4 and Hoxc6. Embryos are shown from the dorsal side, anterior to the top. Hoxb2 expression was expanded posteriorly with Hoxd1 mRNA (A: 75%, n=12), lost with the triple PG1 morpholinos (B: 93%, n=28) and rescued when a combination was injected (C: 48% of embryos had downregulated expression, n=33). The other Hox genes were not affected by the Hoxd1 mRNA overexpression. PG1 MO downregulated expression of Hoxa3 (E: 78%, n=9), Hoxd4 (H: 100%, n=9) and Hoxc6 (K: 78%, n=9) and in the rescues this downregulation was reduced to 34%, 20% and 33% for Hoxa3, Hoxd4 and Hoxc6, respectively (F,I,L).|
|Fig. 8. Effects of single PG1 knockdowns on Hox genes. Embryos injected on the left-hand side (lhs) with the triple PG1 MO combination had reduced Hoxb2 (A: 93%, n=28), Hoxa3 (E: 100%, n=14) and Hoxd4 (I: 79%, n=14). The Hoxa1 MO weakly downregulated Hoxb2 (B: 80%, n=10) and reduced Hoxa3 (F: 54%, n=11), but not Hoxd4 (J: 100%, n=20). The Hoxb1 MO had no effect on any of these genes (C: Hoxb2, 100%, n=10; G: Hoxa3, 100% n=9; K: Hoxd4, 100%, n=18). Hoxd1 MO reduced Hoxd4 expression (L: 100%, n=16), weakly reduced Hoxb2 (D: 43%, n=14), but did not affect Hoxa3 (H: 100%, n=10).|
|Fig. 2. Specificity of PG1 morpholinos. Two morpholinos for each Hox PG1 gene were injected into the left-hand side of the embryo and Krox20 (r3 and r5) and Engrailed-2 (midbrain/hindbrain boundary: * in A-I) expression analysed. The hindbrain region of early tailbud stage embryos (anterior to the top) is shown for non injected controls (NIC; A,D,G), Hoxa1 1st (B: 80%, n=15) and 2nd (C: 70%, n=10) morpholinos, Hoxb1 1st (E: 85%, n=13) and 2nd (F: 70%, n=20) morpholinos and Hoxd1 1st (H: 90%, n=11) and 2nd (I: 80%, n=15) morpholinos. Overexpression of Hoxd1 morpholino insensitive RNA (J) and rescue of Hoxd1 1st and 2nd morpholinos with Hoxd1 RNA is also shown (K,L). Dotted line indicates the midline.|
|Fig. 3. The triple PG1 knockdown phenotype is more severe than the double knockdowns. Morpholinos for each Hox PG1 gene were injected either in double (D: A1/B1 52%, n=25; E: B1/D1 48%, n=29; F: D1/A1 64%, n=33) or triple (C: A1/B1/D1 94%, n=35) combinations into the left-hand side of the embryo and Krox20 and Engrailed-2 (*) expression analysed. The hindbrain region of early tailbud stage embryos is shown, with anterior to the top. Non-injected controls (NIC; A), and embryos injected on the left-hand side of the embryo with control morpholino (B) are also shown. Arrowheads indicate neural crest expression.|
|Fig. 4. Hoxd1 mRNA can partially rescue the triple PG1 morpholino phenotype. Early tailbud stage embryos were injected on the left-hand side (lhs) with all three PG1 morpholinos (A1/B1/D1 MO) either alone (B,C), or in combination with Hoxd1 mRNA (E,F). Non-injected controls (NIC) and control morpholino (Cont MO) are also shown (A,D). Anterior is to the top. G: Krox-20 expression is partially rescued by Hoxd1 RNA. Numbers of embryos with two stripes (E), one broad stripe (F), one narrow stripe (B) or no stripes (C) of Krox20 expression were scored. Co-injection of Hoxd1 RNA partially rescued the Krox20 expression (n=29), which was either absent or very much reduced in embryos injected only with A1/B1/D1MO (n=18).|
|Fig. 5. Triple PG1 knockdown severely affects hindbrain patterning. Morpholinos for each of the Hox PG1 genes (A1/B1/D1MO) were co-injected into the lhs of the embryo (B,D,F,H,J). Non-injected controls are also shown (A,C,E,G,I). The hindbrain region is shown, with anterior to the top. Expression of the neural marker Nrp1 (A,B: 100%, n=11), the anterior marker Otx2 (C,D: 100%, n=10), Gbx2, which is expressed in r1 (E,F: 77%, n=13) and Krox20 (G,H: no stripes 41%; one faint stripe 35%, n=71), was analysed. Dotted line indicates the midline. In later embryos (st. 46), the neural antibody 2-G9 was used to show the morphology of the hindbrain [NIC, I; A1/B1/D1MO, injected on the left-hand side (lhs) *, J: 100%, n=5]. Rhombomere boundaries are marked.|
|Fig. 6. Knockdown of all three PG1 genes affects more posterior Hox genes. Non injected controls (NIC; A,C,E,G,I,K,M,O) and embryos injected on the left-hand side (lhs) with all three Hox PG1 morpholinos (A1/B1/D1MO; B,D,F,H,J,L,N,P) were analysed for Hox gene expression at early tailbud stage. Embryos are shown from the dorsal side, anterior to the top. Hoxb1 expression was lost (A,B: 70%, n=10), and Hoxa2 (C,D: 100%, n=18) and Hoxa3 (E,F: 100%, n=14) expression was almost completely absent. Hoxd3 (G,H: 68%, n=19) and Hoxd4 (I,J: 79%, n=14) neural tube expression was still present, but the pharyngeal arch expression was lost. The more posterior Hox genes (Hoxa5, K,L: 100%, n=10; Hoxc6, M,N: 71%, n=14; Hoxb9, O,P: 93%, n=14) had reduced neural tube expression in the anterior part of the embryo but their posterior expression domains (arrow) and mesodermal expression were unaffected.|
|Fig. 9. Knockdown of all three PG1 genes affects the development of neural crest and its derivatives. Non-injected controls (NIC; A,E,G,I) and embryos injected on the left-hand side with Hoxd1 MO insensitive mRNA (B,J), all three Hox PG1 morpholinos (A1/B1/D1MO; C,F,H,K) or a combination of both (D,L) were analyzed for neural crest gene expression, or stained with alcian blue to show the craniofacial structures. The embryos shown in A-F are early tailbud stage embryos shown from the dorsal side with anterior to the top. With the PG1 MOs, neural crest cells fail to migrate away from the hindbrain, as shown by Xslug (C: 86%, n=29) and Xsnail (F: 100%, n=14) expression. This is rescued (22% of embryos with restricted Xslug expression, n=9) by Hoxd1 overexpression (D) and simple overexpression leads to an expanded Xslug domain (B: 100%, n=10). Slightly later tailbud embryos analyzed for dll4 are shown from the left-hand side, with anterior to the left, and in the triple PG1 knockdown dll4 expression in the branchial arches is restricted to close to the hindbrain (50%) or completely lost (G,H: 44%, n=16). Alcian blue was used to stain the cartilage of stage 49, embryos (I-L) shown from the ventral side to show the craniofacial structures (injected side on the right:*). The complete gill area is missing in the triple morpholino injected embryos (K: 82%, n=11) and this is rescued (26% of embryos with missing gill, n=19) when Hoxd1 mRNA is co-injected (L). C, ceratohyal; G, gill cartilage; M, Meckel's cartilage. Dotted line indicates midline.|
|Fig. 9. Dll4 expression at stg 29-44|