October 1, 2003;
Regulation of apoptosis in theXenopus embryo by Bix3.
Members of the Bix family of homeobox-containing genes are expressed in the vegetal hemisphere
of the Xenopus embryo
at the early gastrula stage
. Misexpression of at least some of the family members causes activation of mesoderm
- and endoderm
-specific genes and it is known that some of the proteins, including Bix2
, interact with Smad proteins via a motif that is also present in the related protein Mixer
. In this paper we study the function of Bix3
. Misexpression of Bix3
, similar to misexpression of other members of the Bix family, causes the activation of a range of mesendodermal genes, but the spectrum of genes induced by Bix3
differs from that induced by Bix1
. More significantly, we find that overexpression of Bix3
also causes apoptosis, as does depletion of Bix3
by use of antisense morpholino oligonucleotides. The ability of Bix3
to causes apoptosis is not associated with its ability to activate transcription and nor with its possession of a Smad interaction motif. Rather, Bix3
lacks a C-terminal motif, which, in Bix1
, acts in cis to inhibit apoptosis. Mutation of this sequence in Bix1
causes the protein to acquire apoptosis-inducing activity.
[+] show captions
Fig. 1. The expression patterns of Bix1 and Bix3. In situ hybridization was performed on bisected embryos at the indicated stages using probes specific for Bix1 and Bix3. In C-H, dorsal is to the right.
(CURATORS NOTE nomenclature changes: bix1 = bix1.3; bix3= bix1.1)
Fig. 4. Apoptosis in Xenopus embryos injected with RNA encoding Bix3. (A-F) One group of embryos was left uninjected whereas another (G-I) was injected with 400 pg of Bix3 RNA at the one-cell stage. At stage 8.5, half of the uninjected embryos were transferred to 0.1 mg/ml cycloheximide (D-F), whereas the other half was allowed to develop normally (A-C). Embryos were collected at the indicated stages and used for TUNEL staining. Positive cells are dark blue. Note that loosely adherent cells in Bix3-injected embryos, as seen in Fig. 2, are lost during the TUNEL procedure. (J) tPARP assay shows that Bix1 does not activate caspase activity whereas induction of apoptosis by Bix3 was first detectable by stage 11. Cycloheximide induces caspase activity from early gastrula stage 10. Intact tPARP has a relative molecular mass of 60 kD (upper arrows in both gels), and this is cleaved into fragments of 36 kD and 24 kD (lower arrows). (K) Human Bcl2 delays the onset of Bix3-induced apoptosis. Caspase assays were performed on control embryos at the indicated stages, embryos injected with RNA encoding Bix3, or embryos injected with Bix3 RNA together with RNA encoding human Bcl2 (500 pg). Note that human Bcl2 delays the onset of caspase activity. (L-N) Partial rescue of the cell adhesion and apoptosis defect by human Bcl2. Bix3 RNA alone (50 pg; L), or in combination with 500 pg of human Bcl2 RNA (M), was injected in the animal pole of one-cell stage embryos. Embryos were left to develop and photographed at mid-gastrula stage 11.5-12. Note that human Bcl2 does not prevent the appearance of darkly pigmented cells in the animal hemisphere but does reduce or delay the incidence of cell disaggregation. Human Bcl2 delayed disaggregation in five out of seven experiments of this sort. (N) Quantitation of an experiment of the type illustrated in (L) and (M). Embryos undergoing cell disaggregation were scored at the indicated stages.
Effects of Bix1 and Bix3 on Xenopus embryos and animal pole regions. Bix1 (E-H) or Bix3 (I-L) RNA (400 pg) was injected into the animal pole regions of Xenopus embryos at the one-cell stage. Other embryos (A-D) were left uninjected. Embryos were fixed at the indicated stages and photographed. (A,E,I) Animal view of stage 9 embryos. (B,F,J) Animal view of stage 12 embryos (except the embryo in the upper-left corner of B, which is viewed from the vegetal pole). (C,G,K) Stage 15 embryos. (D,H,L) Animal caps dissected at stage 8.5 and allowed to develop to the equivalent of early gastrula stage 10.5. Note that animal caps derived from embryos injected with Bix3, but not Bix1, RNA are in the early phases of disaggregation. Arrows in I indicate darkly pigmented cells.
Gene activation by Bix1 and Bix3. Xenopus embryos were injected with RNA (200 pg) encoding Bix1 or Bix3 or were left uninjected. Animal pole regions were dissected at late blastula stage 9 and cultured to the equivalent of early gastrula stage 10.5, when they were assayed for expression of the indicated genes by real-time PCR. Bix3 causes a decrease in expression of the control transcript plakoglobin but activates Sox17α and goosecoid to high levels. It does not induce expression of cerberus or XHex. Bix1 is a powerful inducer of Xhex and also activates expression of cerberus. Inset shows data normalised to levels of plakoglobin expression.
Rescue of apoptosis by injection of 200 pg Bix3 RNA but not Bix1. (A) Uninjected embryos at stage 19. (B) Embryos injected at the one-cell stage with a Bix3 antisense morpholino oligonucleotide begin to undergo apoptosis. (C) Embryos injected at the one-cell stage with a control morpholino oligonucleotide appear normal. (D-F). Embryos injected at the one-cell stage with Bix3 antisense morpholino oligonucleotide together with the indicated amounts of Bix1 RNA which lacks the 5′ untranslated region against which the oligonucleotide is directed. Rescue is not observed at any concentration of RNA. (G-I) Embryos injected at the one-cell stage with Bix3 antisense morpholino oligonucleotide together with the indicated amounts of Bix3 RNA which lacks the 5′ untranslated region against which the oligonucleotide is directed. Apoptosis is 'rescued' by 200 pg RNA (H) but not by 100 pg (G) or 400 pg (I).
An antisense Bix3 morpholino oligonucleotide causes apoptosis in the early Xenopus embryo without affecting regional specification. (A) Group of control embryos at stage 20. (B) Group of embryos injected at the one-cell stage with 50 ng of the control morpholino oligonucleotide described in Fig. 5. (C) Group of embryos injected with 50 ng of the Bix3 antisense morpholino oligonucleotide shown in Fig. 5. Notice the onset of apoptosis in dorsal axial structures. (D) Higher-power view of an embryo injected with the mutated morpholino oligonucleotide. (E) Higher-power view of an embryo injected with a Bix3 antisense morpholino oligonucleotide. (F) High-power view of an embryo injected with a Bix3 antisense morpholino oligonucleotide and stained using the TUNEL technique. Notice apoptotic cells in the neural plate. (G) Transversely bisected late gastrula control embryo stained using the TUNEL technique. Very few apoptotic cells are visible. Inset shows higher-power view of the archenteron roof. (H) Transversely bisected late gastrula embryo previously injected with a Bix3 antisense morpholino oligonucleotide. Note apoptotic cells in the endodermal mass and in dorsal axial structures. Inset shows higher-power view of the archenteron roof, with apoptotic cells in all three germ layers. (I-N) Control animal caps (I) or animal caps derived from embryos injected with RNA encoding activin (6 pg) or noggin (250 pg) together with 50 ng of antisense Bix3 morpholino oligonucleotide (J-N, as indicated) were fixed at the equivalent of stage 11 and stained using the TUNEL technique. Only animal caps treated with activin and injected with Bix3 morpholino undergo apoptosis. (O) Samples equivalent to those illustrated in I-N were subjected to a tPARP assay. Only animal caps treated with activin and injected with Bix3 morpholino undergo apoptosis. (P) A tPARP assay confirms that a Bix3 antisense morpholino oligonucleotide causes premature apoptosis. In control embryos, and in embryos injected with the control morpholino oligonucleotide, caspase activity is first detected at stage 17. In embryos injected with the Bix3 oligonucleotide caspase activity is present at stage 14.
The ability of Bix3 to activate transcription is not correlated with its ability to induce apoptosis. (A) Embryos at the one-cell stage were injected with RNA encoding full-length Bix3 [Bix3(1-389)], truncations of Bix3 (middle three constructs), or a mutation of Bix3 in which two prolines at positions 313 and 314 are replaced by alanines (Bix3 PP313/4AA), thereby mutating the Smad interaction motif. All five constructs induce apoptosis with equal efficiency. (B) Transient transfection assays in cultured cells. The Bix3 constructs described in A were co-transfected into COS cells with a reporter construct containing 6 copies of the P3 site, which is known to bind Bix proteins (Mead et al., 1998), and a reference plasmid. The average of three experiments are shown. In each the luciferase elicited by full-length Bix3 was defined as 100% and activity elicited by the other constructs was normalised relative to this. All transfected constructs used generate protein in approximately equivalent amounts (not shown). Note that the transcription activation activity of Bix3 (1-214) is substantially reduced, but its ability to induce apoptosis resembles that of the full-length protein.
Truncated versions of Bix1 cause apoptosis; the ability to cause apoptosis is not correlated with the ability of the protein to activate transcription but does require DNA binding. (A) RNA encoding truncated and deleted versions of Bix1 was injected into Xenopus embryos and the same constructs were introduced into COS cells to assess their transcriptional activity as described in Fig. 8. All transfected constructs generated protein in approximately equivalent amounts (not shown). Deletion of 25 C-terminal amino acids of Bix1 causes the protein to acquire apoptosis-inducing activity. Further deletions suggest that the apoptosis-inducing domain resides between amino acids 144 and 225 and that optimal apoptotic activity can be obtained in a construct [Bix1 (1-225)] in which 78% of transcriptional activity is lost. (B) Mutation of Q132 to E reduces transcriptional activation, substantially decreases the ability of the protein to cause apoptosis abolishes DNA binding (not shown). The ability of truncated Bix1 to induce apoptosis requires DNA binding. Bix1 (1-376) induces apoptosis, is a powerful activator of transcription and binds DNA (not shown).
A C-terminal domain of Bix1 acts in cis to prevent induction of apoptosis; the N terminus of Bix3 does not possess this activity. Threonine 384 is necessary for the anti-apoptotic function of the Bix1 C-terminal region. (A) A chimeric form of Bix1, in which the 45 C-terminal amino acids are derived from Bix3, induces apoptosis. (B) Comparison of the 45 C terminal amino acids of Bix1 and Bix3. (C) Mutation of threonine 384 of Bix1 to alanine causes the protein to acquire apoptosis-inducing activity, as does mutation of the two glutamic acids of the DWEEN motif.