Targeted knockdown of Zic3 results in axial patterning defects in Xenopus
To examine the role of Zic3 in early embryonic patterning, translational-blocking (TB MO) and splice site blocking (SS MO) morpholinos were designed to knockdown expression in Xenopus (). RT-PCR and subsequent sequencing revealed that Zic3 SS MO altered splicing such that exon 2 is skipped, resulting in a frameshift and premature truncation of the Zic3 protein prior to the DNA binding domains (). To test the function of the Zic3 TB MO, in vitro transcription-translation of a Xenopus zic3 or human HA-ZIC3 expression plasmid was performed with 35S-Met in the presence or absence of TB MO. TB MO specifically inhibited translation of Xenopus zic3 (Xzic3) but not human HA-ZIC3 ().
Fig. 1 Knockdown of Zic3 disrupts axial development in Xenopus. (A) Schematic of zic3 exon-intron structure showing translational-blocking (TB) and splice site blocking (SS) morpholino target sites. (B) RT-PCR analysis of zic3 in control (CTRL) and SS MO injected (more ...)
The Zic3 MOs were independently injected into the marginal zone of both dorsal blastomeres of 4-cell stage embryos to knockdown Zic3 in the dorsal mesoderm where it is highly expressed during gastrulation (Kitaguchi, et al., 2000
; Nakata, et al., 1997
). Zic3 morphants exhibited a variety of axial defects. A scaling system was utilized to assess the phenotypes resulting from a single Zic3 MO dose. The scale ranged from a score of 1, indicating normal axis elongation and closure of the neural tube, to a score of 5, indicating severe dorsal flexion with neural tube defects (). Both Zic3 MOs induced comparable axial defects with average scores of 2.78 (TB MO) and 2.87 (SS MO), as compared to an average score of 1.18 in control (CTRL MO) injected embryos (). The consistency in phenotypes from both Zic3 MOs suggests specificity of knockdown. The Zic3 SS MO was used for additional analyses except where indicated.
Convergent extension defects in Xenopus Zic3 morphants
Defects in C-E are characterized by gastrulation defects, impaired neural tube closure, and failure to extend the A-P axis. Zic3 morphants were analyzed during gastrulation and neurulation to assess the role of Zic3 in C-E. At gastrula stage, approximately 50% of the Zic3 morphants showed delayed blastopore closure, occasionally coupled with irregular shape as assessed at stage 12 (). Some of these embryos eventually closed their blastopores while others failed to do so resulting in an open neural tube, indicative of incomplete gastrulation. Blastopore size was measured to quantify closure at late gastrula stage using brachyury WISH to mark the mesoderm surrounding the blastopore (; within white dotted line). Blastopore area measurements confirmed that Zic3 morphant embryos had blastopore closure defects compared to control MO-injected embryos (). Brachyury expression was not perturbed in the Zic3 morphants.
Fig. 2 Xenopus Zic3 morphants exhibit defective gastrulation and convergent extension morphogenesis. (A) Normal control (CTRL) MO embryos at gastrula stage. (B) Sibling Zic3 morphants at stage 12.25 with delayed and abnormal blastopore closure. (C) The percentage (more ...)
At late gastrula stage, brachyury
also marks the developing notochord, a structure that is characterized by compaction and elongation. If C-E is disrupted, the notochord fails to elongate, resulting in a lower length to width ratio (Wallingford et al., 2001a
). Analysis of the notochord in injected embryos revealed that the average length to width ratio was significantly reduced in Zic3 morphants relative to CTRL MO-injected embryos ().
To assess extension of the A-P axis, embryos were measured from cement gland to tail tip (). Zic3 morphants had a shorter average A-P axis length than CTRL MO embryos (). This finding is consistent with the average C-E phenotype score ~3 seen in the Zic3 morphants (). Taken together, the results demonstrate that Xenopus Zic3 morphants exhibit blastopore closure, notochord elongation and A-P axis defects that are consistent with a disruption of C-E morphogenesis.
The requirement for Zic3 in axis formation is conserved across species
To test whether the role of Zic3 during early development is conserved across species a MO was designed to knockdown zic3 expression in zebrafish. Similar to Xenopus Zic3 morphants, zebrafish Zic3 morphants developed phenotypes associated with C-E defects. During gastrulation, WISH analysis of no tail, a zebrafish ortholog of brachyury, showed that the developing notochord tissue was wide and loosely packed in the majority of Zic3 morphants instead of tightly condensed as seen in the stage-matched controls (). At early somite stages, WISH markers revealed an increased width of somites (white line in ) and the neural plate (red line in ) in Zic3 morphants as compared to controls. Increased somite and neural plate widths are indicative of a failure of cells to converge towards the midline. To quantify this effect, we measured the width of the anterior neural plate (red line in ), which was wider in Zic3 morphants than in controls (). In addition, Zic3 morphants showed defects in A/P axis elongation, as indicated by an increased distance between the head and tail (). At 2 days post-fertilization (dpf), Zic3 morphants exhibited severe curvature and shortening of the A/P body axis with 76% of morphants demonstrating a short body axis as compared to 6% of control embryos (, show representative L images; shows quantitative results). At 5 dpf, A/P axis defects persisted in Zic3 morphants and these embryos often developed edema (). Importantly, co-injection of Zic3 MO with Xenopus zic3 mRNA, which alone had little effect on axial patterning (), was able to rescue the axial defects caused by Zic3 MO (). These results indicate axial phenotypes observed in Zic3 morphants are due specifically to loss of Zic3 function, and that Zic3 plays a conserved role in axis formation.
Fig. 3 Convergent extension and axial defects in Zic3 morphant zebrafish. (A–D) WISH of no tail expression in the notochord at 90% epiboly (A,B) and tailbud (C,D) stages. Control (CTRL) embyos (A,C) showed normal notochord development, whereas Zic3 morphants (more ...)
Fig. 4 Xenopus zic3 rescues axial defects in zebrafish Zic3 morphants. (A–D) Live embryos at 5 days post-fertilization. (A) Uninjected control embryos. (B) Xzic3 mRNA-injected embryos showed subtle phenotypes including a curled tail tip. (C) Zic3 morphant (more ...)
Zic3 has a conserved role in specifying left-right asymmetry
During normal vertebrate development, the heart and gut become asymmetric along the L-R axis. Loss of Zic3 function in humans and mice disrupts organ L-R asymmetry (Gebbia, et al., 1997
; Purandare, et al., 2002
; Ware, et al., 2006b
; Ware, et al., 2004
). In Xenopus
, the cardiac outflow tract stems from the right side of the heart and loops towards the left side (). Xenopus
Zic3 morphants exhibited an increase in heart looping abnormalities compared to control embryos (; Zic3 MO 34%, n= 127; CTRL MO 12%, n= 133). These gross abnormalities included a mirror reversal dextrocardia phenotype and an ambiguous phenotype, characterized by a wide variety of looping or morphologic abnormalities of the outflow tract or heart (). Laterality defects of the gut were also observed in Zic3 morphants. The developing Xenopus
gut originates on the embryo’s right side (right origin, RO), and coils in a counter-clockwise (CCW) direction (). Mirror image reversal (left origin, LO, and clockwise coiling direction, CW) was identified in 7% of Zic3 morphants compared to 4% in control morphants (). Gut anatomy categorized as abnormal included embryos with other combinations of abnormal origin or coiling () and ambiguous embryos (; Zic3 MO 44%, n= 127; CTRL MO 6%, n=133). Abnormal gut coiling, also known as malrotation, is a frequently identified feature of human heterotaxy (Lin et al., 2000
Fig. 5 Reduction of Zic3 expression results in heart and gut abnormalities. (A–C) Ventral view of troponin WISH staining of the heart in Xenopus stage 46 embryos. (A) Normal heart looping with outflow tract from the right side. (B) Mirror phenotype, (more ...)
Zic3 knockdown in zebrafish also resulted in heart and gut laterality defects. In contrast to normal rightward looping of the heart observed in controls, the heart often failed to loop or showed a mirror reversal in Zic3 morphants (; Zic3 MO 48%, n=81; Control 7%, n=58). Xenopus zic3 mRNA was able to rescue these heart laterality defects (), indicating this effect is specific to loss of Zic3 function. In addition, these results demonstrate that, similar to convergent extension defects, Xenopuszic3 mRNA is able to rescue L-R patterning abnormalities.
Fig. 6 Abnormal heart looping and gut formation in zebrafish Zic3 morphants. (A–C) Direction of cardiac looping (arrows) visualized by WISH of cardiac myosin light chain-2 (cmlc2) in 2 dpf embryos. Control embryos showed normal looping (A), whereas Zic3 (more ...)
Zebrafish Zic3 morphants also showed gut laterality and patterning defects (). The most common defect was a mirror reversal of the placement of the liver and pancreas. Ambiguous patterning of liver and pancreas, including multiple embryos with accessory structures, was common. Together, these results demonstrate a conserved role for Zic3 in establishing normal organ laterality.
Laterality of organs is controlled by asymmetric expression of a conserved Nodal-Pitx2 signaling cascade. To determine whether Zic3 functions upstream of Nodal-Pitx2 asymmetry, WISH was used to analyze expression of pitx2 in Xenopus and the Nodal-related gene southpaw (spaw) in zebrafish, both of which are normally expressed in left lateral plate mesoderm. Abnormal pitx2 and spaw expression patterns were identified in a majority of Xenopus and zebrafish Zic3 morphants, confirming the function of Zic3 upstream of Nodal signal transduction ().
Fig. 7 Abnormal L-R molecular marker expression in Xenopus and zebrafish Zic3 morphants. (A) Pitx2 expression in Xenopus Zic3 morphant embryos at stage 30. Embryos were photographed from both left and right sides. Red arrows highlight pitx2 expression. (B) Quantitative (more ...)
Correlation between convergent extension and left-right asymmetry
To understand the potential effect of defective C-E morphogenesis on L-R patterning, a subset of control MO and Zic3 SS MO-injected Xenopus embryos were further analyzed based on their C-E phenotype score and their pitx2 expression pattern. As described, Zic3 morphants exhibited a higher incidence of C-E defects () and L-R defects () than control MO-injected embryos. Analysis of defective C-E morphogenesis and L-R patterning, using pitx2 expression, in the same embryos revealed that disruption of C-E correlated with increased L-R patterning abnormalities (). Interestingly, pitx2 expression was also abnormal in nearly 90% of the small subset of control MO-injected embryos with severe C-E defects. This result suggests that the correlation of L-R patterning abnormalities with severity of C-E defects is generalizable (). In addition, Zic3 morphants with severe C-E phenotypes exhibited an increase in L-R patterning defects compared to Zic3 morphants with normal to mild C-E phenotypes (). Taken together, these results suggest that Zic3 is necessary for both L-R patterning and C-E morphogenesis, and that aberrant C-E morphogenesis exacerbates L-R patterning defects.
Fig. 8 Correlation between C-E morphogenesis and L-R patterning. (A) Percentage of control morphant (CTRL MO) and Zic3 morphant embryos with defective C-E phenotypes. (B) Percentage of control and Zic3 morphants with L-R patterning abnormalities. (C) Comparison (more ...)