XB-ART-50364Dev Biol. December 15, 2015; 408 (2): 328-44.
Xenopus pax6 mutants affect eye development and other organ systems, and have phenotypic similarities to human aniridia patients.
Mutations in the Pax6 gene cause ocular defects in both vertebrate and invertebrate animal species, and the disease aniridia in humans. Despite extensive experimentation on this gene in multiple species, including humans, we still do not understand the earliest effects on development mediated by this gene. This prompted us to develop pax6 mutant lines in Xenopus tropicalis taking advantage of the utility of the Xenopus system for examining early development and in addition to establish a model for studying the human disease aniridia in an accessible lower vertebrate. We have generated mutants in pax6 by using Transcription Activator-Like Effector Nuclease (TALEN) constructs for gene editing in X. tropicalis. Embryos with putative null mutations show severe eye abnormalities and changes in brain development, as assessed by changes in morphology and gene expression. One gene that we found is downregulated very early in development in these pax6 mutants is myc, a gene involved in pluripotency and progenitor cell maintenance and likely a mediator of some key pax6 functions in the embryo. Changes in gene expression in the developing brain and pancreas reflect other important functions of pax6 during development. In mutations with partial loss of pax6 function eye development is initially relatively normal but froglets show an underdeveloped iris, similar to the classic phenotype (aniridia) seen in human patients with PAX6 mutations. Other eye abnormalities observed in these froglets, including cataracts and corneal defects, are also common in human aniridia. The frog model thus allows us to examine the earliest deficits in eye formation as a result of pax6 lesions, and provides a useful model for understanding the developmental basis for the aniridia phenotype seen in humans.
PubMed ID: 25724657
PMC ID: PMC4549229
Article link: Dev Biol.
Grant support: R01 EY017400 NEI NIH HHS , R01 EY018000 NEI NIH HHS , R01 EY022954 NEI NIH HHS , R01EY017400 NEI NIH HHS , EY022954 NEI NIH HHS , EY018000 NEI NIH HHS , P40 OD010997 NIH HHS
Genes referenced: gsx2 ins mab21l1 myc neurog2 nkx2-1 pax6 rax tbx2
Disease Ontologies referenced: aniridia
OMIMs referenced: ANIRIDIA; AN
External Resources: GO Terms referenced: eye development
Article Images: [+] show captions
|Fig. 1. Strategy for targeted mutagenesis of Xenopus tropicalis pax6 locus. (A) Structure of pax6 gene is shown schematically (not to scale). P0, P1, Pα are promoters (shaded in grey). Exons (boxes) are numbered or named on top (0 to 13 or α) and color coded: white, untranslated regions (UTRs); orange, N-terminal region before the paired domain; other colors correspond to color codes for protein domains shown in (B). Two translation start sites are shown (ATG): orange, for full-length isoforms (Pax6 with or without 5a); green, for paired-less isoform (Pax6ΔPD). Positions of mouse Sey and rat rSey2 mutations are indicated by arrows on the bottom. Two pairs of TALEN targets in exon 7 and exon 9 are shown on top. (B) Structures of Pax6 wild-type and representative mutant proteins are shown schematically (not to scale). Color codes of protein domains are shown in the box except the first three amino acids, which are shown in orange. Note that the paired domain is located between amino acids 4–131 and the homeodomain between amino acids 212–272 in X. tropicalis due to insertion of two amino acids between the two domains. The amino acid sequence of each domain is as shown in Walther and Gruss (1991). Most but not all mutations have some non-Pax6 amino acids at the C-terminus. Note that X. tropicalis mutations in exon 7 could theoretically encode Pax6ΔPD if mutations were not simultaneously introduced in exon 9 (not shown in the drawing). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)|
|Fig. 2. Representative phenotypes of F0 embryos injected with TALENs. Lateral (A and B) and dorsal (A′, the same embryo shown in A) views of two embryos injected with a pair of TALENs targeting pax6 exon 9 (200 pg mRNA each). Asterisks in (A) and (A′) indicate the same abnormal eye. The most frequently seen eye phenotype is shown in (B) (arrow). (C) Lateral view of an embryo injected with a pair of TALENs targeting tyr (200 pg mRNA each). The eye is morphologically normal but less pigmented (arrow). (D) Lateral view of an embryo injected with two pair of TALENs targeting pax6 exons 7 and 9 (200 pg mRNA each). The abnormality here (arrow) is similar to what is shown in (B). (E) Mating of mature frogs raised from embryos injected with two pairs of pax6 TALENs. Eyes are missing in one side of each animal (arrows).|
|Fig. 3. Strategy for screening mutations in the pax6 locus edited by two pairs of TALENs targeting exon 7 and exon 9 simultaneously. (A) Schematic drawing of targeting regions, only showing exon 7 (blue) and exon 9 (magenta) for simplicity (not to scale). Small colored arrows indicate different primers for genomic PCR of exon 7 and exon 9 (each sequence is shown on the bottom, from 5′ to 3′). Different combinations of primers can amplify other possible junction regions created by deletion or inversion between exons 7 and 9 as schematically shown. (B) A representative gel pattern of genomic-PCR bands from an embryo injected with both pairs of TALENs mRNAs. In this specific case, one of the possible junctions representing a 3′ inversion site (where a 3′ piece of exon 7 would have been fused inverted to 3′ piece of exon 9, lane 5) was not detected, but in other cases, we have seen such a band. (C) The genomic-PCR band corresponding to what is shown in lane 2 in (B) from 3 pooled embryos was gel-purified, cloned, and sequenced. The results of 6 clones are aligned. Only regions between two target sites (target-Ls, left target; target-Rs, right target in orange) are shown. Corresponding wild-type sequences are shown on top for exon 7 and on the bottom for exon 9. Multiple 10-kb deletion events were confirmed, and junctions where 5′ parts of exon 7 (blue) fused to 3′ parts of exon 9 (magenta) were seen in the expected cleavage regions (in between left and right targets, indicated by scissors) in both exons 7 and 9, suggesting that 10-kbp deletions were due to cleavages by TALENs and thus both TALENs were functional. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)|
|Fig. 4. Xenopus pax6 mutants have abnormal eyes without lenses. (A–D) Offspring from crosses between F0 mosaic mutation carriers (A) or F1 carriers (B–D) showed a consistent phenotype in the eye (red arrows) regardless of the genotype of mutant alleles. F0 crosses (st.716 female #2 x male #3) created a variety of combinations of genotypes in mutant offspring (A), whereas 25% of offspring from F1 crosses (i.e., parents are not mosaic) have a single homozygous (B) or compound heterozygous (C,D) mutant genotype as expected. Note that mutants homozygous for a 13-bp deletion in exon 7 (B) are indistinguishable from other compound heterozygous mutants, regardless of the combination of exon 7 mutations (A, top right embryo), combination of exon 7 and exon 9 mutations (A, bottom right embryo and D), or combination of exon 9 mutations (C). Each embryo shown here was genotyped except for phenotypic wild-type embryos in C and D, thus shown as +/?. Judging from the mutation profiles (see Table S1), we could infer the origins of mutated alleles of offspring from F0 crosses, or in the case of F1 crosses, the genotypes of parents are known in advance, thus we show offspring genotypes in order of parental female locus/male locus except for the top wild-type embryo in (B), where the parental origin of alleles is undeterminable. Actual sequences of mutations can be found in Table S1. (E) Cross sections of embryos marked by red (mutant) and green (wild-type) asterisks in (A) to show morphology of the eyes. The wild-type embryo (green asterisk) has a lens (white arrowhead) and the retina is surrounded by retinal pigment epithelial (RPE) and separated from the brain (green arrowhead), whereas the mutant embryo (red asterisk) has no lens and the retina is disorganized (black arrowhead) and still connected to the brain (red arrowhead). (F) Schematic presentation of proteins encoded by the mutations shown here. Also refer to Fig. 1 for color codes. Exon 7 mutations cause truncation at the C-terminus of the paired domain (located between amino acid residues 4–131), whereas exon 9 mutations cause truncation in the middle of the homeodomain (located between amino acid residues 212–272). Truncations in many cases but not all are followed by unrelated amino acids due to the frameshift. Note that mutant embryos shown in (C) make almost identical mutant proteins with only one amino acid change at position 238, thus this is very close to being a homozygous mutation. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)|
|Fig. 5. Phenotypes of Xenopus pax6 mutants at tadpole stages. (A) Phenotypes of F2 embryos at stage 42 from an intercross of F1 heterozygotes carrying a 13-bp deletion in exon 7. −/−, homozygous mutant; +/−, heterozygous; +/+, wild-type. These are siblings of embryos shown in Fig. 4B. (B) Phenotypes of F2 embryos at stage 47 from cross between an F1 female carrying a 1-bp deletion ( #2) in exon 9 and an F1 male carrying a 13-bp deletion in exon 7. −/−, compound heterozygous mutant (genotype is ex9 δ1-bp#2/ex7 δ13-bp); +/−, heterozygous (genotype, +/ex7 δ13-bp); +/+, wild-type. These are siblings of embryos shown in Fig. 4D. (C and C′) Phenotype of F2 embryos at stage 42 (C) and stage 47 (C′) from mating of an F1 female carrying a 1-bp deletion (#2) in exon 9 and F1 male carrying a 1-bp deletion (#3) in exon 9. Embryos shown in C (mixture of lateral and dorsal views) are the same embryos as shown in C′ (all dorsal views) at the same axial position, the only difference being their ages. −/−, compound heterozygous mutant (genotype is ex9 δ1-bp#2/ex9 δ1-bp#3); +/−, heterozygous wild-type (genotype, ex9 δ1-bp#2/+ for top embryo, +/ex9 δ1-bp#3 for second embryo); +/+, homozygous wild-type. These are siblings of embryos shown in Fig. 4C. Orange arrows show the eye fused to the brain, which is often, but not always, seen in pax6 mutants regardless of genotype. (D–F) Cross sections of the eye region of embryos shown in (B) and (C′). Corresponding embryo and section are marked by the same colored asterisks. Wild-type embryos (D) have well organized retina and lens (white arrowhead), whereas mutants (E and F) do not have lens and the retina is disorganized (black arrowheads). By this stage, mutant eyes were separated from the brain in some embryos (F) or still connected to the brain in other embryos (E, red arrowhead). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)|
|Fig. 6. mab21l1 and myc are downstream of pax6. (A and B) In situ hybridization of wild-type (+/+) and pax6 mutant embryos (−/−, compound heterozygote of complex1 and 10-kbp inversion mutations, see Table S1 for sequence) at st, 21 (A) or st. 15 (B) with mab21l1 (A) or myc (B) probe. Red arrows indicate expression of mab21l1 (A) and myc (B) in the eye region. Expression of both genes is reduced in mutants. (C) In situ hybridization of wild-type (+/?) and rax mutant (−/−, see Fish et al., 2014) embryos with myc probe. Note that myc expression in the presumptive retina region (red arrow) is almost completely lost in the rax mutant as reported previously ( Fish et al., 2014), whereas in the pax6 mutant myc expression is consistently reduced, but to a lesser extent as shown in (B). White dotted circles show the presumptive lens ectoderm (PLE) region, in which expression of myc is indistinguishable between wild-type and rax mutant embryos. By contrast, expression of myc in the PLE of pax6 mutant is reduced compared to the wild-type embryo (B, white dotted circles). (D) In situ hybridization of wild-type embryos with pax6 (top left) and myc (top right) probes at stage 15. The same embryos are bisected at the level shown schematically, clearly showing the expression of pax6 and myc in the PLE (bottom, white dotted circles). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)|
|Fig. 7. pax6 mutants display altered forebrain patterning. In situ hybridization comparison of expression patterns of several genes show altered brain patterns in mutant versus wild-type embryos. pax6 mRNA expression is lost from the pretectum (white arrow) and ventral thalamus of mutants, but appears to be maintained normally in more anterior parts of the forebrain including retina at stage 38 (A). By contrast, as seen in panel (B) the expression domain of nkx2-1 appears to be expanded in the telecephalon (white arrow) of pax6 mutants compared to wild-type. gsx2 expression is lost from the prethalamus (C, white arrow) of the mutant, and neurog2 is greatly reduced in the telencephalon (D, white arrow) of mutant embryos.|
|Fig. 8. insulin (ins) expression is downregulated in pax6 mutants. Wild-type (left) and pax6 mutant (right) embryos at stage 42 were subjected to in situ hybridization with an ins probe. Expression of ins gene in this lateral view of mutant embryos (white arrows) is reduced compared to the level seen in wild-type embryos (orange arrows). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)|
|Fig. 9. Combination of truncated Pax6 protein and Pax6 protein with a small in-frame deletion causes an aniridia-like phenotype in X. tropicalis froglets. (A) An example of a heterozygous genotype (left) of phenotypically wild-type froglet (right). (B) An example of the compound heterozygous genotype (left) of an aniridia-like froglet (right). All other genotypes of aniridia-like froglets are listed in Table S1.|
|Fig. 10. Comparison of aniridia-like frog eyes with human aniridic eyes. For purposes of comparison the images of froglet eyes shown here are approximately half the size of the human eyes. Actual froglet corneal diameter is approximately one-tenth that of human. Normal frog (A) and human (E) eyes have small pupil (indicated by white brackets) and relatively large, pigmented iris. Aniridia-like mutant froglets (B–D) and human aniridia patients (F–H) show variable eye phenotypes including reduced (B–D, H) or absent (F and G) iris tissue. Remaining iris tissue may show focal losses (red arrows D and H). The lenses may be clear (B and F) or have a focal cataract (white arrows in C and G). Also, both frog and human eyes may show peripheral clouding of the cornea due to keratopathy (white arrowheads in B, C and F). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)|
|Fig. 11. General histologic features of aniridia-like mutant froglet eyes. Compared with wild-type eyes (A), these mutants show variable defects in cornea, iris, lens and retina. Most notably, the pupil diameter is much wider in the mutants (B and C). Lenses show variable reduction in size and differentiation consistent with the variable presence of cataracts visible upon gross inspection of eyes of intact animals. While the retina is usually present and grossly normal, it may contain folds and the ganglion cell layer (gcl) is somewhat sparse and irregular. The retina may be partially absent as shown in C. Scale bar in panel A=100 µm.|
|Fig. 12. Xenopus mutants for pax6 exhibit anterior eye defects comparable to those observed in Pax6 mutant mice. Histological sections cut through eyes from either wild-type (Normal) or Pax6 mutant (Mutant) X. tropicalis (Xenopus) and mice (Mouse) showing anterior structures. (A) Wild-type froglet eye. White arrowhead denotes lens epithelial cells. (A′) High magnification view of lens epithelial cells in region of lens adjacent to retina. Dotted line denotes boundary between lens and retina. (B) pax6 Hypomorphic froglet eye. In addition to the eye being smaller, defects in the cornea, iris, and lens are evident. Circle indicates region with likely iridolenticular adhesion. (B′) High magnification view of lens epithelial cells in region of lens adjacent to retina comparable to A′. Dotted line denotes boundary between lens and retina. Note expansion of acellular material between lens and retina compared to wild-type. (C) 8-Week-old adult wild-type mouse (D) 8-week-old adult Pax6Sey-Neu/+ mouse. Black arrowheads denote iridokeratotic adhesion. Open arrowhead denotes iris or ciliary tissues. (D′) Arrow shows iridolenticular adhesion in section cut from different Pax6Sey-Neu/+ mouse. (D′) Arrow shows iris hypoplasia typically observed in these Pax6Sey-Neu/+ mice; this section was cut from a third Pax6Sey-Neu/+ mouse at a plane comparable to that of the wild-type mouse in C. cb, ciliary body; nr, neuroretina. Scale bar in panel A applies to panel B; scale bar in C applies to panel D. Scale bars in A′, B′ and D′ denote 50 μm. Scale bar in D′ denotes 200 μm.|
|Fig. 13. Xenopus pax6 mutant exhibits corneal defects similar to those observed in Pax6 mutant mice and human aniridia. Histology of the cornea of wild-type (Normal) and Pax6 mutant (Mutant) frog (Xenopus), mice (Mouse), and human (Human). (A) Wild-type froglet. (B) pax6 Hypomorphic froglet. (C) 8-Week-old adult wild-type mouse (D) 8-Week-old adult Pax6Sey-Neu/+ mouse. (E) 70-Year-old human. The cornea did not show clinical signs of keratopathy. (F–F′) 49-Year-old individual with classical aniridia. The patient was diagnosed with stage III keratopathy and exhibited full conjunctivalization of the cornea. Arrow denotes blood vessel. Asterisk denotes area of cellular infiltration in the superficial corneal stroma. The samples here were of peripheral cornea. Note: the endothelial layer was not captured in these images of human cornea. bc, basal cells; conj, conjunctival cells, sc, superficial cells; wc, wing cells. Scale bar in panel A applies to panels (B–F). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)|