Click here to close
Hello! We notice that you are using Internet Explorer, which is not supported by Xenbase and may cause the site to display incorrectly.
We suggest using a current version of Chrome,
FireFox, or Safari.
???displayArticle.abstract???
Development of the Xenopus pronephros relies on renal precursors grouped at neurula stage into a specific region of dorso-lateralmesoderm called the kidney field. Formation of the kidney field at early neurula stage is dependent on retinoic (RA) signaling acting upstream of renal master transcriptional regulators such as pax8 or lhx1. Although lhx1 might be a direct target of RA-mediated transcriptional activation in the kidney field, how RA controls the emergence of the kidney field remains poorly understood. In order to better understand RA control of renal specification of the kidney field, we have performed a transcriptomic profiling of genes affected by RA disruption in lateralmesoderm explants isolated prior to the emergence of the kidney field and cultured at different time points until early neurula stage. Besides genes directly involved in pronephric development (pax8, lhx1, osr2, mecom), hox (hoxa1, a3, b3, b4, c5 and d1) and the hox co-factor meis3 appear as a prominent group of genes encoding transcription factors (TFs) downstream of RA. Supporting the idea of a role of meis3 in the kidney field, we have observed that meis3 depletion results in a severe inhibition of pax8 expression in the kidney field. Meis3 depletion only marginally affects expression of lhx1 and aldh1a2 suggesting that meis3 principally acts upstream of pax8. Further arguing for a role of meis3 and hox in the control of pax8, expression of a combination of meis3, hoxb4 and pbx1 in animal caps induces pax8 expression, but not that of lhx1. The same combination of TFs is also able to transactivate a previously identified pax8 enhancer, Pax8-CNS1. Mutagenesis of potential PBX-Hox binding motifs present in Pax8-CNS1 further allows to identify two of them that are necessary for transactivation. Finally, we have tested deletions of regulatory sequences in reporter assays with a previously characterized transgene encompassing 36.5 kb of the X. tropicalis pax8 gene that allows expression of a truncated pax8-GFP fusion protein recapitulating endogenous pax8 expression. This transgene includes three conserved pax8 enhancers, Pax8-CNS1, Pax8-CNS2 and Pax8-CNS3. Deletion of Pax8-CNS1 alone does not affect reporter expression, but deletion of a 3.5 kb region encompassing Pax8-CNS1 and Pax8-CNS2 results in a severe inhibition of reporter expression both in the otic placode and kidney field domains.
Fig. 1. Transcriptomic profiling of RA target genes in lateralmesoderm. A. Experimental workflow. LMZ explants are dissected at early gastrula stage from embryos previously injected with cyp26a1 and GFP mRNA (disruption of RA signaling), or GFP mRNA alone (control). Profiling is performed with batches of explants cultured until early gastrula, mid-gastrula or early neurula stages. B. Overview of DE analysis. Mean-difference plots of log-intensity ratios (differences) versus log-intensity averages (means) of Affymetrix probe sets at early gastrula, mid-gastrula or early neurula stages. Black dots correspond to probe sets with fdr >0.05, and thus excluded from DE analysis. Dots highlighted with blue correspond to probe sets whose expression is down-regulated upon RA disruption (RA-positive targets). Dots highlighted with red are probe sets whose expression is up-regulated upon RA disruption (RA-negative targets).
Fig. 2. Targeted meis3 depletion results in the inhibition of pax8 expression in the kidney field. A. RT-qPCR analysis of meis3 expression in LMZ explants in response to RA depletion. Expression of meis3 is shown as relative expression to odc. Meis3 expression appears to be dependent on RA from midgastrula stage onward (NFst11) and at early neurula stage (NFst14). ∗p < 0.05, ∗∗p < 0.01 (paired Student's t-test). Error bars correspond to SEM. B-D. ISH with meis3 antisense probe. Late neurula stage (NFst18). Lateral (B) and dorsal (C) views. D. Embryo bisected according to the dotted line shown in C. Meis3 is expressed in dorsal and lateralmesoderm (arrowheads), including the kidney field (KF). s: somite.
Fig. 3. Targeted meis3 depletion results in the inhibition of pax8 expression in the kidney field. A. Morpholinos are injected in the V2 blastomere at the 8-cell stage to target the kidney field on the left side of the embryo. Injected embryos are cultured until late neurula stage for ISH analysis of pax8 expression. B. Representative injected embryos. Views of dorsal region, left and right sides of the same embryos are shown. White arrowheads indicate kidney field expression and black arrowhead expression in the otic placode. Pax8 expression is strongly inhibited in the kidney field on the left side of embryos injected with Momeis3 morpholino. It is almost absent or fragmented into small patches. A control morpholino (Cmo) has no effect. C, D. Rescue experiments with a mRNA encoding a version of meis3 transcript not targeted by Momeis3. C. Representative examples of meis3-depleted and meis3 mRNA rescued embryos. D. Cumulative results from 3 independent experiments.
Fig. 4. Targeted meis3 depletion has a moderate effect upon aldh1a2 and lhx1 expression in the kidney field. A, B. Representative injected embryos showing aldh1a2 expression (A) or lhx1 expression (B). A. Adh2a1 expression is very dynamic and can vary from one embyo to the other. Yet comparing Momeis3 injected left side with control right side shows that expression domain in lateralmesoderm (arrowheads) is smaller on the injected side. This mild effect is not observed on embryos injected with a control morpholino (Cmo), although Cmo can slightly affect expression. B. Lhx1 expression is moderately inhibited as a result of Momeis3 injection. Lhx1 expression domain appears narrower. This is not observed after injection of Cmo, although Cmo can also slightly affect lhx1 expression.
Fig. 5. Combined expression of hoxb4, meis3 and pbx1 results in pax8 up-regulation in animal cap assays, but not of lhx1. A. Mixtures of mRNA encoding Xenopus hoxb4, meis3 and pbx1 are injected into the four blastomeres of 4-cell stage embryos close to the animal pole. Animal caps are dissected at mid-blastula stage and cultured until siblings reached late neurula stage, and are processed for RT-qPCR analysis. B-E. RT-qPCR analyses. Expression of the analyzed genes is shown as relative expression to odc. Error bars correspond to s.e.m. B. Expression of pax8 is up-regulated in caps expressing hoxb4 combined to meis3 and pbx1. Expression of meis3 and pbx1 alone has little effect on pax8 expression. The same combination of hoxb4, meis3 and pbx1 neither significantly affect lhx1 expression (C), nor RA pathway components aldh1a2 (D) and cyp26a1 (E). ∗p < 0.05 (paired Student's t-test).
Fig. 6. Transactivation of Pax8-CNS1 by Xenopus hoxb4 combined with meis3 and pbx1. A. Alignment of pax8-CNS1 sequences from the human (hg38, chr2: 113,341,953–113,342,169), mouse (mm10, chr2: 24,539,076–24,539,292), and frog (Xenopus tropicalis; xenTro10, chr3: 125,792,358–125,792,572) with three highlighted conserved putative Pbx-Hox binding motives (blue), and one Meis (orange). Nucleotides identical in two or three sequences are shaded in grey. Mutations tested for every site are shown in red. Note that the putative binding motives mapped here show close similarity to the consensus binding sequences of Pbx-Hox (5′-ATGATTNATNN-3′) or Meis (5′-TGACAGST-3′) (Chang et al., 1996, 1997). B. Reporter expression measured for HEK293 cells transfected with pax8CNS1-luc, and various combinations of plasmids allowing expression of Xenopus hoxb4, meis3 or pbx1 indicated in the bottom of the figure. Values are expressed as fold changes relative to HEK293 cells transfected with pax8CNS1-luc alone. A significant increase of reporter expression is only observed with cells expressing a combination of hoxb4, meis3 and pbx1. Other combinations tested do not elicit any significant increase of reporter expression. ∗∗p < 0.01 (Dunn's multiple comparison test after Kruskal-Wallis test). C. Analysis of mutations of single Pbx-Hox and Meis motifs in transactivation assays. HEK293 cells were transfected with the three plasmids allowing expression of hoxb4, meis3 and pbx1 combined with wild-type pax8CNS1-luc, pax8CNS1mut-lucMeis, pax8CNS1mut-lucHP1, pax8CNS1mut-lucHP2 or pax8CNS1mut-lucHP3. Values are expressed as fold changes relative to HEK293 cells transfected with wild-type pax8CNS1-luc. None of the single mutation tested significantly affected transactivation. D. Analysis of multiple mutations of Pbx-Hox motifs in transactivation assays. HEK293 cells were transfected with the three plasmids allowing expression of hoxb4, meis3 and pbx1 combined with wild-type pax8CNS1-luc, pax8CNS1mut-lucHP1+2, pax8CNS1mut-lucHP1+3, pax8CNS1mut-lucHP2+3 or pax8CNS1mut-lucHP1+2 + 3. Values are expressed as fold changes relative to HEK293 cells transfected with wild-type pax8CNS1-luc. Combining mutation of Pbx-Hox 1 and 3 motifs results in a strong decrease of reporter expression, pointing to these two sites as active hox-Pbx binding sites. ∗p < 0.05 (Dunn's multiple comparison test after Kruskal-Wallis test).
Fig. 7. A genomic region encompassing the CNS1 and CNS2 is required for otic placode- and pronephric kidney field-specific expression of pax8. A. GFP reporter constructs used for transgenesis experiments in X. laevis embryos (left) and bar graphs summarizing GFP expression patterns in the resulting embryos (right). The bar graphs show ratio of embryos showing distinct otic placode- and kidney field-specific expression in both sides (scored as the Category 1), no or faint expression in one side (Category 2), and no expression in both sides (Category 3). B. Representative embryos showing GFP expression scored as the Category 1, Category 2, and Category 3; dorsal view with anterior side to the top. Black and white triangles indicate expression in the otic placode and kidney field, respectively.
Fig. 8. Working model based on present data and previously published results (Cartry et al., 2006; Cirio et al., 2011; Colas et al., 2008; Futel et al., 2015; Le Bouffant et al., 2012) illustrating the mechanisms that control pax8 expression in the kidney field. The present data (1) show that meis3 and hox are RA targets, and are involved in the direct regulation of pax8. They are cooperating with other RA-dependent inputs such as lhx1 (Cirio et al., 2011) (3), and a yet unidentified mechanism dependent on RA-dependent intracellular calcium signaling (Futel et al., 2015) (2). The FGF/erk pathway was previously shown to inhibit pax8 expression (Colas et al., 2008) (5) and can be negatively regulated by MKP3 that is encoded by dups6, another RA target (Le Bouffant et al., 2012)(this study) (4). ER: endoplasmic reticulum.
