The X-linked splicing regulator MBNL3 has been co-opted to restrict placental growth in eutherians

Abstract

Understanding the regulatory interactions that control gene expression during the development of novel tissues is a key goal of evolutionary developmental biology. Here, we show that Mbnl3 has undergone a striking process of evolutionary specialization in eutherian mammals resulting in the emergence of a novel placental function for the gene. Mbnl3 belongs to a family of RNA-binding proteins whose members regulate multiple aspects of RNA metabolism. We find that, in eutherians, while both Mbnl3 and its paralog Mbnl2 are strongly expressed in placenta, Mbnl3 expression has been lost from nonplacental tissues in association with the evolution of a novel promoter. Moreover, Mbnl3 has undergone accelerated protein sequence evolution leading to changes in its RNA-binding specificities and cellular localization. While Mbnl2 and Mbnl3 share partially redundant roles in regulating alternative splicing, polyadenylation site usage and, in turn, placenta maturation, Mbnl3 has also acquired novel biological functions. Specifically, Mbnl3 knockout (M3KO) alone results in increased placental growth associated with higher Myc expression. Furthermore, Mbnl3 loss increases fetal resource allocation during limiting conditions, suggesting that location of Mbnl3 on the X chromosome has led to its role in limiting placental growth, favoring the maternal side of the parental genetic conflict.

Fig 1. Analysis of splicing factor expression enrichment in trophoblastic versus nontrophoblastic tissues.

(A–C) Scatter plots showing trophoblastic expression level and enrichment for 197 splicing regulators in 3 separate comparisons of trophoblastic versus nontrophoblastic tissues. (D) Whole-mount in situ hybridisation analysis of the expression of splicing regulators found to be enriched in trophoblastic tissues (A–C) in mouse embryos at the indicated stages. (E, F) Box plots showing expression levels of the indicated splicing factors in placenta and nonplacental tissues in mouse (E) and in human (Hs), cow (Bt), and opossum (Md) (F). The numerical data underlying this figure can be found in S1 Data. ESC, embryonic stem cell; ICM, inner cell mass; TE, trophectoderm; TPM, transcript per million; TS, trophoblast stem; XEN, extraembryonic endoderm.

Fig 2. Molecular evolution of Mbnl3 in eutherians.

(A) Schematic showing alternative splicing events and transcriptional (arrow) and translational (ATG) start sites leading to long and short MBNL3 isoform production. Splice junctions corresponding to transcripts that code for the long (b) and short (a, c, d) isoforms are indicated. Ancestral (black/gray), eutherian-specific (red) and mouse-specific (orange) TSSs and the location of the DNA transposon MER53 (blue) are also shown. (B) Diagram showing long and short Mbnl3 isoforms and the location of eutherian-specific aa changes. Zinc finger domains are shown in blue. Sequence alignment is shown in S3 Fig. (C) Quantification of the relative usage of the splice junctions highlighted in (A) across a range of mouse tissues. (D) Transgenic analysis of a putative Mbnl3 promoter region. A approximately 8-kb region surrounding the eutherian-specific TSS indicated by the black bar in (A) produced lacZ staining in the trophoblast of 4/10 injected embryos at E7.5. (E) Phylogenetic tree showing the degree of MBNL3 (red) and MBNL2 (blue) protein sequence divergence in eutherian (circles) and noneutherian (triangles) species. The tree was produced by the neighbor-joining method using Kimura corrected distances. (F) RNAcompete-derived sequence logos for the indicated human (Hs), mouse (Mm), opossum (Op), and chicken (Gg) MBNL proteins. (G) Analysis of the Z-score contribution of sequences containing GCUU or GCAGC motifs to the total cumulative Z-score of the top 100 RNAcompete-derived 7-mers for the indicated MBNL proteins. (H) Bar plot showing the Z-score of the top RNAcompete-derived 7-mer for each of the indicated MBNL proteins. (I) Analysis of the Z-score contribution of sequences containing GCG, GCC, GCA, or GCU motifs to the total cumulative Z-score of the RNAcompete-derived 7-mers containing a single GC dinucleotide among the top 100 7-mers for the indicated MBNL proteins. (J) Ratio of the z-score contribution of 7-mer’s containing a pair of GC dinucleotides with a 1-bp spacer (GCxGC) versus a 2-bp spacer (GCxxGC). * in panels G–J indicate chimeric proteins. The numerical data underlying this figure can be found in S1 Data. aa, amino acid; MBNL, Muscleblind-like; TSS, transcription start site.

Fig 3. Mbnl3 restricts placental growth.

(A, E) Analysis of the effects of M2KO, M3KO, or Mbnl2:Mbnl3 DKO on placental (A) and embryo (E) weight at E13.5 and E18.5. (B) Hematoxylin–eosin staining of WT and KO placenta sections obtained at E18.5. In the zoomed images, note the GlyT and SpT cells within the labyrinth zone of the KO placentas (iv, vi, viii). Also note the nuclei undergoing karyorrhexis (black arrow head) and karyolysis (empty arrow head) in the DKO junctional zone (vii). Scale bar for full placental sections = 1 mm. Scale bar for zoomed images = 100 μm. (C, D) Quantification of the proportion of total junctional zone tissue found within the labyrinth (C) and total area of esophilic necrotic tissue (D) at E18.5. A single medial section was analyzed per placenta. (F) Analysis of the effects on embryo weight of placenta specific M3KO with (M2KOM3^m/+) or without (M3^m/+) universal M2KO. (G) Analysis of the effects on embryo weight of epiblast/fetus specific M2KO with (epiDKO) or without (M2epi) universal M3KO. For (A) and (C–G), the number of placentas/embryos analyzed for each genotype are indicated in brackets, and significance levels are calculated by Wilcoxon rank-sum tests. The numerical data underlying this figure can be found in S1 Data. DKO, double knockout; GlyT, glycogen trophoblast; JZ, junctional zone; LZ, labyrinth zone; M2KO, Mbnl2 knockout; M3KO, Mbnl3 knockout; MBNL, Muscleblind-like; SpT, spongiotrophoblast; WT, wild-type.

Fig 4. Effect of Mbnl2, Mbnl3, and double Mbnl2/3 KO on the placental transcriptome.

(A) Bar charts showing the number of differentially spliced exons in M2KO, M3KO, and DKO placentas at the indicated embryonic stages. Exons with increased inclusion in KOs are tallied above the x-axis, while exons with reduced inclusion are tallied below. (B) Venn diagrams showing the overlap in differential spliced exons between M2KO, M3KO, and DKO placentas at E13.5. (C) Boxplots showing the difference in exon inclusion (PSI) in the indicated KOs versus WT placentas for all exons differentially spliced in DKOs versus WTs at E13.5. (D) Boxplots showing the change in PSI in the indicated KO versus WT placentas at E18.5 for all exons with differential inclusion between E13.5 and E18.5 WT placenta. (E, F) Sashimi plots showing usage of Numb exon 9 (E) and Gpr137 exon 3 (F) in placentas with the indicated genotypes at E13.5. The PSI values for the exons in the different genotypes are shown to the right of the plots. (G) Bar charts showing the number of differentially expressed genes in M2KO, M3KO, and DKO placentas at E13.5. Genes with increased expression in KOs are tallied above the x-axis, while genes with reduced expression are tallied below. (H, I) Bar charts showing significance levels for enrichment of TF binding sites among genes up-regulated in M3KO placentas at E11.5. The top 5 most significantly enriched TFs are shown for ChEA (H) and ENCODE (I) data sets. (J, K) Plots showing Myc expression levels in WT and M3KO placentas at E11.5 based on RNA-seq (J) and qPCR (K). In (J), matched placenta pools derived from the same litters are indicated by dashed lines. For (K), values correspond to the fold change between each tested placenta and the average of WT placenta; N = 13 placentas for each genotype. Significance levels were calculated by Wilcoxon rank-sum tests. The numerical data underlying this figure can be found in S1 Data. DKO, double knockout; KO, knockout; MBNL, Muscleblind-like; M2KO, Mbnl2 knockout; M3KO, Mbnl3 knockout; PSI, percentage spliced in; qPCR, quantitative real-time PCR; RNA-seq, RNA sequencing; TF, transcription factor; WT, wild-type.

Fig 5. M3KO reduces the effect of midgestational maternal calorie restriction on fetal growth.

(A, C) Dot plots showing the weights of individual WT (gray) and M3KO (red, M3KO) placentas (A) and embryos (C) harvested from 6 calorie restricted mothers and the corresponding empirical cumulative distribution plot for each genotype. Significance levels were calculated by a permutation test with 1,000 iterations swapping the genotype labels within each litter. (B, D) Analysis of the effects of maternal calorie restriction on the weight of WT and M3KO placentas (B) and embryos (D) at E18.5. For each embryo/placenta, the weight was normalized with respect to the average of WT and KO weights within each litter. Significance levels are calculated by Wilcoxon rank-sum tests. The numerical data underlying this figure can be found in S1 Data. KO, knockout; M3KO, Mbnl3 knockout; WT, wild-type.