Free
Cornea  |   February 2019
Transformation of the Transcriptomic Profile of Mouse Periocular Mesenchyme During Formation of the Embryonic Cornea
Author Affiliations & Notes
  • Justin Ma
    BioSciences Department, Rice University, Houston, Texas, United States
  • Peter Lwigale
    BioSciences Department, Rice University, Houston, Texas, United States
  • Correspondence: Peter Lwigale, BioSciences Department, Rice University, 6100 Main Street, Houston, TX 77025, USA; lwigale@rice.edu
Investigative Ophthalmology & Visual Science February 2019, Vol.60, 661-676. doi:https://doi.org/10.1167/iovs.18-26018
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Justin Ma, Peter Lwigale; Transformation of the Transcriptomic Profile of Mouse Periocular Mesenchyme During Formation of the Embryonic Cornea. Invest. Ophthalmol. Vis. Sci. 2019;60(2):661-676. doi: https://doi.org/10.1167/iovs.18-26018.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose: Defects in neural crest development are a major contributing factor in corneal dysgenesis, but little is known about the genetic landscape during corneal development. The purpose of this study was to provide a detailed transcriptome profile and evaluate changes in gene expression during mouse corneal development.

Methods: RNA sequencing was used to uncover the transcriptomic profile of periocular mesenchyme (pNC) isolated at embryonic day (E) 10.5 and corneas isolated at E14.5 and E16.5. The spatiotemporal expression of several differentially expressed genes was validated by in situ hybridization.

Results: Analysis of the whole-transcriptome profile between pNC and embryonic corneas identified 3815 unique differentially expressed genes. Pathway analysis revealed an enrichment of differentially expressed genes involved in signal transduction (retinoic acid, transforming growth factor-β, and Wnt pathways) and transcriptional regulation.

Conclusions: Our analyses, for the first time, identify a large number of differentially expressed genes during progressive stages of mouse corneal development. Our data provide a comprehensive transcriptomic profile of the developing cornea. Combined, these data serve as a valuable resource for the identification of novel regulatory networks crucial for the advancement of studies in congenital defects, stem cell therapy, bioengineering, and adult corneal diseases.

Corneal development is a complex morphogenetic process that involves coordinated development of three distinct cellular layers, namely the epithelium, stroma, and endothelium, into a transparent tissue essential for vision. The formation of these distinct layers is interdependent and also governed by inductive signals from the surrounding ocular tissues that ensure proper cell migration, proliferation, and differentiation.1,2 The epithelium is derived from the ocular surface ectoderm, whereas the stromal keratocytes and endothelium are generated from the periocular mesenchyme that largely consists of a multipotent embryonic cell population, the neural crest.35 Four major events occur during mouse corneal development: (1) migration of periocular neural crest cells (pNC) into the presumptive corneal region, (2) differentiation of pNC into keratocytes and endothelium, (3) synthesis of stromal extracellular matrix (ECM) and formation of tight junctions and active pump function in the endothelium, and (4) maturation of the surface ectoderm into stratified corneal epithelium.3,57 Misregulation of the molecular cues that promote these events results in various forms of anterior segment dysgenesis.810 
Major signaling pathways including retinoic acid (RA), transforming growth factor beta (TGFβ), and Wnt play critical roles during corneal development. RA is secreted by the optic cup and epithelium into the periocular mesenchyme, where it induces Foxc1 and Pitx2.11 This leads to activation of downstream effectors, such as Tfap2B and vascular endothelial growth factor, that are required for regulating cell fate and establishing angiogenic privilege.12,13 Mutations in the RA pathway leads to congenital anterior dysgenesis linked to Axenfeld-Rieger syndrome or Peters anomaly, characterized by corneal opacity and glaucoma.14,15 TGFβ is expressed by the lens epithelium,16 and it is required for pNC migration and differentiation into corneal endothelium.1618 Although it is hypothesized that the maturation of corneal layers is interdependent, the effect of RA and TGFβ on epithelial maturation is not well studied. The Wnt and Notch signaling pathways are localized in the corneal epithelium where they regulate cell proliferation and stratification.19,20 Cross-talk between these signaling pathways regulates the expression of transcription factors, which play critical roles in imparting cellular identity and function,21 but the mechanisms involved are not well understood. 
In this study, we used high-throughput RNA sequencing (RNA-Seq) to establish a transcriptome profile and analyze the changes in gene expression during mouse corneal development. We analyzed the downstream targets of RA, TGFβ, and Wnt signaling pathways and examined their combined effect on genes involved in modulating key processes, including ECM homeostasis, cell junctions, cell cycle, and neural vascular patterning. Our transcriptome data provide the first progressive expression signature that profiles the genetic landscape of the developing cornea. These findings increase our understanding of the fundamental molecular mechanisms that direct corneal development. In addition, this study identifies several novel genes that may play critical roles during corneal development, which may serve as potential targets for stem cell studies, bioengineering, and advancement of new corneal therapies. 
Materials and Methods
Animals
Only wild-type C57/B6 mouse embryos were used for this study. All animal procedures were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the Institutional Animal Care and Use Committee at Rice University. Timed pregnant mice were obtained from Jackson Laboratory, and embryos were collected at embryonic day (E) 10.5, E14.5, and E16.5 for tissue isolation and histology. 
Dissection of Periocular Mesenchyme and Embryonic Corneas
To obtain pNC, anterior eyes were dissected from E10.5 embryos, incubated in dispase (1.5 mg/ml; Worthington Biochemical, Lakewood, NJ, USA) at 37°C for 5 minutes, and then rinsed in Ringer's solution.The ectoderm/lens vesicles and optic cups were removed and discarded, and pNCs from 26 eyes were pooled into each sample. E14.5 corneas were dissected from surrounding ocular mesenchyme and pooled into 18 corneas per sample. Similarly, E16.5 corneas were dissected at the limbal region and pooled into 12 corneas per sample. Biological triplicates of tissues from each time point were immediately immersed in Trizol reagent (Life Technologies Corp., Grand Island, NY, USA) and flash frozen in liquid nitrogen. 
RNA Sequencing
RNA isolated from a total of nine samples was used for library preparation and sequenced on an Illumina HiSeq 4000 instrument at BGI Genomic Services, United States. Samples were qualified and quantified using an Agilent 2100 bioanalyzer and Step One Plus real-time PCR system. Each sample was assessed for quality by filtering out reads with adaptors, reads that contained a high percentage of unknown bases (>10%), or bases with low sequencing quality (Q < 5).22 The following reads were mapped to reference genes by Bowtie 223 and to the Genome Reference Consortium Mouse Build 38 with Hierarchical Indexing for Spliced Alignment of Transcripts24 (Supplementary Table S1). The average mapping with the reference gene was 77.71%, and the genome mapping ratio was 91.47%. Reads were quantified using RNA-Seq by Expectation Maximization25 and normalized to fragments per kilobase of transcript per million (FPKM) to calculate gene expression levels. Aligned genes with no reads at a particular developmental stage were assigned a FPKM value of 0.01 for differential analysis. Screening of differentially expressed genes (DEGs) was performed through the NOISeq method26 by using the criteria of fold change of ≥1 and divergent probability of ≥0.8 (Supplementary Fig. S1). Deeper analysis into specific pathways followed stricter criteria. Based on log base 2 values, a threshold was set at 2.32 (FPKM = 5). Genes with all values below this threshold were considered not expressed. To reduce the uncertainty of low values, negative values were normalized to a base of 0 (FPKM = 1). Heatmaps were generated using log base 2 values with relative row scaling. 
Data Access
All sequencing data have been deposited in the NCBI's Gene Expression Omnibus database (https://www.ncbi.nlm.nih.gov/geo/browse/, in the public domain) under the accession number GSE121044. 
In Situ Hybridization
Section in situ hybridization was performed as previously described.27 In brief, mouse heads were isolated and fixed in Carnoy's fixative at 4°C overnight. Tissues were embedded in paraffin and sectioned at 8 to 10 μm. Digoxigenin-labeled riboprobes were generated by in vitro transcription with Superscript III. Brightfield images were captured using a Zeiss Axiocam mounted on AxioImager2 microscope (Zeiss, Oberkochen, Baden-Württemberg, Germany). 
Results
Characterization of the Transcriptomes of pNC and Embryonic Corneas
To investigate the transcriptomic profile during corneal development, we performed high-throughput RNA-Seq on pNC isolated at E10.5 and embryonic corneas isolated at E14.5 and E16.5 (Fig. 1A). These time points were selected to capture pNC migration into the corneal region (E10.5), differentiation of corneal epithelium and pNC-derived mesenchyme (E14.5), and postformation of the three cellular layers of the cornea (E16.5). 
Figure 1
 
RNA sequencing analysis of the pNC and embryonic corneal cells. (A) Hematoxylin and eosin staining of mouse embryonic eyes, with highlighted regions within the dotted lines showing the tissue dissected for RNA preparation at E10.5, E14.5, and E16.5. (B) Venn diagram depicts total number of genes categorized between the three stages. (C) Bar plot showing number of differentially regulated genes detected between E10.5 and E14.5, E10.5 and E16.5, and E14.5 and E16.5. (D) Pathway enrichment analysis of DEGs at E10.5 vs. E14.5. Circle size correlates with number of genes and the Rich factor is a representation of the degree of enrichment based on ratio of DEG/non-DEG within the pathway. Scale bars: 50 μm (E10.5 and E14.5) or 100 μm (E16.5). ec, ectoderm; L, lens; oc, optic cup; ep, epithelium; st, stroma; en, endothelium; ey, eyelid.
Figure 1
 
RNA sequencing analysis of the pNC and embryonic corneal cells. (A) Hematoxylin and eosin staining of mouse embryonic eyes, with highlighted regions within the dotted lines showing the tissue dissected for RNA preparation at E10.5, E14.5, and E16.5. (B) Venn diagram depicts total number of genes categorized between the three stages. (C) Bar plot showing number of differentially regulated genes detected between E10.5 and E14.5, E10.5 and E16.5, and E14.5 and E16.5. (D) Pathway enrichment analysis of DEGs at E10.5 vs. E14.5. Circle size correlates with number of genes and the Rich factor is a representation of the degree of enrichment based on ratio of DEG/non-DEG within the pathway. Scale bars: 50 μm (E10.5 and E14.5) or 100 μm (E16.5). ec, ectoderm; L, lens; oc, optic cup; ep, epithelium; st, stroma; en, endothelium; ey, eyelid.
RNA-Seq analysis generated an average of 23,029,819 raw reads. Alignment of reads identified transcripts for 19,391 unique genes, of which reads for 17,038 were detected at all 3 developmental stages (Fig. 1B). Categorizing the transcripts using the NOISeq method revealed 3815 unique DEGs. A total of 1479 genes were differentially expressed between E10.5 and E14.5, of which 536 were downregulated and 943 were upregulated (Fig. 1C). Analysis between E10.5 and E16.5 yielded 3617 DEGs, of which 1922 were downregulated and 1696 were upregulated. We also compared E14.5 and E16.5, which showed that 783 genes were differentially expressed, of which 402 were downregulated and 381 were upregulated. Overall, there was a high number of DEGs between E10.5 and E16.5, which substantially decreased between E10.5 and E14.5, and E14.5 and E16.5 (Fig. 1C). This is supported by hierarchical clustering that indicates higher similarity in transcriptome between E14.5 and E16.5 compared to E10.5 and E14.5 or E10.5 and E16.5 (Supplementary Fig. S1). Further analyses show that 506 genes were enriched only at E10.5, 71 at E14.5, and 355 at E16.5. 
To associate the DEGs to functional roles, we analyzed their distribution by using pathway enrichment analysis based on the KEGG database (Fig. 1D). Several key pathways and processes were significantly enriched, including focal adhesions, ECM-receptor interactions, proteoglycans, and cell adhesion molecules. These pathways and cell processes are important in mediating pNC migration, cell proliferation, matrix assembly, and modulating barrier functions. 
Regulation of Neural Crest Cell (NCC) Markers During Corneal Development
To determine whether genes that are important for establishing NCC identity continue to play a role during corneal development, we analyzed the expression of 46 candidate genes involved in NCC specification, delamination, and early migration.28,29 Based on our threshold value of FPKM of 5, we found that out of the 46 genes, 33 (72%) re-expressed in the pNC, 23 (50%) in the E14.5 corneas, and 18 (39%) in the E16.5 corneas (Fig. 2A). Classification of the 46 NCC genes based on differential regulation (Fig. 2B), revealed that 18 (39%) of genes, including Alx1, Alx4, Pax3, Pax7, Zic1, Zic2, Sox9, and Sox10, are enriched in the pNC. Eleven (24%) genes, including Zeb1, Zeb2, Snai2, Lmo4, and Twist1, maintained nondifferential expression. Four (9%) genes (Tfap2A, Tfap2B, Erg, and Cdh6) are upregulated in the cornea, whereas the remaining 13 (28%) genes, including Axud1, Foxd3, Gbx2, and Rxrg, are not expressed (Supplementary Table S2). To validate our data, we analyzed the spatiotemporal expression of Alx1, Alx4, Snai2, and Tfap2B by in situ hybridization. Alx1 is expressed in the pNC at E10.5, but it is not detected in the corneas at E14.5 and E16.5 (Fig. 2C). Alx4 is expressed in the pNC at E10.5 and stroma at E14.5 but absent in the cornea at E16.5 (Fig. 2D). Snai2 is broadly expressed at all time points and shows strong localization to the corneal epithelium and endothelium at E16.5 (Fig. 2E). Tfap2b is initially expressed in a few pNC cells and ocular ectoderm at E10.5, but it is strongly expressed in the corneal stroma and endothelium at E14.5 and E16.5 (Fig. 2F). 
Figure 2
 
Expression of NCC genes during corneal development. (A) Schematic describes the number of expressed genes at each developmental stage based on threshold value. (B) Heatmap shows relative expression of the transcripts in the pNC, E14.5, and E16.5 corneas. Relative color ranges from white to red based on low (L) or high (H) expression. In addition to the criteria described in the methods, values below threshold were normalized to a log base 2 value of 0. Downregulated genes are highlighted in green, not significantly DEGs in blue, upregulated genes in red, and genes below threshold are not shown (see Supplementary Table S2). (CF) Validation of the expression patterns of Alx1, Alx4, Snai2, and Tfap2b. Black arrows represent regions of enriched expression. Scale bar: 50 μm. co, Cornea; *C-myc expression at E16.5 is excluded.
Figure 2
 
Expression of NCC genes during corneal development. (A) Schematic describes the number of expressed genes at each developmental stage based on threshold value. (B) Heatmap shows relative expression of the transcripts in the pNC, E14.5, and E16.5 corneas. Relative color ranges from white to red based on low (L) or high (H) expression. In addition to the criteria described in the methods, values below threshold were normalized to a log base 2 value of 0. Downregulated genes are highlighted in green, not significantly DEGs in blue, upregulated genes in red, and genes below threshold are not shown (see Supplementary Table S2). (CF) Validation of the expression patterns of Alx1, Alx4, Snai2, and Tfap2b. Black arrows represent regions of enriched expression. Scale bar: 50 μm. co, Cornea; *C-myc expression at E16.5 is excluded.
Regulation of RA Signaling During Corneal Development
We investigated changes to the RA signaling components and found that genes important for metabolism and signaling are differentially regulated (Fig. 3A; Supplementary Table S3).3032 Prometabolic genes, such as Stra6, Raldh1, and Raldh2,33 are not significantly changed between E10.5 and E14.5, but they are downregulated at E16.5. In contrast, Adh1 and Adh7 are upregulated at E16.5. Dhrs3, a metabolic inhibitor that converts retinal back into retinol,33 is upregulated at E14.5. Raldh3 is constitutively expressed at high levels, but its expression is localized to the corneal epithelium.34 The RA-degrading enzyme Cyp26a135 is upregulated at E14.5. Crabp2, which translocates RA from the cytoplasm into the nucleus,30 is downregulated, whereas Crabp1 and Fabp5 are downregulated at E14.5 but upregulated at E16.5. A majority of the nuclear receptors, including Rara, Rarg, Rxra, Rxrb, Nr1h2, and Ppard, are constitutively expressed, but Rarb, Nr2f1, and Nr2f2 are downregulated (Fig. 3B; Supplementary Table S3). Corresponding with these changes, several RA-responsive transcription factors (Sall2, Arnt2, Hes6, and Pitx2)3639 are downregulated at E16.5 (Fig. 3B). RA-induced genes (Egr1 and Btdbd11)40,41 are also substantially decreased at E16.5 (Fig. 3B; Supplementary Table S3). To identify the corneal regions in which RA signaling is regulated, we examined the expression profiles of an RA inhibitor, Cyp26a1, a nuclear receptor, Nr2f2, and a downstream gene, Egr1 (Figs. 3C–E). Our data show that Cyp26a1 is broadly expressed at all time points, with strong localization in the corneal epithelium at E14.5 and E16.5 (Fig. 3C). Nr2f2 is strongly expressed in the pNC at E10.5 and maintained at low levels in the stroma, but it is localized in the corneal epithelium at E14.5 and E16.5 (Fig. 3D). Egr1 is not detectable in the pNC and cornea at E16.5, but it is transiently expressed in the presumptive corneal endothelium at E14.5 (Fig. 3D). 
Figure 3
 
Differential regulation of the RA signaling pathway. (A) Schematic depicts whether components of the RA pathway are upregulated (red), downregulated (green), or not significantly differentially expressed (black). Genes that were upregulated and then downregulated, or vice versa, are represented by blue and orange, respectively. (B) Heatmap summarizes the relative expression of the DEGs. (CE) Validation of the expression patterns of Nr2f2, Egr1, and Cyp26a1. Black arrows represent regions of enriched expression. Scale bar: 50 μm.
Figure 3
 
Differential regulation of the RA signaling pathway. (A) Schematic depicts whether components of the RA pathway are upregulated (red), downregulated (green), or not significantly differentially expressed (black). Genes that were upregulated and then downregulated, or vice versa, are represented by blue and orange, respectively. (B) Heatmap summarizes the relative expression of the DEGs. (CE) Validation of the expression patterns of Nr2f2, Egr1, and Cyp26a1. Black arrows represent regions of enriched expression. Scale bar: 50 μm.
Regulation of TGFβ Signaling During Corneal Development
To examine the mechanisms by which TGFβ signaling regulates corneal development, we investigated the transcription profile of its ligands and downstream genes (Fig. 4A; Supplementary Table S4).42,43 Our data show that TGFβ2 is strongly expressed at E10.5 and E14.5 but downregulated at E16.5, and TGFβ3 is upregulated at E14.5 and E16.5. Interestingly, TGFβR2 is upregulated at E14.5 and E16.5, but its associated receptor TGFβR142 is downregulated. In addition, multiple inhibitors (Bambi, Strap, Smad7, Tgif, and Evi1) and an activator (Msg1) of TGFβ signaling through Smad2/3 regulation are downregulated. Overall, a large number of DEGs favors enrichment of the TGFβ pathway. Accordingly, genes repressed by the TGFβ pathway (Cdk2, Cdk4, C-myc, Id2, and Id3) are downregulated, and TGFβ-induced genes (Rbl2, Aebp1, and Creb3l1) are upregulated. The observed differential regulation aligns with TGFβ function in cell cycle regulation, differentiation, and ECM synthesis.4449 We also observed that several TGFβ-induced epithelial-mesenchymal transition genes, including Hey1 and Prrx2,50,51 were downregulated, possibly due to regulation through other pathways. 
Figure 4
 
Differential regulation of the TGFβ signaling pathway. (A) Schematic depicts whether components of the TGFβ pathway are upregulated (red), downregulated (green), or not significantly differentially expressed (black). (B) Heatmap summarizes the relative expression of the DEGs. (CE) Validation of the expression patterns of Hmga2, C-ski, and Creb3l1. Black arrows represent regions of enriched expression. Scale bar: 50 μm.
Figure 4
 
Differential regulation of the TGFβ signaling pathway. (A) Schematic depicts whether components of the TGFβ pathway are upregulated (red), downregulated (green), or not significantly differentially expressed (black). (B) Heatmap summarizes the relative expression of the DEGs. (CE) Validation of the expression patterns of Hmga2, C-ski, and Creb3l1. Black arrows represent regions of enriched expression. Scale bar: 50 μm.
To analyze how TGFβ regulates corneal development, we examined the expression of downstream targets Hmga2,52 nuclear repressor c-Ski,53 and mediator of collagen synthesis Creb3l148 (Figs. 4C–E). The observed expression patterns are consistent with our dataset and show that Hmga2 is initially ubiquitously expressed at E10.5 and E14.5, but it localizes to the corneal epithelium at E16.5 (Fig. 4C). c-Ski is constitutively expressed in the pNC and cornea (Fig. 4D). Creb3l1 is not expressed at E10.5, but it was strongly expressed in the corneal stroma and endothelium at E14.5 and E16.5 (Fig. 4E). In addition to the changes observed in the canonical TGFβ signaling, we discovered differential regulation of other members of the TGFβ superfamily. Bmp4, Bmp5, Acvr2b, and Gdf11 are all downregulated, whereas Bmp1, Bmp3, Avcr2a, and Gdf10 are upregulated (Fig. 4B). 
Regulation of the Wnt Signaling During Corneal Development
Next, we investigated the mechanisms by which the Wnt pathway is modulated during corneal development (Fig. 5A).54,55 Our data reveal that several Wnt genes (Wnt2, Wnt2b, Wnt3, Wnt3a, Wnt4, Wnt5a, Wnt6, Wnt7b, Wnt9b, Wnt10a, Wnt10b, Wnt11, and Wnt16) are upregulated (Fig. 5B). However, Frizzled receptors are either upregulated (Fzd6 and Fzd10) or downregulated at E14.5 (Fzd3 and Fzd4) and E16.5 (Fzd1 and Fzd2) (Fig. 5B; Supplementary Table S5). This is consistent with previous reports56 and further identifies the novel expression of Wnt ligands and receptors. We found that many inducers of canonical activity, including Prrx2, HMG family, Bambi, Strap, Sox11, Frat2, Pclaf, and Ezh2 are downregulated, whereas the repressors Wif1, Dkk1, Dkk2, Dkk3, Notum, Ndrg1, Nfat5, and Sox6 are upregulated (Figs. 5A, 5B). Spatiotemporal analysis confirmed the expression of candidate Wnt modulators. The Wnt activator Mta157 is localized in the periocular mesenchyme and all cellular layers of the cornea (Fig. 5C), whereas the Wnt activator Sox1158 is initially strongly expressed in the pNC at E10.5, but it is not detectable at E14.5 and E16.5 (Fig. 5D). The upregulated Wnt inhibitor Ndrg159,60 is not detectable in the pNC at E10.5, but it is later localized to the corneal epithelium at E16.5 (Fig. 5E). 
Figure 5
 
Differential regulation of the Wnt signaling pathway. (A) Schematic depicts whether components of the Wnt pathway are upregulated (red), downregulated (green), or not significantly differentially expressed (black). (B) Heatmap summarizes the relative expression of the DEGs. (CE) Validation of the expression patterns of Mta1, Sox11, and Ndrg1. Black arrows represent regions of enriched expression. Scale bar: 50 μm. AA, ambiguously associated.
Figure 5
 
Differential regulation of the Wnt signaling pathway. (A) Schematic depicts whether components of the Wnt pathway are upregulated (red), downregulated (green), or not significantly differentially expressed (black). (B) Heatmap summarizes the relative expression of the DEGs. (CE) Validation of the expression patterns of Mta1, Sox11, and Ndrg1. Black arrows represent regions of enriched expression. Scale bar: 50 μm. AA, ambiguously associated.
Next, we analyzed how Wnt downstream genes are modulated. Our data revealed that several downstream targets, particularly those related to proliferation (C-myc, N-myc, Ccnd1, and Birc5) were downregulated (Figs. 5A, 5B). However, we also identified upregulation of a few genes that are activated by the canonical pathway (Wisp1 and Irx3). Our data also indicate that genes involved in the Wnt/planar cell polarity (PCP) and Wnt/Ca2+ pathways were upregulated (Gpc4, Pk3, Plcb4, Camk2, and Prkcb). 
Crosstalk Between Signaling Pathways Is Critical for Corneal Development
To examine how the cross talk between RA, TGFβ, and Wnt signaling pathways regulates corneal development, we analyzed the differential expression of their downstream transcription factors. Out of 1755 transcription factors, we found a total of 1118 genes expressed above the threshold. Of these genes, 143 were upregulated, 218 were downregulated, and 757 were not differentially expressed (Fig. 6A). Next, we annotated their association with the signaling pathways based on published data. From those upregulated genes, 62 are associated with RA signaling, 65 with TGFβ signaling, 74 with Wnt signaling, and the data are insufficient for 40. From those downregulated genes, 69 are associated with RA signaling, 92 with TGFβ signaling, 106 with Wnt signaling, and the data are insufficient for 82. The top 20 upregulated and downregulated transcription factors are summarized in Table 1 and are in a full list in Supplementary Table S6. Several transcription factors, such as Pax6 or Foxc2, are involved in more than one pathway, indicating potential cross talk during corneal development. 
Figure 6
 
Categorization of differentially expressed transcription factors into the RA, TGFb, and Wnt signaling pathways. (A) Allocation of transcription factors from a total pool of 1755 (compiled from Riken's mouse database and self-annotated). (B, C) Venn diagrams showing overlap between differentially expressed transcription factors. Values outside the circles represent the genes that are not well characterized or not studied within the relevant pathways.
Figure 6
 
Categorization of differentially expressed transcription factors into the RA, TGFb, and Wnt signaling pathways. (A) Allocation of transcription factors from a total pool of 1755 (compiled from Riken's mouse database and self-annotated). (B, C) Venn diagrams showing overlap between differentially expressed transcription factors. Values outside the circles represent the genes that are not well characterized or not studied within the relevant pathways.
Table 1
 
Top Differentially Expressed Transcription Factors
Table 1
 
Top Differentially Expressed Transcription Factors
The net regulatory effect of the above transcription factors determines corneal morphogenesis. This includes formation of the collagen ultrastructure, proliferation and differentiation of the cellular layers, and neurovascular patterning.5,6,61 Therefore, we analyzed the expression of critical components of corneal development, including genes for the ECM, matrix remodeling proteins, ECM receptors, cell junction proteins, epithelial development, cell cycle, and neurovascular patterning. The top DEGs are reported in Tables 2 to 5 and full lists are in Supplementary Tables S7 to S14. Our data indicate that the majority of the ECM and matrix remodeling proteins, including collagens, laminins, and thrombospondins, are upregulated and expressed at high levels (Table 2, Supplementary Tables S7S9), indicating that a large number of components contribute to establishing the ultrastructure. We also observed downregulated genes, such as Vtn, Emilin2, and Nid2, that may play critical roles during early corneal development. In addition, several extracellular matrix receptors are upregulated (Itga11, Itga3, ItgaV, Itgb4, Dag1, Ddr1, and Cd44) or downregulated (Itga4, Itga8, and Itga9) (Table 2). Although not differentially expressed, transcripts for Itga5, Itga6, Itgb1, and Itgb5 are detected at high levels (Supplementary Table S10), and they may form heterodimers with differentially expressed integrins.62 Expression of cell junction genes, such as Gja1, Tjp1, and Ocln, are similarly enriched (Table 2, Supplementary Table S11). This is accompanied by the expression of genes involved in differentiation of the corneal epithelium, including Pax6, Klf4, and Klf5,6366 as well as epithelial structural genes, such as Krt5, Krt12, Krt14, and Krt1567 (Table 3, Supplementary Table S12). 
Table 2
 
Top Differentially Expressed ECM and Junction-Associated Genes
Table 2
 
Top Differentially Expressed ECM and Junction-Associated Genes
Table 3
 
Top Differentially Expressed Epithelial-Associated Genes
Table 3
 
Top Differentially Expressed Epithelial-Associated Genes
Table 4
 
Top Differentially Expressed Cell Cycle-Associated Genes
Table 4
 
Top Differentially Expressed Cell Cycle-Associated Genes
Table 5
 
Top DEGs Associated With Angiogenesis and Axon Guidance
Table 5
 
Top DEGs Associated With Angiogenesis and Axon Guidance
Our data also show a high number of cell cycle genes are downregulated and cell cycle inhibitors are upregulated (Table 4, Supplementary Table S13), suggesting an overall reduction in cell proliferation. We also observed that genes involved in angiogenesis and axon guidance were differentially regulated (Table 5, Supplementary Table S14) and have potential roles in establishing the neurovascular patterns that lead to high innervation and corneal avascularity. 
We validated the spatiotemporal expression of several genes identified in our data. Fbln2, which encodes an ECM glycoprotein,68 is expressed at low levels in the pNC at E10.5, strongly expressed in the corneal mesenchyme at E14.5, and sparsely expressed in the stroma and endothelium at E16.5 (Fig. 7A). Serpinh1, which is involved in collagen biosynthesis,69 is expressed in the pNC at E10.5 and maintained in the corneal mesenchyme at E14.5 and in the stroma and endothelium at E16.5 (Fig. 7B). Cell junction protein Emp170 shows broad expression at all time points but is enriched in the epithelium at E16.5 (Fig. 7C). Antiangiogenic protein Pedf71 is also broadly expressed at all time points but shows strong localization to the posterior stroma and endothelium at E16.5 (Fig. 7D). 
Figure 7
 
Spatiotemporal expression of genes involved in corneal morphogenesis. (A) Fbln2 is expressed at low levels at E10.5 and E16.5 and is enriched throughout the stroma at E16.5. (B) Serpinh1 is expressed in the pNC, stroma, and epithelium. At E10.5 it is enriched in the temporal mesenchyme, at E14.5 enriched in the anterior stroma, and E16.5 enriched in the endothelium. (C) Emp1 is expressed in all stages and layers, and enriched in the epithelium at E16.5. (D) Pedf is expressed in all stages and layers and enriched in the posterior cornea at E16.5. Black arrows represent regions of enriched expression. Scale bar: 50 μm.
Figure 7
 
Spatiotemporal expression of genes involved in corneal morphogenesis. (A) Fbln2 is expressed at low levels at E10.5 and E16.5 and is enriched throughout the stroma at E16.5. (B) Serpinh1 is expressed in the pNC, stroma, and epithelium. At E10.5 it is enriched in the temporal mesenchyme, at E14.5 enriched in the anterior stroma, and E16.5 enriched in the endothelium. (C) Emp1 is expressed in all stages and layers, and enriched in the epithelium at E16.5. (D) Pedf is expressed in all stages and layers and enriched in the posterior cornea at E16.5. Black arrows represent regions of enriched expression. Scale bar: 50 μm.
Discussion
Corneal development occurs during a critical period when the adjacent presumptive lens and retinal tissues undergo morphogenic changes and gene expression.5,72 These changes in the ocular environment play a crucial role in directing differentiation of both the NCC- and ectoderm-derived corneal progenitors.34,72 In this study, we provide the first detailed analysis of the transcriptome profiles of corneal cells during development. We have identified genes that are enriched at E10.5, E14.5, and E16.5, which may respectively be involved in pNC migration and proliferation, differentiation of the corneal layers, and organization of the ECM and cell-cell junctions. We link these data to genes involved in key signaling pathways and transcriptional regulation of cell behavior. 
NCC contribution to the corneal endothelium and stromal keratocytes comprises the largest proportion of the cornea.5 Due to their dynamic and multipotential characteristics, NCCs are primed to respond to new signals from surrounding environments during their migration from the neural tube and aggregation into the periocular region.73 A majority of the candidate NCC genes are expressed in the periocular mesenchyme at E10.5, which could be important for maintenance of multipotency, which is required for subsequent differentiation into various ocular tissues, including the cornea, iris, and the orbital bones and cartilage.74,75 Consistent with this observation, our data revealed progressive downregulation of the NCC genes, such as Sox9 and Sox10, which are involved in chondrogenesis and neural differentiation, respectively.76,77 The NCC genes that were expressed in the cornea, such as Snai2 and Twist1, may either maintain their roles or take on different functions during differentiation. Twist1 is involved in craniofacial development and is an inhibitor of Sox9 and Sox10,7880 suggesting a potential role in inhibiting these genes in the cornea.81 Snai2 is sustained in the adult corneal epithelium during wound healing by TGFβ and plays a role in epithelial-mesenchymal transition,82 cell proliferation, migration, and differentiation,83 but its function in the corneal endothelium and stroma remain unclear. 
RA signaling is a major factor during organogenesis of various tissues, including the central nervous system, ear, gut, heart, and the eye.84 RA signals in the periocular mesenchyme and presumptive cornea are either autocrine or derived from the ectoderm, optic cup, or lens.72,85 Our data indicate that both pNC and embryonic corneas have the potential for retinol uptake and RA metabolism, but these processes are strictly regulated. We observed that Raldh3, which is expressed in the corneal epithelium,34 may be the major source of RA synthesis at E16.5. All cellular RA binding proteins were significantly downregulated at E14.5. This, coupled with elevated expression of Cyp26a1, suggests a decrease in RA-mediated signaling. Differential expression of modulators of RA signaling is crucial for proper development of various tissues and organs.35 Cyp26a1 mutant mice exhibit patterning defects in limbs and the central nervous system due to an elevation of RA signaling.86 Strong expression of Cyp26a1 in the corneal epithelium suggests its involvement in moderating the RA signaling to levels that permit cell differentiation. Upregulation of RA metabolizing enzymes Adh1 and Adh7, along with Crabp1 and Fabp5, may represent increased signaling through alternative pathways. Crabp1-RA interaction activates Erk1/2, which triggers a signaling cascade that regulates cell cycle and promotes differentiation.87,88 
TGFβ signaling has been implicated in driving cell migration and differentiation and formation of the collagen ultrastructure during corneal development.17,89 Our data show elevated TGFβ2 transcripts concomitant with the formation of the corneal endothelium,5 followed by its rapid downregulation. Combined with the previous observation that the corneal endothelium is absent in TGFβ2 knockout mice,18 our data suggest that high levels of TGFβ2 are required for its formation. We also observed upregulation of TGFβ3, which stimulates matrix assembly in vitro.90,91 Upregulation of TGFβR2 is in line with its function as the primary facilitator of TGFβ signaling. TGFβR2 mutants recapitulate TGFβ2 knockout mice phenotypes.89 In addition, they are unable to phosphorylate Smad2, misexpress Foxc1 and Pitx2, and display abnormal keratocyte differentiation and collagen synthesis.89 Canonically, TGFβ interacts with TGFβR2 to recruit and phosphorylate TGFβR1, which activates Smad2/3 signaling.42,92 Although the downregulation of the interacting partner TGFβR1 was unexpected, TGFβR2 can also form a complex with TGFβR3, which has higher specificity for TGFβ2.93,94 Combined with the downregulation of Smad2/3 inhibitors, this indicates an increased activity of TGFβ signaling. Along with the induction of lumican and keratocan,18 TGFβ signaling may mark the transition from highly proliferative pNC toward induced differentiation. Our data also show upregulation of Aebp1 and Creb3l1, which are important for collagen synthesis.47,48 These genes may cooperate with other sources of collagen synthesis and maturation, such as Bmp3 and Bmp1.95,96 
The Wnt/β-catenin pathway is required for the proper development of the cornea.19,97,98 During mouse corneal development, Wnt ligands are expressed throughout the presumptive epithelium.56 Increased expression of Wntless and Porc indicate that Wnt signaling may also exert paracrine effects to the stroma. This is supported by reports of expression of Fzd receptors and activation of Wnt signaling in the stromal mesenchyme and corneal endothelium.56,99 Although Wnt ligands were uniformly upregulated, there was a clear distinction in the differential expression of Fzd. Fzd4 and Fzd10 are associated with the Wnt/β-catenin pathway, whereas Fzd3 and Fzd6 are involved in the Wnt/PCP pathway.100103 Fzd4 is required for retinal angiogenesis and implicated in corneal neovascularization.104,105 Fzd3 is involved in neural crest induction and migration.106108 Reduced expression of Fzd4 and Fzd3 and upregulation of Fzd10 and Fzd6 may be required for corneal cell differentiation and avascularity.100103,109111 Despite upregulation of Wnt ligands and receptors, our data suggest that Wnt/β-catenin signaling is inhibited at multiple levels. This complements previous observations that active Wnt/β-catenin signaling is absent in the corneal epithelium at E14.5 and E16.5, and it is progressively reduced in the stroma until postnatal day 3.99 This downregulation is critical for proper development of the cornea.19,97,98 In contrast, our data suggest that noncanonical Wnt pathways are upregulated. The Wnt/PCP and Wnt/Ca pathways have been studied during the formation of the eye field and retinogenesis,112 but their roles in the cornea are not clear. Our data indicate an increase in the components of the Wnt/PCP pathway, including Wnt4, Wnt5a, and Fzd6.100,113,114 In adults, the Wnt/PCP pathway is important for corneal homeostasis and also guides directional migration of epithelial cells during wound healing.110 Wnt/PCP signaling is also involved in cell differentiation, collagen orientation, cell alignment, and axon guidance,115117 all of which are required for proper corneal development. 
Our data suggest multiple novel connections between the RA, TGFβ, and Wnt signaling pathways. It is well established that the RA induction of Pitx2 suppresses Wnt signaling through upregulation of Dkk2,97,118 and we also observe this pattern. In addition, misregulation of either Wnt or TGFβ greatly impacts Pitx2 levels, suggesting that the different pathways interact for proper signaling control.89,97 Potential crosstalk is observed in the upregulation of genes associated with RA signaling (Sox6 and Hic1), which suppress Wnt signaling.119123 The Wnt activating genes (Prrx2 and Hmga2) are upregulated by TGFβ,51,52 and we observed that Hmga2 localizes to the corneal epithelium where Wnt expression is dominant.56 Strap activates Wnt but represses TGFβ,124,125 and its downregulation may play an important role in balancing these pathways. Our data indicate that the reduction in proliferation occurs at E14.5 and progresses during corneal development. It is likely the RA and TGFβ pathways modulate the cell proliferation promoted by Wnt signaling, which may occur through regulation of Lin28, C-myc, Id2, and Id3.45,97,126129 Proper regulation of Wnt signaling is crucial, as gain of function in epithelial β-catenin and DKK mutants show increased proliferation, impaired differentiation, and reduced ECM in the epithelium and stroma.97,98,130,131 This arrangement may change in the postnatal cornea as the epithelium undergoes stratification.7 Verification of these associations during corneal development will require additional studies. 
The expression of ECM proteins is abundant and critical for the coordinated fibrillogenesis of the cornea. The absence of either collagens or regulatory proteoglycans causes dysfunctional fibrillogenesis and corneal opacity.10,132 Our results confirm a high expression of transcripts and upregulation of well-known corneal ECM proteins, including decorin, lumican, keratocan, and collagen I. Interestingly, several of the downregulated genes (Vtn, Vcan, Has2, and Tgfbi) are involved in neural crest induction and migration.133137 Several matricellular genes are upregulated at E14.5 and downregulated at E16.5 (Fbln2, Spp1, and Ecm1), suggesting that they are required for cell migration, differentiation, or the initial organization of the corneal ECM. Upregulation of matrix remodeling genes from the cathepsin, matrix metalloproteinase, and a disintegrin and metalloproteinase families may be required for cellular positioning and collagen alignment, which are crucial for establishing a lattice structure and transparency.138140 Genes that regulate cell junctions follow a similar trajectory and may be important for intercellular communications and establishing the epithelial and endothelial barrier.141 This coincides with genes that regulate epithelial differentiation and elevation of epithelial markers, suggesting that the maturation and function of the epithelium develop simultaneously.142,143 Interestingly, our data indicate that epithelial genes, such as Emp1, Gsto1, Gsta4, and Glut1, are also expressed in the pNC. This could indicate the epithelial origin of the pNC or a functional role in the mesenchyme. Some epithelial genes, such as Slurp1 and Psca, are not expressed at E14.5 and E16.5, indicating that they are required at later stages of corneal development and are involved in maintenance of the epithelial layer.144146 
Cross talk between the ECM and resident cells is mediated through cellular receptors, mostly composed of integrins. Itga3b1 and Itga11b1 are required for collagen deposition and matrix assembly.147,148 ItgaVb1 and ItgaVb5 may affect neural guidance149 or interact with latent TGFβ, which may affect ECM assembly.150,151 Itga4b1, Itga8b1, and Itga9b1 are receptors for fibronectin, and they mediate cell adhesion and migration.152155 Their downregulation implies reduced motility. Upregulation of Ddr1, which is regulated by collagen, may provide a feedback mechanism to retain high activity of matrix remodeling genes.156,157 Our data show that both Agrin and its receptor Dag1 are upregulated in the cornea.158 Misregulation of either Agrin or Dag1 causes similar corneal defects, suggesting they may interact during corneal development.158161 
Angiogenesis and neurogenesis are two closely related processes that require intricate orchestration of signals to generate a highly innervated yet avascular cornea. Our previous studies indicated that these two processes are separated during early corneal development.162,163 Our data reveal that multiple factors common to neurovascular patterning were highly expressed or upregulated concomitantly with antiangiogenic factors. This includes the class 3 semaphorins (Sema3A, Sema3C, and Sema3F) that we studied.162166 We also observed upregulation of an extremely potent antiangiogenic factor, Pedf, and its receptor Plxdc2.167 Pedf protects against neovascularization in disease and wound healing in the retina and cornea,71,168170 and it is likely to play a similar role during corneal development. 
Conclusions
Here, we report the first transcriptome analysis of the early development of the mouse cornea. Our data identify a large number of differentially regulated genes during corneal development. We describe the genetic landscape of corneal morphogenesis and provide novel insights of how cross talk between the RA, Wnt, and TGFβ pathways regulates transcription factors involved in cell migration, proliferation, and differentiation. This data will serve as a valuable resource for identifying novel genes essential for corneal development and potential targets for corneal therapies. 
Acknowledgments
The authors thank the members of the Lwigale laboratory for the critical reading of and helpful comments on the manuscript. 
Supported by the National Institute of Health (NIH) grant EY027048. The authors alone are responsible for the content and writing of this paper. 
Disclosure: J. Ma, None; P. Lwigale, None 
References
Collomb E, Yang Y, Foriel S, Ebastien Cadau S, Pearton DJ, Dhouailly D. The corneal epithelium and lens develop independently from a common pool of precursors. Dev Dyn. 2013; 242: 401–413.
Sevel D, Isaacs R. A re-evaluation of corneal development. Trans Am Ophthalmol Soc. 1988; 86: 178–207.
Hay ED, Revel JP. Fine structure of the developing avian cornea. Monogr Dev Biol. 1969; 1: 1–144.
Lwigale PY, Cressy PA, Bronner-Fraser M. Corneal keratocytes retain neural crest progenitor cell properties. Dev Biol. 2005; 288: 284–293.
Pei YF, Rhodin JA. The prenatal development of the mouse eye. Anat Rec. 1970; 168: 105–125.
Haustein J. On the ultrastructure of the developing and adult mouse corneal stroma. Anat Embryol. 1983; 168: 291–305.
Zieske JD. Corneal development associated with eyelid opening. Int J Dev Biol. 2004; 48: 903–911.
Evans AL, Gage PJ. Expression of the homeobox gene Pitx2 in neural crest is required for optic stalk and ocular anterior segment development. Hum Mol Genet. 2005; 14: 3347–3359.
Li J, Qin Y, Zhao F-K, et al. Anterior segment dysgenesis correlation with epithelial-mesenchymal transition in Smad4 knockout mice. Int J Ophthalmol. 2016; 9: 943–947.
Sun M, Chen S, Adams SM, et al. Collagen V is a dominant regulator of collagen fibrillogenesis: dysfunctional regulation of structure and function in a corneal-stroma-specific Col5a1-null mouse model. J Cell Sci. 2011; 124: 4096–4105.
Seo S, Chen L, Liu W, et al. Foxc1 and Foxc2 in the neural crest are required for ocular anterior segment development. Invest Ophthalmol Vis Sci. 2017; 58: 1368–1377.
Chen L, Martino V, Dombkowski A, Williams T, West-Mays J, Gage PJ. AP-2β is a downstream effector of PITX2 required to specify endothelium and establish angiogenic privilege during corneal development. Invest Opthalmol Vis Sci. 2016; 57: 1072–1081.
Seo S, Singh HP, Lacal PM, et al. Forkhead box transcription factor FoxC1 preserves corneal transparency by regulating vascular growth. Proc Natl Acad Sci U S A. 2012; 109: 2015–2020.
Tümer Z, Bach-Holm D. Axenfeld-Rieger syndrome and spectrum of PITX2 and FOXC1 mutations. Eur J Hum Genet. 2009; 17: 1527–1539.
Carmona S, da Luz Freitas M, Froufe H, et al. Novel de novo FOXC1 nonsense mutation in an Axenfeld-Rieger syndrome patient. Am J Med Genet A. 2017; 173: 1607–1610.
Banh A, Deschamps PA, Gauldie J, Overbeek PA, Sivak JG, West-Mays JA. Lens-specific expression of TGF-beta induces anterior subcapsular cataract formation in the absence of Smad3. Invest Ophthalmol Vis Sci. 2006; 47: 3450–3460.
Flügel-Koch C, Ohlmann A, Piatigorsky J, Tamm ER. Disruption of anterior segment development by TGF-β1 overexpression in the eyes of transgenic mice. Dev Dyn. 2002; 225: 111–125.
Saika S, Saika S, Liu C-Y, et al. TGFβ2 in corneal morphogenesis during mouse embryonic development. Dev Biol. 2001; 240: 419–432.
Zhang Y, Yeh L-K, Zhang S, et al. Wnt/β-catenin signaling modulates corneal epithelium stratification via inhibition of Bmp4 during mouse development. Development. 2015; 142: 3383–3393.
Djalilian AR, Namavari A, Ito A, et al. Down-regulation of Notch signaling during corneal epithelial proliferation. Mol Vis. 2008; 14: 1041–1049.
D'Alessio AC, Fan ZP, Wert KJ, et al. A systematic approach to identify candidate transcription factors that control cell identity. Stem Cell Reports. 2015; 5: 763–775.
Illumina, Inc. Quality Scores for Next-Generation Sequencing; 2011. Available at: https://www.illumina.com/documents/products/technotes/technote_Q-Scores.pdf.
Langmead B, Trapnell C, Pop M, Salzberg SL. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol. 2009; 10: R25.
Kim D, Langmead B, Salzberg SL. HISAT: a fast spliced aligner with low memory requirements. Nat Methods. 2015; 12: 357–360.
Li B, Dewey CN. RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinformatics. 2011; 12: 323.
Tarazona S, Garcia-Alcalde F, Dopazo J, Ferrer A, Conesa A. Differential expression in RNA-seq: a matter of depth. Genome Res. 2011; 21: 2213–2223.
Etchevers HC, Vincent C, Le Douarin NM, Couly GF. Thecephalic neural crest provides pericytes and smooth muscle cells to allblood vessels of the face and forebrain. Development. 2001; 128: 1059–1068.
Simoes-Costa M, Tan-Cabugao J, Antoshechkin I, Sauka-Spengler T, Bronner ME. Transcriptome analysis reveals novel players in the cranial neural crest gene regulatory network. Genome Res. 2014; 24: 281–290.
Simoes-Costa M, Bronner ME. Establishing neural crest identity: a gene regulatory recipe. Development. 2015; 142: 242–257.
Kedishvili NY. Enzymology of retinoic acid biosynthesis and degradation. J Lipid Res. 2013; 54: 1744–1760.
Rhinn M, Dollé P. Retinoic acid signalling during development. Development. 2012; 139: 843–858.
Connolly RM, Nguyen NK, Sukumar S. Molecular pathways: current role and future directions of the retinoic acid pathway in cancer prevention and treatment. Clin Cancer Res. 2013; 19: 1651–1659.
Shannon SR, Moise AR, Trainor PA. New insights and changing paradigms in the regulation of vitamin A metabolism in development. Wiley Interdiscip Rev Dev Biol. 2017; 6: e264.
Matt N, Dupé V, Garnier J-M, et al. Retinoic acid-dependent eye morphogenesis is orchestrated by neural crest cells. Development. 2005; 132: 4789–4800.
White RJ, Nie Q, Lander AD, Schilling TF. Complex regulation of cyp26a1 creates a robust retinoic acid gradient in the zebrafish embryo. PLoS Biol. 2007; 5: e304.
Hermosilla VE, Hepp MI, Escobar D, et al. Developmental SALL2 transcription factor: a new player in cancer. Carcinogenesis. 2017; 38: 680–690.
Hao N, Bhakti VLD, Peet DJ, Whitelaw ML. Reciprocal regulation of the basic helix-loop-helix/Per-Arnt-Sim partner proteins, Arnt and Arnt2, during neuronal differentiation. Nucleic Acids Res. 2013; 4: 5626–5638.
Kim SC, Kim C-K, Axe D, et al. All-trans-retinoic acid ameliorates hepatic steatosis in mice by a novel transcriptional cascade. Hepatology. 2014; 59: 1750–1760.
Bohnsack BL, Kasprick DS, Kish PE, Goldman D, Kahana A. A zebrafish model of axenfeld-rieger syndrome reveals that pitx2 regulation by retinoic acid is essential for ocular and craniofacial development. Invest Ophthalmol Vis Sci. 2012; 53: 7–22.
Edwards SA, Darland T, Sosnowski R, Samuels M, Adamson ED. The transcription factor, Egr-1, is rapidly modulated in response to retinoic acid in P19 embryonal carcinoma cells. Dev Biol. 1991; 148: 165–173.
Merrill RA, Ahrens JM, Kaiser ME, Federhart KS, Poon VY, Clagett-Dame M. All-trans retinoic acid-responsive genes identified in the human SH-SY5Y neuroblastoma cell line and their regulated expression in the nervous system of early embryos. Biol Chem. 2004; 385: 605–614.
Chaudhury A, Howe PH. The tale of transforming growth factor-beta (TGFbeta) signaling: a soigné enigma. IUBMB Life. 2009; 61: 929–939.
Massagué J. TGFβ signalling in context. Nat Rev Mol Cell Biol. 2012; 13: 616–630.
Ewen ME, Sluss HK, Whitehouse LL, Livingston DM. TGF beta inhibition of Cdk4 synthesis is linked to cell cycle arrest. Cell. 1993; 74: 1009–1020.
Kowanetz M, Valcourt U, Bergström R, Heldin C-H, Moustakas A. Id2 and Id3 define the potency of cell proliferation and differentiation responses to transforming growth factor beta and bone morphogenetic protein. Mol Cell Biol. 2004; 24: 4241–4254.
Shi J, Zhuang Y, Liu XK, Zhang YX, Zhang Y. TGF-beta induced RBL2 expression in renal cancer cells by down-regulating miR-93. Clin Transl Oncol. 2014; 16: 986–992.
Tumelty KE, Smith BD, Nugent MA, Layne MD. Aortic carboxypeptidase-like protein (ACLP) enhances lung myofibroblast differentiation through transforming growth factor β receptor-dependent and -independent pathways. J Biol Chem. 2014; 289: 2526–2536.
Chen Q, Lee C-E, Denard B, Ye J. Sustained induction of collagen synthesis by TGF-β requires regulated intramembrane proteolysis of CREB3L1. PLoS One. 2014; 9: e108528.
Frederick JP, Liberati NT, Waddell DS, Shi Y, Wang X-F. Transforming growth factor beta-mediated transcriptional repression of c-myc is dependent on direct binding of Smad3 to a novel repressive Smad binding element. Mol Cell Biol. 2004; 24: 2546–2559.
Zavadil J, Cermak L, Soto-Nieves N, Böttinger EP. Integration of TGF-beta/Smad and Jagged1/Notch signalling in epithelial-to-mesenchymal transition. EMBO J. 2004; 23: 1155–1165.
Juang Y-L, Jeng Y-M, Chen C-L, Lien H-C. PRRX2 as a novel TGF-β-induced factor enhances invasion and migration in mammary epithelial cell and correlates with poor prognosis in breast cancer. Mol Carcinog. 2016; 55: 2247–2259.
Thuault S, Valcourt U, Petersen M, Manfioletti G, Heldin C-H, Moustakas A. Transforming growth factor-β employs HMGA2 to elicit epithelial–mesenchymal transition. J Cell Biol. 2006; 174: 175–183.
Suzuki H, Yagi K, Kondo M, Kato M, Miyazono K, Miyazawa K. c-Ski inhibits the TGF-β signaling pathway through stabilization of inactive Smad complexes on Smad-binding elements. Oncogene. 2004; 23: 5068–5076.
MacDonald BT, Tamai K, He X. Wnt/beta-catenin signaling: components, mechanisms, and diseases. Dev Cell. 2009; 17: 9–26.
Clevers H, Nusse R. Wnt/β-catenin signaling and disease. Cell. 2012; 149: 1192–1205.
Liu H, Mohamed O, Dufort D, Wallace VA. Characterization of Wnt signaling components and activation of the Wnt canonical pathway in the murine retina. Dev Dyn. 2003; 227: 323–334.
Yan D, Avtanski D, Saxena NK, Sharma D. Leptin-induced epithelial-mesenchymal transition in breast cancer cells requires β-catenin activation via Akt/GSK3- and MTA1/Wnt1 protein-dependent pathways. J Biol Chem. 2012; 287: 8598–8612.
Bhattaram P, Penzo-Méndez A, Kato K, et al. SOXC proteins amplify canonical WNT signaling to secure nonchondrocytic fates in skeletogenesis. J Cell Biol. 2014; 207: 657–671.
Cai K, El-Merahbi R, Loeffler M, Mayer AE, Sumara G. Ndrg1 promotes adipocyte differentiation and sustains their function. Sci Rep. 2017; 7: 7191.
Liu W, Xing F, Iiizumi-Gairani M, et al. N-myc downstream regulated gene 1 modulates Wnt-β-catenin signalling and pleiotropically suppresses metastasis. EMBO Mol Med. 2012; 4: 93–108.
Ojeda AF, Munjaal RP, Lwigale PY. Knockdown of CXCL14 disrupts neurovascular patterning during ocular development. Dev Biol. 2017; 423: 77–91.
Hynes RO. Integrins: bidirectional, allosteric signaling machines. Cell. 2002; 110: 673–687.
Tsui S, Wang J, Wang L, Dai W, CTCF-mediated Lu L. and Pax6-associated gene expression in corneal epithelial cell-specific differentiation. PLoS One. 2016; 11: e0162071.
Ouyang H, Xue Y, Lin Y, et al. WNT7A and PAX6 define corneal epithelium homeostasis and pathogenesis. Nature. 2014; 511: 358–361.
Swamynathan SK, Katz JP, Kaestner KH, Ashery-Padan R, Crawford MA, Piatigorsky J. Conditional deletion of the mouse Klf4 gene results in corneal epithelial fragility, stromal edema, and loss of conjunctival goblet cells. Mol Cell Biol. 2007; 27: 182–194.
Kenchegowda D, Harvey SAK, Swamynathan S, Lathrop KL, Swamynathan SK. Critical role of Klf5 in regulating gene expression during post-eyelid opening maturation of mouse corneas. PLoS One. 2012; 7: e44771.
Bragulla HH, Homberger DG. Structure and functions of keratin proteins in simple, stratified, keratinized and cornified epithelia. J Anat. 2009; 214: 516–559.
Olijnyk D, Ibrahim AM, Ferrier RK, et al. Fibulin-2 is involved in early extracellular matrix development of the outgrowing mouse mammary epithelium. Cell Mol Life Sci. 2014; 71: 3811–3828.
Ito S, Nagata K. Biology of Hsp47 (Serpin H1), a collagen-specific molecular chaperone. Semin Cell Dev Biol. 2017; 62: 142–151.
Bangsow T, Baumann E, Bangsow C, et al. The epithelial membrane protein 1 is a novel tight junction protein of the blood—brain barrier. J Cereb Blood Flow Metab. 2008; 28: 1249–1260.
Ren J-G, Jie C, Talbot C. How PEDF prevents angiogenesis: a hypothesized pathway. Med Hypotheses. 2005; 64: 74–78.
Cvekl A, Wang W-L. Retinoic acid signaling in mammalian eye development. Exp Eye Res. 2009; 89: 280–291.
Bronner ME. Formation and migration of neural crest cells in the vertebrate embryo. Histochem Cell Biol. 2012; 138: 179–186.
Le Douarin NM, Creuzet S, Couly G, Dupin E. Neural crest cell plasticity and its limits. Development. 2004; 131: 4637–4650.
Creuzet S, Couly G, Le Douarin NM. Patterning the neural crest derivatives during development of the vertebrate head: insights from avian studies. J Anat. 2005; 207: 447–459.
Hino K, Saito A, Kido M, et al. Master regulator for chondrogenesis, Sox9, regulates transcriptional activation of the endoplasmic reticulum stress transducer BBF2H7/CREB3L2 in chondrocytes. J Biol Chem. 2014; 289: 13810–13820.
Pozniak CD, Langseth AJ, Dijkgraaf GJP, Choe Y, Werb Z, Pleasure SJ. Sox10 directs neural stem cells toward the oligodendrocyte lineage by decreasing Suppressor of Fused expression. Proc Natl Acad Sci U S A. 2010; 107: 21795–21800.
Vincentz JW, Firulli BA, Lin A, Spicer DB, Howard MJ, Firulli AB. Twist1 controls a cell-specification switch governing cell fate decisions within the cardiac neural crest. PLoS Genet. 2013; 9: e1003405.
Fakhouri WD, Metwalli K, Naji A, et al. Intercellular genetic interaction between Irf6 and Twist1 during craniofacial development. Sci Rep. 2017; 7: 7129.
Gu S, Boyer TG, Naski MC. Basic helix-loop-helix transcription factor Twist1 inhibits transactivator function of master chondrogenic regulator Sox9. J Biol Chem. 2012; 287: 21082–21092.
Chen Z, Huang J, Liu Y, et al. FGF signaling activates a Sox9-Sox10 pathway for the formation and branching morphogenesis of mouse ocular glands. Development. 2014; 141: 2691–2701.
Taneyhill LA, Coles EG, Bronner-Fraser M. Snail2 directly represses cadherin6B during epithelial-to-mesenchymal transitions of the neural crest. Development. 2007; 134: 1481–1490.
Aomatsu K, Arao T, Abe K, et al. Slug is upregulated during wound healing and regulates cellular phenotypes in corneal epithelial cells. Invest Opthalmol Vis Sci. 2012; 53: 751–756.
Kam RKT, Deng Y, Chen Y, Zhao H. Retinoic acid synthesis and functions in early embryonic development. Cell Biosci. 2012; 2: 11.
Duester G. Retinoic acid synthesis and signaling during early organogenesis. Cell. 2008; 134: 921–931.
Sakai Y, Meno C, Fujii H, et al. The retinoic acid-inactivating enzyme CYP26 is essential for establishing an uneven distribution of retinoic acid along the anterio-posterior axis within the mouse embryo. Genes Dev. 2001; 15: 213–225.
Persaud SD, Lin Y-W, Wu C-Y, Kagechika H, Wei L-N. Cellular retinoic acid binding protein I mediates rapid non-canonical activation of ERK1/2 by all-trans retinoic acid. Cell Signal. 2013; 25: 19–25.
Wei L-N. Non-canonical activity of retinoic acid in epigenetic control of embryonic stem cell. Transcription. 2013; 4: 158–161.
Ittner LM, Wurdak H, Schwerdtfeger K, et al. Compound developmental eye disorders following inactivation of TGFbeta signaling in neural-crest stem cells. J Biol. 2005; 4: 11.
Karamichos D, Hutcheon AEK, Zieske JD. Transforming growth factor-β3 regulates assembly of a non-fibrotic matrix in a 3D corneal model. J Tissue Eng Regen Med. 2011; 5: e228–e238.
Karamichos D, Rich CB, Zareian R, et al. TGF-β3 stimulates stromal matrix assembly by human corneal keratocyte-like cells. Invest Opthalmol Vis Sci. 2013; 54: 6612–6619.
Neuzillet C, de Gramont A, Tijeras-Raballand A, et al. Perspectives of TGF-β inhibition in pancreatic and hepatocellular carcinomas. Oncotarget. 2014; 5: 78–94.
López-Casillas F, Wrana JL, Massagué J. Betaglycan presents ligand to the TGF beta signaling receptor. Cell. 1993; 73: 1435–1444.
Sarraj MA, Escalona RM, Western P, Findlay JK, Stenvers KL. Effects of TGFbeta2 on wild-type and Tgfbr3 knockout mouse fetal testis. Biol Reprod. 2013; 88: 66.
Zhou X, Tao Y, Liang C, Zhang Y, Li H, Chen Q. BMP3 alone and together with TGF-β promote the differentiation of human mesenchymal stem cells into a nucleus pulposus-like phenotype. Int J Mol Sci. 2015; 16: 20344–20359.
Hartigan N, Garrigue-Antar L, Kadler KE. Bone morphogenetic protein-1 (BMP-1). Identification of the minimal domain structure for procollagen C-proteinase activity. J Biol Chem. 2003; 278: 18045–18049.
Gage PJ, Qian M, Wu D, Rosenberg KI. The canonical Wnt signaling antagonist DKK2 is an essential effector of PITX2 function during normal eye development. Dev Biol. 2008; 317: 310–324.
Mukhopadhyay M, Gorivodsky M, Shtrom S, et al. Dkk2 plays an essential role in the corneal fate of the ocular surface epithelium. Development. 2006; 133: 2149–2154.
Wang Y, Mahesh P, Wang Y, et al. Spatiotemporal dynamics of canonical Wnt signaling during embryonic eye development and posterior capsular opacification (PCO). Exp Eye Res. 2018; 175: 148–158.
Corda G, Sala A. Non-canonical WNT/PCP signalling in cancer: Fzd6 takes centre stage. Oncogenesis. 2017; 6: e364.
Galli LM, Munji RN, Chapman SC, et al. Frizzled10 mediates WNT1 and WNT3A signaling in the dorsal spinal cord of the developing chick embryo. Dev Dyn. 2014; 243: 833–843.
Abidin BM, Owusu Kwarteng E, Heinonen KM. Frizzled-6 regulates hematopoietic stem/progenitor cell survival and self-renewal. J Immunol. 2015; 195: 2168–2176.
Chesnutt C, Burrus LW, Brown AMC, Niswander L. Coordinate regulation of neural tube patterning and proliferation by TGFβ and WNT activity. Dev Biol. 2004; 274: 334–347.
Paes KT, Wang E, Henze K, et al. Frizzled 4 is required for retinal angiogenesis and maintenance of the blood-retina barrier. Invest Opthalmol Vis Sci. 2011; 52: 6452.
Lu Y, Tai PWL, Ai J, et al. Transcriptome profiling of neovascularized corneas reveals miR-204 as a multi-target biotherapy deliverable by rAAVs. Mol Ther Nucleic Acids. 2018; 10: 349–360.
Yanfeng WA, Tan C, Fagan RJ, Klein PS. Phosphorylation of frizzled-3. J Biol Chem. 2006; 281: 11603–11609.
Deardorff MA, Tan C, Saint-Jeannet JP, Klein PS. A role for frizzled 3 in neural crest development. Development. 2001; 128: 3655–3663.
Chang C-H, Tsai R-K, Tsai M-H, Lin Y-H, Hirobe T. The roles of Frizzled-3 and Wnt3a on melanocyte development: in vitro studies on neural crest cells and melanocyte precursor cell lines. J Dermatol Sci. 2014; 75: 100–108.
Panzica DA, Findlay AS, van Ladesteijn R, Collinson JM. The core planar cell polarity gene, Vangl2, maintains apical-basal organisation of the corneal epithelium. J Anat. 2019; 234: 106–119.
Findlay AS, Panzica DA, Walczysko P, et al. The core planar cell polarity gene, Vangl2, directs adult corneal epithelial cell alignment and migration. R Soc Open Sci. 2016; 3: 160658.
Kawakami Y, Wada N, Nishimatsu S, Komaguchi C, Noji S, Nohno T. Identification of chick frizzled-10 expressed in the developing limb and the central nervous system. Mech Dev. 2000; 91: 375–378.
Fuhrmann S. Wnt signaling in eye organogenesis. Organogenesis. 2008; 4: 60–67.
Heinonen KM, Vanegas JR, Lew D, Krosl J, Perreault C. Wnt4 enhances murine hematopoietic progenitor cell expansion through a planar cell polarity-like pathway. PLoS One. 2011; 6: e19279.
Martineau X, Abed É, Martel-Pelletier J, Pelletier J-P, Lajeunesse D. Alteration of Wnt5a expression and of the non-canonical Wnt/PCP and Wnt/PKC-Ca2+ pathways in human osteoarthritis osteoblasts. PLoS One. 2017; 12: e0180711.
Goodyear RJ, Lu X, Deans MR, Richardson GP. A tectorin-based matrix and planar cell polarity genes are required for normal collagen-fibril orientation in the developing tectorial membrane. Development. 2017; 144: 3978–3989.
Li Y, Li A, Junge J, Bronner M. Planar cell polarity signaling coordinates oriented cell division and cell rearrangement in clonally expanding growth plate cartilage. Elife. 2017; 6: e23279.
Fenstermaker AG, Prasad AA, Bechara A, et al. Wnt/planar cell polarity signaling controls the anterior-posterior organization of monoaminergic axons in the brainstem. J Neurosci. 2010; 30: 16053–16064.
Gage PJ, Kuang C, Zacharias AL. The homeodomain transcription factor PITX2 is required for specifying correct cell fates and establishing angiogenic privilege in the developing cornea. Dev Dyn. 2014; 243: 1391–1400.
Leow SC, Poschmann J, Too PG, et al. The transcription factor SOX6 contributes to the developmental origins of obesity by promoting adipogenesis. Development. 2016; 143: 950–961.
Valenta T, Lukas J, Doubravska L, Fafilek B, Korinek V. HIC1 attenuates Wnt signaling by recruitment of TCF-4 and β-catenin to the nuclear bodies. EMBO J. 2006; 25: 2326–2337.
Kormish JD, Sinner D, Zorn AM. Interactions between SOX factors and Wnt/beta-catenin signaling in development and disease. Dev Dyn. 2010; 239: 56–68.
Hamada-Kanazawa M, Ogawa D, Takano M, Miyake M. Sox6 suppression induces RA-dependent apoptosis mediated by BMP-4 expression during neuronal differentiation in P19 cells. Mol Cell Biochem. 2016; 412: 49–57.
Burrows K, Antignano F, Chenery A, et al. HIC1 links retinoic acid signalling to group 3 innate lymphoid cell-dependent regulation of intestinal immunity and homeostasis. PLoS Pathog. 2018; 14: e1006869.
Yuan G, Zhang B, Yang S, et al. Novel role of STRAP in progression and metastasis of colorectal cancer through Wnt/β-catenin signaling. Oncotarget. 2016; 7: 16023–16037.
Datta PK, Moses HL. STRAP and Smad7 synergize in the inhibition of transforming growth factor beta signaling. Mol Cell Biol. 2000; 20: 3157–3167.
Kumar S, Duester G. Retinoic acid signaling in perioptic mesenchyme represses Wnt signaling via induction of Pitx2 and Dkk2. Dev Biol. 2010; 340: 67–74.
Pietenpol JA, Holt JT, Stein RW, Moses HL. Transforming growth factor beta 1 suppression of c-myc gene transcription: role in inhibition of keratinocyte proliferation. Proc Natl Acad Sci U S A. 1990; 87: 3758–3762.
Xu B, Zhang K, Huang Y. Lin28 modulates cell growth and associates with a subset of cell cycle regulator mRNAs in mouse embryonic stem cells. RNA. 2009; 15: 357–361.
Park JT, Kato M, Lanting L, et al. Repression of let-7 by transforming growth factor-β1-induced Lin28 upregulates collagen expression in glomerular mesangial cells under diabetic conditions. Am J Physiol Physiol. 2014; 307: F1390–F1403.
Zhang Y, Call MK, Yeh L-K, et al. Aberrant expression of a -catenin gain-of-function mutant induces hyperplastic transformation in the mouse cornea. J Cell Sci. 2010; 123: 1285–1294.
Mizoguchi S, Suzuki K, Zhang J, et al. Disruption of eyelid and cornea morphogenesis by epithelial β-catenin gain-of-function. Mol Vis. 2015; 21: 793–803.
Kao WW-Y, Liu C-Y. Roles of lumican and keratocan on corneal transparency. Glycoconj J. 2002; 19: 275–285.
Delannet M, Martin F, Bossy B, Cheresh DA, Reichardt LF, Duband JL. Specific roles of the alpha V beta 1, alpha V beta 3 and alpha V beta 5 integrins in avian neural crest cell adhesion and migration on vitronectin. Development. 1994; 120: 2687–2702.
Casini P, Nardi I, Ori M. Hyaluronan is required for cranial neural crest cells migration and craniofacial development. Dev Dyn. 2012; 241: 294–302.
Dutt S, Kléber M, Matasci M, Sommer L, Zimmermann DR. Versican V0 and V1 guide migratory neural crest cells. J Biol Chem. 2006; 281: 12123–12131.
Kim HR, Ingham PW. The extracellular matrix protein TGFBI promotes myofibril bundling and muscle fibre growth in the zebrafish embryo. Dev Dyn. 2009; 238: 56–65.
Wang F, Hu W, Xian J, Ohnuma S, Brenton JD. The Xenopus Tgfbi is required for embryogenesis through regulation of canonical Wnt signalling. Dev Biol. 2013; 379: 16–27.
Meek KM, Knupp C. Corneal structure and transparency. Prog Retin Eye Res. 2015; 49: 1–16.
Zhou H-Y, Cao Y, Wu J, Zhang W-S. Role of corneal collagen fibrils in corneal disorders and related pathological conditions. Int J Ophthalmol. 2017; 10: 803–811.
Zhou C, Petroll WM. MMP regulation of corneal keratocyte motility and mechanics in 3-D collagen matrices. Exp Eye Res. 2014; 121: 147–160.
Mantelli F, Mauris J, Argüeso P. The ocular surface epithelial barrier and other mechanisms of mucosal protection: from allergy to infectious diseases. Curr Opin Allergy Clin Immunol. 2013; 13: 563–568.
Gu L-H, Coulombe PA. Keratin function in skin epithelia: a broadening palette with surprising shades. Curr Opin Cell Biol. 2007; 19: 13–23.
Swamynathan SK, Swamynathan SK. Ocular surface development and gene expression. J Ophthalmol. 2013; 2013: 103947.
Swamynathan S, Delp EE, Harvey SAK, Loughner CL, Raju L, Swamynathan SK. Corneal expression of SLURP-1 by age, sex, genetic strain, and ocular surface health. Invest Opthalmol Vis Sci. 2015; 56: 7888–7896.
Norman B, Davis J, Piatigorsky J. Postnatal gene expression in the normal mouse cornea by SAGE. Invest Opthalmol Vis Sci. 2004; 45: 429–440.
Swamynathan S, Buela K-A, Kinchington P, et al. Klf4 regulates the expression of Slurp1, which functions as an immunomodulatory peptide in the mouse cornea. Invest Ophthalmol Vis Sci. 2012; 53: 8433–8446.
Doane KJ, Howell SJ, Birk DE. Identification and functional characterization of two type VI collagen receptors, alpha 3 beta 1 integrin and NG2, during avian corneal stromal development. Invest Ophthalmol Vis Sci. 1998; 39: 263–275.
Byström B, Carracedo S, Behndig A, Gullberg D, Pedrosa-Domellöf F. α11 integrin in the human cornea: importance in development and disease. Invest Opthalmol Vis Sci. 2009; 50: 5044–5053.
Myers JP, Santiago-Medina M, Gomez TM. Regulation of axonal outgrowth and pathfinding by integrin-ECM interactions. Dev Neurobiol. 2011; 71: 901–923.
Stepp MA. Corneal integrins and their functions. Exp Eye Res. 2006; 83: 3–15.
Munger JS, Harpel JG, Giancotti FG, Rifkin DB. Interactions between growth factors and integrins: latent forms of transforming growth factor-beta are ligands for the integrin alphavbeta1. Mol Biol Cell. 1998; 9: 2627–2638.
Mould AP, Askari JA, Craig SE, Garratt AN, Clements J, Humphries MJ. Integrin alpha 4 beta 1-mediated melanoma cell adhesion and migration on vascular cell adhesion molecule-1 (VCAM-1) and the alternatively spliced IIICS region of fibronectin. J Biol Chem. 1994; 269: 27224–27230.
Bieritz B, Spessotto P, Colombatti A, Jahn A, Prols F, Hartner A. Role of α8 integrin in mesangial cell adhesion, migration, and proliferation. Kidney Int. 2003; 64: 119–127.
Bazigou E, Xie S, Chen C, et al. Integrin-alpha9 is required for fibronectin matrix assembly during lymphatic valve morphogenesis. Dev Cell. 2009; 17: 175–186.
San Martin R, Pathak R, Jain A, et al. Tenascin-C and integrin α9 mediate interactions of prostate cancer with the bone microenvironment. Cancer Res. 2017; 77: 5977–5988.
Vogel W, Gish GD, Alves F, Pawson T. The discoidin domain receptor tyrosine kinases are activated by collagen. Mol Cell. 1997; 1: 13–23.
Castro-Sanchez L, Soto-Guzman A, Guaderrama-Diaz M, Cortes-Reynosa P, Salazar EP. Role of DDR1 in the gelatinases secretion induced by native type IV collagen in MDA-MB-231 breast cancer cells. Clin Exp Metastasis. 2011; 28: 463–477.
Bassat E, Mutlak YE, Genzelinakh A, et al. The extracellular matrix protein agrin promotes heart regeneration in mice. Nature. 2017; 547: 179–184.
Fuerst PG, Rauch SM, Burgess RW. Defects in eye development in transgenic mice overexpressing the heparan sulfate proteoglycan agrin. Dev Biol. 2007; 303: 165–180.
Gupta V, Kawahara G, Gundry SR, et al. The zebrafish dag1 mutant: a novel genetic model for dystroglycanopathies. Hum Mol Genet. 2011; 20: 1712–1725.
Satz JS, Philp AR, Nguyen H, et al. Visual impairment in the absence of dystroglycan. J Neurosci. 2009; 29: 13136–13146.
McKenna CC, Munjaal RP, Lwigale PY. Distinct roles for neuropilin1 and neuropilin2 during mouse corneal innervation. PLoS One. 2012; 7: e37175.
McKenna CC, Ojeda AF, Spurlin J, Kwiatkowski S, Lwigale PY, Lwigale PY. Sema3A maintains corneal avascularity during development by inhibiting Vegf induced angioblast migration. Dev Biol. 2014; 391: 241–250.
Buehler A, Sitaras N, Favret S, et al. Semaphorin 3F forms an anti-angiogenic barrier in outer retina. FEBS Lett. 2013; 587: 1650–1655.
Maione F, Molla F, Meda C, et al. Semaphorin 3A is an endogenous angiogenesis inhibitor that blocks tumor growth and normalizes tumor vasculature in transgenic mouse models. J Clin Invest. 2009; 119: 3356–3372.
Yang W-J, Hu J, Uemura A, Tetzlaff F, Augustin HG, Fischer A. Semaphorin-3C signals through neuropilin-1 and plexinD1 receptors to inhibit pathological angiogenesis. EMBO Mol Med. 2015; 7: 1267–1284.
Cheng G, Zhong M, Kawaguchi R, et al. Identification of PLXDC1 and PLXDC2 as the transmembrane receptors for the multifunctional factor PEDF. Elife. 2014; 3: e05401.
Bouck N. PEDF: anti-angiogenic guardian of ocular function. Trends Mol Med. 2002; 8: 330–334.
Dawson DW, Volpert OV, Gillis P, et al. Pigment epithelium-derived factor: a potent inhibitor of angiogenesis. Science. 1999; 285: 245–248.
Gao S, Li C, Zhu Y, et al. PEDF mediates pathological neovascularization by regulating macrophage recruitment and polarization in the mouse model of oxygen-induced retinopathy. Sci Rep. 2017; 7: 42846.
Figure 1
 
RNA sequencing analysis of the pNC and embryonic corneal cells. (A) Hematoxylin and eosin staining of mouse embryonic eyes, with highlighted regions within the dotted lines showing the tissue dissected for RNA preparation at E10.5, E14.5, and E16.5. (B) Venn diagram depicts total number of genes categorized between the three stages. (C) Bar plot showing number of differentially regulated genes detected between E10.5 and E14.5, E10.5 and E16.5, and E14.5 and E16.5. (D) Pathway enrichment analysis of DEGs at E10.5 vs. E14.5. Circle size correlates with number of genes and the Rich factor is a representation of the degree of enrichment based on ratio of DEG/non-DEG within the pathway. Scale bars: 50 μm (E10.5 and E14.5) or 100 μm (E16.5). ec, ectoderm; L, lens; oc, optic cup; ep, epithelium; st, stroma; en, endothelium; ey, eyelid.
Figure 1
 
RNA sequencing analysis of the pNC and embryonic corneal cells. (A) Hematoxylin and eosin staining of mouse embryonic eyes, with highlighted regions within the dotted lines showing the tissue dissected for RNA preparation at E10.5, E14.5, and E16.5. (B) Venn diagram depicts total number of genes categorized between the three stages. (C) Bar plot showing number of differentially regulated genes detected between E10.5 and E14.5, E10.5 and E16.5, and E14.5 and E16.5. (D) Pathway enrichment analysis of DEGs at E10.5 vs. E14.5. Circle size correlates with number of genes and the Rich factor is a representation of the degree of enrichment based on ratio of DEG/non-DEG within the pathway. Scale bars: 50 μm (E10.5 and E14.5) or 100 μm (E16.5). ec, ectoderm; L, lens; oc, optic cup; ep, epithelium; st, stroma; en, endothelium; ey, eyelid.
Figure 2
 
Expression of NCC genes during corneal development. (A) Schematic describes the number of expressed genes at each developmental stage based on threshold value. (B) Heatmap shows relative expression of the transcripts in the pNC, E14.5, and E16.5 corneas. Relative color ranges from white to red based on low (L) or high (H) expression. In addition to the criteria described in the methods, values below threshold were normalized to a log base 2 value of 0. Downregulated genes are highlighted in green, not significantly DEGs in blue, upregulated genes in red, and genes below threshold are not shown (see Supplementary Table S2). (CF) Validation of the expression patterns of Alx1, Alx4, Snai2, and Tfap2b. Black arrows represent regions of enriched expression. Scale bar: 50 μm. co, Cornea; *C-myc expression at E16.5 is excluded.
Figure 2
 
Expression of NCC genes during corneal development. (A) Schematic describes the number of expressed genes at each developmental stage based on threshold value. (B) Heatmap shows relative expression of the transcripts in the pNC, E14.5, and E16.5 corneas. Relative color ranges from white to red based on low (L) or high (H) expression. In addition to the criteria described in the methods, values below threshold were normalized to a log base 2 value of 0. Downregulated genes are highlighted in green, not significantly DEGs in blue, upregulated genes in red, and genes below threshold are not shown (see Supplementary Table S2). (CF) Validation of the expression patterns of Alx1, Alx4, Snai2, and Tfap2b. Black arrows represent regions of enriched expression. Scale bar: 50 μm. co, Cornea; *C-myc expression at E16.5 is excluded.
Figure 3
 
Differential regulation of the RA signaling pathway. (A) Schematic depicts whether components of the RA pathway are upregulated (red), downregulated (green), or not significantly differentially expressed (black). Genes that were upregulated and then downregulated, or vice versa, are represented by blue and orange, respectively. (B) Heatmap summarizes the relative expression of the DEGs. (CE) Validation of the expression patterns of Nr2f2, Egr1, and Cyp26a1. Black arrows represent regions of enriched expression. Scale bar: 50 μm.
Figure 3
 
Differential regulation of the RA signaling pathway. (A) Schematic depicts whether components of the RA pathway are upregulated (red), downregulated (green), or not significantly differentially expressed (black). Genes that were upregulated and then downregulated, or vice versa, are represented by blue and orange, respectively. (B) Heatmap summarizes the relative expression of the DEGs. (CE) Validation of the expression patterns of Nr2f2, Egr1, and Cyp26a1. Black arrows represent regions of enriched expression. Scale bar: 50 μm.
Figure 4
 
Differential regulation of the TGFβ signaling pathway. (A) Schematic depicts whether components of the TGFβ pathway are upregulated (red), downregulated (green), or not significantly differentially expressed (black). (B) Heatmap summarizes the relative expression of the DEGs. (CE) Validation of the expression patterns of Hmga2, C-ski, and Creb3l1. Black arrows represent regions of enriched expression. Scale bar: 50 μm.
Figure 4
 
Differential regulation of the TGFβ signaling pathway. (A) Schematic depicts whether components of the TGFβ pathway are upregulated (red), downregulated (green), or not significantly differentially expressed (black). (B) Heatmap summarizes the relative expression of the DEGs. (CE) Validation of the expression patterns of Hmga2, C-ski, and Creb3l1. Black arrows represent regions of enriched expression. Scale bar: 50 μm.
Figure 5
 
Differential regulation of the Wnt signaling pathway. (A) Schematic depicts whether components of the Wnt pathway are upregulated (red), downregulated (green), or not significantly differentially expressed (black). (B) Heatmap summarizes the relative expression of the DEGs. (CE) Validation of the expression patterns of Mta1, Sox11, and Ndrg1. Black arrows represent regions of enriched expression. Scale bar: 50 μm. AA, ambiguously associated.
Figure 5
 
Differential regulation of the Wnt signaling pathway. (A) Schematic depicts whether components of the Wnt pathway are upregulated (red), downregulated (green), or not significantly differentially expressed (black). (B) Heatmap summarizes the relative expression of the DEGs. (CE) Validation of the expression patterns of Mta1, Sox11, and Ndrg1. Black arrows represent regions of enriched expression. Scale bar: 50 μm. AA, ambiguously associated.
Figure 6
 
Categorization of differentially expressed transcription factors into the RA, TGFb, and Wnt signaling pathways. (A) Allocation of transcription factors from a total pool of 1755 (compiled from Riken's mouse database and self-annotated). (B, C) Venn diagrams showing overlap between differentially expressed transcription factors. Values outside the circles represent the genes that are not well characterized or not studied within the relevant pathways.
Figure 6
 
Categorization of differentially expressed transcription factors into the RA, TGFb, and Wnt signaling pathways. (A) Allocation of transcription factors from a total pool of 1755 (compiled from Riken's mouse database and self-annotated). (B, C) Venn diagrams showing overlap between differentially expressed transcription factors. Values outside the circles represent the genes that are not well characterized or not studied within the relevant pathways.
Figure 7
 
Spatiotemporal expression of genes involved in corneal morphogenesis. (A) Fbln2 is expressed at low levels at E10.5 and E16.5 and is enriched throughout the stroma at E16.5. (B) Serpinh1 is expressed in the pNC, stroma, and epithelium. At E10.5 it is enriched in the temporal mesenchyme, at E14.5 enriched in the anterior stroma, and E16.5 enriched in the endothelium. (C) Emp1 is expressed in all stages and layers, and enriched in the epithelium at E16.5. (D) Pedf is expressed in all stages and layers and enriched in the posterior cornea at E16.5. Black arrows represent regions of enriched expression. Scale bar: 50 μm.
Figure 7
 
Spatiotemporal expression of genes involved in corneal morphogenesis. (A) Fbln2 is expressed at low levels at E10.5 and E16.5 and is enriched throughout the stroma at E16.5. (B) Serpinh1 is expressed in the pNC, stroma, and epithelium. At E10.5 it is enriched in the temporal mesenchyme, at E14.5 enriched in the anterior stroma, and E16.5 enriched in the endothelium. (C) Emp1 is expressed in all stages and layers, and enriched in the epithelium at E16.5. (D) Pedf is expressed in all stages and layers and enriched in the posterior cornea at E16.5. Black arrows represent regions of enriched expression. Scale bar: 50 μm.
Table 1
 
Top Differentially Expressed Transcription Factors
Table 1
 
Top Differentially Expressed Transcription Factors
Table 2
 
Top Differentially Expressed ECM and Junction-Associated Genes
Table 2
 
Top Differentially Expressed ECM and Junction-Associated Genes
Table 3
 
Top Differentially Expressed Epithelial-Associated Genes
Table 3
 
Top Differentially Expressed Epithelial-Associated Genes
Table 4
 
Top Differentially Expressed Cell Cycle-Associated Genes
Table 4
 
Top Differentially Expressed Cell Cycle-Associated Genes
Table 5
 
Top DEGs Associated With Angiogenesis and Axon Guidance
Table 5
 
Top DEGs Associated With Angiogenesis and Axon Guidance
Supplement 1
Supplement 2
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×