1. Neuroscience
Download icon

Downregulation of ribosome biogenesis during early forebrain development

  1. Kevin F Chau
  2. Morgan L Shannon
  3. Ryann M Fame
  4. Erin Fonseca
  5. Hillary Mullan
  6. Matthew B Johnson
  7. Anoop K Sendamarai
  8. Mark W Springel
  9. Benoit Laurent
  10. Maria K Lehtinen  Is a corresponding author
  1. Boston Children’s Hospital, United States
  2. Harvard Medical School, United States
Research Article
  • Cited 3
  • Views 2,139
  • Annotations
Cite this article as: eLife 2018;7:e36998 doi: 10.7554/eLife.36998

Abstract

Forebrain precursor cells are dynamic during early brain development, yet the underlying molecular changes remain elusive. We observed major differences in transcriptional signatures of precursor cells from mouse forebrain at embryonic days E8.5 vs. E10.5 (before vs. after neural tube closure). Genes encoding protein biosynthetic machinery were strongly downregulated at E10.5. This was matched by decreases in ribosome biogenesis and protein synthesis, together with age-related changes in proteomic content of the adjacent fluids. Notably, c-MYC expression and mTOR pathway signaling were also decreased at E10.5, providing potential drivers for the effects on ribosome biogenesis and protein synthesis. Interference with c-MYC at E8.5 prematurely decreased ribosome biogenesis, while persistent c-MYC expression in cortical progenitors increased transcription of protein biosynthetic machinery and enhanced ribosome biogenesis, as well as enhanced progenitor proliferation leading to subsequent macrocephaly. These findings indicate large, coordinated changes in molecular machinery of forebrain precursors during early brain development.

https://doi.org/10.7554/eLife.36998.001

Introduction

Neural tube closure (neurulation) is a fundamental milestone of early brain development, yet relatively little is known about the cellular and molecular transitions occurring in neural precursor cells before and after this process due to experimental challenges inherent to investigating this nascent organ (Greene and Copp, 2014; Massarwa and Niswander, 2013; Wallingford et al., 2013; Wilde et al., 2014). Prior to neural tube closure, the neural plate is home to multipotent neural stem cells, including forebrain neurectodermal precursor cells. After neural tube closure, these neurectodermal precursors become progressively lineage restricted as neuroepithelial cells, and then radial glial cells, ultimately giving rise to all neurons and glia in the adult forebrain (Bjornsson et al., 2015). As these progenitors proliferate, their spatial patterning serves as a blueprint for the maturing brain (Rallu et al., 2002; Sur and Rubenstein, 2005). While genes involved in driving the more mature stages of forebrain development are becoming better understood, remarkably little is known about the key genes orchestrating the function of earlier neurectodermal precursors.

While transcriptional regulation is essential for the specification and maturation of the early forebrain, less is known about the dynamics of protein biosynthesis at this early stage. Recent studies have begun to explore how regulated protein synthesis is critical for the successful construction and function of healthy cells and organs (Fujii et al., 2017; Kondrashov et al., 2011; Pilaz et al., 2016; Roko Rasin and Silver, 2016; Shi and Barna, 2015). In turn, the regulation of protein biosynthetic machinery has emerged as a tunable program that can instruct cellular transitions between stem cell dormancy, proliferation, and differentiation (DeBoer et al., 2013; Fujii et al., 2017; Khajuria et al., 2018; Kraushar et al., 2016; Sanchez et al., 2016; Scognamiglio et al., 2016). Mutations in genes encoding ribosomal proteins are associated with neural tube closure defects (NTD; Greene and Copp, 2014; Wilde et al., 2014), suggesting that regulation of protein biosynthesis is critical during the earliest stages of forebrain development as well. Proteomic analyses have also revealed that ribosomal and translational proteins are elevated in amniotic fluid (AF) prior to neurulation, and are substantially decreased in nascent cerebrospinal fluid (CSF) following neurulation (Chau et al., 2015). However, the mechanisms leading to these changes in the AF and CSF proteomes remain incompletely understood, as this developmental stage precedes choroid plexus development and its secretion of factors into the CSF (Hunter and Dymecki, 2007; Lehtinen et al., 2011; Lun et al., 2015).

Here, we used RNA sequencing to reveal the transcriptomic signature of presumptive forebrain precursor cells before and after neurulation. High expression of the protein biosynthetic machinery together with elevated protein synthesis emerged as a signature of early neural precursors. These transcriptional and cell biological changes closely mirrored proteomic changes in the adjacent AF and CSF. Many genes that were downregulated after neurulation are known, direct targets of the transcription factor c-MYC (hereafter MYC) in other cell types (Ben-Porath et al., 2008; Zeller et al., 2003). Accordingly, MYC modulated ribosome biogenesis in forebrain precursors. Its forced, persistent expression in neural progenitors by mouse genetics approaches increased transcription of protein biosynthetic machinery and was accompanied by increased proliferation of radial glial progenitors leading to macrocephaly by birth. Taken together, our data identify regulation of protein biosynthetic machinery as an important signature of early forebrain development.

Results

Transcriptome signature of early forebrain neuroepithelium

To define the identity and biology of developing forebrain neuroepithelial cells, we microdissected the neuroepithelium away from the adjacent mesenchyme and surface ectoderm in E8.5 and E10.5 embryos (Figure 1A, Chau et al., 2015), and performed next-generation RNA sequencing (RNAseq) analysis (Figure 1). Gene expression analysis identified 3898 genes (q < 0.05) with significantly different expression patterns between the two ages, with 2375 genes enriched in E8.5 neuroepithelium, and 1523 genes enriched in E10.5 neuroepithelium (Figure 1B, Figure 1—figure supplement 1A).

Figure 1 with 1 supplement see all
Transcriptome analysis of microdissected forebrain epithelium reveals downregulation of genes encoding protein biosynthetic machinery.

(A) Schematic of E8.5 embryo with open forebrain neural tube (left) and E10.5 embryo (right). Shaded regions encircled by dotted line denote developing forebrain epithelium microdissected for RNAseq. (B) Heatmap and hierarchical clustering of ~3900 differentially expressed genes (q < 0.05): 2375 genes were enriched in E8.5 and 1523 genes were enriched in E10.5. Each biological replicate contained tissue pooled from one litter of embryos. Red and green indicate relatively higher and lower expression, with gene FPKM values log2 transformed, centered and scaled by rows for display purposes. (C, D) MA plot displaying genes encoding secreted factors (C), and receptors (D). Each dot represents a single gene. Red dots denote differentially expressed genes as identified by Cuffdiff (q < 0.05). Genes below blue line (y = 0) are enriched in E8.5. (E) Functional annotation clustering of E8.5 neuroepithelium enriched genes revealed overrepresentation of genes encoding ribosomal proteins, ribosome biogenesis and translation factors. The top five enriched functional clusters are shown. (F) Functional annotation clustering of E10.5 neuroepithelium enriched genes shows overrepresentation of genes needed for neuron differentiation. The top five enriched functional clusters are shown. (G–I) GSEA of E8.5 versus E10.5 neuroepithelium for gene sets involved in ribosome biogenesis and translation. Broad Institute Molecular Signatures Database Identifiers: KEGG_RIBOSOME (G), GO_RIBOSOME_BIOGENESIS (H), and TRANSLATION (I). Each line represents a single gene in the gene set. Genes on the right side are enriched in E8.5. (J–L) Correlation plots of average expression (log2 transformed FPKM) at E8.5 and E10.5 for ribosomal proteins (J), translation factors (K), and all differentially expressed genes (L). In all cases correlation was significant; ribosomal proteins (J), Spearman R = 0.91, p<0.0001; translation factors (K) Spearman R = 0.98, p<0.0001; and DEG (L), Spearman R = 0.82, p<0.0001.

https://doi.org/10.7554/eLife.36998.002

Among the differentially expressed genes, many were secreted factors and receptors involved in signaling pathways with cardinal roles in brain development including WNT and BMP/TGFβ (Figure 1C,D, Supplementary file 1; Monuki, 2007; Sur and Rubenstein, 2005; Wilde et al., 2014). Some secreted factors (e.g. BMP1 and SHH) were enriched in both E10.5 progenitors and CSF, suggesting their secretion into the adjacent fluid (Supplementary file 1; Chau et al., 2015), while factors known to be involved in organismal development and neural tube closure including Wnt5a and Pax3 were enriched in E8.5 (Supplementary file 1). Differential gene expression was further validated by quantitative RT-PCR (qRT-PCR) on 81 genes including transcription factors, cell surface receptors, and secreted factors, many of which showed an overall positive correlation (Figure 1—figure supplement 1B). Expression of Glast and Blbp were enriched in E10.5 progenitors, indicating the transition from neuroepithelial cells to radial glial cells (Figure 1—figure supplement 1C).

We next determined the biological functions of the most differentially expressed genes at each age. Consistent with the progressive lineage restriction of progenitors, initiation of neurogenesis, and patterning of the brain, the most enriched gene category at E10.5 was related to neuronal differentiation (e.g. Ngn1, Blbp, Glast, Tbr2, Bmp4; Figure 1F, Supplementary file 1). However, unexpectedly, the three most enriched gene categories in E8.5 neuroepithelium were related to protein biosynthetic machinery (Figure 1E, Supplementary file 1) and included genes encoding ribosomal proteins (e.g. Rpl24), genes involved in ribosome biogenesis (e.g. Fbl, Dkc1), and translation factors (e.g. Eif4e). Gene set enrichment analysis (GSEA; Subramanian et al., 2005) further confirmed that genes involved in ribosome biogenesis and protein synthesis were significantly enriched in E8.5 progenitors (Figure 1G–I). MA plots (expression ratio [M] vs. average intensity [A], log transformed) provided an overview of the expression changes of individual genes, revealing that the majority of genes encoding ribosomal proteins (Figure 2A), ribosome biogenesis (Figure 2B), and translation factors (Figure 3A), were enriched in E8.5 neuroepithelium. Expression of ribosomal protein or translation factor genes at E10.5 vs. E8.5 showed a positive correlation (R = 0.91 and 0.98 respectively; Figure 1J,K), indicating that despite downregulation of most ribosomal and translation factor genes, their stoichiometry remained similar at the two ages. Expression levels of differentially expressed genes at E10.5 vs. E8.5 also showed a positive correlation (R = 0.82, Figure 1L). There was no correlation between the average expression levels of genes and their fold changes between the two ages (R = −0.1696). Collectively, our data provide transcriptomic signatures of developing forebrain precursors and uncover an overall downregulation of genes encoding protein biosynthetic machinery during the inception of the mammalian forebrain.

Ribosome biogenesis decreases from E8.5 to E10.5.

(A, B) MA plot displaying genes encoding ribosomal proteins (A), ribosome biogenesis factors (B). Each dot represents a single gene. Red dots denote differentially expressed genes as identified by Cuffdiff (q < 0.05). Genes below blue line (y = 0) are enriched in E8.5. (C) Immunohistochemistry of the nucleolar protein Fibrillarin (green) in E8.5, E10.5 and E14.5 neuroepithelium. Scale bar = 20 μm. (D) Example of z-stack image of Fibrillarin staining (left) and 3D reconstruction of nucleoli using Imaris (right). (E) Quantification of nucleolar volume using Imaris. Each data point represents one nucleolus. *p≤0.05, **p≤0.01, Welch’s ANOVA with Games-Howell post-hoc test. Sample size, E8.5: n = 135 from three embryos; E10.5: n = 139 from three embryos; E14.5: n = 146 from three embryos. (F, G) Representative images of fluorescent in situ hybridization of 5.8S pre-rRNA (F, red) and 5.8S total rRNA (G, green). (H) Y10b immunostaining shows higher levels of 5.8S rRNA in E8.5 than E10.5 neuroepithelium. Scale bar = 20 μm. (I) Quantification of 5.8S pre-rRNA signal shows larger nucleolar area in E8.5 compared to E10.5 neuroepithelium. Each data point represents one nucleolus. ****p≤0.0001, Welch’s t-test. Sample size, E8.5: n = 150 from three embryos; E10.5: n = 202 from three embryos. (J) Average nucleolar area in E8.5 vs. E10.5 embryo. **p≤0.01, unpaired t-test, n = 3 embryos. (K, L) RPL11 (K, red) and RPS12 (L, red) were more highly expressed along the apical surface of E8.5 than E10.5 neuroepithelium. Phospho-Vimentin (P-Vim, green) labels dividing progenitors. Scale bar = 20 μm. (M) Immuoblotting shows similar expression of RPL10A between E8.5 and E10.5. (N) Representative images of TEM in neuroepithelial cells at E8.5 and R10.5. (O–Q) Quantification of TEM ribosomal number per standardized field of view (FOV), 78,736 nm2, (O), ribosomal density in cytoplasm (P), and percent of the standard FOV occupied by membrane-bound organelles (Q). *p≤0.05, **p≤0.01, Unpaired t-test.

https://doi.org/10.7554/eLife.36998.004
Downregulation of mTOR signaling pathway and decreased protein synthesis in E10.5 forebrain progenitors.

(A) MA plot displaying genes encoding translation factors. Each dot represents a single gene. Red dots denote differentially expressed genes as identified by Cuffdiff (q<0.05). Genes below blue line (y = 0) are enriched in E8.5. (B) Immunostaining of developing forebrain progenitors shows higher expression of the translation initiation factor EIF3η (green) in E8.5 versus E10.5 neuroepithelium. Scale bar = 20 μm. (C) Immunostaining of developing forebrain neuroepithelium shows decreased phosphorylation of 4E-BP1 (red) in E10.5 neuroepithelium. Scale bar 20 μm. (D) Immunoblotting shows decreased expression and phosphorylation of 4E-BP1 at E10.5. (E) Immunoblotting shows decreased phosphorylation of S6K at E10.5. (F) Immunostaining of developing forebrain neuroepithelium shows decreased phosphorylation of ribosomal protein S6 (red) in E10.5 neuroepithelium. Scale bar 20 μm. (G) Immunoblotting shows decreased phosphorylation of ribosomal protein S6 at E10.5. (H) Schematic of OPP injection into pregnant dams and incorporation into translating polypeptides in the embryos. (I) OPP incorporation assay in E8.5 and E10.5 developing forebrain neuroepithelium. (J) Quantification of OPP fluorescence intensity using Image J shows decreased rate of protein synthesis at E10.5. **p≤0.01 Welch’s t-test. For each age, n = 9 embryos from three litters.

https://doi.org/10.7554/eLife.36998.005

Decreased ribosome biogenesis and protein synthesis in E10.5 neuroepithelium

The higher expression of genes associated with ribosomes, ribosome biogenesis, and protein translation in early E8.5 precursors compared to more committed forebrain progenitors at E10.5 suggested that the protein biosynthetic machinery may be differentially regulated during early forebrain development. Ribosomal RNA (rRNA) transcription and initial assembly of pre-ribosomes occurs in nucleoli. As increased ribosome biogenesis is associated with larger nucleoli (Silvera et al., 2010), nucleolar volume provides a proxy for ribosome biogenesis (Baker, 2013; Sanchez et al., 2016). We visualized nucleoli with Fibrillarin (Figure 2C), acquired z-stack images of the developing neural tissue, and performed 3D-reconstructions of individual nucleoli in neural precursors (Figure 2D). Quantification of nucleolar volume revealed that E8.5 forebrain precursors had larger nucleoli compared to more mature forebrain progenitors at E10.5 (Figure 2E). No further reduction in nucleolar volume was observed between E10.5 neuroepithelial cells and E14.5 radial glial progenitors of the cerebral cortex (Figure 2E), suggesting that the E8.5 to E10.5 transition represents an important regulatory stage for ribosome biogenesis in the early forebrain.

Focusing on E8.5 and E10.5 neuroepithelia, we observed higher 5.8S pre-rRNA levels in E8.5 vs. E10.5 progenitors by fluorescence in situ hybridization (FISH; Figure 2F), and a modest decrease in 5.8S total rRNA at E10.5 (Figure 2G) that was supported by Y10b immunostaining (Figure 2H). Quantification of 5.8S pre-rRNA signal showed larger nucleolar area in E8.5 progenitors (Figure 2I,J), consistent with the fibrillarin quantification (Figure 2E). In agreement with these findings, ribosomal proteins including RPL11 and RPS12, which have important roles in the assembly of ribosomal subunits, were more highly expressed at E8.5 vs. E10.5 (Figure 2K,L; also Chau et al., 2015). On the other hand, expression of RPL10A protein was similar between the two ages (Figure 2M) despite higher Rpl10a RNA expression at E8.5 (FPKM: E8.5=1962.03; E10.5 = 1242.73), suggesting the involvement of post-transcriptional mechanisms. Transmission electron microscopic (TEM) analyses revealed that E8.5 precursors had more ribosomes than E10.5 progenitors per field of view (Figure 2N,O). While ribosome density within the free cytoplasmic space was not different between these two ages (Figure 2P), more E10.5 cytoplasm than E8.5 cytoplasm was occupied by other organelles including endoplasmic reticulum, mitochondria, and Golgi (Figure 2Q), indicating an overall shift in organelle landscape at this age.

Gene expression analyses demonstrated the parallel downregulation of translational machinery from E8.5 to E10.5 progenitors (Figure 3A), including decreased expression of eukaryotic initiation factors (eIFs) such as EIF3η (Figure 3B). Activation of the growth promoting mTOR signaling pathway is linked to increased ribosome biogenesis and protein translation, a function mediated by the mTORC1 complex (Laplante and Sabatini, 2012). While Mtor expression itself was not changed from E8.5 to E10.5 (Mtor FPKM: E8.5=21.15; E10.5 = 25.91), components of the mTOR signaling pathway were differentially expressed and/or activated at these two ages (Supplementary file 1). For example, 4EBP1, a direct target of mTOR, showed increases in both expression and phosphorylation in the E8.5 neuroepithelium (Figure 3C,D). S6K1, a direct mTORC1 target that was similarly expressed at the two ages was also more highly phosphorylated at E8.5 compared to E10.5 (Figure 3E). Finally, S6 ribosomal protein, a substrate of S6K1, was more highly phosphorylated at E8.5 (Figure 3F,G). Taken together, these data demonstrate differential mTOR pathway activation in E8.5 compared to E10.5 neuroepithelium.

E8.5 neural progenitors showed higher 35S-methionine incorporation in vitro compared to E10.5 progenitors (counts per million cells, shown as E8.5 fold change normalized to E10.5 progenitors: Expt. 1 = 2.0 fold; Expt. 2 = 1.4 fold; Expt. 3 = 1.1 fold), indicative of a higher protein synthesis rate in the younger forebrain progenitor cells. We next visualized actively elongating nascent polypeptides in vivo at the single-cell level using O-propargyl-puromycin (OPP; Liu et al., 2012) delivered maternally by intraperitoneal injection (Figure 3H). OPP incorporation was higher in E8.5 compared to E10.5 neuroepithelial cell bodies (Figure 3I,J), consistent with their larger nucleolar volumes and higher 35S-methionine incorporation. Collectively, these data demonstrate that presumptive forebrain progenitors have higher levels of ribosome biogenesis and protein synthesis compared to more mature progenitor cells of the developing forebrain.

Downregulation of protein biosynthetic machinery matches AF and CSF proteomes

The early developing forebrain is bathed first by amniotic fluid (AF) and following neural tube closure, by cerebrospinal fluid (CSF). As neural progenitors can release signaling factors and membrane particles directly into the CSF (Arbeille et al., 2015; Marzesco et al., 2005), we tested the extent to which the changes observed in the forebrain transcriptome (Figure 1) reflected concurrent changes in the AF and CSF proteomes (Chau et al., 2015). We identified 691 proteins present in the AF and CSF that were also expressed by the developing forebrain neuroepithelium. Within this group of 691 proteins, the availability of 493 proteins matched gene expression patterns observed in the forebrain tissue: 395 proteins were enriched in E8.5 AF and more highly expressed by E8.5 neuroepithelium, Figure 4A, lower left quadrant; 98 proteins were enriched in E10.5 CSF and more highly expressed by E10.5 neuroepithelium, Figure 4A, upper right quadrant. Gene ontology analysis showed that, among proteins and genes enriched in E8.5 AF and neuroepithelium, the most highly represented functional category was ribosomes/translation (Figure 4B). Further analysis revealed that nearly all ribosomal proteins and translation factors were enriched in both E8.5 AF and E8.5 neuroepithelium (Figure 4C,D), supporting the model that these fluid proteins can originate in the forebrain tissue. Not only were ribosomal proteins less abundant in E10.5 CSF, but many were no longer detected therein (Figure 4E; Chau et al., 2015). Together, these data demonstrate that the changes in the AF and CSF proteomes during early forebrain development match the downregulation of protein biosynthetic machinery in the adjacent neuroepithelium, thereby providing a developmental biomarker signature of concurrent cell biological changes in the developing forebrain.

Downregulation of protein biosynthetic machinery during early forebrain development matches the AF and CSF proteomes.

(A) Plot showing all proteins/genes that are detected in both AF/CSF and the neighboring neuroepithelium. Each dot represents a single protein/gene. Red dots denote differentially expressed genes between E8.5 and E10.5 epithelium (q<0.05). Genes left of x = 0 were enriched in E8.5 epithelium whereas proteins below y = 0 were enriched in E8.5 AF. Therefore, genes/proteins in lower left quadrant (shaded) were enriched in both E8.5 epithelium and AF. MS = mass spectrometry. (B) Functional annotation clustering of genes/proteins enriched in both E8.5 epithelium and AF (genes/proteins in shaded quadrant in (A)) shows that ribosomes/translation is the most overrepresented category. (C, D) Comparison of AF/CSF proteomes with neuroepithelium transcriptome showed that most ribosomal proteins and translation factors enriched in E8.5 AF were enriched in age-matched epithelium (shaded quadrants). (E) Schematics depicting the specific ribosomal protein subunits that were detected in E8.5 AF (left) and E10.5 CSF (right). Subunits with blue and orange were detected in both fluid and tissue, whereas those in orange were only detected in tissue.

https://doi.org/10.7554/eLife.36998.006

MYC modulates ribosome biogenesis in developing forebrain

In other cell types, the transcription factor MYC regulates genes encoding ribosomal proteins, proteins involved in ribosome biogenesis, and translation initiation and elongation factors (van Riggelen et al., 2010). Analyses of differentially expressed transcription factors between E8.5 and E10.5 neuroepithelium revealed that Myc expression was approximately ten-fold higher in E8.5 neuroepithelium (Figure 5A; Myc FPKM: E8.5=28.73, E10.5 = 2.76), suggesting MYC as a candidate regulator of ribosome biogenesis in the developing forebrain. There was no reciprocal compensatory suppression of Mycn or Mycl (Mycn FPKM: E8.5=28.03, E10.5 = 20.91; Mycl FPKM: E8.5=6.58, E10.5 = 10.34). We confirmed the high level of MYC expression in E8.5 neuroepithelium and its decreased expression in E10.5 neuroepithelium (Figure 5B, Figure 5—figure supplement 1A–C; see also Shannon et al., 2018). Once downregulated in E10.5 neural progenitors, MYC expression remained low throughout cerebral cortical development (Figure 5B).

Figure 5 with 1 supplement see all
MYC modulates ribosome biogenesis in the developing forebrain.

(A) MA plot displaying genes encoding transcription factors in E8.5 and E10.5 neuroepithelium. Each dot represents a single gene. Red dots denote differentially expressed genes identified by Cuffdiff (q<0.05). Genes below blue line (y = 0) enriched in E8.5. Myc (arrow) expression is ~10 fold higher in E8.5 epithelium (FPKM: E8.5=28.73, E10.5 = 2.76). (B) MYC expression was enriched in E8.5 neuroepithelium. Once downregulated at E10.5, MYC expression remained low throughout cortical development. Scale bar = 20 μm. (C, D) GSEA of E8.5 versus E10.5 neuroepithelium for gene sets containing genes up-regulated by MYC and whose promoters are bound by MYC (C), and E-box containing MYC target genes (D). Broad Institute Molecular Signatures Database Identifiers: DANG_MYC_TARGETS_UP (C), BENPORATH_MYC_TARGETS_WITH _EBOX (D). Each line represents a single gene in the gene set; genes on the right side enriched in E8.5. (E) Quantification of nucleolar volume of E8.5 embryos treated with vehicle control or KJ-Pyr-9 for 24 hr. Each data point represents one nucleolus. ***p≤0.001, Welch’s t-test. Sample size, vehicle: n = 140 from three embryos; KJ-Pyr-9: n = 140 from three embryos. (F) Quantification of nucleolar volume of Myc-/- compared to controls (wild type and heterozygous littermates) in E8.5 neuroepithelium. *p≤0.05 Unpaired t-test. Sample size, controls: n = 238 from five embryos; Myc-/-: n = 97 from two embryos. (G) Immunostaining shows overexpression of MYC (red) in the developing cortex of E12.5 MYC-OE (right) embryos from the Nestin-cre x StopFLMYC cross. TUJ1 (green) staining labels neurons. (H) Functional annotation clustering of the 105 MYC-OE enriched genes shows overrepresentation of genes encoding ribosome constituents. The top five enriched functional clusters are shown. (I, J) GSEA of WT versus MYC-OE apical progenitors for gene sets involved in ribosome biogenesis. Broad Institute Molecular Signatures Database Identifiers: KEGG_RIBOSOME (I), and GO_RIBOSOME_ BIOGENESIS (J). Each line represents a single gene in the gene set, genes on the right side are enriched in MYC-OE. (K) Heatmap of the 43 ribosomal protein genes that are differentially expressed between MYC-OE and WT apical progenitors (* q < 0.3, **q < 0.1). All ribosomal proteins are more highly expressed in MYC-OE. Red and green indicate relatively higher and lower expression, with gene FPKM values log2 transformed. (L) MA plot displaying genes encoding ribosomal proteins in E13.5 apical progenitors. Each dot represents a single gene. Red dots denote differentially expressed genes as identified by Cuffdiff (q<0.1). Genes above blue line (y = 0) are enriched in MYC-OE. (M) Quantification of nucleolar volume of WT and MYC-OE (Foxg1-cre driven) forebrain progenitors at E11.5. Each data point represents one nucleolus. ****p≤0.0001, Welch’s t-test. Sample size, WT: n = 194 from four embryos; MYC-OE: n = 144 from three embryos. (N) Quantification of nucleolar volume of WT and MYC-OE (Nestin-cre driven) apical progenitors at E13.5. Each data point represents one nucleolus. ***p≤0.001, Welch’s t-test. Sample size, WT: n = 234 from five embryos; MYC-OE: n = 248 from five embryos.

https://doi.org/10.7554/eLife.36998.007

GSEA demonstrated that many known MYC target genes were enriched in E8.5 compared to E10.5 neuroepithelium (Figure 5C,D; Ben-Porath et al., 2008; Zeller et al., 2003), and some of these target genes were associated with ribosome biogenesis and translation (e.g. Ncl, Rps13, and Eif4e). To test if interfering with MYC activity regulates ribosome biogenesis, we exposed wild type embryos to the MYC inhibitor, KJ-Pyr-9 (Hart et al., 2014) in utero, and observed smaller nucleoli compared to vehicle-injected controls (Figure 5E). In agreement with previous studies (Davis et al., 1993; Zinin et al., 2014), we confirmed that Myc-deficient embryos showed a triad of developmental defects including smaller size, neural tube closure defects, and developmental delay (Figure 5—figure supplement 1D). Nucleolar volume was also decreased in Myc-deficient embryos compared to developmentally stage-matched controls (Figure 5F).

MYC has important roles in cell cycle regulation (Dang, 2013). Therefore, its rapid downregulation by E10.5 was unexpected given that E10.5 represents a stage of continued progenitor proliferation and the start of forebrain neurogenesis. To determine the consequences of persistent MYC expression on cerebral cortical development, we genetically forced MYC expression by crossing StopFLMYC mice (Calado et al., 2012) with Foxg1-Cre (Hébert and McConnell, 2000) or Nestin-Cre (Tronche et al., 1999) mice (Figure 5G, Figure 5—figure supplement 1E–G). We purified Pax6-positive cortical progenitors at E13.5 (from Nestin-Cre cross, Figure 5—figure supplement 1H), and analyzed gene expression by RNA-seq. We identified 135 differentially expressed genes between WT and MYC-overexpressing (MYC-OE) embryos (q < 0.1), with 105 genes activated and 30 genes repressed in the MYC-OE progenitors (Figure 5—figure supplement 1I, Supplementary file 2). A cross-comparison between the 105 MYC activated genes with our early E8.5-E10.5 RNA-seq dataset (Figure 1) revealed 53 genes that were also enriched in E8.5 progenitors when MYC expression is naturally high (Supplementary file 2). Functional annotation clustering using DAVID revealed ribosomes as the most enriched gene category among the MYC-upregulated genes (Figure 5H, Supplementary file 2). GSEA further revealed that genes encoding ribosome components (Figure 5I), genes involved in ribosome biogenesis (Figure 5J), along with other known MYC target genes (Figure 5—figure supplement 1J; Zeller et al., 2003) were upregulated in the MYC-OE. Among the selective subset of ribosomal proteins that were differentially expressed in MYC-OE mice (q < 0.1), all were upregulated (Figure 5K,L, Supplementary file 2), even those subjected to less-stringent statistical parameters (q < 0.3, Figure 5K). These gene expression changes were accompanied by a modest increase in ribosome biogenesis in both Foxg1-Cre and Nestin-Cre MYC-OE mice (Figure 5M,N). Despite this upregulation of ribosome biogenesis and the expression of genes encoding translational machinery (Figure 5—figure supplement 1K), changes in protein synthesis at progenitor cell bodies were not consistently observed in either Myc-deficient or MYC-OE studies (data not shown). Taken together, these findings demonstrate that Myc modulates ribosome biogenesis in the developing forebrain, and that additional, as yet unidentified mechanisms participate in the regulation of protein biosynthesis at this developmental stage.

Persistent MYC expression increases progenitor proliferation, leading to macrocephaly

Genes with known functions in regulating cerebral cortical neurogenesis were also upregulated in MYC-OE progenitors including Insulin-like growth factor 2 (Igf2, Figure 6—figure supplement 1A, Supplementary file 2), which is typically not highly expressed by apical progenitors and is instead delivered by the CSF to regulate proliferation of progenitors (Lehtinen et al., 2011), and Insulinoma-Associated 1 (Insm1, Supplementary file 2), which accelerates cortical development by promoting delamination of apical progenitor cells (Farkas et al., 2008; Tavano et al., 2018). The coordinated effects of MYC activation of several of these pathways resulted in a large brain phenotype that emerged by E14.5 in both Foxg1-MYC and Nestin-MYC mice (Figure 6A–C, Figure 6—figure supplement 1B–D), and was well defined by birth in Nestin-MYC mice (Figure 6D–G). No viable pups were recovered from the Foxg1-Cre cross (eight litters examined), indicating embryonic lethality between E14.5 and birth. This outcome may be due to the combinatorial effects of MYC overexpression and Foxg1 heterozygosity (Hébert and McConnell, 2000), perhaps in tissues outside the brain. No differences in body weight were observed at P0 in Nestin-MYC mice (body weight [g]±SEM: WT = 1.35 ± 0.03, n = 16; MYC-OE = 1.33 ± 0.02, n = 15; unpaired t-test, p=0.57). By P8, MYC-OE and control brains were similar in size (brain weight [g]±SEM: WT = 0.38 ± 0.01, n = 9; MYC-OE = 0.39 ± 0.02, n = 6; p=0.72, unpaired t-test). However the MYC-OE mice had much smaller body size (body weight [g]±SEM: WT = 4.86 ± 0.09, n = 13; MYC-OE = 3.20 ± 0.22, n = 9; p<0.0001, Welch’s t-test), leading to a sustained difference in their brain-body ratio (brain weight/body weight: WT: 0.081 ± 0.001, n = 9; MYC-OE: 0.127 ± 0.006, n = 6; p=0.0003, Welch’s t-test).

Figure 6 with 1 supplement see all
Persistent MYC expression in cortical progenitors leads to macrocephaly.

(A) Representative images of E14.5 brains from WT and MYC-OE from the Nestin-cre x StopFLMYC cross. Scale bar = 2 mm. (B) Quantification of E14.5 cortical length (olfactory bulb excluded). *p≤0.05, unpaired t-test, WT: n = 6 from two litters, MYC-OE: n = 10 embryos from two litters. (C) Quantification of E14.5 cortical area. Cortical area of one hemisphere was measured (olfactory bulb excluded). p>0.05, unpaired t-test, WT: n = 6 from two litters, MYC-OE: n = 10 embryos from two litters. (D) Representative images of P0 brains from WT and MYC-OE from the Nestin-cre x StopFLMYC cross. Scale bar = 2 mm. (E) Quantification of P0 brain weight. Olfactory bulb, medulla and pons were excluded from measurements. ****p≤0.0001, unpaired t-test, No outliers, WT: n = 10 pups from three litters, MYC-OE: n = 12 pups from three litters. (F) Quantification of P0 cortical length as in (B). ****p≤0.0001, unpaired t-test, outlier excluded by ROUT method, WT: n = 10 pups from three litters, MYC-OE: n = 11 pups from three litters. (G) Quantification of P0 cortical area as in (C). ****p≤0.0001, Welch’s t-test, outlier excluded by ROUT method, WT: n = 10 pups from three litters, MYC-OE: n = 11 pups from three litters. (H) Percent PAX6-positive progenitors that were also BrdU-positive after a 2 hr BrdU pulse at E15.5. *p≤0.05, Welch’s t-test, n = 5 embryos from three litters. (I) Representative H and E staining of WT and MYC-OE forebrain at P0. (J) Quantification of cortical thickness of P0 cortex. Thickness is measured from the ventricular surface to the pial surface in the dorsal-lateral cortex. p>0.05, unpaired t-test, n = 6 pups from five litters. (K) MYC-OE had increased number of CUX1-positive upper layer neurons at P0. **p≤0.01; paired t-test, n = 4 litters, 1–2 pairs of embryos per litter were quantified. (L) Examples of 100 μm wide cortical columns at P0 used for cell counting. CUX1: upper layer neurons (red), CTIP2: lower layer neurons (green).

https://doi.org/10.7554/eLife.36998.009

While no tumors were observed at the ages examined in this study, histological analyses suggested that MYC-OE by the Nestin promoter increased the size of the entire brain. A two-hour BrdU pulse delivered at E15.5 showed a larger proportion of Pax6-positive apical progenitors in S-phase in MYC-OE mice (Figure 6H), contributing to increased cortical thickness in MYC-OE mice by birth (Figure 6I,J). MYC-OE cortices had increased Cux1-positive staining cells destined for the upper layers of the cerebral cortex (Figure 6K,L), which contributed to the increased overall number of cells in the cerebral cortex (cell number±SEM: WT: 2,348 ± 199.3; MYC-OE: 2,518 ± 174.3, n = 4, p=0.07, paired t-test). On the other hand, no difference was observed in the number of Ctip2-positive lower layer neurons (cell number±SEM: WT: 651.5 ± 35.4; MYC-OE: 639 ± 34.9, n = 4, p=0.82, paired t-test). Together, our findings support the model that MYC overexpression in the Nestin lineage affects multiple pathways and that their convergence influences the development of the brain and the entire organism.

Discussion

Our study reveals major changes in expression of the protein biosynthetic pathway during early specification of the mammalian forebrain. This work (1) demonstrates that enhanced biogenesis of ribosomes and protein synthetic machinery serve as transcriptional and cell biological signatures defining early forebrain precursor cells; (2) reveals that the changing proteomes of AF and CSF provide a biomarker signature that matches the concurrent, normal development of the adjacent forebrain; (3) identifies MYC as a contributor to the regulation of ribosome biogenesis in the developing forebrain; and (4) shows that persistent MYC expression leads to increased ribosome biogenesis, enhanced cortical progenitor proliferation, and macrocephaly. We conclude that, as in other stem cells, neural progenitor cells dynamically regulate protein biosynthetic machinery to meet their changing needs, and that this process is regulated in part by MYC.

The DNA transcriptome is an essential starting point for our understanding of tissue regionalization, patterning, and individual cell identities in the mammalian central nervous system. Nevertheless, not all mRNAs are selected for protein translation, and our discovery of temporal regulation of the protein biosynthetic machinery during early specification of the forebrain uncovers a new layer of regulation fundamental to the early construction of the brain. Regulation of the protein biosynthetic machinery provides a tunable molecular program harnessed by cells to guide transitions between stem cell states (DeBoer et al., 2013; Fujii et al., 2017; Khajuria et al., 2018; Kraushar et al., 2016; Sanchez et al., 2016; Scognamiglio et al., 2016). Cell cycle in the forebrain lengthens over the course of development (Caviness and Takahashi, 1995). As such, the higher rates of ribosome biogenesis and protein synthesis observed in neurectodermal precursors relative to post-neurulation progenitors are consistent with a model in which rapidly dividing cells synthesize more proteins to support their proliferation (Buszczak et al., 2014). Genes such as Pelo and Abce1 are downregulated in E10.5 progenitors (Supplementary file 1), suggesting that additional levels of translational control including ribosome recycling may be engaged during this developmental time window (Dever and Green, 2012).

Disruptions in ribosome structure and function are linked to a number of genetically inherited ribosomopathies such as Diamond-Blackfan anemia (Boria et al., 2010; Choesmel et al., 2007; Ebert and Lipton, 2011). Nucleolar size, ribosome biogenesis, and protein translation have been implicated in aging and longevity (Buchwalter and Hetzer, 2017; Tiku et al., 2016). In the central nervous system, disruptions in mRNA processing and translation can eventually impair the form and function of delicate neural circuitry. Changes in mRNA binding proteins are linked to neurodevelopmental disorders including autism spectrum disorder (Kraushar et al., 2014; Popovitchenko et al., 2016), and stem cell-derived neural progenitors from schizophrenia patients have altered levels of protein synthesis (Topol et al., 2015). During later stages of neurogenesis in the cerebral cortex, subcellular transport of mRNA by binding proteins including FMRP (Fragile-X mental retardation protein) ferry mRNA to sites of local translation in more polarized cells (Kwan et al., 2012; Pilaz et al., 2016; Pilaz and Silver, 2017). It is tempting to speculate that in the developing forebrain, the assembly of specialized ribosomes could enable unique or localized translation in developing precursor cells, fine-tuning cellular identities and tailoring individualized developmental programs as in other tissues (Bortoluzzi et al., 2001; Fujii et al., 2017; Shi et al., 2017; Simsek et al., 2017).

While ribosomal protein expression and ribosome biogenesis decrease as the embryo develops, we did not observed any difference in cytoplasmic ribosome density from EM analysis (Figure 2). This might be due to the long half-life of ribosomes (Hirsch and Hiatt, 1966; Nikolov et al., 1983), and thus ribosomes generated earlier at E8.5 would likely still be present at E10.5. Using OPP and methionine incorporation, we provided evidence that protein synthesis is also downregulated as the embryo develops. However, we do not know whether differential protein synthesis is driven by changes in ribosome biogenesis or by overall changes in transcription dynamics. Furthermore, OPP was administered intraperitoneally into the pregnant dams, and it is possible that availability of OPP to the embryonic progenitors might be different between E8.5 and E10.5. Indeed, increased ribosomal protein expression does not always result in increased translation because not all ribosomal proteins are associated with polysomes (Kraushar et al., 2015). It is possible that some ribosomal proteins perform extraribosomal functions independent of translation (Warner and McIntosh, 2009; Zhou et al., 2015). For instance, RPL11 is recruited to the promoter regions of p53 target genes during nucleolar stress to promote p53 transcriptional activity (Mahata et al., 2012). Therefore, additional evidence is needed to confirm that changes in ribosome biogenesis directly cause differential protein synthesis in the early neural progenitors.

Neural progenitors depend on their adjacent fluid environment for appropriate fluid pressure and instructive signals (Lun et al., 2015). Developing neural tissue also releases membrane bound vesicles into the adjacent fluid environment (Cossetti et al., 2014; Marzesco et al., 2005). We found that the protein biosynthetic changes occurring in the forebrain neuroepithelium were reflected in the proteomic content of the adjacent AF and CSF (Figure 4). The CSF is commonly sampled for biomarkers of neurologic diseases. Our data demonstrate that during early forebrain development, the proteomic signature of the early brain fluids provides a biomarker signature of the normal, healthy forebrain, opening a new ‘window’ into this stage of early brain development. Whether the ribosomal and translational machinery found in the AF and CSF are equipped to actively synthesize proteins within the fluid environment remains to be elucidated. Alternatively, the fluids might serve as a channel for intercellular transfer of ribosomes and other proteins (Cossetti et al., 2014; Court et al., 2008). Future studies will also reveal whether maturation-associated release of protein biosynthetic machinery into the developing brain fluids is an active or passive process, and whether this process shares features with membrane shedding that occurs in other cell types, such as at the maturing red blood-cell surface (Gautier et al., 2016).

The swift downregulation of MYC following neurulation could be due to chromatin modifications, epigenetic mechanisms, and/or inhibition of RNA polymerase II elongation. Myc-deficiency (Kerosuo and Bronner, 2016) as well as ground-level changes in DNA methylation, histone modifications, and nucleosome positioning are associated with NTD (reviewed in Greene and Copp, 2014; Wilde et al., 2014). Cross-referencing our data (Figure 1) with NTD Wiki, a repository of genes required for neurulation (www.ntdwiki.wikispaces.com), revealed that a number of MYC targets are associated with NTD (data not shown). Complex gene-environment interactions have long been appreciated to underlie NTD. Despite modern successes in reducing the incidence of NTD by dietary fortification (e.g. folate) and increased awareness of adverse consequences of maternal exposures (e.g. alcohol and drug use) on the developing fetus, NTD continue to represent one of the most common birth defects worldwide (Wallingford et al., 2013). Neurulation varies along the anterior-posterior axis, and specific cell types (e.g. hinge points, neural fold cells) have distinct roles in this process (Massarwa et al., 2014). Thus, while our study investigated anterior forebrain development, variation in expression of the protein biosynthetic machinery along the anterior-posterior and dorsal-ventral/medial-lateral axes could differentially affect neurulation along the entire body axis. Overall, the identification of molecular pathways regulating protein biosynthetic machinery during neurulation may provide new opportunities to seek answers to these complex conditions.

Aberrant regulation of the signaling pathways examined in this study in cortical progenitors are associated with cortical overgrowth syndromes such as hemimegalencephaly, a brain malformation characterized by unilateral enlargement of one hemisphere (D'Gama et al., 2017; Poduri et al., 2012). Increased MYC expression has been reported in hemimegalencephaly (Yu et al., 2005), though to our knowledge, mutations in MYC itself have not been shown to drive the pathogenesis of this malformation.

While perhaps best known for its role as an oncogene, we did not observe any cortical tumors in Nestin:MYC brains. Context-dependent effects of MYC have been reported, with age- and tissue-dependent effects on cellular phenotypes including proliferation and cell growth (Gabay et al., 2014; Zinin et al., 2014). The tumorigenic consequences of persistent MYC expression of this model emerged later in adult mice as choroid plexus carcinoma and ciliary body medulloepithelioma (Shannon et al., 2018), exposing the select vulnerability of certain subtypes of epithelial cells in the Nestin lineage to tumorigenesis. Such selectivity of MYC-associated pathologies may be determined by the epigenetic landscape of differentiated cells in adult tissues. MYC may act as a universal amplifier of expressed genes, promoting proliferation in already dividing cells (Lin et al., 2012; Nie et al., 2012). However, in more differentiated cells, target genes may be confined to heterochromatin and inaccessible to MYC (Kress et al., 2015). Certain cell types may also require a genetic double-hit such as concomitant p53-deficiency in the cortex (Momota et al., 2008), or particular gene-environment triggers, for transformation.

Overall, cellular identity and health reflect the net equation between a cell’s transcriptional and translational output (Buszczak et al., 2014; Fujii et al., 2017; Holmberg and Perlmann, 2012; Khajuria et al., 2018; Sanchez et al., 2016). These processes require multiple regulatory steps that are vulnerable to disruptions accumulating from cell-intrinsic genetic programs, and/or cell-extrinsic environmental cues. In the developing brain, environmental signals can entail disturbances of local gradients diffusing through tissues (e.g. Toyoda et al., 2010) or altered delivery of growth-promoting factors by the adjacent AF or CSF (Chau et al., 2015; Lehtinen et al., 2011). All of these signaling activities are susceptible to exogenous maternal exposures including illness, substance abuse, and environmental toxins. Thus, our findings provide a new paradigm for understanding brain development through investigation of molecular pathways regulating the biosynthetic machinery in forebrain progenitors.

Materials and methods

Key resources table
Reagent type
(species) or resource
DesignationSource or referenceIdentifiersAdditional information
Strain, strain background (Mus Musculus)Gt(ROSA)26Sortm13(CAG-MYC,-CD2*)Rsky(referred as StopFLMYC)The Jackson Laboratory
(Bar Harbor, ME)
MGI:5444670Maintained on a C57Bl/6 background
Strain, strain background (Mus Musculus)Tg(Nes-cre)1Kln (referred as Nestin-cre)The Jackson LaboratoryMGI:2176173Maintained on a C57Bl/6 background
Strain, strain background (Mus Musculus)Foxg1tm1(cre)Skm(referred as Foxg1-cre)The Jackson LaboratoryMGI:1932522Maintained on a C57Bl/6 background
Strain, strain background (Mus Musculus)Myc-deficient mice (c-myc-/-)Provided by Troy BaudinoBaudino et al. (2002)Maintained on a C57Bl/6 background
Strain, strain background (Mus Musculus)CD-1 IGS Mouse (referred as CD-1)Charles River
(Wilmington, MA)
Strain code: 022Wildtype timed pregnant mice
AntibodyRabbit anti-4E-BP1Cell Signaling
(Danvers, MA)
96441:1000
AntibodyMouse anti-5.8S ribosomal RNA [Y10B]Abcam
(United Kingdom)
ab1711191:50; antigen retrieval with steaming in citric acid
AntibodyMouse anti-ACTBCell Signaling122621:2000
AntibodyRat anti-BrdUBiorad (Hercules, CA)MCA20601:200; antigen retrieval with steaming in citric acid
AntibodyRabbit anti-MYCAbcamab320721:100 for IHC, antigen retrieval with steaming in citric acid; 1:2000 for WB
AntibodyRat anti-CTIP2Abcamab184651:200
AntibodyRabbit anti-CUX1Santa Cruz Biotechnology
(Dallas, TX)
sc130241:200
AntibodyMouse anti-EIF3ηSanta Cruz Biotechnologysc1372141:100; antigen retrieval with steaming in citric acid
AntibodyMouse anti-FibrillarinAbcamab45661:250; antigen retrieval with steaming in citric acid
AntibodyMouse anti-GAPDHCell Signaling971661:1000
AntibodyRabbit anti-p4E-BP1Cell Signaling28551:200 for IHC; 1:1000 for WB
AntibodyRabbit anti-PAX6Biolegend
(San Diego, CA)
9013011:100; antigen retrieval with steaming in citric acid; 1:1000 for FACS
AntibodyRabbit anti-pS6Cell Signaling53641:200 for IHC; 1:1000 for WB
AntibodyRabbit anti-pS6KCell Signaling92341:1000
AntibodyMouse anti-pVimentinEnzo Bioscience
 (Farmingdale, NY)
ADI-KAM-CC249-E1:400
AntibodyMouse anti-RPL10ANovusbio (Littleton, CO)H00004736-M011:500
AntibodyRabbit anti-RPL11Santa Cruz Biotechnologysc503631:50
AntibodyRabbit anti-RPS12Proteintech
(Chicago, IL)
16490–1-AP1:50
AntibodyRabbit anti-S6Cell Signaling22171:1000
AntibodyRabbit anti-S6KCell Signaling92021:1000
AntibodyMouse anti-TUJ1Biolegend8012021:100 for IHC; 1:1000 for FACS
AntibodyRabbit anti-VinculinCell Signaling139011:1000
Recombinant DNA reagentQuaser 570 coupled 5.8S pre-rRNA FISH probeProvided by Debra Silver1:200
Recombinant DNA reagentQuaser 670 coupled 5.8S total rRNA FISH probeProvided by Debra Silver1:200
Commercial assay or kitRecoverAll Total Nucleic Acid Isolation Kit for FFPEAmbion (Foster City, CA)AM1975Manufacturer's protocol
Commercial assay or kitOvation RNA-Seq System V2Nugen (San Carlos, CA)7102Manufacturer's protocol
Commercial assay or kitOvation Ultralow System V2 1–16Nugen0344Manufacturer's protocol
Commercial assay or kitTruSeq RNA Library Prep Kit v2Illumina (San Diego, CA)RS-122Manufacturer's protocol
Commercial assay or kitRneasy Micro KitQiagen (Germany)74004Manufacturer's protocol
Commercial assay or kitPierce BCA Protein Assay KitThermo Fisher
Scientific (Waltham, MA)
23227
Commercial assay or kitClick-iT plus OPP protein synthesis assay kitThermo Fisher ScientificC10456
Chemical compound, drugO-propargyl-puromycin (OPP)Life Technologies
(Carlsbad, CA)
C10459IP injection, dosage: 50 mg/kg
Chemical compound, drugKJ-Pyr-9Tocris (United Kingdom)5306IP injection, dosage: 10 mg/kg
Chemical compound, drug35S-MethioninePerkin Elmer
 (Waltham, MA)
NEG709A51 μCi
Chemical compound, drug5-Bromo-2′-deoxyuridine (BrdU)Sigma (St. Louis, MO)B5002IP injection, 50 mg/kg
Software, algorithmTopHathttps://ccb.jhu.edu/software/tophat/index.shtmlv2RNAseq analysis
Software, algorithmCufflinkshttp://cole-trapnell-lab.github.io/cufflinks/v2RNAseq analysis
Software, algorithmDAVIDhttps://david.ncifcrf.gov/v6.7, 6.8RNAseq analysis
Software, algorithmGSEAhttp://software.broadinstitute.org/gsea/index.jspv2RNAseq analysis
Software, algorithmR StudioRstudio, Inc.v0.99RNAseq analysis
Software, algorithmPrismGraphPadv7Statistical analysis
Software, algorithmFIJI (Image J)https://fiji.sc/#v1Image analysis
Software, algorithmImarisBitplaneImage analysis

Mice

Timed pregnant CD1 dams were obtained from Charles River Laboratories. Myc-deficient mice (Baudino et al., 2002) were maintained in a C57BL/6J genetic background. StopFLMYC mice (JAX: 020458) were maintained in a C57BL/6J genetic background and crossed with Nestin-cre line (JAX: 003771) or Foxg1-cre line (JAX: 004337) to generate MYC-OE mice, in which human MYC transgene is selectively expressed in neural progenitor cells. All analyses were carried out using male and female mice. All animal experimentation was carried out under protocols approved by the IACUC of Boston Children’s Hospital.

E8.5 and E10.5 forebrain epithelium RNAseq

Forebrain epithelium at E8.5 and E10.5 was dissected as described (Chau et al., 2015). Each sequenced sample comprised forebrain epithelial tissues pooled across one litter. Total RNA was isolated using the RNeasy Micro Kit (Qiagen), converted to cDNA, and preamplified using the Ovation RNA-seq System V2 (NuGEN) following the manufacturer’s instructions. cDNA was converted to Illumina paired-end sequencing libraries following the standard protocol (TruSeq v2) and sequenced on a Illumina HiSeq 2000 instrument to a depth of ~20–60 million pass-filter reads per library, after standard quality control filters. The 50 base pair paired-end reads were mapped to the UCSC mm9 mouse reference genome using TopHat v2, and fragments per kilobase per million reads (FPKM) values were estimated using cufflinks v2, and differentially expressed genes (DEG) were identified using cuffdiff v2 with q value < 0.05 (Trapnell et al., 2012).

FACS of neural progenitors

E13.5 dorsal telencephalon was microdissected, avoiding the lateral ganglionic eminence and structures ventral to it. The cortex was separated from the meninges, and cortices from samples of the same genotype were pooled and sliced into small, uniformly sized pieces. Tissues were digested with 2.5% Trypsin (Invitrogen, Carlsbad, CA), then dissociated into single cells by repeated pipetting. Cells were fixed in 4% paraformaldehyde (PFA), incubated with primary antibodies, and then secondary antibodies. Each step was carried out in 4°C for 30 mins, with rotation. RNAsin (NEB, Ipswich, MA) was added to buffers to prevent RNA degradation (Hrvatin et al., 2014). Cells were sorted using FACS Aria IIU (BD).

Antibodies: Rabbit anti-PAX6 (Biolegend 901301, 1:1000), Mouse anti-TUJ1 (Biolegend 801202, 1:1000)

E13.5 neural progenitor RNAseq

RNA was extracted from sorted neural progenitors using RecoverAll Total Nucleic Acid Isolation Kit (Ambion), then reverse transcribed into cDNA and pre-amplified using Ovation RNA-Seq System V2 (Nugen 7102). Libraries were prepared using Ovation Ultralow System V2 1–16 (Nugen 0344), and sequenced (Illuminia HiSeq 2500) to a depth of ~25–40 million reads per library. The 50 base pair single-end reads were mapped to the UCSC mm10 mouse reference genome using TopHat v2, FPKM values were estimated using cufflinks v2, and DEG were identified using cuffdiff v2 with q value < 0.1 (Trapnell et al., 2012).

RNAseq data analysis

All analyses were performed using genes with FPKM >1, which we considered as the threshold of expression. Hierarchical clustering and heatmaps of differentially expressed genes were generated in R using the heatmap.2 command in ‘gplots’ package, FPKM values were log2 transformed, and centered and scaled by rows for display purposes. Distance was calculated using the ‘Maximum’ method whereas clustering was performed using the ‘Complete’ method. Functional annotation clustering was performed using DAVID v6.7 and v6.8 (https://david.ncifcrf.gov/home.jsp; Huang et al., 2009). Gene set enrichment analysis was performed using GSEA v2 Subramanian et al., 2005), gene sets were obtained from the Broad Institute Molecular Signatures Database (http://software.broadinstitute.org/gsea/msigdb). MA plots were created in R using the ma.plot command in the ‘affy’ package, and MS vs RNAseq plots were created using the plot command.

Tissue processing

Samples were fixed in 4% PFA. For cryosectioning, samples were incubated in the following series of solutions: 10% sucrose, 20% sucrose, 30% sucrose, 1:1 mixture of 30% sucrose and OCT (overnight), and OCT (1 hr). Samples were frozen in OCT. For microtome sectioning, samples were paraffin embedded in the histology core at Beth Israel Deaconess Medical Center.

Immunohistochemistry

Cryosections were blocked and permeabilized (0.3% Triton-X-100 in PBS; 5% serum), incubated in primary antibodies overnight and secondary antibodies for 2 hr. Sections were counterstained with Hoechst 33342 and mounted using Fluoromount-G (SouthernBiotech, Birmingham, AL). The following primary antibodies were used: anti-5.8S rRNA (Y10b; Abcam, ab171119, 1:50), anti-BrdU (Biorad, MCA2060, 1:200), anti-cMYC (Abcam, ab32072, 1:100), anti-CTIP2 (Abcam, ab18465, 1:200), anti-Cux1 (Santa Cruz Biotechnology, sc13024, 1:200), anti-EIF3η (Santa Cruz Biotechnology, sc-137214, 1:100), anti-Fibrillarin (Abcam, ab4566, 1:250), anti-p4E-BP1 (Cell Signaling, 2855, 1:200), anti-Pax6 (Biolegend, 901301, 1:100), anti-pS6 (Cell Signaling, 5364, 1:200), anti-pVimentin (Enzo Bioscience, ADI-KAM-CC249-E, 1:400), anti-Rpl11 (Santa Cruz Biotechnology, sc50363 1:50), anti-Rps12 (Proteintech, 16490–1-AP, 1:50), anti-Tuj1 (Biolegend, 801202, 1:100). Secondary antibodies were selected from the Alexa series (Invitrogen, 1:500). For BrdU, Fibrillarin, Pax6, cMyc, 5.8S rRNA, and EIF3η staining, antigen retrieval/denaturation was performed before the blocking step: A food steamer (Oster 5712) was filled with water and preheated until the chamber was approximately 100°C, sections were immersed in boiling citric acid buffer (10 mM sodium citrate; 0.05% Tween 20; pH = 6) and placed in steamer for 20 min. Sections were cooled to room temperature. H&E staining was carried out according to standard procedures (Shannon et al., 2018).

Immunoblotting

Tissues were homogenized in RIPA buffer supplemented with protease and phosphatase inhibitors. Protein concentration was determined by BCA assay (Thermo Scientific 23227). Samples were denatured in 2% SDS by heating at 95°C for 5 min. Equal amounts of proteins were loaded and separated by electrophoresis in a 4–15% gradient polyacrylamide gel, transferred to nitrocellulose blot (250mA, 1.5 hr), blocked in 5% BSA or milk, incubated with primary antibodies overnight at 4°C followed by HRP conjugated secondary antibodies (1:5000) for 1 hr, and visualized with ECL substrate. For phosphorylation analysis, the phospho-proteins were probed first, and then blots were stripped (Thermo Scientific 21059) and reprobed for total proteins. The following primary antibodies were used: anti-4E-BP1 (Cell Signaling, 9644, 1:1000), anti-ACTB (Cell Signaling, 12262, 1:2000), anti-cMYC (Abcam, ab32072, 1:2000), anti-GAPDH (Cell Signaling, 97166, 1:1000), anti-p4E-BP1 (Cell Signaling, 9459, 1:1000), anti-pS6 (Cell Signaling, 5364, 1:1000), anti-pS6K (Cell Signaling, 9234, 1:1000), anti-RPL10A (Novusbio, H00004736-M01, 1:500), anti-S6 (Cell Signaling, 2217, 1:1000), anti-S6K (Cell Signaling, 9202, 1:1000), anti-Vinculin (Cell Signaling, 13901, 1:1000).

Fluorescent in situ hybridization

Cryosections were permeabilized in 0.5% Triton-X-100 for 20 min, incubated with probes (Pilaz et al., 2016) overnight at 37°C, counterstained with Hoechst, and mounted using Fluoromount-G (SouthernBiotech). Probes: Quaser 570 coupled 5.8S pre-rRNA, Quaser 670 coupled 5.8S total rRNA.

35S-Methionine labeling

E8.5 and E10.5 forebrain neuroepithelium was dissected as described (Chau et al., 2015) and trypsinized. Cells were serum starved in methionine-free DMEM for 1 hr at 37°C, then incubated with 51 μCi 35S-Methionine (Perkin Elmer NEG709A) at 37°C for an additional hour. Cycloheximide (50 μg/ml) was added to stop translation. 35S-Methionine incorporation was measured using scintillation counter.

Nucleolar volume quantification

Nucleolar volume was quantified according to published methods using Imaris (Bitplane; Baker, 2013; Sanchez et al., 2016; Shannon et al., 2018; Silvera et al., 2010). To ensure fair representation, randomly selected nucleoli were selected for quantification across the image field. When quantifying nucleolar volume embryonically, we specifically quantified cells close to the ventricular surface. Therefore, at E14.5 the quantified cells should represent radial glia in the ventricular zone. For relative nucleolar volume, each volume value was normalized to the average nucleolar volume of the controls in the corresponding litter. 5.8S rRNA signal area (nucleolar area) was quantified using FIJI (Image J).

Neuroepithelium OPP quantification

OPP quantification was performed as described by Liu et al. (2012). Pregnant dams received intraperitoneal OPP injections (50 mg/kg OPP; Life Technologies). One hour later, developing tissues were obtained and sectioned to a thickness of 7 μm using a cryostat. OPP signals were detected using the Click-iT plus OPP protein synthesis assay kits (Life Technologies) according the manufacturer’s suggested procedures. Images were taken at 20X (Zeiss Axio Observer D1 inverted microscope) and fluorescence intensity was quantified using FIJI (ImageJ). For each sample, OPP intensity from six independent regions of interest (185 μm2) along the ventricular surface was measured and averaged.

MYC inhibitor injection

KJ-Pyr-9 (Hart et al., 2014) was dissolved in Tween 80:DMSO:5% dextrose (1:1:8) and injected at a dosage of 10 mg/kg into pregnant dams at E7.5. Samples were collected 24 hr later for analysis.

BrdU cell proliferation assay

BrdU (50 mg/kg) was injected intraperitoneally into pregnant dams 2 hr prior to tissue collection. Brains were cryosectioned (7 μm thickness) and stained with BrdU and Pax6 antibodies. Images were acquired at 20X (Zeiss LSM 700 laser scanning confocal microscope). Cells were counted in a 100 μm wide column in the dorsal-lateral cortex. For each sample, 4–6 sections along the anterior/posterior axis of the forebrain were counted and averaged. The proliferation index was defined as the percentage of Pax6-positive cells that were also BrdU-positive.

P0 cortical neuron counting

14 μm thick cryosections were stained with antibodies, and images were acquired at 20X (Zeiss Axio Observer D1 inverted microscope). Counting was performed using FIJI (Image J) on 100 μm wide columns in the dorsal-lateral cortex in the region just anterior to the hippocampus.

P0 cortical thickness measurement

Measurements were performed on H and E-stained coronal sections. Thickness was defined as the length extending from the ventricular zone up to the pial surface in the dorsal-lateral cortex.

Quantitative RT-PCR

RNA was isolated using Trizol extraction protocol or RecoverAll Total Nucleic Acid Isolation Kit (Ambion), and reverse-transcribed into cDNA. Gene expression was measured by Taqman qPCR (Life Technologies), using Tbp as an internal control.

Transmission electron microscopy

All tissue processing, sectioning, and imaging was carried out at the Conventional Electron Microscopy Facility at Harvard Medical School. E8.5 and E10.5 tissues were fixed in 2.5% Glutaraldehyde/2% Paraformaldehyde in 0.1 M sodium cacodylate buffer (pH 7.4). They were then washed in 0.1M cacodylate buffer and postfixed with 1% Osmiumtetroxide (OsO4)/1.5% Potassiumferrocyanide (KFeCN6) for one hour, washed in water three times and incubated in 1% aqueous uranyl acetate for one hour. This was followed by two washes in water and subsequent dehydration in grades of alcohol (10 min each; 50%, 70%, 90%, 2 × 10 min 100%). Samples were then incubated in propyleneoxide for one hour and infiltrated overnight in a 1:1 mixture of propyleneoxide and TAAB Epon (Marivac Canada Inc. St. Laurent, Canada). The following day, the samples were embedded in TAAB Epon and polymerized at 60 degrees C for 48 hr. Ultrathin sections (about 80 nm) were cut on a Reichert Ultracut-S microtome, and picked up onto copper grids stained with lead citrate. Sections were examined in a JEOL 1200EX Transmission electron microscope or a TecnaiG² Spirit BioTWIN. Images were recorded with an AMT 2 k CCD camera.

Ribosomal quantification was performed using Imaris (Bitplane). For 20 images per individual (N = 3 at each age), ribosomal density was calculated within a 280.5 nm x 280.5 nm box in an inverted color image that contained only cytoplasm and ribosomes (no membrane bound organelles). Ribosomes were counted by the Imaris software using the ‘spots’ tool, with estimated diameter of 250px and with automatic background subtraction ‘on’, and quality above the automatic threshold. The number of ribosomes per field of view (FOV) was calculated by multiplying the above calculate density by the cytoplasmic area. The cytoplasmic area was calculated by creating a hand-drawn surface in Imaris around the free cytoplasmic space in the standard FOV (2692 nm x 1762.6 nm). The % FOV occupied by organelles was calculated by subtracting the free cytoplasmic area from the total area to arrive at the organelle-occupied area.

Statistical analysis

Biological replicates (N) were defined as samples from distinct individuals analyzed either in the same experiment or within multiple experiments. Samples were pooled across multiple litters so as to reduce inter-litter variability. Statistical analyses were performed using Prism seven or R. Outliers were excluded using ROUT method (Q = 1%). Appropriate statistical tests were selected based on the distribution of data, homogeneity of variances, and sample size. F tests or Bartlett’s tests were used to assess homogeneity of variances between data sets. Parametric tests (T test, ANOVA) were used only if data were normally distributed and variances were approximately equal. Otherwise, nonparametric alternatives were chosen. Data are presented as means ± standard errors of the mean (SEMs). Please refer to figure legends for statistical tests used and sample size. P values < 0.05 were considered significant (*p≤0.05, **p≤0.01, ***p≤0.001, ****p≤0.0001)

References

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
  11. 11
  12. 12
  13. 13
  14. 14
  15. 15
  16. 16
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
  22. 22
  23. 23
  24. 24
  25. 25
  26. 26
  27. 27
  28. 28
  29. 29
    Turnover of liver ribosomes in fed and in fasted rats
    1. CA Hirsch
    2. HH Hiatt
    (1966)
    The Journal of Biological Chemistry 241:5936–5940.
  30. 30
  31. 31
  32. 32
  33. 33
  34. 34
  35. 35
  36. 36
  37. 37
  38. 38
  39. 39
  40. 40
  41. 41
  42. 42
  43. 43
  44. 44
  45. 45
  46. 46
  47. 47
  48. 48
  49. 49
  50. 50
  51. 51
  52. 52
  53. 53
  54. 54
  55. 55
  56. 56
  57. 57
  58. 58
  59. 59
  60. 60
  61. 61
  62. 62
  63. 63
  64. 64
  65. 65
  66. 66
  67. 67
  68. 68
  69. 69
  70. 70
  71. 71
  72. 72
  73. 73
  74. 74
  75. 75
  76. 76
  77. 77
  78. 78
  79. 79
  80. 80
  81. 81
  82. 82
  83. 83

Decision letter

  1. Marianne Bronner
    Reviewing Editor; California Institute of Technology, United States

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

[Editors’ note: a previous version of this study was rejected after peer review, but the authors submitted for reconsideration. The first decision letter after peer review is shown below.]

Thank you for submitting your work entitled "Dynamic changes in protein biosynthesis during cell fate restriction of neural progenitors and in tumorigenesis" for consideration by eLife. Your article has been evaluated by a Senior Editor and three reviewers, one of whom is a member of our Board of Reviewing Editors.

Our decision has been reached after consultation between the reviewers. Based on these discussions and the individual reviews below, we regret to inform you that your work will not be considered further for publication in eLife.

The reviewers expressed enthusiasm for the conceptual advance and the timeliness of the studies. They acknowledge how these studies add to emerging work on the importance of translational regulation in the control of mammalian development and new concepts related to the control imposed on cell and tissue formation through specific ribosomal subunits. They also appreciate the variety of state of the art techniques and the initial transcriptional and proteomic experiments. However, in its current state, the paper is quite descriptive, somewhat correlative, the follow up analysis focusing on translation and the role of c-Myc in the process are lacking, and the manuscript is unfocused at the end.

The full reviews are attached below. A synopsis of the major critiques is outlined in the following 4 points.

1) The authors only directly examine whether protein synthesis is decreased post-neurulation (by the incorporation of OPP in Figure 2H) one time in the manuscript. The authors often use instead increased nucleolar volume (as a proxy) and EM analysis. The EM data need annotations and ideally immuno-gold to support the conclusion that punctae are actually ribosomes. It would be ideal if the authors could support their conclusion with standard molecular analysis of biogenesis – for example, are rRNA levels and/or processing different across development – and/or other direct measurements of protein synthesis and functional ribosomes (monosomes vs. polysomes).

2) Myc OE in apical progenitors is employed to support the idea that dynamic regulation of ribosome biogenesis and protein synthesis is needed in neural progenitors over development. However, these data are largely correlative and it remains unclear mechanistically whether alterations in ribosomes and protein synthesis explain the Myc OE phenotypes. Although Myc has been previously linked to ribosome biogenesis, this transcription factor has other functions, for example in associating with and impacting cell cycle regulators. Thus, the described alterations in brain size, Pax6+ S phase cells, and Cux1 number could be due to cell cycle alterations, potentially independent of ribosome regulation. Consistent with this possibility, Myc OE embryonic brains exhibit only a slight increase in ribosome biogenesis and no significant alterations in translation. The JQ1 experiments are problematic in interpretation of a specific effect on Myc, as JQ1 is a broad epigenetic (bromodomain) inhibitor and alterations in a number of epigenetic regulators can lead to neural tube defects. Can the authors build a case that better supports the title of the paper with additional ways of manipulating translational machinery directly to investigate their hypothesis that down-regulation of protein synthesis is needed for neurulation? Further experiments are needed to mechanistically link Myc modulation to ribosome biogenesis and/or to better link ribosome/translational regulation with cell fate restriction, in order to make a more impactful paper.

3) The Myc OE mouse model complements previous LOF findings that Myc controls brain size and progenitors (albeit using a human transgene). The authors conclude that Myc OE increases proliferation and in their title link dynamic changes in protein synthesis to "cell fate restriction." However, Myc OE phenotypes are only reported at later stages with analysis of only one neuronal marker, limiting the strength of the conclusions about cell fate restriction and proliferation. Are all neuronal layers affected? It would be important to demonstrate that Myc OE impacts proliferation and fate restriction at earlier relevant stages, E10.5-E12.5, when expression of Myc and ribosome/translation factors is reduced. Foxg1-cre/myc mice are also used which initiate c-myc expression early at e9.5 but there is little analysis of these embryos. In addition the data for Nestin-cre mice do not support the hypothesis that c-myc is driver of translation.

4) The reviewers would prefer that results related to Myc OE drives tumors in the choroid plexus is removed. These data diffuse the focus of the paper as the phenotypes and biology here are vastly different from early development. More attention on the early stages and requested experiments above would better strengthen the manuscript.

Despite the enthusiasm, the requirements suggested to flesh out the story could take considerable time to complete. If you feel it is possible to address these concerns, we encourage you to resubmit a more complete story for full review at a later date.

Reviewer #1:

This manuscript conceptually builds from the authors previous findings (Chau et al., 2015) that showed that the proteome of the amniotic fluid and early cerebrospinal fluid contains factors that promote placode and neural differentiation, respectively. The idea that arises in the current manuscript is that these factors derive from the developing neuroepithelium and the composition changes over time. Using RNA-seq, the authors profile the emerging forebrain in the mouse at E8.5 and E10.5. An added value of their analysis is the ability to compare the transcriptional datasets with their previous proteome analysis of the amniotic fluid and cerebrospinal fluid to make some inferences about signaling/reception and protein synthesis. The focus of the current manuscript is on the strong expression of ribosomal proteins at E8.5 and their downregulation by E10.5 as neurogenesis proceeds. It explores the role of MYC in this process using loss and gain of Myc expression. Together the manuscript presents novel and convincing data. Ultimately it is still somewhat correlative but the authors have done a nice job in highlighting the overall importance of MYC in the regulation of early development of the neuroepithelium and differentiation. These studies add to emerging work on the importance of translational regulation in the control of mammalian development and new concepts related to the exquisite control imposed on cell and tissue formation through specific ribosomal subunits.

The authors focus in on MYC as a possible upstream regulator. Most of the data in Figure 3 are solid but the JQ1 experiments are problematic in interpretation of a specific effect on Myc, as JQ1 is a broad epigenetic (bromodomain) inhibitor and alterations in a number of epigenetic regulators can lead to neural tube defects.

The results end on somewhat of a tangent and turns toward a cancer phenotype (which would be expected by MYC-OE) within the choroid plexus. Despite leading away from the main point, it is OK to retain this data as it shows a link between developmental and cancer pathways (again, not a surprise). However, since MYC overexpression is persistent through development, the following statement is incorrect or at best misleading: "The appropriate downregulation of Myc after neurulation is essential to prevent early missteps of development that can manifest as later-onset diseases of the nervous system."

Reviewer #2:

Here, Lehtinen and colleagues describe expression of ribosomal components and translation in neuroepithelial progenitors, which they associate with developmental restrictions in cell fate specification. Transcriptome analyses show down-regulation of these components between E8.5 and E10.5, which matches transcriptomes of the choroid plexus. This is parallel with changes in protein synthesis and proposed ribosome biogenesis. The authors suggest that reduced protein synthesis is required for differentiation to occur. They show that expression of Myc, a transcription factor implicated in ribosome biogenesis, correlates with these developmental profiles and demonstrate that in vivo Myc overexpression (Myc OE) perturbs brain development, increasing brain size and affecting neuronal markers. Myc OE in progenitors also drives tumors and alterations in translational machinery in the adult choroid plexus.

The main interesting concept put forth in this paper that ribosome biogenesis is developmentally regulated. Several additional concepts are intriguing: that CSF acquire translational components secreted from the neuroepithelium, and ribosome heterogeneity in progenitors (which has been suggested previously but poorly described). The data are of interest for those investigating neurulation and regulation of progenitor fate. However, in its current state, the paper is quite descriptive and somewhat unfocused, and the concepts described above could be better fleshed out.

1) Myc OE in apical progenitors is employed to support the idea that dynamic regulation of ribosome biogenesis and protein synthesis is needed in neural progenitors over development. However, these data are largely correlative and it remains unclear mechanistically whether alterations in ribosomes and protein synthesis explain the Myc OE phenotypes. Although Myc has been previously linked to ribosome biogenesis, this transcription factor has other functions, for example in associating with and impacting cell cycle regulators. Thus, the described alterations in brain size, Pax6+ S phase cells, and Cux1 number could be due to cell cycle alterations, potentially independent of ribosome regulation. Consistent with this possibility, Myc OE embryonic brains exhibit only a slight increase in ribosome biogenesis and no significant alterations in translation. Can the authors build a case that better supports the title of the paper with additional ways of manipulating translational machinery, for example? Further experiments are needed to mechanistically link Myc modulation to ribosome biogenesis and/or to better link ribosome/translational regulation with cell fate restriction, in order to make a more impactful paper.

2) The Myc OE mouse model complements previous LOF findings that Myc controls brain size and progenitors (albeit using a human transgene). The authors conclude that Myc OE increases proliferation and in their title link dynamic changes in protein synthesis to "cell fate restriction." However Myc OE phenotypes are only reported at later stages with analysis of only one neuronal marker, limiting strength of conclusions about cell fate restriction and proliferation. Are all neuronal layers affected? It would be important to demonstrate that Myc OE impacts proliferation and fate restriction at earlier relevant stages, E10.5-E12.5, when expression of Myc and ribosome/translation factors is reduced.

3) The last two figures of the paper demonstrate Myc OE drives tumors in the choroid plexus and modifies translation markers in these adult stages. While I appreciate this helps support the link between Myc and ribosome regulation, given the phenotypes and biology here are vastly different from early development, these data diffuse the focus of the paper.

4) The conclusion that E8.5 neuroepithelial progenitors have higher ribosome biogenesis than E10.5 is based upon increased nucleolar volume (as a proxy) and EM analysis. The EM data need annotations and ideally immuno-gold to support the conclusion that punctae are actually ribosomes. It would be ideal if the authors could support their conclusion with standard molecular analysis of biogenesis- for example, are rRNA levels and/or processing different across development?

Reviewer #3:

Chau et al. performed transcriptome analysis on forebrain e8.5 and e10.5 neuroepithelium and discovered that the protein biosynthetic machinery is down-regulated after neurulation. The authors then explore if c-myc regulates the expression of ribosomes and other translational machinery genes during neurulation. The work is extensive and the authors use a variety of state of the art techniques. Differential regulation of ribosomes and other translational machinery as a driver of neurulation is novel and could have a broad impact on our understanding of this important and complex morphogenetic process. While the initial transcriptional and proteomic experiments are performed well, the follow up analysis focusing on translation and the role of c-myc in the process are lacking.

The authors only directly examine whether protein synthesis is decreased post-neurulation (by the incorporation of OPP in Figure 2H) one time in the manuscript. The authors often use instead nucleolar data, which is interesting, but it is only correlative. A more direct examination of protein synthesis should also be performed on the c-myc knock-out and c-myc over-expression mice. Additional direct measurements of protein synthesis and functional ribosomes (monosomes vs. polysomes) are also needed in all instances to substantiate their findings.

The necessity of c-myc in coordinating the down-regulation of the protein synthesis machinery is also lacking. The authors use a conventional knock-out when a conditional knock-out might be more informative. The authors use JQ1 as an inhibitor of c-myc but concerns exist about the specificity of JQ1 for only c-myc. The authors could also use more standard and complementary systems that are likely to be more specific to test the necessity of c-myc. Unfortunately without these data, the author's conclusion (subsection “Critical timing of MYC expression during early forebrain development”, last paragraph) is not supported.

The over-expression studies are also lacking. The authors use an innovative c-myc line and two cre drivers (Nestin and Foxg1) but perform a detailed analysis for only the Nestin-cre driver. Unfortunately this does not seem to be the best mouse since c-myc is expression commences days after neurulation is complete. Instead the Foxg1-cre/myc mice seem better since c-myc expression initiates at e9.5 but the necessary studies are not performed. In addition the data for Nestin-cre mice do not support the hypothesis that c-myc is driver of translation and neurulation

C-myc also likely regulates many pathways, just not the protein synthesis machinery. Thus, the authors should also manipulate the translational machinery directly to investigate their hypothesis that down-regulation of protein synthesis is needed for neurulation.

Finally, a more minor concern, but still a possible significant problem, is that the difference in protein machinery expression could vary along the A-P and D-V/M-L axes. Since neurulation varies along the A-P axis and there are specific cell types along the M-L/D-V axis (hinge points, neural fold cells), differences could exist. Some data (Figure 2C) support this possibility. This caveat should be stated explicitly in the manuscript.

[Editors’ note: what now follows is the decision letter after the authors submitted for further consideration.]

Thank you for submitting your article "Downregulation of ribosome biogenesis and protein synthesis during early forebrain development" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by Marianne Bronner as the Senior and Reviewing Editor. The following individual involved in review of your submission has agreed to reveal her identity: Lee Niswander (Reviewer #2).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission. The reviewers agree that the manuscript is much improved but needs a few changes prior to acceptance.

Summary:

In this paper, Chau et al. delve into the important topic of developmental ribosome biogenesis. he authors demonstrate that ribosomal mRNA and proteins are higher at an early stage of forebrain development (E8.5) and decrease by the onset of neurogenesis (E10.5). Moreover, they make the case that c-MYC positively regulates ribosome biogenesis and that a consequence of c-MYC overexpression is an increase neuronal production and macrocephaly.

Essential revisions:

1) Please add to the Discussion some technical caveats of monitoring translation in vivo (some of which were mentioned in the response to reviewers) and as discussed in detail in the next paragraph. Also in the Discussion please stage whether there could be a function of altered ribosome biogenesis independent of modulating translation.

While authors have strong evidence that ribosome signature (and biogenesis) changes during development (as described in a manuscript that was not cited Kraushar et al., 2015, but also in the cited Kraushar et al., 2016), authors do not provide strong evidence that they disrupted dynamics of protein synthesis due to ribosome and/or MYC changes. For example, RPL11 is known to sit at the monosome but not to associate with polysome, while during stress conditions it binds to the DNA promoter of p53 oncogene (Slavov, Cell Reports, 2015; Mahata et al., Ocogene, 2015). In addition, critical experiments testing protein synthesis (e.g. OPP injection or methionine analysis), in current state do not exclude possibility that different transcription dynamics play role in the observed OPP and methionine changes. Here, authors may consider an additional line of evidence – they may want to discuss discrepancies between levels of mRNA expression and matching protein these mRNAs encode (e.g. using qRT-PCR and western).

2) Please remove "protein synthesis" from the title. Since OPP was injected intraperitoneally, the availability of OPP to progenitors may be different between E8.5 and 10.5 (e.g. similar to decrease of a number of proteins in AF and CSF by E10.5). Overall, there is no strong experimental evidence that protein synthesis is downregulated in early progenitors, while it may also be easily upregulated for selective number of other mRNAs whose protein expression was not tested. For example, total levels of ribosomal proteins can decrease but their association with polysomes may increase (Kraushar et al., 2015, 2016). The same concern goes for not seeing consistent, significant differences in OPP incorporation in Myc-/- mice. Thus, it is critical to either remove "protein synthesis" from their title or perform additional experiments (e.g. experiments focusing on protein synthesis where transcription is transiently inhibited at its minimum).

3) Using immunohistochemistry, the authors show significant decrease for a number of ribosome biogenesis players and ribosomal proteins. Is there a ribosome protein that did not change or maybe even increased? These possible additional findings would be in line with EM pictures and EM analysis that clearly show large number of ribosomes in apical progenitors.

https://doi.org/10.7554/eLife.36998.017

Author response

[Editors’ note: the author responses to the first round of peer review follow.]

The reviewers expressed enthusiasm for the conceptual advance and the timeliness of the studies. They acknowledge how these studies add to emerging work on the importance of translational regulation in the control of mammalian development and new concepts related to the control imposed on cell and tissue formation through specific ribosomal subunits. They also appreciate the variety of state of the art techniques and the initial transcriptional and proteomic experiments. However, in its current state, the paper is quite descriptive, somewhat correlative, the follow up analysis focusing on translation and the role of c-Myc in the process are lacking, and the manuscript is unfocused at the end.

The full reviews are attached below. A synopsis of the major critiques is outlined in the following 4 points.

1) The authors only directly examine whether protein synthesis is decreased post-neurulation (by the incorporation of OPP in Figure 2H) one time in the manuscript. The authors often use instead increased nucleolar volume (as a proxy) and EM analysis. The EM data need annotations and ideally immuno-gold to support the conclusion that punctae are actually ribosomes.

In the revised manuscript, we provide larger, annotated, representative EM images of from E8.5 and E10.5 neural progenitor cells (Figure 2K). We quantified the number of ribosomes and other organelles occupying the cytoplasmic space in neural precursors and found that the E8.5 cells contain more ribosomes per defined area of cytoplasm than E10.5 progenitors (Figure 2L), though ribosome density within the free cytoplasmic space, per se, was not different between these two ages (Figure 2M). In addition, more E10.5 cytoplasm than E8.5 cytoplasm was occupied by other organelles including endoplasmic reticulum, mitochondria, and Golgi (Figure 2N), indicating an overall shift in organelle landscape at this age.

We are confident that the punctae in our images are ribosomes, and based on morphological characteristics, are very unlikely to be any other structures. For example glycogen, which we often observe in our EM analyses of other cell types such as choroid plexus epithelial cells, is larger, more irregular in shape, aggregates more, and is typically more electron dense. We consulted with additional experts on ribosomes and EM in the HMS community including:

Maria Ericsson (manager, Harvard Medical School EM Facility, where we perform our EM analyses)

Mark Fleming, MD, DPhil (Boston Children’s Hospital; e.g. Nguyen et al., Science 2017)

Wei-Chung Lee, PhD (HMS/BCH Neurobiology; Lee et al., Nature 2016)

These experts agree that the punctae in our samples are ribosomes. In addition, there was a general consensus that because our samples are so packed with ribosomes, immuno-gold EM would be problematic to interpret since the overall signal would likely be too strong/saturated. Together with our revised representative images and quantification (see above), we hope the reviewers will agree that ribosome biogenesis is decreased in the E8.5 to E10.5 forebrain.

It would be ideal if the authors could support their conclusion with standard molecular analysis of biogenesis – for example, are rRNA levels and/or processing different across development – and/or other direct measurements of protein synthesis and functional ribosomes (monosomes vs. polysomes).

We reached out to Debby Silver (Duke University), to discuss this point and to get her advice on reagents that may work well in the developing forebrain at these early ages. The Silver lab kindly shared FISH probes against 5.8S pre-rRNA and 5.8S total rRNA (Pilaz et al.,), which reveal more rRNA signal in E8.5 than E10.5 progenitors. These new data are shown in new Figure 2F, 2G.

We purchased the Y10b antibody used by many labs in the field (e.g. Kondrashov et al., 2011) and found that it, too, showed stronger immunofluorescence signal in E8.5 progenitors than E10.5 progenitors. These new data are shown in new Figure 2H.

We quantified 35S-methionine incorporation in vitro in forebrain progenitors, and observed higher 35S-methionine incorporation at E8.5 compared to E10.5 (counts per million cells, presented as E8.5 fold change normalized to E10.5 forebrain: Expt. 1 = 2.0-fold; Expt. 2 = 1.4-fold; Expt. 3 = 1.1-fold), indicative of a higher protein synthesis rate in E8.5 forebrain progenitor cells. These new data are described in the last paragraph of the subsection “Decreased ribosome biogenesis and protein synthesis in E10.5 neuroepithelium”. We have also added a new author to our study, Anoop Sendamarai (Fleming lab, BCH/HMS), as we collaborated with a lab with specific expertise in performing this experiment and that had a radioactivity license.

We reached out to experts in the Harvard Medical School community regarding the technical requirements for analyzing monosomes and polysomes. From our pilot dissections and discussions with Drs. Dan Finley and Mark Fleming (e.g. Nguyen et al., Science 2017), we have concluded that these analyses are technically not feasible given the very limited amounts of E8.5 presumptive forebrain that can be micro-dissected across similarly staged embryos and pooled across multiple litters of mice. However, as we provide new data for the rest of the points raised above, we hope that the reviewers agree that our manuscript is strengthened by the inclusion of these new data and that polysome profiling is not necessary to demonstrate that ribosome biogenesis is downregulated from E8.5 to E10.5.

2) Myc OE in apical progenitors is employed to support the idea that dynamic regulation of ribosome biogenesis and protein synthesis is needed in neural progenitors over development. However, these data are largely correlative and it remains unclear mechanistically whether alterations in ribosomes and protein synthesis explain the Myc OE phenotypes. Although Myc has been previously linked to ribosome biogenesis, this transcription factor has other functions, for example in associating with and impacting cell cycle regulators. Thus, the described alterations in brain size, Pax6+ S phase cells, and Cux1 number could be due to cell cycle alterations, potentially independent of ribosome regulation. Consistent with this possibility, Myc OE embryonic brains exhibit only a slight increase in ribosome biogenesis and no significant alterations in translation.

We agree that MYC has many roles in the cells, and that the large brain phenotype observed is very likely due to a combination of MYC’s effects on multiple signaling pathways. For example, MYC overexpression also led to the upregulation of genes involved with cell proliferation. For example, we observed upregulation of Igf2 and Insm1 in MYC-expressing progenitors. IGF2 regulates proliferation of cortical progenitor cells (e.g. Lehtinen et al., 2011), and Insm1 is a transcription factor with roles in promoting the delamination of apical progenitors cells, thereby accelerating cortical development by promoting the formation of basal progenitors (e.g. Farkas et al., 2008; Tavano et al., 2018). We clarify this point in the first paragraph of the subsection “Persistent MYC expression increases progenitor proliferation, leading to macrocephaly”.

In the revised manuscript, we have added new experiments quantifying ribosome biogenesis (via Fibrillarin staining) in Foxg1:MYC mice as well as wild type mice treated with MYC inhibitor or vehicle control. These new, gain- and loss-of-function data, shown in new Figure 5E, 5M, confirm MYC’s modest but consistent regulation of ribosome biogenesis at this early stage of forebrain development. We performed analyses of OPP incorporation in Foxg1:MYC and Nestin:MYC as well as Myc-/- mice. In these analyses, we did not see consistent, significant differences in OPP incorporation, demonstrating MYC-regulation of ribosome biogenesis does not drive protein synthesis in these cells and that other mechanisms exist to regulate these processes. We mention these data in the last paragraph of the subsection “MYC modulates ribosome biogenesis in developing forebrain”.

The JQ1 experiments are problematic in interpretation of a specific effect on Myc, as JQ1 is a broad epigenetic (bromodomain) inhibitor and alterations in a number of epigenetic regulators can lead to neural tube defects.

We removed the previous JQ1 data from the revised manuscript. We present instead new data using a more specific MYC inhibitor (KJ-Pyr-9;Hart et al., 2014)and found that intraperitoneal delivery to the pregnant dam reduced nucleolar volume quantified by Fibrillarin staining. These new data are presented in new Figure 5E.

Can the authors build a case that better supports the title of the paper with additional ways of manipulating translational machinery directly to investigate their hypothesis that down-regulation of protein synthesis is needed for neurulation?

We have rewritten the text to clarify that our goal is not to investigate the process of neurulation. We also expand our studies to include new analyses of the mTOR signaling pathway, which is also downregulated between E8.5 and E10.5. These new data are presented in new Figure 3. In parallel, we also tested the consequences of maternal rapamycin injection during this early stage of development. In our injections of n=5 litters, we did not observe neural tube defects, though cannot rule out that they may occur. We have revised the Discussion section in the manuscript and hope that our revised Discussion clarifies these important points to the reader.

Further experiments are needed to mechanistically link Myc modulation to ribosome biogenesis and/or to better link ribosome/translational regulation with cell fate restriction, in order to make a more impactful paper.

We performed additional experiments to demonstrate MYC’s contributions to ribosome biogenesis by using (1) Selective MYC inhibitor KJ-Pyr-9 (see above), and (2) quantifying ribosome biogenesis in Foxg1-MYC mice, which has larger nucleoli. We clarified in the text that changes in ribosome biogenesis and MYC expression take place concurrently as neural progenitor progress to more lineage-committed fates. In the longer term, we observe a large brain phenotype in the mice, which may be due to a combination of MYCs actions on downstream signaling pathways (e.g. ribosomes but also enhanced expression of factors with known roles in development including Igf2 and Insm1; see above).

3) The Myc OE mouse model complements previous LOF findings that Myc controls brain size and progenitors (albeit using a human transgene). The authors conclude that Myc OE increases proliferation and in their title link dynamic changes in protein synthesis to "cell fate restriction." However, Myc OE phenotypes are only reported at later stages with analysis of only one neuronal marker, limiting the strength of the conclusions about cell fate restriction and proliferation. Are all neuronal layers affected?

As requested, we performed additional analyses on cell cycle at earlier ages (E11.5) in Foxg1-MYC mice, but did not see significant differences in proliferation, consistent with the modest increase in brain size at E14.5. We also analyzed the cell identities contributing to the cerebral cortex in Nestin-MYC mice at P0 and P8. We binned the cortex into equal segments, and quantified the numbers of cells expressing the following laminar markers: Cux1 and Ctip2. At P0, we observed significantly more Cux1+ upper-layer neurons in the cortex of MYC-OE (Figure 6K, p≤0.01) but no difference in the number of Ctip2+ lower layer neurons (p=0.82). At P8, we also observed more Cux1+ neurons in MYC-OE, although the difference did not reach statistical significance (p=0.22). We did not observe differences in numbers of Ctip2+ neurons at P8 (p=0.96). These data suggest that the combinatorial effects of MYC overexpression most readily influence the later stages of neurogenesis, when upper layer neurons of the cerebral cortex are formed. However, we cannot rule out more subtle effects of cortical development that are not evaluated by these types of assays, or at the ages examined in this study. Our new data are included in the last paragraph of the subsection “Persistent MYC expression increases progenitor proliferation, leading to macrocephaly”. In the revised manuscript, we have focused on ribosome biogenesis during early stages of development and have shifted focus away from cell fate restriction.

It would be important to demonstrate that Myc OE impacts proliferation and fate restriction at earlier relevant stages, E10.5-E12.5, when expression of Myc and ribosome/translation factors is reduced. Foxg1-cre/myc mice are also used which initiate c-myc expression early at e9.5 but there is little analysis of these embryos. In addition the data for Nestin-cre mice do not support the hypothesis that c-myc is driver of translation.

We carried out the requested quantification of cell proliferation in Foxg1-MYC mice at E11.5, but did not observe a significant difference in proliferation, consistent with the modest increase in brain size at E14.5. Foxg1-MYC mice are embryonic lethal beginning soon after E14.5, preventing further analyses at later ages. While the Foxg1-MYC mice did show increased ribosome biogenesis, similar to the Nestin-MYC mice, they did not show differences in OPP incorporation. We explain this point in the subsection “Persistent MYC expression increases progenitor proliferation, leading to macrocephaly”. These data demonstrate that regulation of ribosome biogenesis does not necessarily lead to increases in protein synthesis.

4) The reviewers would prefer that results related to Myc OE drives tumors in the choroid plexus is removed. These data diffuse the focus of the paper as the phenotypes and biology here are vastly different from early development. More attention on the early stages and requested experiments above would better strengthen the manuscript.

We removed the last 2 figures of the original manuscript on choroid plexus tumors and placed more emphasis on the early stages of forebrain development in order to strengthen the manuscript. These changes are also reflected in our revised title, “Downregulation of ribosome biogenesis and protein synthesis during early forebrain development.”

Reviewer #1:

[…] The authors focus in on MYC as a possible upstream regulator. Most of the data in Figure 3 are solid but the JQ1 experiments are problematic in interpretation of a specific effect on Myc, as JQ1 is a broad epigenetic (bromodomain) inhibitor and alterations in a number of epigenetic regulators can lead to neural tube defects.

We removed the previous JQ1 data from the revised manuscript. We present instead new data using a more specific MYC inhibitor(KJ-Pyr-9; see above)and found that intraperitoneal delivery to the pregnant dam reduced nucleolar volume quantified by Fibrillarin staining. These new data are presented in Figure 5E in the revised manuscript.

The results end on somewhat of a tangent and turns toward a cancer phenotype (which would be expected by MYC-OE) within the choroid plexus. Despite leading away from the main point, it is OK to retain this data as it shows a link between developmental and cancer pathways (again, not a surprise). However, since MYC overexpression is persistent through development, the following statement is incorrect or at best misleading: "The appropriate downregulation of Myc after neurulation is essential to prevent early missteps of development that can manifest as later-onset diseases of the nervous system."

As described above in our response to Major critique #4, we have removed the choroid plexus tumor portion of the study from the revised manuscript.

Reviewer #2:

[…] 1) Myc OE in apical progenitors is employed to support the idea that dynamic regulation of ribosome biogenesis and protein synthesis is needed in neural progenitors over development. However, these data are largely correlative and it remains unclear mechanistically whether alterations in ribosomes and protein synthesis explain the Myc OE phenotypes. Although Myc has been previously linked to ribosome biogenesis, this transcription factor has other functions, for example in associating with and impacting cell cycle regulators. Thus, the described alterations in brain size, Pax6+ S phase cells, and Cux1 number could be due to cell cycle alterations, potentially independent of ribosome regulation.

As mentioned above in response to Major critique #2, we agree that MYC has many roles in the cells, and that the large brain phenotypes observed is very likely due to a combination of MYC’s effects on multiple signaling pathways. For example, MYC overexpression also led to the upregulation of genes involved with cell proliferation. We observed upregulation of Igf2 and Insm1 in MYC-expressing progenitors. IGF2 regulates proliferation of cortical progenitor cells (e.g. Lehtinen et al., Neuron 2011), and Insm1 is a transcription factor with roles in promoting the delamination of apical progenitors cells, thereby accelerating cortical development by promoting the formation of basal progenitors (e.g. Farkas et al., 2008; Tavano et al., 2018). We explain this point in the first paragraph of the subsection “Persistent MYC expression increases progenitor proliferation, leading to macrocephaly”.

Consistent with this possibility, Myc OE embryonic brains exhibit only a slight increase in ribosome biogenesis and no significant alterations in translation. Can the authors build a case that better supports the title of the paper with additional ways of manipulating translational machinery, for example? Further experiments are needed to mechanistically link Myc modulation to ribosome biogenesis and/or to better link ribosome/translational regulation with cell fate restriction, in order to make a more impactful paper.

In the revised manuscript, we have added new experiments quantifying ribosome biogenesis (via Fibrillarin staining) in Foxg1-MYC mice as well as wild type mice treated with MYC inhibitor or vehicle control. These new, gain- and loss-of-function data, show in new Figure 5E, 5F, 5M, 5N, confirm MYC’s consistent regulation of ribosome biogenesis at this early stage of forebrain development. We performed analyses of OPP incorporation in Foxg1:MYC and Nestin:MYC as well as Myc-/- mice. In these analyses, we did not see consistent, significant differences in OPP incorporation, demonstrating MYC-regulation of ribosome biogenesis does not drive protein synthesis in these cells and that other mechanisms exist to regulate and perhaps protect cells from these processes. We believe it is particularly important to present these surprising findings to the field. We mention these data in the last paragraph of the subsection “MYC modulates ribosome biogenesis in developing forebrain”. In light of all our new data and analyses, we have revised the title of the manuscript to better reflect our findings, “Downregulation of ribosome biogenesis and protein synthesis during early forebrain development.”

2) The Myc OE mouse model complements previous LOF findings that Myc controls brain size and progenitors (albeit using a human transgene). The authors conclude that Myc OE increases proliferation and in their title link dynamic changes in protein synthesis to "cell fate restriction." However Myc OE phenotypes are only reported at later stages with analysis of only one neuronal marker, limiting strength of conclusions about cell fate restriction and proliferation. Are all neuronal layers affected? It would be important to demonstrate that Myc OE impacts proliferation and fate restriction at earlier relevant stages, E10.5-E12.5, when expression of Myc and ribosome/translation factors is reduced.

We performed additional analyses on Foxg1-MYC mice at E10.5-E11.5. These data did not uncover increased proliferation. However, in the Foxg1-MYC mice, the brain phenotype is only beginning to emerge at E14.5. These data suggest that any shifts in cell proliferation are very subtle and as mentioned above, MYC overexpression acts likely by multiple mechanisms to increase brain size.

3) The last two figures of the paper demonstrate Myc OE drives tumors in the choroid plexus and modifies translation markers in these adult stages. While I appreciate this helps support the link between Myc and ribosome regulation, given the phenotypes and biology here are vastly different from early development, these data diffuse the focus of the paper.

As described above in our response to Major critique #4, we removed the choroid plexus tumor portion of the study from the revised manuscript.

4) The conclusion that E8.5 neuroepithelial progenitors have higher ribosome biogenesis than E10.5 is based upon increased nucleolar volume (as a proxy) and EM analysis. The EM data need annotations and ideally immuno-gold to support the conclusion that punctae are actually ribosomes.

As mentioned above in response to Major critique #1, in the revised manuscript, we provide larger, annotated, representative EM images of from E8.5 and E10.5 neural progenitor cells (Figure 2K). We quantified the number of ribosomes and other organelles occupying the cytoplasmic space in neural precursors and found that the E8.5 cells contain more ribosomes per defined area of cytoplasm than E10.5 progenitors (Figure 2L), though ribosome density within the free cytoplasmic space, per se, was not different between these two ages (Figure 2M). In addition, more E10.5 cytoplasm than E8.5 cytoplasm was occupied by other organelles including endoplasmic reticulum, mitochondria, and Golgi (Figure 2N), indicating an overall shift in organelle landscape at this age.

We are confident that the punctae in our images are ribosomes, and based on morphological characteristics, are very unlikely to be any other structures. For example glycogen, which we often observe in our EM analyses of other cell types such as choroid plexus epithelial cells, is larger, more irregular in shape, aggregates more, and is typically more electron dense. We consulted with additional experts on ribosomes and EM in the HMS community including:

Maria Ericsson (manager, Harvard Medical School EM Facility, where we perform our EM analyses)

Mark Fleming, MD, DPhil (Boston Children’s Hospital; e.g. Nguyen et al., Science 2017)

Wei-chun Lee, PhD (HMS/BCH Neurobiology; Lee et al., Nature 2016)

These experts agree that the punctae in our samples are ribosomes. In addition, there was a general consensus that because our samples are so packed with ribosomes, immuno-gold EM would be problematic to interpret since the overall signal would likely be too strong/saturated. Together with our revised representative images and quantification (see above), we hope the reviewers will agree that ribosome biogenesis is decreased in the E8.5 to E10.5 forebrain.

It would be ideal if the authors could support their conclusion with standard molecular analysis of biogenesis- for example, are rRNA levels and/or processing different across development?

As mentioned above in response to Major critique #1, we reached out to Debby Silver (Duke University), to discuss this point and to get her advice on reagents that may work well in the developing forebrain at these early ages. The Silver lab kindly shared FISH probes against 5.8S pre-rRNA and 5.8S total rRNA (Pilaz et al., 2016), which reveal more intense rRNA signal in E8.5 than E10.5 progenitors. These new data are shown in new Figure 2F, 2G.

We purchased the Y10b antibody used by many labs in the field (e.g. Kondrashov et al., 2011) and found that it, too, showed stronger immunofluorescence signal in E8.5 progenitors than E10.5 progenitors. These new data are shown in new Figure 2H.

We quantified 35S-methionine incorporation in vitro in forebrain progenitors, and observed higher 35S-methionine incorporation at E8.5 compared to E10.5 (counts per million cells, presented as E8.5 fold change normalized to E10.5 forebrain: Expt. 1 = 2.0-fold; Expt. 2 = 1.4-fold; Expt. 3 = 1.1-fold), indicative of a higher protein synthesis rate in E8.5 forebrain progenitor cells. These new data are described in the last paragraph of the subsection “Decreased ribosome biogenesis and protein synthesis in E10.5 neuroepithelium”.

We reached out to experts in the Harvard Medical School community regarding the technical requirements for analyzing monosomes and polysomes. From our pilot dissections and discussions with Drs. Dan Finley and Mark Fleming (e.g. Nguyen et al., Science 2017), we have concluded that these analyses are technically not feasible given the very limited amounts of E8.5 presumptive forebrain that can be micro-dissected across similarly staged embryos and pooled across multiple litters of mice. However, as we provide new data for the rest of the other points raised above, we hope that the reviewer agrees that our manuscript is strengthened by the inclusion of these new data and that polysome profiling is not necessary to demonstrate that ribosome biogenesis is downregulated from E8.5 to E10.5.

Reviewer #3:

Chau et al. performed transcriptome analysis on forebrain e8.5 and e10.5 neuroepithelium and discovered that the protein biosynthetic machinery is down-regulated after neurulation. The authors then explore if c-myc regulates the expression of ribosomes and other translational machinery genes during neurulation. The work is extensive and the authors use a variety of state of the art techniques. Differential regulation of ribosomes and other translational machinery as a driver of neurulation is novel and could have a broad impact on our understanding of this important and complex morphogenetic process. While the initial transcriptional and proteomic experiments are performed well, the follow up analysis focusing on translation and the role of c-myc in the process are lacking.

The authors only directly examine whether protein synthesis is decreased post-neurulation (by the incorporation of OPP in Figure 2H) one time in the manuscript. The authors often use instead nucleolar data, which is interesting, but it is only correlative. A more direct examination of protein synthesis should also be performed on the c-myc knock-out and c-myc over-expression mice. Additional direct measurements of protein synthesis and functional ribosomes (monosomes vs. polysomes) are also needed in all instances to substantiate their findings.

As mentioned above in response to Major critique #1, we quantified 35S-methionine incorporation in vitro in forebrain progenitors, and observed higher 35S-methionine incorporation at E8.5 compared to E10.5 (counts per million cells, presented as E8.5 fold change normalized to E10.5 forebrain: Expt. 1 = 2.0-fold; Expt. 2 = 1.4-fold; Expt. 3 = 1.1-fold), indicative of a higher protein synthesis rate in E8.5 forebrain progenitor cells. These new data are described in the last paragraph of the subsection “Decreased ribosome biogenesis and protein synthesis in E10.5 neuroepithelium”.

We have invested tremendous effort to perform the OPP analyses in the gain and loss-of-function models in lab. First, we have bred more Myc-deficient mice for this analysis at E8.5. However, these studies have been very challenging. The mice breed poorly, have very small litters, and as Myc-deficiency leads to small size, neural tube closure defects, and developmental delay, we have been faced with limited samples that are at similar developmental stages for our analyses. From the samples that we have collected, we have not observed reliable differences in OPP incorporation. Second, we have performed additional OPP quantification experiments in Foxg1:MYC mice. As we showed in the initial manuscript submission (with Nestin:MYC mice), MYC overexpression also did not result in clear upregulation of protein synthesis. We describe these observations in the last paragraph of the subsection “MYC modulates ribosome biogenesis in developing forebrain”. These observations differ from our hypothesis, which was formed based on findings in the cancer literature. Given our findings, we have focused the contributions of Myc expression in our study to be on ribosome biogenesis.

We also reached out to experts in the Harvard Medical School community regarding the technical requirements for analyzing monosomes and polysomes. From our pilot dissections and discussions with Drs. Dan Finley and Mark Fleming (e.g. Nguyen et al., Science 2017), we have concluded that these analyses are technically not feasible given the very limited amounts of E8.5 presumptive forebrain that can be micro-dissected across similarly staged embryos and pooled across multiple litters of mice. However, as we provide new data for the rest of the points raised above, we hope that the reviewers agree that our manuscript is improved and that polysome profiling is not necessary to demonstrate that ribosome biogenesis is downregulated from E8.5 to E10.5.

The necessity of c-myc in coordinating the down-regulation of the protein synthesis machinery is also lacking. The authors use a conventional knock-out when a conditional knock-out might be more informative. The authors use JQ1 as an inhibitor of c-myc but concerns exist about the specificity of JQ1 for only c-myc. The authors could also use more standard and complementary systems that are likely to be more specific to test the necessity of c-myc. Unfortunately without these data, the author's conclusion (subsection “Critical timing of MYC expression during early forebrain development”, last paragraph) is not supported.

As reviewer #1 raised concerns about the JQ1 MYC inhibitor due to its broader role as a bromodomain inhibitor, we have removed the previous JQ1 data from the revised manuscript. We present new data using KJ-Pyr-9, a more specific MYC inhibitorand found that intraperitoneal delivery to the pregnant dam reduced nucleolar volume quantified by Fibrillarin staining. These new data are presented in Figure 5E in the revised manuscript. These findings are in agreement with our data showing the Myc-deficient mice have smaller nucleoli as well (Figure 5F). While we do have the conditional Myc knockout mouse in lab, the field lacks a Cre driver line that would induce Cre recombination specifically in the neuroepithelium at the critical time point in development, just before neurulation. In the revised manuscript, we have removed the section regarding neural tube closure.

The over-expression studies are also lacking. The authors use an innovative c-myc line and two cre drivers (Nestin and Foxg1) but perform a detailed analysis for only the Nestin-cre driver. Unfortunately this does not seem to be the best mouse since c-myc is expression commences days after neurulation is complete. Instead the Foxg1-cre/myc mice seem better since c-myc expression initiates at e9.5 but the necessary studies are not performed. In addition the data for Nestin-cre mice do not support the hypothesis that c-myc is driver of translation and neurulation.

As requested, we have now performed additional analyses of nucleolar size and OPP incorporation in the Foxg1:MYC mice. We found that MYC expression in this line also resulted in increased nucleolar size (new Figure 5M), but as mentioned above, we did not observe differences in OPP incorporation (included in the last paragraph of the subsection “MYC modulates ribosome biogenesis in developing forebrain”). We agree with the reviewer that based on these findings, MYC does not appear to be a driver of translation in this system, although there may be other regulatory steps including accessibility to chromatin contributing to these results. We discuss this point in the Discussion section as well.

C-myc also likely regulates many pathways, just not the protein synthesis machinery. Thus, the authors should also manipulate the translational machinery directly to investigate their hypothesis that down-regulation of protein synthesis is needed for neurulation.

We agree that Myc has many roles and likely contributes to brain development in number of ways. We have rewritten the text to clarify that our goal is not to investigate the process of neurulation. We also expand our studies to include new analyses of the mTOR signaling pathway, which is also downregulated between E8.5 and E10.5. These new data are presented in new Figure 3. In parallel, we also tested the consequences of maternal rapamycin injection to inhibit the mTOR signaling pathway during this early stage of development. In our injections of n=5 litters, we did not observe neural tube defects, though cannot rule out that they may occur with different doses or other modifications to the protocol. We have revised the Discussion section in the manuscript and hope that our revised Discussion clarifies these important points to the reader.

Finally, a more minor concern, but still a possible significant problem, is that the difference in protein machinery expression could vary along the A-P and D-V/M-L axes. Since neurulation varies along the A-P axis and there are specific cell types along the M-L/D-V axis (hinge points, neural fold cells), differences could exist. Some data (Figure 2C) support this possibility. This caveat should be stated explicitly in the manuscript.

This is an interesting and important point. We have included this in the fifth paragraph of the Discussion section.

[Editors' note: the author responses to the re-review follow.]

Essential revisions:

1) Please add to the Discussion some technical caveats of monitoring translation in vivo (some of which were mentioned in the response to reviewers) and as discussed in detail in the next paragraph. Also in the Discussion please stage whether there could be a function of altered ribosome biogenesis independent of modulating translation.

While authors have strong evidence that ribosome signature (and biogenesis) changes during development (as described in a manuscript that was not cited Kraushar et al., 2015, but also in the cited Kraushar et al., 2016), authors do not provide strong evidence that they disrupted dynamics of protein synthesis due to ribosome and/or MYC changes. For example, RPL11 is known to sit at the monosome but not to associate with polysome, while during stress conditions it binds to the DNA promoter of p53 oncogene (Slavov, Cell Reports, 2015; Mahata et al., Ocogene, 2015). In addition, critical experiments testing protein synthesis (e.g. OPP injection or methionine analysis), in current state do not exclude possibility that different transcription dynamics play role in the observed OPP and methionine changes. Here, authors may consider an additional line of evidence – they may want to discuss discrepancies between levels of mRNA expression and matching protein these mRNAs encode (e.g. using qRT-PCR and western).

We have added a new paragraph to the Discussion section to address these concerns.

2) Please remove "protein synthesis" from the title. Since OPP was injected intraperitoneally, the availability of OPP to progenitors may be different between E8.5 and 10.5 (e.g. similar to decrease of a number of proteins in AF and CSF by E10.5). Overall, there is no strong experimental evidence that protein synthesis is downregulated in early progenitors, while it may also be easily upregulated for selective number of other mRNAs whose protein expression was not tested. For example, total levels of ribosomal proteins can decrease but their association with polysomes may increase (Kraushar et al., 2015, 2016). The same concern goes for not seeing consistent, significant differences in OPP incorporation in Myc-/- mice. Thus, it is critical to either remove "protein synthesis" from their title or perform additional experiments (e.g. experiments focusing on protein synthesis where transcription is transiently inhibited at its minimum).

We have removed “protein synthesis” from the title.

3) Using immunohistochemistry, the authors show significant decrease for a number of ribosome biogenesis players and ribosomal proteins. Is there a ribosome protein that did not change or maybe even increased? These possible additional findings would be in line with EM pictures and EM analysis that clearly show large number of ribosomes in apical progenitors.

Based on our RNAseq data, ribosomal proteins that are not differentially expressed between E8.5 and E10.5 include:

IDE8.5 FPKME10.5 FPKMFold_change (log2)Sig
Rpl13a 481.511718.4520.577324no
Rpl7a 1405.231227.52-0.195062no
Rpsa 368.641292.213-0.335194no
Rpl4 1519.981168.83-0.378986no

Using immunoblotting, we found that RPL10A, which has higher RNA expression at E8.5 (Rpl10a FPKM: E8.5 = 1962.03; E10.5 = 1242.73), has comparable protein expression between E8.5 and E10.5. The difference between mRNA and protein levels suggests the presence of additional post-transcriptional mechanisms that regulate of expression of this gene. This new blot is now shown in new Figure 2M.

Despite the down-regulation of the majority of ribosomal proteins, we did not observe changes in ribosome density by EM. These observations could also be explained by the long half-life of ribosomes, which can be as long as 5 days. We have also addressed this in the Discussion section.

https://doi.org/10.7554/eLife.36998.018

Article and author information

Author details

  1. Kevin F Chau

    1. Department of Pathology, Boston Children’s Hospital, Boston, United States
    2. Program in Biological and Biomedical Sciences, Harvard Medical School, Boston, United States
    Contribution
    Conceptualization, Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-7899-3247
  2. Morgan L Shannon

    Department of Pathology, Boston Children’s Hospital, Boston, United States
    Contribution
    Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing—review and editing
    Competing interests
    No competing interests declared
  3. Ryann M Fame

    Department of Pathology, Boston Children’s Hospital, Boston, United States
    Contribution
    Conceptualization, Data curation, Formal analysis, Supervision, Validation, Visualization, Methodology, Writing—review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-8244-2624
  4. Erin Fonseca

    Department of Pathology, Boston Children’s Hospital, Boston, United States
    Contribution
    Data curation, Formal analysis, Visualization
    Competing interests
    No competing interests declared
  5. Hillary Mullan

    Department of Pathology, Boston Children’s Hospital, Boston, United States
    Contribution
    Data curation, Validation, Visualization
    Competing interests
    No competing interests declared
  6. Matthew B Johnson

    Division of Genetics, Boston Children’s Hospital, Boston, United States
    Contribution
    Resources, Data curation, Software, Formal analysis, Supervision
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-6909-5712
  7. Anoop K Sendamarai

    Department of Pathology, Boston Children’s Hospital, Boston, United States
    Contribution
    Formal analysis, Supervision, Investigation, Writing—review and editing
    Competing interests
    No competing interests declared
  8. Mark W Springel

    Department of Pathology, Boston Children’s Hospital, Boston, United States
    Contribution
    Formal analysis, Validation, Investigation
    Competing interests
    No competing interests declared
  9. Benoit Laurent

    1. Division of Newborn Medicine and Epigenetics Program, Department of Medicine, Boston Children’s Hospital, Boston, United States
    2. Department of Cell Biology, Harvard Medical School, Boston, United States
    Contribution
    Data curation, Formal analysis, Supervision, Investigation
    Competing interests
    No competing interests declared
  10. Maria K Lehtinen

    1. Department of Pathology, Boston Children’s Hospital, Boston, United States
    2. Program in Biological and Biomedical Sciences, Harvard Medical School, Boston, United States
    Contribution
    Conceptualization, Supervision, Funding acquisition, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing
    For correspondence
    maria.lehtinen@childrens.harvard.edu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-7243-2967

Funding

National Science Foundation (Graduate Research Fellowship)

  • Kevin F Chau

National Institutes of Health (R01 NS088566)

  • Maria K Lehtinen

Pediatric Hydrocephalus Foundation (Research grant)

  • Maria K Lehtinen

Simons Foundation (SFARI Pilot Grant #402089)

  • Maria K Lehtinen

New York Stem Cell Foundation (Robertson Investigator)

  • Maria K Lehtinen

National Institutes of Health (NIH T32 HL110852)

  • Kevin F Chau
  • Ryann M Fame

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

We thank members of the Lehtinen and Fleming labs, C Harwell, A LaMantia, S Dymecki, A Lassar, and R Segal for helpful discussions, T Baudino for Myc knockout mice, D Silver for FISH probes, J Steen for RPS12 and RPL11 antibodies, M Baizabal, M Ericsson, A Malesz, and P Schmidt for experimental advice and assistance. We are grateful for the following support: NSF Graduate Research Fellowship (KFC), NIH T32 HL110852 (KFC and RMF), Pediatric Hydrocephalus Foundation, Simons Foundation SFARI Pilot Grant #402089, NIH R01 NS088566 (MKL), BCH IDDRC 1U54HD090255, and the New York Stem Cell Foundation. MK Lehtinen is a New York Stem Cell Foundation – Robertson Investigator.

Ethics

Animal experimentation: All animal experimentation was carried out under protocols approved by the IACUC of Boston Children's Hospital (protocol number 17-10-3547R).

Reviewing Editor

  1. Marianne Bronner, California Institute of Technology, United States

Publication history

  1. Received: March 28, 2018
  2. Accepted: May 9, 2018
  3. Accepted Manuscript published: May 10, 2018 (version 1)
  4. Version of Record published: June 1, 2018 (version 2)

Copyright

© 2018, Chau et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 2,139
    Page views
  • 354
    Downloads
  • 3
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, Scopus, PubMed Central.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)