November 15, 2017;
Phosphorylation Dynamics Dominate the Regulated Proteome during Early Xenopus Development.
The earliest stages of animal development are largely controlled by changes in protein phosphorylation mediated by signaling pathways and cyclin-dependent kinases. In order to decipher these complex networks and to discover new aspects of regulation by this post-translational modification, we undertook an analysis of the X. laevis phosphoproteome at seven developmental stages beginning with stage VI oocytes and ending with two-cell embryos. Concurrent measurement of the proteome and phosphoproteome enabled measurement of phosphosite occupancy as a function of developmental stage. We observed little change in protein expression levels during this period. We detected the expected phosphorylation of MAP kinases, translational regulatory proteins, and subunits of APC/C that validate the accuracy of our measurements. We find that more than half the identified proteins possess multiple sites of phosphorylation that are often clustered, where kinases work together in a hierarchical manner to create stretches of phosphorylated residues, which may be a means to amplify signals or stabilize a particular protein conformation. Conversely, other proteins have opposing sites of phosphorylation that seemingly reflect distinct changes in activity during this developmental timeline.
P41 GM108538 NIGMS NIH HHS
, R01 GM096767 NIGMS NIH HHS
, R01 HD084399 NICHD NIH HHS
GO Terms referenced:
[+] show captions
Figure 1. Quantitative changes of the X. laevis proteome over seven developmental stages. (A) Micrographs of X. laevis at the developmental stages taken for proteomic analyses. (B) Individual protein abundance presented on a log2 axis across developmental time points. Data are normalized to the mean value of each protein. (C) Relative standard deviation (RSD) for all proteins in panel (B) at individual experimental time points. (D) Heatmap showing the change in protein expression across the developmental time points. The heatmap was generated from the log2 normalized data using the default parameters in Matlab. Six groups representing significant trends of protein expression were manually selected for analysis. Proteins with the 15% greatest quantitative change are presented and are considered significant at a FDR of 0.20 using a Benjamini-Hochberg test. Neither clustering nor dendrogram generation was performed along the development stage axis of the heatmaps. These points are defined by the biology of development, and their differences are fixed by that biology.
Figure 2. Quantitative changes of proteins that mediate oocyte maturation. Protein abundance normalize to stage VI oocyte are plotted at the seven experimental time points. An average of protein intensity for biological replicates is shown. Standard deviations for each protein are less than 0.2 for all time points. Proteins that were not detected have had their abundance set to zero. Mos was normalized to experimental time point 2.
Figure 3. Galaxy plot correlating changes in protein levels with changes in phosphorylation. Individual protein abundances (plotted on the log2 axis) normalized to amount present in stage VI oocytes are plotted relative to phosphorylation site occupancy of that protein.
Figure 4. Heatmap displaying changes in absolute occupancy of individual phosphorylation sites. Nine groups were manually selected from the top 25% of sites having the greatest variation in phosphorylation across the seven experimental time points. The heatmap was generated using the MatLab Euclidian algorithm from the log2 normalized data. Individual cluster nodes were manually excised for further analysis. All values included are significant at a FDR of 0.15 using a Benjamini-Hochberg test.
Figure 5. Phosphorylation events during progesterone-dependent oocyte maturation. (A) Outline of the major members of the signaling pathway for oocyte maturation initiated by progesterone adapted from87. Phosphorylation events are represented in green and dephosphorylation in red with arrows indicating activation and bars indicating repression. Pathways with intermediates not shown are indicated with a broken line. (B) The occupancy of individual phosphorylation sites in proteins within the pathway is presented relative to the seven experimental time points. Changes are normalized to the mean of the individual phosphopeptide. Inserts of each panel identify the detected sites of phosphorylation.
Figure 6. Consensus phosphorylation sites. GproX was used for unsupervised and unstandardized clustering of all identified phosphopeptide sequences. Upper and lower threshold log2 values of 0.26 and −0.32 were used, corresponding to ratios of 1.2 and 0.8, respectively. This generated 22 unique clusters according to changes in phosphorylation as a function of developmental time. Expression profile and corresponding consensus sequences generated using WebLogo for (A) cluster 20 and (B) cluster 21.
Figure 7. Clustered sites of phosphorylation. A selection of multiple-phosphorylated peptides is presented. Phosphorylation sites indicated in red had a minimum fractional occupancy of >0.1. Predicted kinase substrate sites were identified using PhosphoMotif Finder86. GRK (G protein-coupled receptor kinase), GSK3 (glycogen synthase kinase 3), CK2 (casein kinase 2), PKC (protein kinase C), PKA (protein kinase A), CDK2 (cyclin-dependent kinase 2), CK1 (casein kinase 1). CK2* indicates that phosphorylation at this site requires prior phosphorylation of a proximal priming site.
Figure S1. Protein change during the seven developmental time
points. Related to Figure 1. Box and whisker plot demonstrating the fold
change of the proteins relative to their adjacent stage. Outliers were
determined by 0DWODE¶V default Box plot algorithms and are marked by the
µ+¶One outlier at 11.92 fold change was removed for the Stage VI / 45
Figure S2. Unpacked galaxy plot of the development. Related to Figure
3. Each of the individual time points shown in relation to protein change.
Figure 3 is a combination of the six galaxy plots overlaid in the manuscript.
The grey background for each panel is of the other time points. The turquois
blue describes the expression of the titled time point protein vs.
Figure S3. Heat map of all phosphorylation
site changes. Related to Figure 4. (A) Full heat
map of all phosphorylation sites corrected for
their protein level. Figure 4 only includes the top
60% of phosphorylation site variance. Protein
mean is determined based on the normalized
signal intensity of all the proteins in the sample
for the respective time point
Figure S3. Heat map of all phosphorylation
site changes. Related to Figure 4. &RKHQ¶V'
Analysis of individual phosphorylation sites
between time points, A delta greater than or equal to 0.25 is considered
Figure S4. Kinase consensus sequence profiles. Related to Figure 6. All
kinase sequences and their respective amino acid sequence are presented
here. Figure 6 shows cluster 4 and 20.
Figure S4 Continued.