Phosphoproteomic response to epidermal growth factor in native rat inner medullary collecting duct

Abstract

Epidermal growth factor (EGF) has important effects in the renal collecting duct to regulate salt and water transport. To identify elements of EGF-mediated signaling in the rat renal inner medullary collecting duct (IMCD), we carried out phosphoproteomic analysis. Biochemically isolated rat IMCD suspensions were treated with 1 µM of EGF or vehicle for 30 min. We performed comprehensive quantitative phosphoproteomics using tandem mass tag (TMT)-labeling of tryptic peptides followed by protein mass spectrometry. We present a data resource reporting all detected phosphorylation sites and their changes in response to EGF. For a total of 29,881 unique phosphorylation sites, 135 sites were increased and 119 sites were decreased based on stringent statistical analysis. The data are provided to users at https://esbl.nhlbi.nih.gov/Databases/EGF-phospho/. The analysis demonstrated that EGF signals through canonical EGF pathways in the renal IMCD. Analysis of Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways in which EGF-regulated phosphoproteins are over-represented in native rat IMCD cells confirmed mapping to RAF-MEK-extracellular signal-regulated kinase (ERK) signaling but also pointed to a role for EGF in the regulation of protein translation. A large number of phosphoproteins regulated by EGF contained PDZ domains that are key elements of epithelial polarity determination. We also provide a collecting duct EGF-network map as a user-accessible web resource at https://esbl.nhlbi.nih.gov/Databases/EGF-network/. Overall, the phosphoproteomic data presented provide a useful resource for experimental design and modeling of signaling in the renal collecting duct.NEW & NOTEWORTHY EGF negatively regulates transepithelial water and salt transport across the kidney collecting duct. This study identified phosphoproteins affected by EGF stimulation in normal rat collecting ducts, providing insights into global cell signaling mechanisms. Bioinformatic analyses highlighted enhanced canonical ERK signaling alongside a diminished activity in the PI3K-Akt pathway, which is crucial for cell proliferation and survival. This EGF response differs somewhat from prior studies where both pathways were prominently activated.

Keywords: growth factor; kidney; mass spectrometry; phosphorylation; protein kinase.

Graphical abstract

Figure 1.

Immunoblot of rat IMCD suspensions using phosphospecific antibody recognizing the EGFR protein phosphorylated at Y1091, a site of autophosphorylation expected to increase with ligand-induced EGFR activation. IMCD suspensions were treated with 1 µM EGF or vehicle for 30 min. Immunoblotting was carried out using an antibody recognizing total EGFR or phospho-specific antibody recognizing pY 1091 site. Densitometry analysis was performed on both bands. The data were normalized so that the average density of the vehicle-treated samples was established as 100%. EGF significantly increased pY1091-EGFR. Values are means ± SE of band densities normalized to control. EGFR, epidermal growth factor receptor; IMCD, inner medullary collecting duct. *P <0.05, unpaired Student’s t test (n = 3).

Figure 2.

Volcano plot for all phosphopeptides identified from quantitative phosphoproteomics experiments (A) and for total protein (B). A: x-axis depicts log₂ ratio of isobaric labeling signal of phosphosites measured in EGF-treated samples over the control samples. The y-axis depicts the negative of the log₁₀ of P value determined from paired t tests (n = 3). Phosphosites that satisfied dual criteria of P < 0.1 and |log₂(EGF/Control)| > 0.3428 were used for further analysis. A log₂(EGF/Control) significance cut-off level at 0.3428 was determined as the 99% confidence interval (2.58 × SD) over all phosphosites in three sample pairs. Specific phosphophorylation sites that were significantly changed by EGF (orange dots) can be identified by hovering over individual points in online version of this graph accessible at https://esbl.nhlbi.nih.gov/Databases/EGF-phospho/Volcano%20plot.html. The proteins labeled in red font are members of the Kyoto Encyclopedia of Genes and Genomes (KEGG) dataset “ERBB Pathway “(rno04012). B: changes in total protein abundances in response to EGF in inner medullary collecting duct (IMCD) suspensions. Total protein levels were determined from at least two peptides that map to a given protein.

Figure 3.

MS2 spectrum of EGF-induced phosphorylated ERK2 (MAPK1) and ERK1 (MAPK3) encompassing the T-E-Y activation loop. A: top: an extracted MAPK1 spectrum with phosphorylation of T183 and Y185; bottom: the tandem mass tag (TMT)-isobaric tag intensities from three pairs of EGF- and vehicle-treated samples. B, top: an extracted MAPK3 spectrum with phosphorylation of T202 and Y205; bottom: the TMT-isobaric tag intensities from three pairs of EGF- and vehicle-treated samples. CTR, vehicle-treated control samples; EGF, EGF-treated samples.

Figure 4.

Bioinformatic analysis of EGF-regulated phosphoproteins using the Database for Annotation, Visualization and Integrated Discovery (DAVID). Classifiers identified significant overrepresentation of regulated phospho-proteins in A: KEGG Pathways, B: Gene Ontology Biological Processes, C: Protein Domains. The official gene symbols of EGF-regulated phosphoproteins were uploaded to DAVID website using all identified phosphoproteins as background for functional annotation clustering. Bar heights show significance level calculated as −log₁₀(P) for Fisher Exact tests. The integers within the bar indicate the number of phosphoproteins that match each classifier. KEGG, Kyoto Encyclopedia of Genes and Genomes.

Figure 5.

Phosphopeptide motif analysis using in-house Java application PTM-Logo. A: mono-phosphosites upregulated in response to EGF were used to create a sequence preference motif using all monophosphosites as background. B: MARK1 substrate preference logo generated from Poll et al. (47). The sequence from A was input into KinasePredictor (https://esbl.nhlbi.nih.gov/Databases/Kinase_Logos/KinasePredictor.html) to identify kinases most likely responsible and the resulting motif was downloaded from https://esbl.nhlbi.nih.gov/Databases/Kinase_Logos/. C: mono-phosphosites downregulated in response to EGF were used to create a sequence preference motif using all monophosphosites as background. D: Camk2δ substrate preference logo generated from Poll et al. (47). EGF, epidermal growth factor.

Figure 6.

Effect of EGF on intracellular Ca²⁺ of rat IMCD measured by Fluo-4 fluorescence. Freshly isolated rat IMCD segments were microdissected and loaded with calcium indicator dye, Fluo-4. The same tubule was treated in order with vehicle, 0.1 µM EGF, and 100 µM carbachol with a brief wash in between treatments (n = 3). A–C: representative fluorescence images in response to vehicle, EGF, and carbachol stimulation. D: plotting of fluorescence intensity as a function of time acquired at the tubule level. E–G: plotting of fluorescence intensity as a function of time acquired at the single-cell level. Injection of vehicle did not alter Ca signal, whereas injection of 0.1 µM EGF induced a small, transient, increase in intracellular Ca²⁺ that was lower in magnitude than that induced by 100 µM carbachol (see Supplemental Video S1).EGF, epidermal growth factor; IMCD, inner medullary collecting duct.

Figure 7.

Mapping proteins with significant phosphorylation change by EGF to cellular structures or functions. The identified EGF-regulated phosphoproteins were mapped to Gene Ontology cellular component or molecular function terms (gray boxes) using Automated Bioinformatics Extractor (ABE, https://esbl.nhlbi.nih.gov/ABE/). Proteins with increased phosphorylation change at specific sites by EGF were labeled as red boxes. Proteins with decreased phosphorylation change at specific sites by EGF were labeled as blue boxes. Plotting was achieved in an Excel sheet, then exported and edited in Cytoscape. EGFR, epidermal growth factor receptor; mTOR, mammalian target of rapamycin.

Figure 8.

A causal map summarizing the effects of EGF in native rat IMCD. The network of nodes (identified phosphosites) was connected by edges (directions of effect) based on literature search and phosphoproteomic data obtained from the present study. Nodes that are presented as red color (significantly increased phosphorylation by EGF) and blue color (significantly decreased phosphorylation by EGF) have both P value < 0.1 and |log₂(EGF/Control)| > 0.3428. Nodes that are presented as light red or light blue have a P value < 0.1 but with a |log₂(EGF/Control)| value lower than the cut-off mark of 0.3428. Nodes that are presented as gray color are either not significantly changed or not identified in the present study. The mapping shows that following EGF stimulation, autophosphorylation of EGFR at Y1091 and Y1171 induced the tyrosine kinase activity of the receptor to phosphorylate several adaptor proteins that are known to interact with EGFR. Subsequent downstream effectors were found predominantly in activation of RAF-MEK-ERK signaling pathway, whereas effects on AKT-mTOR pathway lead to translation repression. EGF has also affected many proteins associated with cytoskeleton arrangement. Also shown are several ERK downstream effectors identified in the present study (ABI1, TEX264, ETV6, EIF4EBP1, RAF1). AKT, α serine/threonine-protein kinase; EGFR, epidermal growth factor receptor; IMCD, inner medullary collecting duct; mTOR, mammalian target of rapamycin.

Figure 9.

A scatter plot comparing the phosphorylation change of the same phosphosites identified in both EGF (current) and dDAVP (previous study). To compare the current dataset with the previous dDAVP dataset (53), we first converted the t test P values of all identified phosphosites in the current study to the P_joint value that was used in the previous study (53). X-axis depicts log₂(EGF/Control) values and y-axis depicts log₂(dDAVP/Control) of the same phosphosites. The black circles shows the phosphosites that are significantly changed in either EGF or dDAVP study. The red circles show the phosphosites that are significantly changed in both EGF and dDAVP studies. Among these, EPB41|4B (S401) exhibits an opposite change in phosphorylation in response to EGF and dDAVP stimulation, whereas POC5 (S80) was increased, and KAZN(S357) and LIMD1 (S233) were decreased, in phosphorylation in both EGF and dDAVP studies. EGF, epidermal growth factor.