Phosphoproteome-based kinase activity profiling reveals the critical role of MAP2K2 and PLK1 in neuronal autophagy

Abstract

Recent studies have demonstrated that dysregulation of macroautophagy/autophagy may play a central role in the pathogenesis of neurodegenerative disorders, and the induction of autophagy protects against the toxic insults of aggregate-prone proteins by enhancing their clearance. Thus, autophagy has become a promising therapeutic target against neurodegenerative diseases. In this study, quantitative phosphoproteomic profiling together with a computational analysis was performed to delineate the phosphorylation signaling networks regulated by 2 natural neuroprotective autophagy enhancers, corynoxine (Cory) and corynoxine B (Cory B). To identify key regulators, namely, protein kinases, we developed a novel network-based algorithm of in silico Kinome Activity Profiling (iKAP) to computationally infer potentially important protein kinases from phosphorylation networks. Using this algorithm, we observed that Cory or Cory B potentially regulated several kinases. We predicted and validated that Cory, but not Cory B, downregulated a well-documented autophagy kinase, RPS6KB1/p70S6K (ribosomal protein S6 kinase, polypeptide 1). We also discovered 2 kinases, MAP2K2/MEK2 (mitogen-activated protein kinase kinase 2) and PLK1 (polo-like kinase 1), to be potentially upregulated by Cory, whereas the siRNA-mediated knockdown of Map2k2 and Plk1 significantly inhibited Cory-induced autophagy. Furthermore, Cory promoted the clearance of Alzheimer disease-associated APP (amyloid β [A4] precursor protein) and Parkinson disease-associated SNCA/α-synuclein (synuclein, α) by enhancing autophagy, and these effects were dramatically diminished by the inhibition of the kinase activities of MAP2K2 and PLK1. As a whole, our study not only developed a powerful method for the identification of important regulators from the phosphoproteomic data but also identified the important role of MAP2K2 and PLK1 in neuronal autophagy.

Keywords: autophagy; corynoxine; kinase activity; phosphoproteome; phosphorylation; protein kinase.

Figure 1.

Flowchart of the identification of the quantitative phosphoproteome in neuronal autophagy. According to the weights of the labeled essential amino acids, N2a cells were divided into 3 groups, including “heavy” (¹³C₆¹⁵N₂-Lysine, ¹³C₆¹⁵N₄-Arginine), “middle” (¹³C₄¹⁴N₂-Lysine, ¹³C₆¹⁴N₄-Arginine) and “light” (¹²C₆¹⁴N₂-Lysine, ¹²C₆¹⁴N₄-Arginine). After checking the labeling efficiency (> 96%), the “light” and “middle” groups were treated with 15 μM Cory or Cory B for 3 h (Fig. S1), and the “heavy” group, which acted as a control group, was treated with 0.1% (v:v) DMSO. The cell lysates were mixed at a ratio of 1:1:1 and reduced with 10 mM dithiothreitol for 1 h at 56°C and then alkylated with 20 mM iodoacetamide for 45 min at room temperature and protected from light. Then, proteins were digested in solution with trypsin gold at a ratio of 1:50 (trypsin to protein) overnight and 1:100 (trypsin to protein) for another 4 h. After fractionation by high-pH reverse-phase HPLC, peptides were divided into 12 fractions. Finally, enriched phosphopeptides were subjected to LC-MS/MS for the identification of the quantitative phosphoproteome.

Figure 2.

A summary of the identified and quantified phosphoproteome and p-sites. (A) The distribution of MS/MS spectral counts (≤ 30) of phosphopeptides identified in this study. (B) The numbers of p-Ser, p-Thr and p-Tyr residues. (C) The number distribution of p-sites in quantified phosphoproteins. (D) The overlap of p-sites identified in this study compared with known p-sites in public databases. (E) The distribution of SILAC ratios for quantified p-sites in Cory- and Cory B-treated N2a cells against the control group. (F) The number of significantly upregulated (≥ 2-fold) or downregulated (≤ 0.5-fold) p-sites in the Cory or Cory B groups relative to the control. (G) The overlap of phosphoproteins with at least one ≥ 2-fold changed p-site against curated autophagy regulators in the THANATOS database. (H) The GO-based enrichment analysis of biologic processes that are differentially regulated by Cory (left) and Cory B (right) using GSEA (p-value < 0.05).

Figure 3.

The iKAP algorithm to profile the change in kinase activity. (A) The in silico prediction of ssKSRs and the integration of protein-protein interactions between kinases and substrates were performed using our previously reported iGPS 1.0 method. (B) A WKSPN was modeled using the ssKSRs and the quantification information of each phosphopeptide. (C) All single-kinase networks were extracted out and each was further separated into 2 subnetworks, including the upregulated network (ratio > 1) and downregulated network (ratio < 1). (D) The Chi-square test was performed to calculate the significant changes in kinase activity.

Figure 4.

The phosphoproteome-based identification of protein kinases potentially involved in neuronal autophagy. We used the iKAP algorithm for the prediction of differentially regulated kinases (p-value < 0.05) and adopted mouse proteins annotated in THANATOS to filter potentially false positive hits for the (A) Cory and (B) Cory B groups. (C) The list of kinases differentially regulated by Cory. (D) The differentially regulated kinases induced by Cory B. (E) The total (t-) protein and phosphorylation (p-) levels of RPS6KB1 were measured to probe the RPS6KB1 activity dynamics upon Cory or Cory B treatment. (F) Cory but not Cory B decreases the phosphorylation level and the kinase activity of RPS6KB1 (***p < 0.001, Cory vs. Ctrl). (G) The total protein and phosphorylation levels of MAP2K2 were measured. (H) Cory increases the phosphorylation level and kinase activity of MAP2K2 (*p < 0.05; ns, not significant). (I) The total protein and phosphorylation dynamics of PLK1. (J) Cory increases the PLK1 activity (*p < 0.01, Cory vs. Ctrl).

Figure 5.

MAP2K2 and PLK1 are involved in the regulation of compound-induced neuronal autophagy. (A) The changes in LC3B-II upon the Cory and Cory B treatments, after silencing MAP2K2 and PLK1. (B) Quantifications of the LC3B-II:ACTB ratio in the Cory or (C) Cory B group. (D) The changes in SQSTM1 treated with Cory and Cory B, after the silencing of MAP2K2 and PLK1. (E) Quantifications of the SQSTM1:ACTB ratio in the Cory or (F) Cory B group. *p < 0.5; ***p < 0.001; ns, not significant; NT, nontargeting.

Figure 6.

The clearance of overexpressed APP and SNCA by Cory was greatly diminished through the inhibition of MAP2K2 or PLK1. (A) U0126 inhibits the phosphorylation and kinase activity of MAP2K2. (B) BI2356 reduces the PLK1 activity. (C) The changes in LC3B-II and SQSTM1 treated with Cory, after the inhibition of MAP2K2 or PLK1. (D) Quantifications of the SQSTM1:ACTB ratio. (E) Quantifications of the LC3B-II:ACTB ratio. (F) The clearance of FI-APP by Cory, after the inhibition of MAP2K2 or PLK1. (G) Quantifications of the FI-APP:ACTB ratio. (H) The clearance of mutant SNCA by Cory, after the inhibition of MAP2K2 or PLK1. (I) Quantifications of the SNCA:ACTB ratio. **p < 0.01; ***p < 0.001.

Figure 7.

A model of Cory-induced neuroprotective autophagy through the upregulation of MAP2K2 and PLK1 kinase activity. MAP2K2 is important for the induction of autophagy, whereas PLK1 might participate in maturation of the autophagosome.