Helen Frankenthaler Foundation

Phosphopeptide research substrate

Dynamic Equilibrium of Protein Phosphorylation by Kinases and Phosphatases Visualized by Phos-Tag SDS-PAGE

Abstract

The phosphorylation state of 20 types of intracellular proteins in the presence of the protein phosphatase 1 (PP1)- and PP2A-specific Ser/Thr phosphatase inhibitor calyculin A or the Tyr phosphatase inhibitor pervanadate was visualized by Phos-tag SDS-PAGE followed by immunoblotting. All blots showed a Phos-tag pattern indicating increased phosphorylation in the presence of one or both phosphatase inhibitors. The increase in phosphorylation stoichiometry per protein tends to be greater for Ser/Thr phosphatase inhibition than for Tyr phosphatase inhibition. This is consistent with the fact that the number of Ser/Thr kinase genes in the human genome is greater than that of Tyr kinases and with the fact that the phospho-Ser/phospho-Thr ratio in the actual human phosphoproteome is far greater than that of phospho-Tyr ratio. This suggests that cellular proteins are routinely and randomly phosphorylated by different kinases with no biological significance, simply depending on the frequency of substrate encounters. Phosphatase is responsible for routinely removing these unwanted phosphate groups systematically and maintaining the dynamic equilibrium of physiological protein phosphorylation. Phos-tag SDS-PAGE visualized that the kinase reaction involves many incidental phosphorylation and that phosphatases play broader roles besides being strict counterparts to kinases.

1. Introduction

Protein phosphorylation plays a critical role in the regulation of fundamental cellular processes in all living cells. The phosphorylation state of proteins is constantly maintained in dynamic equilibrium by the action of kinases and phosphatases. Protein kinases are well characterized, whereas protein phosphatases have been less discussed. This may be because phosphorylation by kinases is a signaling switch-on reaction, whereas phosphatases are thought to have the passive function of signaling switch-off by dephosphorylation.

Previously, our phosphorylation analysis of intracellular proteins using Phos-tag two-dimensional fluorescence difference gel electrophoresis (2D DIGE) showed that many proteins switch to a hyperphosphorylated state in the presence of the Ser/Thr phosphatase inhibitor calyculin A. Calyculin A is a cell-permeable inhibitor of protein phosphatase 1 (PP1) and protein phosphatase 2A (PP2A) which binds to the catalytic subunit of PP1 and PP2A. The IC 50 s for PP1 and PP2A are 2 nM and 0.5–1 nM, respectively, with potent cytotoxicity at the nanomolar level. Among the Ser/Thr phosphatases, PP1 and PP2A are ubiquitously and abundantly expressed in various types of eukaryotic cells and are involved in many of the universal biological activities, such as the cell cycle, metabolism, cytoskeletal regulation, ion channel and membrane receptor regulation, transcription, cell signaling, and cell differentiation. 2D DIGE suggested that many proteins, despite routinely, randomly, and accidentally acting as substrates for various Ser/Thr kinases, are returned to their normal phosphorylation state by PP1 and PP2A, which are responsible for removing unintended phosphate groups. Mass spectrometry of gel spots after 2D DIGE showed that cytoskeletal proteins such as lamins, keratins, and vimentin are hyperphosphorylated by calyculin A. While 2D DIGE only visualized proteins with high intracellular abundance, the present study used Phos-tag SDS-PAGE followed by immunoblotting to investigate the changes in phosphorylation states of a larger number of proteins with low intracellular abundance, such as proteins involved in signal transduction. In this study, we discuss the broader role of PP1 and PP2A, beyond simply switching off proteins activated by kinases.

To further discuss the role of Tyr phosphatase, changes in the phosphorylation state of intracellular proteins by the Tyr phosphatase inhibitor pervanadate were also examined in a similar manner. Protein tyrosine phosphorylation regulates cellular signaling pathways underlying a wide range of fundamental physiological processes. Tyr kinases and Tyr phosphatases work in a coordinated manner to regulate reversible phosphorylation reactions that occur within seconds to minutes. The human genome encodes 90 Tyr kinases and 107 putative protein Tyr phosphatases. The almost identical number of Tyr kinase and Tyr phosphatase genes suggests similar levels of complexity between the two families. The diversity of Tyr phosphatases suggests that they are highly specific in function and substrate recognition in the regulation of signaling. Protein tyrosine phosphatases are a highly diverse family of enzymes, defined by the active-site signature motif His-Cys-X-X-X-X-X-Arg, in which the Cys acts as a nucleophile and is essential for catalysis. In vivo Tyr phosphatase activity is irreversibly inhibited by pervanadate, which oxidizes Cys at the catalytic site. In this paper, we discuss the role of Tyr phosphatases in contrast to the role of Ser/Thr phosphatases.

2. Results

2.1. Selection of Target Proteins

The total lysate of HeLa cells was subjected to conventional SDS-PAGE and Phos-tag SDS-PAGE followed by immunoblotting with different types of specific antibodies for cellular proteins. Twenty antibodies that specifically detected the protein of interest with an excellent signal-to-noise ratio were selected. Therefore, although the proteins analyzed in this study were randomly selected, they were categorized into the following nine types: (1) proteins related to the MAPK pathway, namely, A-Raf, ATF2, JNK1, MAPKAPK2, MEK1, p38 MAPK, and p42 MAPK; (2) proteins related to the JAK-STAT pathway, namely, JAK2, STAT1, STAT3, and STAT6; (3) proteins related to the mTOR pathway, namely, mTOR, Raptor, and Rictor; (4) a Wnt signaling pathway-related protein, β-catenin; (5) a cell cycle-related protein, CDK2, (6) a glycogen synthesis-related protein, GSK-3β; (7) a tumor suppressor protein, p53; (8) a phosphoinositide 3-kinase, PI3 kinase p110β; and (9) serum response factor, SRF. The number of phosphorylation sites and the estimated upstream kinases in human cells registered in the online database of post-transcriptional modifications, PhosphoSitePlus, for the 20 proteins are summarized in Table 1.

Table 1.

Number of potential phosphorylation sites and putative upstream kinases in human cells registered in PhosphoSitePlus database 1, for the 20 proteins analyzed.

2.2. Phosphorylation State of MAPK Pathway-Related Proteins in the Presence of the Phosphatase Inhibitor

The Phos-tag SDS-PAGE patterns of MAPK pathway-related proteins, namely, A-Raf, ATF2, JNK1/3, MAPKAPK2, MEK1, p38 MAPK, and p42 MAPK, in the presence of calyculin A or pervanadate are shown in Figure 1. Each blot is described below.

Figure 1.

Phosphorylation state of MAPK pathway-related proteins in the presence of phosphatase inhibitor. Total lysates of HeLa cells treated with calyculin A (lane 1) or pervanadate (lane 2) were subjected to Phos-tag SDS-PAGE together with untreated control lysate (lane C). The antibody used for the blot is indicated below each panel. Rf values indicating mobility are shown to the left of each panel. The arrowhead indicates the position of the non-phosphorylated form. The arrows on the left or right of each panel indicate newly observed bands or changes in band intensity in lane 1 or lane 2, respectively, compared to lane C. Phos-tag SDS-PAGE (8% w/v polyacrylamide, 20 µM Zn 2+–Phos-tag) was performed using a neutral pH gel system buffered with Bis-Tris–HCl.

A-Raf: A-Raf, B-Raf, and C-Raf are the Ser/Thr kinases, which are the main effectors recruited by GTP-bound Ras to activate the MEK-MAPK pathway. In untreated cells (lane C), multiple bands were detected, indicating that several sites are constitutively phosphorylated under homeostatic conditions. In calyculin A-treated cells (lane 1), only a few bands remained in the same position as in control cells, and several significantly up-shifted bands were detected. This suggests that multiple sites are phosphorylated by multiple Ser/Thr kinases. Putative upstream Ser/Thr kinases in vivo have not been deposited in PhosphoSitePlus (see Table 1). However, since A-Raf is similar in sequence and function to C-Raf, and several kinases are involved in the phosphorylation of the activation sites, it is likely that there are several upstream Ser/Thr kinases similar to C-Raf. In the pervanadate-treated cells (lane 2), the banding pattern was almost the same as in the control cells. This suggests that A-Raf is not phosphorylated by any Tyr kinases.

ATF2: cyclic AMP-dependent transcription factor 2 (ATF2) interacts with a variety of viral oncoproteins and cellular tumor suppressors and is a target of the SAPK/JNK and p38 MAP kinase signaling pathways. ATF2 is phosphorylated by activated p38 MAP kinase. In untreated cells (lane C), multiple bands were detected, indicating that several sites are constitutively phosphorylated under homeostatic conditions. In cyclin A-treated cells (lane 1), few bands remained in the same position as in control cells, and one exaggeratedly shifted band was detected. This suggests that multiple sites are phosphorylated by multiple Ser/Thr kinases. In pervanadate-treated cells (lane 2), three up-shifted bands were detected. The Tyr phosphorylation sites were not registered in PhosphoSitePlus (Table 1), whereas the protein may contain potential substrates for multiple tyrosine kinases and phosphorylated by multiple Tyr kinases.

JNK1: Stress-activated protein kinase (SAPK)/Jun-amino-terminal kinase (JNK) is potently and preferentially activated by a variety of environmental stresses. Activation of JNK1 occurs via the phosphorylation of Thr183/Tyr185 by MKK4 and MKK7. In untreated cells (lane C), distinct bands due to the non-phosphorylated 46 kDa and 54 kDa isoforms and several up-shifted low-intensity bands were detected. In the calyculin A-treated cells (lane 1), several newly appearing phosphorylated bands were detected. This suggests that multiple sites are phosphorylated by multiple Ser/The kinases. In the pervanadate-treated cells (lane 2), one new phosphorylated band was detected. JNK1 would be phosphorylated by several Tyr kinases, whereas the stoichiometry of Tyr phosphorylation at a particular site is less than 10%, as estimated from the intensity of the phosphorylated band in lane 2.

MAPKAPK2: MAP kinase-activated protein kinase 2 (MAPKAPK2) is rapidly phosphorylated by p38 MAPK and activated in response to cytokines, stress, and chemotactic factors. Activation of MAPKAPK2 occurs through the phosphorylation of Thr222, Ser272, and Thr334 by p38 MAPK. In untreated cells (lane C), multiple bands were detected, indicating that several sites are constitutively phosphorylated under homeostatic conditions. In the calyculin A-treated cells (lane 1), few bands remained in the same position as in control cells and two newly phosphorylated bands were detected. In the pervanadate-treated cells (lane 2), four newly appeared phosphoryla