Reactive sulfur species (RSS) modulate protein functions via S-polysulfidation of reactive Cys residues. Here, we report that Ca2+/calmodulin (CaM)-dependent protein kinase IV (CaMKIV) was reversibly inactivated by RSS via polysulfidation of the active-site Cys residue. CaMKIV is phosphorylated at Thr196 by its upstream CaMK kinase (CaMKK), resulting in the induction of its full activity. In vitro incubation of CaMKIV with the exogenous RSS donors Na2Sn (n = 2–4) resulted in dose-dependent inhibition of the CaMKK-induced phospho-Thr196 and consequent inactivation of the enzyme activity. Conversely, mutated CaMKIV (C198V) was refractory to the Na2Sn-induced enzyme inhibition. A biotin-polyethylene glycol-conjugated maleimide capture assay revealed that Cys198 in CaMKIV represents a target for S-polysulfidation. Furthermore, phosho-Thr196 and CaMKIV activity were inhibited by incubation with cysteine hydropersulfide, a newly identified RSS that is generated from cystine by cystathionine-γ-lyase. In transfected cells expressing CaMKIV, ionomycin-induced CaMKIV phosphorylation at Thr196 was decreased upon treatment with either Na2S4 or the endoplasmic reticulum (ER) stress inducer thapsigargin, whereas cells expressing mutant CaMKIV (C198V) were resistant to this treatment. In addition, the ionomycin-induced phospho-Thr196 of endogenous CaMKIV was also inhibited by treatment either with Na2S4 or thapsigargin in Jurkat T lymphocytes. Taken together, these data define a novel signaling function for intracellular RSS in inhibiting CaMKIV activity via S-polysulfidation of its Cys198 during the response to ER stress.
Reactive sulfur species (RSS) comprise a group of sulfur-containing molecules that play regulatory roles in biological systems. Among various RSS, cysteine hydropersulfide (CysSSH) and cysteine hydropolysulfide (CysSSnH) are produced by cystathionine-γ-lyase (CSE) using cystine (CysSSCys) as a substrate in cultured cells and in vivo . These RSS act as potent scavengers for reactive oxygen species and show a strong redox signaling regulatory function via electrophile thiolation. Notably, the chemical reactivity of RSS is thought to be involved in the catalytic activity of particular proteins. In particular, a wide variety of proteins are regulated by the RSS-induced modification S-polysulfidation, including actin, tubulin, GAPDH, nuclear factor-κB, protein tyrosine phosphatase 1B, phosphatase with sequence homology to tensin, and mitogen-activated protein kinase kinase 1 [2–6]. Thus, the identification of novel S-polysulfidated proteins has become an emerging theme in the analysis of the biological significance of RSS.
Ca2+/CaM (calmodulin)-dependent kinases (CaMKs) play critical roles in processing transduction by enhancing intracellular Ca2+ levels [7,8]. Among them, multifunctional CaMKs, such as CaMKII and members of the CaMK cascade, are present in most mammalian tissues, but are particularly abundant in the brain, where they phosphorylate and regulate numerous protein substrates . In addition to CaMKII, the CaMK cascade consists of CaMK kinase (CaMKK) and its downstream substrates CaMKI and CaMKIV. Binding of Ca2+/CaM to CaMKI, CaMKII, and CaMKIV allows autophosphorylation of CaMKII or phosphorylation of CaMKI and CaMKIV by the upstream CaMKK. This phosphorylation increases the Ca2+/CaM-dependent activity of CaMKI and CaMKIV but not of CaMKII. In turn, CaMKII and CaMKIV, but not CaMKI, can also exhibit significant Ca2+-independent activity [8,10]. This autonomous activity of CaMKII and CaMKIV is thought to be critical in several physiological situations, especially for the potentiation of synaptic transmission during learning and memory [11,12]. Thus, the regulation of CaMKs by conventional means requires binding of Ca2+/CaM to and phosphorylation of the enzymes to initiate the process.
CaMKII activities also increase in oxidant cellular environments [13–15] and are directly affected by oxidative stress . Previously, we have reported other mechanisms of CaMK regulation, such as S-glutathionylation of CaMKI at Cys179  and S-nitrosylation of CaMKII at Cys6/30 . In addition, the activity of CaMKI, but not CaMKIV, is inhibited by nitric oxide . It had been reported that CaMKIV activities also increase in oxidant cellular environments ; however, the direct mechanism affecting CaMKIV remained unknown. In the present study, we investigated the molecular mechanisms of RSS-dependent regulation of CaMKIV in CaMKIV-expressing cells. Taken together, our results demonstrated that S-polysulfidation at the active cysteine residue mediates inactivation of the CaMKIV enzyme.
Recombinant GST-fused CaMKIV , CaMKKα , and rat CSE  were expressed in Escherichia coli DH5α and purified as described previously. Purification of GST-cleaved CaMKIV was performed using PreScission protease and CaM-agarose chromatography as described previously . Recombinant rat CaM was expressed in E. coli BL21 (DE3) using pET-CM, kindly provided by Dr Nobuhiro Hayashi (Department of Life Science, Graduate School of Bioscience and Biotechnology, Tokyo Institute of Technology, Yokohama, Japan) . The expression vector for hemagglutinin (HA)-tagged CaMKIV (pME18s-HA-CaMKIV) was kindly provided by Dr Hiroshi Tokumitsu (Okayama University, Okayama, Japan). The monoclonal antibody specific to phosphorylated CaMKIV (p-CaMKIV, threonine 196) was provided by Dr Naohito Nozaki (MAB Institute, Inc., Hokkaido, Japan). The rabbit anti-CaMKIV polyclonal antibody (pAb) was obtained from Cell Signaling Technology (Danvers, MA, U.S.A.). The rabbit anti-CSE pAb was provided by Dr Isao Ishii (Department of Health Chemistry, Showa Pharmaceutical University, Tokyo, Japan) . Na2S2, Na2S3, Na2S4, and 3′,6′-di(O-thiosalicyl) fluorescein (SSP4) were obtained from Dojindo Laboratories (Kumamoto, Japan). NaHS, dimethyl trisulfide (DMTS), and STO-609, a selective inhibitor of CaMKK, were purchased from Sigma–Aldrich (St Louis, MO, U.S.A.). ECL prime (enhanced chemiluminescence) immunoblotting detection reagents were from GE Healthcare (Piscataway, NJ, U.S.A.). All other materials and reagents were of the highest quality available from commercial suppliers.
Inactivation of CaMKIV by RSS
To generate CysSSH, recombinant CSE (100 µg/ml) was incubated in 30 mM HEPES buffer (pH 7.5) containing 50 μM pyridoxal phosphate with buffer alone or 1 mM cystine for 60 min at 37°C. Upon removing CSE from the products resulting from the reaction above, we obtained pass-through fractions from the ultrafiltration membrane device (Amicon Ultra-0.5 ml 30K centrifugal filters; Millipore, Darmstadt, Germany). Recombinant CaMKIV (20 µg/ml) was incubated with one-tenth of the products above for 30 min at 30°C in 30 mM HEPES (pH 7.5) and 400 µg/ml bovine serum albumin. Recombinant CaMKIV (20 µg/ml) was also incubated with increasing amounts of polysulfides (Na2Sn, n = 2–4) (0–1 mM) or NaHS (0–10 mM) for 10 min at 30°C in 30 mM HEPES (pH 7.5) and 400 µg/ml bovine serum albumin. Then, pretreated CaMKIV (10 µg/ml) was activated with CaMKKα (2 µg/ml) for 15 min at 30°C in 80 mM HEPES (pH 7.5), 5 mM MgCl2, 1 mM CaCl2, 10 µM CaM, and 400 µM ATP. An aliquot of reaction mixture was then removed and analyzed for activity. Kinase activity was measured at 30°C for 10 min in 40 mM HEPES (pH 7.5), 10 mM MgCl2, 1.5 mM CaCl2, 1 µM CaM, 10 µM [γ-32P] ATP, 50 µM synthetic peptide Syntide-2 (PLARTLSVAGLPCKK), and CaMKIV (5 µg/ml) in a final volume of 25 µl. 32P incorporation was determined by spotting 20 µl of aliquots onto Whatman P-81 phosphocellulose paper followed by washing in 75 mM phosphoric acid .
Determination of enzymatically synthesized CysSSH
Levels of enzymatically synthesized CysSSH were determined with the use of SSP4 in a qualitative assessment. Samples of interest were reacted with 50 µM SSP4 in 20 mM Tris–HCl (pH 7.4) in the presence of 1 mM cetyltrimethylammonium bromide in the dark for 10 min at room temperature. Fluorescence intensities of the resultant solutions were determined using a microplate reader (Varioskan Flash, Thermo Fisher Scientific, Waltham, MA, U.S.A.) with excitation wavelength 482 nm and emission wavelength 518 nm.
Construction of plasmids
The CaMKIV mutant, C198V (in which Cys198 was replaced with Val), was generated using pGEX6P and pME18s-HA CaMKIV plasmid DNA and a QuickChange II site-directed mutagenesis kit (Stratagene, La Jolla, CA, U.S.A.). The C198V mutant was confirmed by sequencing analysis.
Cell culture, transfection, and stimulation
HEK293 cells were maintained in Dulbecco's modified Eagle medium containing 10% fetal calf serum and subcultured for 24 h in 6-cm dishes in a humidified atmosphere at 37°C. They were then transfected with the pME18s-HA-CaMKIV construct (1 μg) using LipofectAMINE LTX (Life Technologies, Inc., Carlsbad, CA, U.S.A.). After 24–36 h incubation, the cells were serum-starved for 18 h and preincubated with buffer alone, Na2S4 for 10 min, DMTS for 10 min or 1 µM thapsigargin for 18 h, and stimulated with or without 10 µM A23187 for 3.5 min.
The human T-lymphocyte cell line Jurkat was maintained in RPMI 1640 medium containing 10% fetal calf serum in 10-cm dishes in a humidified atmosphere at 37°C. Cells were washed and resuspended in serum-free RPMI 1640. A 1-ml aliquot containing 1.25 × 106 cells was added to microcentrifuge tubes and warmed at 37°C for at least 1 h prior to the start of the experiment. Cells were treated with buffer alone, Na2S4 for 10 min, or 1 µM thapsigargin for 8 h, and stimulated with or without 10 µM A23187 for 3.5 min. The tubes were then centrifuged in a microcentrifuge for 30 s and the supernatants were removed. Cell pellets were then resuspended in lysis buffer.
Preparation of cerebellar granule cells (CGCs) was conducted in compliance with the Guideline for Animal Experimentation at Osaka Prefecture University, with an effort to minimize the number of animals used and their suffering. CGCs were obtained from 7- to 10-day-old Wistar rats as described previously . Cells were plated at a density of 2.0 × 106 cells per 6-cm dish to prepare cell extracts and maintained in modified Eagle's medium containing 25 mM K+ (HK-MEM) with 5% fetal calf serum, 5% horse serum, and 1% penicillin/streptomycin in a humidified atmosphere at 37°C. After 14 days in vitro, cells were used for experiments. CGCs were preincubated for 60 min with HK-MEM and then washed with Low K+ Ringer's solution [140 mM NaCl, 5 mM KCl, 1.2 mM KH2PO4, 2.5 mM CaCl2, 1.2 mM MgSO4, 11 mM glucose, and 15 mM HEPES–NaOH (pH 7.4)]. After being stimulated with or without Na2S4 for 10 min, the solution was changed to High K+ (25 mM KCl) or Low K+ (5 mM KCl) Ringer's solution and incubated for 10 min in a humidified atmosphere at 37°C.
S-polysulfidated CaMKIV using a biotin-polyethylene glycol-conjugated maleimide capture method
S-polysulfidated proteins were detected using a biotin-polyethylene glycol-conjugated maleimide (biotin-PEG-MAL) capture method. Briefly, cells were lysed with ice-cold RIPA buffer [50 mM Tris–HCl (pH 7.5), 150 mM NaCl, 1 mM phenylmethane sulfonyl fluoride, 10 µg/ml aprotinin, 1 mM sodium orthovanadate, 25 mM sodium fluoride, 10 mM sodium pyrophosphate, 5 mM EDTA, 0.5% sodium deoxycholate, 0.1% SDS, and 1% Nonidet P40] containing 1 mM biotin-PEG-MAL  and a proteinase inhibitor cocktail. After whole biotinylated proteins in the lysates were captured and enriched with streptavidin agarose (Thermo Fisher Scientific, MA, U.S.A.), polysulfidated proteins were collected and subjected to western blotting for CaMKIV.
Recombinant CSE (100 µg/ml) was incubated in 30 mM HEPES buffer (pH 7.5) containing 50 μM pyridoxal phosphate and 1 mM cystine for 60 min at 37°C. The peptide CaMKIV194-203 was incubated with buffer alone or one-tenth of the above reaction mixture for 30 min at 30°C in 30 mM HEPES (pH 7.5) and then mixed with 5 mM iodoacetamide and incubated at 37°C for 60 min. The samples were loaded onto a C18 Sep-Pack cartridge (Waters, Milford, MA, U.S.A.) equilibrated and washed with 2% acetonitrile containing 0.1% formic acid to remove the non-peptide components. The peptide was eluted with 65% acetonitrile containing 0.1% formic acid and subjected to an Agilent 6500 Q-TOF MS (Santa Clara, CA, U.S.A.). We set the search parameters as follows: variable modifications, S-acetamide (+57 Da), SS-acetamide (+89 Da), and SS-cysteine (+119 Da).
Western blot analysis
For preparation of lysates, cells were homogenized by sonication in RIPA buffer. Total cell lysates were centrifuged at 15 000×g for 15 min at 4°C. Supernatants (20 µg) were separated by SDS–polyacrylamide gel electrophoresis (PAGE) and blotted onto polyvinylidene fluoride membranes. The membranes were incubated for 1 h at room temperature with anti-phospho-CaMKIV (1 : 20), anti-CaMKIV (1 : 1000), or anti-CSE (1 : 1000) antibodies, followed by anti-horseradish peroxidase-linked secondary antibodies. Blots were developed using ECL prime detection reagent. Band intensity was quantified using the FUJIFILM MultiGauge software (Tokyo Japan).
All results are represented as the means ± SE of at least three determinations. The statistical evaluation was performed using a one-way ANOVA test. We considered P < 0.05 to be statistically significant.
CaMKIV is inactivated by treatment with RSS
For full activation, CaMKIV requires the phosphorylation of a crucial activation loop at Thr196 by an upstream kinase, CaMKK . In initial experiments, we determined the effects of RSS on CaMKIV activation and phosphorylation of Thr196 using an anti-phospho-Thr196 monoclonal antibody. Both the activation and Thr196 phosphorylation of CaMKIV in the absence of CaMKK are both negligible (Supplementary Figure S1). Recombinant mouse CaMKIV (20 ng/µl) was incubated with 100 µM Na2S4 for 10 min and then activated by CaMKK. The activity of CaMKIV was measured using a synthetic peptide, Syntide-2, as a substrate. As shown in Figure 1A, both the activation and Thr196 phosphorylation of CaMKIV by CaMKK were almost completely inhibited by Na2S4. The addition of 20 mM dithiothreitol (DTT), a small molecule-reducing agent, completely reversed these inhibitions. As cells contain various sulfur species including persulfides and polysulfides , we next clarified which sulfur species were actually involved in the regulation of CaMKIV in cells by carrying out comparative studies between the effects of Na2S2, Na2S3, and Na2S4 on CaMKIV activation. CaMKIV (20 ng/µl) was incubated with Na2S2–4 (0.01–1 mM) or NaHS (0.01–10 mM), a H2S donor, for 10 min and then activated by CaMKK. All the sulfur species inhibited the activity of CaMKIV with approximate IC50 values of 20 µM for Na2S4, 50 µM for Na2S3, 200 µM for Na2S2, and 2 mM for NaHS (Figure 1B).
CaMKIV is reversibly inactivated by Na2S4.
CaMKIV is inactivated by treatment with an enzymatically synthesized CysSSH
CSE is the last key enzyme in the transsulfuration pathway for the biosynthesis of cysteine. In addition, this enzyme is capable of directly generating CysSSH or H2S using either cystine or cysteine as a substrate, respectively  (Figure 2A). We incubated CSE either with cystine or cysteine and then examined the CysSSH level with a sulfane sulfur-specific fluorescent probe, SSP4 (Figure 2B). There was a significant increase in SSP4 fluorescence when CSE was incubated with cystine but not with cysteine as its substrate. Since it is not expected for CSE to produce RSS using cysteine as a substrate, no significant increase in SSP4 fluorescence was observed when CSE was incubated with cysteine. Next, we assessed whether this enzymatically synthesized CysSSH may be involved in the regulation of CaMKIV activity. As shown in Figure 2B, cystine alone had minimal effects on both CaMKK-induced activation and Thr196 phosphorylation of CaMKIV. However, when CaMKIV was incubated with the products resulting from the CSE/cystine but not from CSE/cysteine reaction, there was a significant effect on inhibition (32.76 ± 1.64% of control activity and 35.16 ± 5.23% of control phospho-Thr196). Conversely, the addition of 20 mM DTT completely reversed cystine- and CSE/cystine-induced inhibition (Supplementary Figure S2). To test whether CSE is required to be in close proximity to its targets, CaMKIV and/or CaMKK, for delivering RSS, we removed CSE from the products resulting from the CSE/cystine reaction using an ultrafiltration technique prior to the incubation with CaMKIV. As shown in Figure 2C, removal of CSE from the reaction mixture had no effects on the inhibition of phosho-Thr196 and enzyme activity, indicating that the binding of CSE to RSS targets is not necessary for the inhibition of CaMKIV activity. Next, we tested whether CaMKK-induced phosphorylation at Thr196 of CaMKIV affects enzymatically synthesized CysSSH-induced inactivation of the enzyme. We first pretreated CaMKIV either with or without a CaMKK-containing reaction mixture and analyzed enzymatically synthesized CysSSH-induced CaMKIV inhibition. As shown in Figure 2D, prior phosphorylation of CaMKIV with CaMKK prevented subsequent enzymatically synthesized CysSSH-induced inhibition.
Inhibition of CaMKIV activity by an enzymatically synthesized CysSSH.
Cys198 is an essential site for inactivation of CaMKIV by RSS
Based on the above observations, CaMKIV and/or CaMKK may be responsible for the targeted effects of RSS. We previously showed that Ca2+/CaM-dependent protein kinase I is fully and reversibly inactivated by oxidative S-glutathionylation at Cys179 . Cys179, which is located in subdomain VIII, is considered as a key protein kinase catalytic domain indicator  that is highly conserved among several kinases including CaMKIV (i.e. Cys198, based on the mouse CaMKIV sequence). Therefore, we generated a mouse CaMKIV in which Val was substituted for Cys198 (C198V) and characterized both its activity and sensitivity to inhibition by Na2S4. We constructed and purified CaMKIVs by site-specific separation of the GST tag from proteins expressed using pGEX6P vectors. Both CaMKIVs were at least 90% pure and gave a major band at ∼60 kDa on SDS–PAGE with Coomassie Brilliant Blue staining (Figure 3A). The C198V mutant was similar to wild type in terms of activation by CaMKK (167.26 ± 0.57 and 145.53 ± 2.75 nmol/min/mg protein for the C198V mutant and wild type, respectively). However, a minimal decrease in activity was noted with the C198V mutant with increasing amounts of Na2S4 (0–300 µM), whereas the wild-type protein was significantly inactivated (Figure 3B). The treatment of CaMKIV wild type with Na2S4 (100 or 300 µM) led to a decrease in Thr196 phosphorylation by CaMKK (Figure 3B). In contrast, the C198V mutant proved resistant to Na2S4 treatment.
Generation of an RSS-resistant CaMKIV mutant.
Cys198 represents the site of S-polysulfidation on CaMKIV in cells
The above experiments clearly demonstrate the susceptibility of Cys198 in CaMKIV to RSS in vitro. It was therefore of great interest to examine the ability of CaMKIV to be inhibited within the cellular environment under RSS-loading conditions. To determine the ability of RSS to inhibit CaMKIV activity, HEK-293 cells, transfected with an HA-tagged wild-type or C198V CaMKIV, were treated with the combination of Na2S4 and A23187. HA-tagged CaMKIV was then immunoprecipitated, and Ca2+/CaM-dependent activity was measured using a synthetic peptide, Syntide-2, as a substrate. A23187-induced phosphorylation of Thr196 was also determined using an anti-phospho-Thr196 CaMKIV antibody (Figure 4A). The Ca2+/CaM-dependent activity of immunoprecipitated HA-tagged wild type or C198V CaMKIV was greatly enhanced by stimulation with A23187, indicative of CaMKK activity in HEK293 cells. The Ca2+-induced HA-tagged wild-type CaMKIV activation is suppressed by Na2S4. In contrast, cells expressing C198V proved resistant to Na2S4. A23187-induced phosphorylation of Thr196 was also evident in both cells. Pretreatment of the cells expressing wild-type CaMKIV with Na2S4 caused an inhibition of A23187-induced Thr196 phosphorylation. In contrast, cells expressing C198V proved resistant to Na2S4. We also tested dose-dependent effects of Na2S4 on CaMKIV activity using an anti-phospho-Thr196 CaMKIV antibody. Pretreatment of the cells expressing wild-type CaMKIV with Na2S4 caused a dose-dependent inhibition of A23187-induced Thr196 phosphorylation (Figure 4B). In contrast, cells expressing C198V proved resistant to Na2S4. We also employed garlic-derived sulfur compounds, DMTS, as the RSS. Pretreatment of the cells with DMTS appeared to exert similar effects on the A23187-induced phosphorylation at Thr196 (Figure 4C).
Effect of RSS donors on the activity and the phospho-Thr196 of CaMKIV in cells.
S-polysulfidation is frequently found in proteins, which may explain the diverse protein functions of the specific enzymes, which include both activation and inactivation. We examined whether Cys198 is accessible to S-polysulfidation by the treatment with Na2S4 using a biotin-PEG-MAL capture method. The treatment of cells with Na2S4 led to an increase in CaMKIV S-polysulfidation (Figure 4D). C198V displayed a reduction in CaMKIV S-polysulfidation by Na2S4 relative to the wild-type enzyme, suggesting that Cys198 constitutes one of the major S-polysulfidation sites in CaMKIV in cells.
Mass spectrometry analysis to identify S-polysulfidation at Cys198 of CaMKIV
We used a high-resolution quadrupole time of flight (QTOF) mass spectrometer to analyze recombinant CaMKIV. Because S-polysulfidated peptides are metastable , they were alkylated by iodoacetamide prior to the analysis by mass spectrometry (MS). Recombinant CaMKIV protein (∼2 μg each of untreated, CSE/cystine-treated, or Na2S3-treated) was mixed with 5 mM iodoacetamide at 37°C for 60 min and then digested overnight with chymotrypsin (Roche Diagnostics, Penzberg, Germany) (1 : 20 w/w) at 25°C for 16 h. The digests thus produced were introduced by electrospray ionization into a QTOF mass spectrometer. Electrospray ionization MS spectra of the peptide (CaMKIV194–203: MKTVCGTPGY) from untreated CaMKIV shows the +2 charged state of the S-acetamide peptide (m/z 557.25). An expected mass of S-S-acetamide peptide (m/z 573.24) from CSE/cystine-treated or Na2S3-treated CaMKIV was not observed. The peak intensity of the S-acetamide peptide from Na2S3-treated CaMKIV was decreased to a level ≈10% of the untreated enzyme (Supplementary Table 1). The CaMKIV183–193 peptide (GLSKIVEHQVL: m/z 611.86) from either untreated or Na2S3-treated CaMKIV was adequately recovered. Thus, Na2S3-treated CaMKIV appeared to be resistant against chymotrypsin digestion. Therefore, we then analyzed the synthetic peptide corresponding to CaMKIV194–203 (MKTVC198GTPGY). Using the CaMKIV peptides treated with the products resulting from the CSE/cystine reaction mixture, three products with mass spectra of 1113, 1145, and 1175 Da were detected, corresponding to CaMKIV194–203–S-acetamide, CaMKIV194–203–SS-acetamide, and CaMKIV194–203–SS-Cys adducts. Notably, S-cysteinylation of the peptide (CaMKIV194–203–SS-Cys) was also reliably detectable and was evident when the peptide was treated with cystine alone (Table 1).
|QTOF MS of CaMKIV194–203 .|
|Sample .||Modification .||Expected (m/z) .||Observed (m/z) .||Relative abundance (%) .|
|QTOF MS of CaMKIV194–203 .|
|Sample .||Modification .||Expected (m/z) .||Observed (m/z) .||Relative abundance (%) .|
Effects of RSS on the phosphorylation of CaMKIV at Thr196 in intact cells
The expression of CaMKIV is restricted primarily to discrete regions of the brain, T-lymphocytes, and post-meiotic germ cells [29–31]. We therefore analyzed whether the RSS-induced inactivation of endogenous CaMKIV could be observed in CGCs and in the human T-lymphocyte cell line Jurkat. The treatment of CGCs with High K+ (25 mM KCl) increased the level of phospho-Thr196 in endogenous CaMKIV (Figure 5A), whereas pretreatment with Na2S4 significantly blocked High K+-induced Thr196 phosphorylation. Similarly, A23187-induced phospho-Thr196 in endogenous CaMKIV was inhibited by Na2S4 treatment in Jurkat cells (Figure 5B). We also examined whether cAMP response element-binding protein (CREB) phosphorylation (Ser-133), a well-described nuclear target of CaMKIV, was also inhibited by Na2S4 treatment in Jurkat cells (Figure 5C). The treatment of Jurkat cells with A23187 increased the level of phospho-Ser133 in CREB, whereas pretreatment with Na2S4 significantly blocked this phosphorylation. When cells were pretreated with STO-609 , a selective inhibitor of CaMKK, A23187-induced phospho-Ser133 in CREB was inhibited. Even though CaMKI and CaMKII phosphorylate CREB in vitro, CaMKI is not translocated into the nucleus, and CaMKII-mediated phospho-Ser133 in CREB is not dependent on CaMKK. It is noteworthy that A23187-induced/STO-609-sensitive phosphorylation (Thr-172) of AMP-activated protein kinase (AMPK), a target of CaMKK, was increased by Na2S4. This finding is consistent with the results of previous studies, demonstrating that NaHS increases AMPK activity through the CaMKK pathway [33,34].
Effect of Na2S4 on the phosphorylation of endogenous CaMKIV at Thr196 in cells.
CaMKIV inhibition occurs via its Cys198 modification under ER stress
Recently, a novel signaling cascade was reported to regulate the ER stress response through RSS-induced S-polysulfidation of the protein tyrosine phosphatase . We therefore analyzed whether thapsigargin-induced ER stress inhibited CaMKIV activity. In HEK293 cells expressing CaMKIV, pretreatment with thapsigargin resulted in a decrease in A23187-induced Thr196 phosphorylation (Figure 6A), whereas cells expressing C198V were refractory to the thapsigargin-induced inhibition. We also observed a significant increase in CSE expression in the cells within 18 h of induction of ER stress. In Jurkat cells, pretreatment with thapsigargin resulted in a decrease in A23187-induced Thr196 phosphorylation of endogenous CaMKIV (Figure 6B). However, we could observe a significant increase in S-polysulfidated CaMKIV when we exposed Jurkat cells to Na2S4 but not to thapsigargin (Figure 6C).
Effect of thapsigargin-induced ER stress on the phosphorylation of CaMKIV at Thr196 in cells.
Here, we show for the first time that a member of the CaMK family, CaMKIV, is sensitive to inhibition by RSS. Our data indicate that Cys198 in CaMKIV represents a major target of ER stress signaling and exogenously applied RSS and that S-polysulfidation at this site inhibits the enzyme activity. These results suggest that CSE generates RSS to S-polysulfate CaMKIV rather than directly participating in the polysulfidation process (Figure 2C). This enzyme is also capable of producing cysteine and H2S, which have been shown to influence multidirectional physiological functions. However, neither cysteine nor the products resulting from the CSE/cysteine reaction mixture inhibited CaMKIV activity (Figure 2B). Conversely, a H2S donor, NaHS, led to CaMKIV inhibition with IC50 values of ∼2 mM (Figure 1B). Notably, all commercial reagents (Na2S or NaHS) are thought to contain significant amounts of inorganic polysulfides [35,36]. Therefore, to test which RSSs were actually responsible for the CaMKIV inhibition, we carried out comparative studies between the effects of a series of polysulfides (Na2Sn, n = 2–4). We showed that increasing the number of sulfur atoms led to decreasing IC50 values (Figure 1B), consistent with a recent report showing that increasing the number of consecutive sulfur atoms would lead to increased chemical reactivity owing to the increased nucleophilicity . Meanwhile, it is expected that the speciation of these polysulfides in aqueous solution is similar, regardless of which salt is used to make the reagent due to fast disproportionation equilibria . It might be therefore more likely that the larger sulfur atoms effect was due to overall larger concentration of the total sulfane sulfur content when the longer polysulfide chain salts were used.
It is well known that some protein –SH groups can undergo reversible S-polysulfidation reactions, leading to regulatory allosteric effects. However, a prior S-polysulfidation reaction with CaMKIV at Cys198 prevents subsequent phosphorylation of Thr196 (Figure 2D). Conversely, prior phosphorylation of CaMKIV at Thr196 by its upstream CaMKK resulted in loss of subsequent CaMKIV inhibition, presumably through its S-polysulfidation at Cys198, by RSS (Figure 2D). In contrast, inhibition of CaMKI by oxidation at its homologous residue, Cys179, appears to be dominant over activation of the kinase by phosphorylation at Thr177 . Recently, exogenous NaHS has been shown to confer CaMKK activation in various cells [33,34]; however, our data clearly show that the inhibition of CaMKIV appears to be dominant over activation of CaMKK by RSS in cells (Figures 4 and 5). In particular, Na2S4 treatment of Jurkat cells activates AMPK, a target of CaMKK, whereas it inhibits the phosphorylation of CaMKIV at Thr196 by ionomycin (Figure 5C and Supplementary Figure S3).
We generated the CaMKIV mutant, C198V (in which Cys198 was replaced with Val). A Cys was changed to a to Ser or Cys to a Ala mutant. The activity of mutant C179A or C179S of CaMKI was decreased to a level ≈30% (C179A) or ≈10% (C179S) as much as the wild-type enzyme. Since both the wild-type and the C198V mutant CaMKIV were activated by CaMKK to similar extents, we generated a mutant CaMKIV in which Val was substituted for Cys198 (C198V). The structural difference between valine and alanine/serine may be significant in the conformation of the mutated protein.
Several enzymes have been observed to be S-polysulfidated in vitro and in intact cells, as demonstrated by a modified biotin switch assay and/or MS. The modified biotin switch technique was shown not to be the ideal method to detect protein per/polysulfide species . We then developed a novel method to detect protein polysulfidation using a PEG-MAL method . By densitometric analysis of CaMKIV polysulfidated in a biotin-PEG-MAL capture method, we showed that Cys198 in CaMKIV represents a target for S-polysulfidation, leading to a decrease in its enzyme activity (Figures 4D and 6C). Notably, S-cysteinylation of the peptide (CaMKIV194–203–SS-Cys) was also reliably detectable and was evident when the peptide was treated with cystine alone (Table 1). Thus, the cystine-induced CaMKIV inactivation in the present study (Figure 2B,D) most probably was a consequence of S-cysteinylation (CaMKIV–SS-Cys). We could not detect chymotryptic peptide containing Cys198 from CSE/cystine- or Na2S3-treated CaMKIV by MS (Supplementary Table 1). Several proteins have been observed to be S-polysulfidated, as demonstrated by mass differences between trypsinized [2,4] or chymotrypsinized  native and NaHS-treated enzymes. Thus, resistance to hydrolysis might be a general issue for kinases after S-polysulfidation or just specific to chymotrypsin for S-polysulfidated CaMKIV, and thereby Cys198 modification might interfere with subsequent phosphorylation or dephosphorylation at Thr196 of CaMKIV.
To address the cellular conditions under which CaMKIV may be S-polysulfidated and inactivated, we exposed HEK293 cells expressing CaMKIV or Jurkat cells to thapsigargin to trigger ER stress and observed a significant inhibition of ionomycin-induced activation of CaMKIV (Figure 6A,B). A marked increase was observed in the expression of the ER chaperone-binding immunoglobulin protein and the CCAAT enhancer-binding protein homologous protein, which confirmed the induction of the ER stress response and ATF4 transcriptional response, respectively (Supplementary Figure S4). Activation of PKR-like ER kinase and inositol-requiring enzyme 1α was also revealed by the reduced mobility of their phosphorylated forms compared with the inactive unphosphorylated forms within 2 h of ER stress induction (Supplementary Figure S4). In addition, it has been shown that CSE protein synthesis is up-regulated via the eIF2α–ATF4 pathway . Consistent with this, we could detect an increase in the abundance of CSE in HEK293 cells under the induction of ER stress (Figure 6A). Furthermore, when we treated Jurkat cells with Na2S3, the level of ionomycin-induced phosphorylation of CREB, a well-described nuclear target of CaMKIV, declined progressively with kinetics comparable to those of CaMKIV inhibition (Supplementary Figure S3). Thus, it is conceivable that as the S-polysulfidation and inactivation of CaMKIV are involved in response to ER stress, they may presumably be due to an induction of CSE protein synthesis and may lead to the suppression of CREB-dependent transcriptional activation. However, overexpression of CSE failed to inhibit CaMKIV activity in HEK293 cells (data not shown). In addition, we could not detect an increase in the abundance of CSE in Jurkat cells under the induction of ER stress (data not shown). Alternatively, according to a recent study, intracellular RSS levels could be maintained by the cystine uptake machinery propelled by the appropriate amino acid transporters, such as cystine/glutamate transporter (xCT) . As ATF4 was shown to induce the expression of xCT , activation of the ATF4/xCT module might affect ER stress-induced CaMKIV inactivation that is associated with the ability to generate RSS rather than consequential to the effects of high CSE expression. Although thapsigargin clearly induces a consistent ER stress phenotype (Supplementary Figure S4), the ER stress pathway may not be directly connected with CaMKIV signal in cells.
In summary, enzymatically generated CysSSH and exogenous RSS donors are capable of inhibiting the CaMKK-induced phosho-CaMKIV at Thr196 and thereby mediating the inactivation of enzyme activity in cells. The resulting S-polysulfidation at Cys198 and inactivation of CaMKIV were facilitated during the response to ER stress. Thus, in addition to Ca2+/CaM-dependent regulation of CaMKIV function, polysulfidation may also serve to mediate CaMKIV inactivation as a response to ER stress via the suppression of CREB-dependent transcriptional activation. The possibility also exists that ER stress-induced CaMKIV inactivation might occur through means other than via CSE expression levels. These results provide a better understanding of the mechanisms underlying the biological effects of RSS and indicate that CaMKs may represent novel targets for basic and translation research related to ER stress. Thapsigargin is interesting on two grounds: (1) it is known to increase cytoplasmic calcium levels and (2) a recent report suggested that via the activation of a calcium-sensitive NADPH oxidase, Dual oxidase 1, it also increases cytoplasmic peroxide levels too . The treatment of recombinant CaMKIV with diamide, a thiol-oxidizing agent, either in the presence or absence of glutathione, results in inactivation of the enzyme, and mutated CaMKIV (C198V) was refractory to the inhibition (data not shown). Thus, CaMKIV is inhibited via its S-oxidation at Cys198, but neither by its S-glutathionylation nor by its S-nitrosylation. We are initiating a study to determine what kind of –SH modification is responsible for the observed CaMKIV inhibition aside from its S-polysulfidation. Since CaMKI activity is reversibly inactivated by Na2S4 (data not shown), we are also initiating a study to determine whether RSS inactivate CaMKI via S-polysulfidation of its active-site cysteine residue. In light of a recent report suggesting a major role for the thioredoxin machinery in the reduction of protein per/polysulfide species , it would be particularly interesting to see how the CaMKIV inactivation process is affected/mediated by the thioredoxin system in the near future.
AMP-activated protein kinase
biotin-polyethylene glycol-conjugated maleimide
Ca2+/calmodulin-dependent protein kinase
cerebellar granule cells
cAMP response element-binding protein
polyacrylamide gel electrophoresis
quadrupole time of flight
reactive sulfur species
T.T. and Y.W. designed the research. T.T., H.I., and N.H. performed the research. T.A. provided reagents for the polysulfide-specific biotin-labeling assay. T.T., Y.T., T.A., and Y.W. analyzed the data. T.T. and Y.W. wrote the paper.
This work was supported, in part, by a Grant-in-Aid for Scientific Research on Innovative Areas ‘Oxygen Biology: a new criterion for integrated understanding of life’ [No. 26111008] (Y.W.), Young Scientists B [No. 15K18994] (Y.T.), the Program for the Strategic Research Foundation at Private Universities [No. S1311012] (Y.W.) of The MEXT, Japan, and Grants-in-Aid from the Showa Pharmaceutical University for Young Scientists [No. H28-2] (T.T.).
We thank Dr I. Ishii for his stimulating discussions. We also thank S. Tojo, D. Saito, and T. Kusanagi for their technical assistance. We also thank Editage for English language editing.
The Authors declare that there are no competing interests associated with the manuscript.