RESEARCH ARTICLE Cytoskeleton, December 2014 71:707–718 (doi: 10.1002/cm.21204) C V

2014 Wiley Periodicals, Inc.

Evidence for Thiol/Disulfide Exchange Reactions Between Tubulin and Glyceraldehyde-3-Phosphate Dehydrogenase Lisa M. Landino,* Tara D. Hagedorn, and Kelly L. Kennett Department of Chemistry, The College of William and Mary, Williamsburg, Virginia

Received 27 July 2014; Revised 29 November 2014; Accepted 16 December 2014 Monitoring Editor: George Bloom

While thiol redox reactions are a common mechanism to regulate protein structure and function, protein disulfide bond formation is a marker of oxidative stress that has been linked to neurodegeneration. Both tubulin and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) contain multiple cysteines that have been identified as targets for oxidation to disulfides, Snitrosation and S-glutathionylation. We show that GAPDH is one of three prominent brain microtubuleassociated proteins (MAPs), in addition to MAP-2 and tau, with reactive cysteines. We detected a threefold to fourfold increase in tubulin cysteine oxidation by hydrogen peroxide in the presence of rabbit muscle GAPDH by 5-iodoacetamidofluorescein labeling and by Western blot detection of higher molecular weight inter-chain tubulin disulfides. In thiol/disulfide exchange experiments, tubulin restored 50% of oxidized GAPDH cysteines and the equilibrium favored reduced GAPDH. Further, we report that oxidized GAPDH is repaired by the thioredoxin reductase system (TRS). Restoration of GAPDH activity after reduction by both tubulin and the TRS was time-dependent suggesting conformational changes near the active site cysteine149. The addition of brain MAPs to oxidized tubulin reduced tubulin disulfides and labeling of MAP-2 and of GAPDH decreased. Because the extent of tubulin repair of oxidized GAPDH was dependent Additional Supporting Information may be found in the online version of this article. Abbreviations used: ABTS, 2,20 -azino-bis(3-ethylbenzothiazoline-6sulphonic acid); BCA, bicinchoninic acid; DTT, dithiothreitol; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HOCl, hypochlorous acid; IAF, 5-iodoacetamido-fluorescein; IAM, iodoacetamide; LDH, lactate dehydrogenase; MAPs, microtubule-associated proteins; PB, phosphate buffer; PME, 0.1 M PIPES pH 6.9, 1 mM MgSO4, 2 mM EGTA; TCEP, tris(2-carboxyethyl)phosphine; Trx, thioredoxin; TrxR, thioredoxin reductase. *Address correspondence to: Lisa M. Landino, Department of Chemistry, The College of William and Mary, P.O. Box 8795, Williamsburg, VA 23187-8795, USA. E-mail: [email protected] Published online 28 December 2014 in Wiley Online Library (wileyonlinelibrary.com).

on buffer strength, we conclude that electrostatics influence thiol/disulfide exchange between the two proteins. The novel interactions presented herein may protect GAPDH from inhibition under oxidative stress conditions. V 2014 Wiley Periodicals, Inc. C

Key Words:

GAPDH; tubulin; thioredoxin; thioredoxin reductase; cysteine oxidation; thiol/disulfide exchange; microtubule-associated proteins

Introduction

M

icrotubule proteins including tubulin, tau and microtubule-associated protein-2 (MAP-2) have been the focus of our in vitro work on protein thiol reactivity. We showed that cysteine oxidation of microtubule proteins to disulfides by physiologically-relevant oxidants including ONOO2, hypochlorous acid (HOCl) and HOSCN is associated with decreased microtubule polymerization [Landino et al., 2002, 2004b,c, 2011; Clark et al., 2014]. Tubulin, a heterodimer composed of similar 50 kDa a- and b-subunits, contains 20 reduced cysteines (12 in a-tubulin and eight in btubulin) [Luduena and Roach, 1991; Nogales et al., 1998]. We hypothesize that microtubule protein thiols may protect other cellular targets from oxidation because some tubulin cysteine oxidation (1–2 mol cys) is tolerated before microtubule polymerization is compromised [Landino et al., 2002, 2004a]. This hypothesis is reinforced by our in vitro studies showing that the disulfides in tubulin, tau and MAP2 are repaired by the thioredoxin and glutaredoxin reductase systems thereby restoring polymerization activity [Landino et al., 2004a,b,c,d]. Because tubulin is so abundant in neurons, comprising 10–15% of all protein, it is very likely that it interacts with other proteins in addition to well-characterized microtubule-associated proteins (MAPs) like tau and MAP2 [Anderson, 1979]. Recent reports implicate the glycolytic enzymes, GAPDH and enolase, in neuronal cell death. In the case of GAPDH, the mechanism of cell death involves oxidative stress and disulfide-bonded GAPDH aggregates [Nakajima et al., 2007; Butterfield et al., 2010]. Enolase is linked not only to glucose metabolism in brain but also 707 䊏

Fig. 1. Detection of GAPDH in porcine brain MAPs fraction. A: Western blot detection of GAPDH. Lane 1: rabbit muscle GAPDH; lane 2: PC-MAPs; lane 3: heat stable MAPs. B: Coomassie-stained SDS-PAGE of MAPs fractions. Lane 1: MW standards; lane 2: rabbit muscle GAPDH; lane 3: PC-MAPs; lane 4: heat stable MAPs. C: IAF-labeled PC-MAPs. Lane 1: 75 kDa standard; lane 2: PC-MAPs treated with TCEP; lane 3, PC-MAPs in the absence of TCEP.

hypoxia, ischemia and the development of Alzheimer’s disease [Butterfield and Bader Lange, 2009]. A number of reports dating from the 1980s to the present show that glycolytic enzymes are often associated with microtubules and, in one case, tubulin is reported to be an endogenous inhibitor of GAPDH-mediated membrane fusion [Durrieu et al., 1987; Muronetz et al., 1994; Volker and Knull, 1997; Glaser et al., 2002]. GAPDH, with its four cysteines per monomer, undergoes the same cysteine modifications as tubulin including Snitrosation, S-glutathionylation and oxidation to form disulfides [Mohr et al., 1999; Nakajima et al., 2007]. GAPDH is also of interest because a number of non-glycolytic functions have been identified [Sirover, 2011]. Given its reported electrostatic interaction with tubulin and our considerable work on protein thiol modifications, we were interested in potential redox interactions between tubulin and GAPDH. Furthermore, in a proteomics study using a neuronal cell line, cysteines of tubulin and GAPDH were identified as targets of oxidative stress [Cumming et al., 2004]. Herein we present evidence for thiol/disulfide exchange between the tubulin and GAPDH and examine effects of each on the other’s function. We examined the ability of the thioredoxin reductase system (TRS) to reduce GAPDH cysteines that were oxidized by H2O2 and to determine if tubulin or the TRS restored GAPDH activity lost due to oxidation. The TRS, composed of thioredoxin reductase (TrxR), thioredoxin (Trx) and NADPH, reduces a diverse group of both small-molecule and protein disulfides [Nordberg and Arner, 2001]. TrxR catalyzes the NADPHdependent reduction of an active site disulfide in oxidized Trx, Trx-S2, to the dithiol, Trx-(SH)2. Reduced Trx then undergoes thiol/disulfide exchange with oxidized protein substrates, protein-S2 [Nordberg and Arner, 2001; Holmgren and Lu, 2010].

Results and Discussion We routinely purify tubulin from porcine brain by temperature-dependent cycles of polymerization and

䊏 708

Landino et al.

depolymerization. The high concentration of tubulin in mammalian brain makes this an efficient method to obtain tubulin and its associated proteins [Williams and Lee, 1982]. Tubulin and MAPs were separated by cation exchange chromatography on PC; thus, the MAPs fraction is designated as PC-MAPs. We were interested in assaying brain PC-MAPs for GAPDH because numerous reports had identified it as a MAP. In addition to PC-MAPs, we prepared a heat-stable protein fraction composed primarily of MAP2 and tau (hsMAPs) [Vallee, 1986]. Detection of GAPDH in Brain MAPs

PC-MAPs were assayed for GAPDH by Western blot using an anti-GAPDH antibody and by measuring GAPDH activity. In Fig. 1A, we observed anti-GAPDH reactivity with rabbit muscle GAPDH (lane 1) and with porcine brain PC-MAPs (lane 2). The greatly decreased antibody reactivity with the hsMAPs (lane 3) was expected because GAPDH is not heat stable like MAP2 and tau and would precipitate. The anti-GAPDH reactivity corresponded with the stained Coomassie gel results in Fig. 1B. Bands of control rabbit muscle GAPDH (lane 1) and of porcine brain GAPDH (lane 2) are present at 36 kDa. The hsMAPs in lane 3 showed only a very faint band of the appropriate size. The gel of IAF-labeled proteins in Fig. 1C and the stained PC-MAPs in Fig. 1B revealed three prominent proteins: (1) MAP2 at 280 kDa (MAP2a and 2b), (2) the tau proteins centered near 50 kDa and (3) GAPDH. The amount of GAPDH in PC-MAPs was estimated to be 25% of total protein from the Coomassie-stained gel and by IAF labeling coupled with SDS-PAGE to detect those MAPs with reactive cysteines. MAPs fractions were assayed for GAPDH activity by monitoring the reduction of NAD1 in the presence of GAP. The GAPDH activity of the PC-MAPs fraction was 8.2 3 1027 mol NADH/min/mg protein (n 5 4) vs. 1.6 3 1026 mol NADH/min/mg (n 5 4) for rabbit muscle GAPDH (equal amounts of total protein were used). The hsMAPs fraction contained no detectable GAPDH activity. CYTOSKELETON

Fig. 2. Enhanced oxidation of tubulin by H2O2 in the presence of GAPDH. A: Western blot detection of inter-chain tubulin disulfides. Tubulin (1.5 mM protein, 30 mM cys) was treated with 500 mM H2O2 or 100 mM HOCl in the presence or absence of GAPDH (6.9 mM protein, 28 mM cys). Lanes 1, 3 and 5 contain tubulin only whereas lanes 2, 4 and 6 contain both tubulin and GAPDH. The b-tubulin antibody complex was detected with a 2˚ antibody-HRP conjugate and a chemiluminescent substrate. B: Tubulin and GAPDH cysteine oxidation, detection by IAF labeling. Reactions (20 ml) contained 3 mM tubulin (60 mM cys) and 500 mM H2O2. Lane 1: control; lane 2: tubulin 1 H2O2; lanes 3–5: tubulin and H2O2 1 10, 15 and 20 mM GAPDH (40–80 mM cys). C: Time-dependence of tubulin and GAPDH cysteine oxidation; detection by IAF labeling. Lane 1: control; lane 2: tubulin 1 H2O2; lanes 3–5: oxidized tubulin treated with 15 mM GAPDH (60 mM cys) for 10, 20 or 30 min at 22 C. Samples (lanes 2–5) were treated with H2O2 for 30 min at 22 C. Catalase was added to all samples to quench H2O2 prior to GAPDH addition. Following oxidation and GAPDH treatment, IAF was added (1.2 mM) for 30 min at 37 C. Reactions were quenched with loading buffer containing b-ME. D: Reaction conditions were identical to C. Plot shows mean band intensities from three independent experiments.

Enhanced Tubulin Cysteine Oxidation by H2O2 in the Presence of GAPDH

The tubulin preparation used in Figs. 2–4 was treated with dithiothreitol (DTT) and then desalted to remove it and other small molecules present during purification including unbound GTP, glycerol, EGTA and Mg21. Tubulin was exchanged into PB pH 7.4 because this buffer does not react with oxidants [Radi, 1996]. Likewise, commercial rabbit muscle GAPDH was treated with DTT to reduce cysteines and desalted into PB 7.4. These methods, including the buffers and ratios of oxidant to tubulin cys are consistent with those we have performed in our tubulin oxidation studies [Landino et al., 2011; Clark et al., 2014]. Identical protein samples containing tubulin alone or a mixture of GADPH and tubulin were treated with H2O2 or HOCl, separated by SDS-PAGE under nonreducing conditions and probed with an anti-b-tubulin antibody. As shown in Fig. 2A, we observed enhanced oxidation of tubulin cysteines by H2O2 to higher molecular weight species when GAPDH CYTOSKELETON

was present (lane 4) relative to when GAPDH was absent (lane 3). The arrows mark the dimers and tetramers that are routinely observed when tubulin cysteines are oxidized [Landino et al., 2004a, 2011; Clark et al., 2014]. There are multiple, detectable oligomers of slightly different molecular weight because of the diversity of a- and b-tubulin gene products in mammalian tubulin [Luduena, 1998; McKean et al., 2001]. Controls typically contain traces of tubulin dimers due to air oxidation. Published work from our laboratory confirmed that the addition of reducing agents such as b-ME, DTT and TCEP abolish all higher molecular weight tubulin species including those in the sample wells [Landino et al., 2002, 2004a; Clark et al., 2014]. Unlike H2O2, HOCl oxidation of tubulin vs. a mixture of tubulin and GAPDH yielded greater tubulin oxidation in the absence of GAPDH (Fig. 2A, lane 5) vs. in the presence of GAPDH (lane 6). HOCl is an excellent tubulin cysteine oxidant and given the shorter incubation period and lower HOCl concentration, this result was expected [Landino et al., 2011]. HOCl is a more potent thiol oxidant than H2O2 toward all Evidence for Thiol/Disulfide Exchange Reactions

709 䊏

Fig. 3. Thiol/disulfide exchange between tubulin and GAPDH. A: Tubulin reduction of H2O2-oxidized GAPDH; detection by IAF labeling GAPDH (20 mM, 80 mM cys) was treated with 500 mM H2O2 for 15 min at 22 C. The GAPDH oxidation reaction was quenched with the addition of 500 U catalase. Tubulin (4 mM, 80 mM cys) was added as described. Lane 1: GAPDH; lane 2: GAPDH 1 500 mM H2O2; lanes 3–5: H2O2-oxidized GAPDH 1 tubulin for 10, 20, 30 min; lane 6; tubulin and GAPDH, no oxidant; lane 7: tubulin, no oxidant. Protein samples were subjected to SDS-PAGE under reducing conditions. B: Western blot detection of inter-chain tubulin disulfides Samples were prepared as described in A. Lane 1: control tubulin (nonreducing); lane 2: tubulin and GAPDH, no oxidant; lanes 3–5: H2O2-oxidized GAPDH 1 tubulin for 10, 20, 30 min; lane 6: tubulin 1 500 mM H2O2; lane 7: tubulin 1 b-ME. Protein samples were subjected to SDS-PAGE under nonreducing conditions. The b-tubulin antibody complex was detected with a 2˚ antibody-HRP conjugate and a chemiluminescent substrate. C: Reactions conditions were identical to A. Plots show mean band intensities for three independent experiments. D: GAPDH reduction of HOCl-oxidized tubulin; detection by IAF labeling. Lane 1: tubulin (4 mM, 80 mM cys); lanes 2–4: HOCl-oxidized tubulin 1 GAPDH (20 mM, 80 mM cys) for 10, 20, 30 min; lane 5: tubulin 1 HOCl; lane 6: GAPDH; lane 7: tubulin 1 GAPDH, no oxidant. E: Reactions conditions were identical to D. Plot shows mean band intensities from three independent experiments.

thiols and therefore exhibits no selectivity [Peskin and Winterbourn, 2001]. IAF Labeling of Tubulin and GAPDH Cysteines

To further examine tubulin cysteine oxidation by H2O2, we used IAF, a reagent that only reacts with reduced cysteines, to monitor changes to cysteines of both tubulin and GAPDH. Figure 2B shows decreased labeling of both aand b-tubulin in the presence of H2O2 and increasing concentrations of GAPDH (lanes 3–5). No subunit specificity was observed consistent with prior tubulin oxidation studies [Landino et al., 2002, 2007, 2011]. Previously, we reported that H2O2 is a poor tubulin cysteine oxidant and this was confirmed by the modest 15% decrease in tubulin labeling in lane 2 relative to control (lane 1) [Landino et al., 2002]. The intensity of IAF-labeled GAPDH increased because its concentration increased in lanes 3–5. In a time course experiment, we observed decreased tubulin labeling in the presence of constant GAPDH (Fig. 2C, lanes 3–5). All GAPDH-containing samples showed decreased tubulin

䊏 710

Landino et al.

labeling relative to tubulin treated with H2O2 in the absence of GAPDH (lane 2) even though the sample in lane 3 reacted for only 10 vs. 30 min in lane 2. The intensity of the GAPDH bands also decreased indicative of increased GAPDH oxidation. Tubulin band intensities from multiple experiments were integrated and are summarized in Fig. 2D. Several control experiments were performed to ensure the validity of the IAF labeling assay and our interpretation of band intensity changes (Supporting Information Figs. 1S and 2S). Whereas tubulin labeling decreased to 75% of control when treated with H2O2 (Supporting Information Fig. 1S, lanes 1 vs. 2), GAPDH labeling decreased to 36% of control (lanes 3 vs. 4). For this reason, we conclude that H2O2 is a better GAPDH cysteine oxidant. When tubulin and GAPDH were combined and labeled, in the absence of any oxidant treatment (Supporting Information Fig. 1S, lane 5), the intensity of each increased by 5–8% relative to when each was labeled separately. Addition of a neutral protein such as BSA did not increase labeling of either protein. A specific electrostatic interaction between CYTOSKELETON

Fig. 4. Effect of buffer strength on thiol/disulfide exchange between tubulin and GAPDH. Tubulin reduction of H2O2-oxidized GAPDH; detection by IAF labeling. A: GAPDH (20 mM, 80 mM cys) was treated with 500 mM H2O2 for 15 min at 22 C in 10 mM PB. The GAPDH oxidation reaction was quenched with the addition of 500 U catalase. Tubulin (4 mM, 80 mM cys) and varying concentrations of PB were added as described. Lane 1: GAPDH; lanes 2 and 3: control GAPDH 1 tubulin; lane 4: GAPDH 1 500 mM H2O2; lanes 5–8: H2O2-oxidized GAPDH 1 tubulin at the indicated [buffer]. Protein samples were subjected to SDS-PAGE under reducing conditions. B: Control and oxidized GAPDH samples were prepared and analyzed as in A. Lanes 1 and 2: control GAPDH; lanes 3 and 4: control tubulin; lanes 5 and 6: oxidized GAPDH; lanes 7 and 8: oxidized GAPDH 1 tubulin. C: Reactions conditions were identical to A. These data represent the mean band intensities 6 standard error (n 5 3–5). One-way ANOVA showed that all conditions decreased IAF labeling relative to control, P < 0.001. Based on Tukey post hoc tests, GAPDH oxidized at 10 and 100 mM equivalent < repair at 100 and 70 mM PB equivalent < repair at 40 and 10 mM PB equivalent. D: Western blot detection of inter-chain tubulin disulfides. Samples were prepared as described in A. Lane 1: tubulin (reducing); lane 2: tubulin (nonreducing) lanes 3–6: H2O2-oxidized GAPDH 1 tubulin at the indicated [buffer] for 30 min. Protein samples were subjected to SDS-PAGE under nonreducing conditions. The b-tubulin antibody complex was detected with a 2˚ antibody-HRP conjugate and a chemiluminescent substrate.

tubulin and GAPDH may alter protein conformation in a manner that exposes more cysteines to IAF. The addition of TCEP increased labeling of both proteins because they are susceptible to air oxidation (Supporting Information Fig. 1S, lane 7). Because TCEP is a phosphine, not a thiol-based reducing agent, it can be present during the IAF labeling step. An additional IAF labeling experiment (Supporting Information Fig. 2S) also confirmed that GAPDH oxidation by H2O2 decreased in the presence of tubulin. Pretreatment of GAPDH with iodoacetamide (IAM) to block cysteines abolished its ability to enhance tubulin oxidation by H2O2. Effects of LDH on Tubulin Cysteine Oxidation by H2O2

Though GAPDH cysteine reactivity and interaction with tubulin were our primary focus, we assayed LDH to CYTOSKELETON

determine if it could enhance tubulin cysteine oxidation by H2O2. We chose LDH because, like GAPDH, it has a basic pI (8.4 vs. 8.5 for GAPDH) and five reduced cysteines. Brain tubulin species have pI values ranging from 5.2 to 5.8 and therefore would be expected to interact electrostatically with GAPDH and LDH [Williams et al., 1999]. The addition of LDH increased tubulin cysteine oxidation by H2O2 as evidenced by decreased IAF labeling (Supporting Information Fig. 3S). We suspected that GAPDH cysteines were oxidized by H2O2 to disulfides and reacted with tubulin cysteines by thiol/disulfide exchange to regenerate reduced GAPDH. GAPDH could then be re-oxidized by H2O2 and undergo subsequent exchange with additional tubulin cysteines. The 30 min incubation period with H2O2 would allow for continued GAPDH oxidation and exchange with tubulin thereby yielding greater tubulin Evidence for Thiol/Disulfide Exchange Reactions

711 䊏

Fig. 5. Repair of HOCl-oxidized tubulin by PC-MAPs. A: PC-MAPs reduction of HOCl-oxidized tubulin; detection by IAF labeling. Lane 1: tubulin (10 mM, 200 mM cys); lane 2: tubulin 1 100 mM HOCl; lane 3: tubulin 1 PC-MAPs (7 mg), no oxidant; lanes 4–5: HOCl-oxidized tubulin 1 PC-MAPs (7 mg) for 15 and 30 min. B: Western blot detection of inter-chain tubulin disulfides Samples were identical to those in A. The b-tubulin antibody complex was detected with a 2˚ antibody-HRP conjugate and a chemiluminescent substrate. C: Reactions conditions were identical to A. Plot shows mean band intensities from three independent experiments.

cysteine oxidation. HOCl is a more effective tubulin cysteine oxidant; therefore, tubulin disulfides form readily and cannot exchange with oxidized GAPDH [Landino et al., 2011]. Tubulin Repair of H2O2-Oxidized GAPDH

We postulate that the specific cysteines of tubulin and GAPDH that interact via thiol/disulfide exchange are those that are surface-exposed due to steric limitations. Our published work, using a variety of physiologically relevant oxidants, does not support selective oxidation of tubulin cysteines [Landino et al., 2011]. To test our thiol/disulfide exchange hypothesis, GAPDH was treated with H2O2 and tubulin was added. Control assays were performed using HRP and ABTS to ensure that all H2O2 would be scavenged by catalase prior to the addition of tubulin. All samples including controls were treated with catalase to ensure consistency. Figure 3A shows that GAPDH cysteines were oxidized by H2O2 (lanes 1 vs. 2). Increasing GAPDH band intensity, relative to lane 2, was observed in lanes 3–5 indicative of GAPDH cysteine reduction by tubulin over time. However, after 30 min, GAPDH cysteines were only restored to 68% of control. Thiol/disulfide exchange is an equilibrium process which would likely be slowed due to steric constraints between proteins. While GAPDH cysteines may be accessible to oxidants, newly formed disulfides may not be accessible to tubulin cysteines. While GAPDH band intensities increased, it was less clear that tubulin band intensities decreased in Fig. 3A. For

䊏 712

Landino et al.

this reason, we analyzed identical samples for tubulin interchain disulfides by Western blot. In Fig. 3B, we observed increasing oxidation over time in lanes 3–5 and a concomitant decrease in the main b-tubulin band at 50 kDa confirming that tubulin cysteines were oxidized as GAPDH cysteines were reduced (Fig. 3A). This blot further confirmed that H2O2 is a poor tubulin oxidant (lane 6) and that addition of b-ME abolished all higher molecular weight species (lane 7). A Coomassie stained gel of oxidized tubulin run under nonreducing conditions is shown in Supporting Information Fig. 4S. The a- and b-tubulin bands do not resolve and the intensity of the main tubulin “region” near 50 kDa decreased in intensity as the concentration of oxidant increased. Likewise, greater smearing occurs at higher molecular weights. As for H2O2-oxidized GAPDH, the addition of tubulin to HOCl-oxidized GAPDH restored some, but not all, GAPDH labeling (Supporting Information Fig. 5S, lanes 2 vs. 4). STI, which has a pI of 4.5 but no reduced cysteines, had no effect on GAPDH labeling (Supporting Information Fig. 5S, lanes 4 vs. 5 and Fig. 3D).Changes in GAPDH band intensities that summarize the results in Fig. 3A are plotted in Fig. 3C. GAPDH Repair of HOCl-Oxidized Tubulin

To confirm that thiol/disulfide exchange occurred in the reverse direction, HOCl-oxidized tubulin was treated with reduced GAPDH. The intensity of the tubulin bands CYTOSKELETON

increased with time after GAPDH addition (Fig. 3D, lanes 2–4) relative to oxidized tubulin in lane 5. Whereas tubulin restored 50% of the GAPDH labeling that had been lost due to oxidation (Figs. 3A and 3C), GAPDH restored only 20% of tubulin labeling (Figs. 3D and 3E). While attempts were made to equalize the concentrations of reactive cysteines of each protein, steric constraints may affect protein interactions with reactive, oxidizible cysteines. Likewise, oxidation of either may alter protein structure such that a molecule like IAF will label, but the thiol/disulfide interactions between the proteins are no longer possible. Lastly, the thiol/disulfide exchange equilibrium may favor reduced GAPDH (Figs. 3A–3C) and oxidized tubulin. This is consistent with our hypothesis because H2O2 is a better GAPDH than tubulin oxidant (Supporting Information Figs. 1S and 2S) and GAPDH would then be reduced by tubulin. Effects of Buffer Concentration on Tubulin Repair of Oxidized GAPDH

We hypothesized that electrostatics played a role in the thiol/ disulfide interactions between tubulin and GAPDH. At 100 mM PB, charge interactions between the proteins would be limited due to preferential interaction of the proteins with the buffer ions. In Fig. 4A, we show that the intensity of the GAPDH band decreased (Fig. 4A, lanes 5–8) as the buffer concentration increased. At all buffer concentrations tested, the intensity of the GAPDH band was greater than that of oxidized GAPDH (lane 4) consistent with tubulin repair of oxidized GAPDH observed in Fig. 3A. Though the GAPDH band intensity was dependent on buffer concentration, supportive of tubulin repair, little to no change in the tubulin bands was observed as buffer concentration increased from 10 to 100 mM in Fig. 4A. To address this, controls were performed to determine if buffer concentration influenced IAF labeling of unmodified tubulin and GAPDH. While Fig. 4B shows distinct changes in both the tubulin and GAPDH band shapes, only tubulin bands had different intensities at the two buffer concentrations (based on integration). The different control tubulin band intensities in lanes 3 and 4 of Fig. 4B are noteworthy. At both buffer concentrations, the intensities of the tubulin bands were decreased when combined with oxidized GAPDH (lanes 7 and 8) relative to tubulin alone (lanes 3 and 4). GAPDH band intensities from multiple experiments are summarized in Fig. 4C. There was no significant difference in GAPDH oxidation by H2O2 at 10 and 100 mM PB. The GAPDH band intensities increased as the concentration of buffer decreased suggestive of increased proteinprotein interaction as lower buffer strength and therefore greater repair by tubulin. Statistical analysis showed a significant difference between the higher and lower buffer concentrations. These ionic strength effects on IAF labeling may explain why changes in tubulin band intensities were not observed CYTOSKELETON

in Fig. 3A even though the Western blot results in Fig. 3B showed time-dependent increases in tubulin cysteine oxidation. To confirm that tubulin cysteine oxidation occurred in a manner consistent with GAPDH band intensity changes in Fig. 4A, tubulin samples were analyzed by Western blot under nonreducing conditions. Figure 4D shows the characteristic higher molecular weight species that were detected following incubation with oxidized GAPDH. In this case, the buffer concentration decreased in lanes 3–6 and the extent of tubulin cysteine oxidation increased. More tubulin tetramers are observed in lanes 5 and 6 and the b-tubulin band intensity decreased in lane 6. The results in Figs. 4A–4D are supportive of a role for electrostatics in the GAPDH/tubulin interaction. However, given the effect of buffer concentration on control tubulin IAF labeling, we conclude that IAF labeling has limitations when making quantitative arguments about changes in tubulin cysteines. In particular, for tubulin, the local environment influences cysteine reactivity with iodoacetamides and maleimides [Britto et al., 2002]. Britto et al. reported that 6–7 of the 20 cysteines of tubulin react rapidly with these reagents due to the presence of positive charge that enhances the formation of the more reactive thiolate anion. The presence of negative charge, even near surface-accessible cysteines, decreased reactivity with thiol reagents. Repair of HOCl-Oxidized Tubulin by PC-MAPs

In Fig. 1, we provide evidence that GAPDH is a prominent component of brain MAPs. However, rabbit muscle GAPDH was used in our oxidation and exchange studies (Figs. 2 and 3). Thus, we sought to determine if brain GAPDH and possibly the other major MAPs, tau and MAP2, reacted with oxidized tubulin. In Fig. 5A, the addition of PC-MAPs to HOCl-oxidized tubulin (lane 2), restored IAF labeling of both tubulin subunits in a timedependent manner (lanes 4 and 5). Likewise, the intensities of both the MAP2 and GAPDH bands decreased in lanes 4 and 5 relative to lane 3 (MAPs, but no oxidant). However, it was impossible to observe tau band changes because the tau species co-migrate with tubulin on SDS-PAGE. Indeed, tau IAF labeling likely contributes to the increased tubulin band intensities in lane 3 (vs. lane 1). Additional thiol/disulfide exchange assays of oxidized tubulin and PC-MAPs consistently showed greater reactivity of MAP2 compared to brain GAPDH based on IAF labeling intensity changes. These data are summarized in Fig. 5C. PC-MAPs contain a higher concentration of MAP2 vs. GAPDH (Figs. 1 and 5A) so it is not surprising that MAP2 was more reactive. MAP2, composed of MAP2a and 2b, contains seven cysteines vs. four in GAPDH. MAP2 primarily interacts with tubulin electrostatically via the extreme C-termini of a- and b-tubulin which contain Evidence for Thiol/Disulfide Exchange Reactions

713 䊏

Fig. 6. Repair of oxidized GAPDH by the TRS. A: TRS reduction of H2O2-oxidized GAPDH; detection by IAF labeling GAPDH (25 mM, 100 mM cys) was oxidized with 1 mM H2O2 for 30 min at 37 C. The GAPDH oxidation reaction was quenched with the addition of 500 U catalase. Lane 1: GAPDH; lane 2: GAPDH 1 H2O2; lane 3: H2O2-oxidized GAPDH 1 5 mM E. coli Trx and 100 nM TrxR; lanes 4 and 5: H2O2-oxidized GAPDH 1 5 mM E. coli Trx, 100 nM TrxR and 500 mM (lane 4) or 1 mM (lane 5) NADPH. After IAF labeling for 30 min at 37 C, samples were subjected to SDS-PAGE under reducing conditions. The IAF-labeled protein samples were stained with Coomassie blue. B: IAF labeling samples were prepared as in A. GAPDH activity was monitored at 340 nm in a 96-well plate at 25 C. Reactions (200 ml total) contained 0.1 M PB pH 8.0, 2.0 mM DL-GAP, 1.0 mM NAD1 and a 2 ml aliquot (0.25 mM GAPDH final) of the samples prepared as described above.

multiple negative charges [Lewis et al., 1988]. Thioldisulfide exchange between oxidized tubulin and MAP2 has not been described and is worthy of future investigation because the C-termini contain no cysteines [Luduena, 2013]. Tubulin disulfide reduction by PC-MAPs was confirmed by Western blot in Fig. 5B. Incubation of oxidized tubulin with PC-MAPs for 15 and 30 min (lanes 4 and 5) resulted in decreased tubulin oligomers and increased intensity of the b-tubulin band relative to oxidized tubulin alone in lane 2. Based on the results in Figs. 3D and 3E showing limited reduction of tubulin disulfides by rabbit muscle GAPDH, we were not surprised to observe that brain GAPDH was less reactive than MAP2 especially given its lower concentration of reactive cysteines. The thiol/disulfide exchange equilibrium favors reduced GAPDH and oxidized tubulin even when concentrations of GAPDH and tubulin cysteines are equal (Figs. 3A–3C). Effect of Tubulin on GAPDH Activity

Muronetz et al. reported that tubulin inhibited GAPDH enzymatic activity in a phosphate-dependent manner [Muronetz et al., 1994]. Under our conditions, tubulin, present at the ratios used in Figs. 2 and 3, did not inhibit control GAPDH activity. Instead, we consistently observed a 10–15% increase in GAPDH activity. In those assays,

䊏 714

Landino et al.

20 mM GAPDH was incubated with 4 mM tubulin (cys equivalent) prior to dilution for the activity assay. No tubulin effect was observed at 2 mM tubulin. We attribute this to reduction of air-oxidized GAPDH cysteines by tubulin thereby increasing activity (Supporting Information Figs. 1S and 2S). DTT also increased control GAPDH activity by 10–15%. As reported, oxidation of GAPDH with H2O2 and HOCl inhibited activity in a concentration-dependent manner [Little and O’Brien, 1969; Peskin and Winterbourn, 2006]. The GAPDH active site contains cys149 which is essential for activity [Sirover, 2011]. Hwang et al. showed H2O2 oxidized the active site cysteine (cys 152 of human GAPDH) to cysteic acid or to a disulfide with cys 156 [Hwang et al., 2009]. Likewise, Martyniuk et al. showed that modification of the active site cysteine by a,bunsaturated carbonyls correlated with GAPDH inhibition [Martyniuk et al., 2011]. They reported that the active site cysteine was in a microenvironment that lowered its pKa value to 6. We assayed H2O2-oxidized GAPDH samples that had been incubated with tubulin for 10–30 min for GAPDH activity. Even though our IAF labeling and Western blot results showed that tubulin reduced some GAPDH disulfides (Figs. 3A–3C), tubulin treatment restore less than 10% of GAPDH activity after 30 min. While cys149 is CYTOSKELETON

Fig. 7. Restoration of GAPDH activity by tubulin and by the TRS. GAPDH (10 ml, 25 mM, 100 mM cys) was oxidized with 0.5 mM H2O2 for 30 min at 37 C. Oxidized GAPDH was combined with tubulin (total volume 5 20 ml, 7.5 mM tubulin, 150 mM cys) or with the TRS (total volume 5 20 ml, 5 mM E. coli Trx, 100 nM TrxR and 500 mM NADPH). Aliquots (2 ml) of the GAPDH/tubulin or GAPDH/TRS repair reactions were assayed for GAPDH activity at the indicated times. These data summarize the results of three independent experiments.

accessible to oxidants and to IAF, it may not be readily accessible to tubulin cysteines that could reduce it and restore activity. Peskin and Winterbourn reported that 70% of GAPDH activity was restored when oxidized GAPDH was treated with DTT for 30 min [Peskin and Winterbourn, 2006]. GAPDH Reduction by the TRS

To further investigate the reversibility of GAPDH oxidation and inhibition, we examined its reactivity with the TRS. We chose the TRS because we were interested in a protein-based repair system to compare with tubulin. Components of the TRS were assayed for their ability to reduce GAPDH cysteines oxidized by H2O2. Figure 6A shows IAF labeling of oxidized GAPDH treated with either Trx and TrxR or with the complete TRS composed of Trx, TrxR and NADPH. For Trx and TrxR treatment (Fig. 6A, lane 3), a modest increase in labeling was observed whereas the TRS (lanes 4 and 5) restored labeling to 92 and 94% of control. Trx contains an active site CXXC motif that undergoes thiol/disulfide exchange with oxidized protein substrates [Aslund et al., 1997]. In the absence of NADPH (lane 3), exchange would be an equilibrium process similar to what we observed with oxidized GAPDH and tubulin. When we substituted Trx that had been treated with H2O2 to oxidize the cysteines of the CXXC motif to a disulfide, no increase in GAPDH labeling was observed. However, the combination of oxidized Trx, TrxR and NADPH did restore IAF labeling. To our knowledge, the only published CYTOSKELETON

interactions of components of the TRS with GAPDH are in the model plant Arabidopsis thaliana and with the photosynthetic GAPDH [Sparla et al., 2002; Bedhomme et al., 2012]. Figure 6B summarizes the change in GAPDH band intensities coupled with the results of the GAPDH activity assay. Despite nearly complete restoration of IAF labeling to control levels, TRS-treated samples had only 40% of control GAPDH activity after 30 min. While treatment of oxidized GAPDH with Trx and TrxR restored only 10% of IAF labeling, 30% of GAPDH activity was detected. Although oxidized GAPDH was incubated with the TRS for 30 min prior to IAF addition, some repair is likely to continue during the 30 min IAF labeling step (before all cysteines are modified). Thus, it is difficult to determine the exact time of repair. These data, coupled with a study by Nakajima et al. that examined changes in GAPDH secondary structure following cysteine oxidation prompted us to consider the time dependence of restoration of GAPDH activity [Nakajima et al., 2007]. Based on CD spectral changes of oxidized and control GAPDH, they reported that b-sheet content increased upon oxidation; however, they did not address the reversibility of the conformational change. Thus, even if GAPDH were oxidized and subsequently reduced by tubulin or the TRS, activity may be compromised due to slow conformational changes near the active site. Therefore, the activity of oxidized GAPDH treated with either tubulin or the TRS was measured over a 90 min period. Figure 7 shows that the restoration of activity was time-dependent for both tubulin and the TRS. This is important because it suggests that cysteine reduction alone is not sufficient to restore activity but that a slower protein conformational change might be necessary. If H2O2 or other oxidants were produced in vivo at a low, but steady rate, and critical cysteines of GAPDH were oxidized, tubulin, the TRS and possibly other repair systems might prevent irreversible inhibition. In support of this, Broniowska and Hogg reported different mechanisms of GAPDH inhibition by NO donors in cell-free vs. cellular systems [Broniowska and Hogg, 2010]. H2O2-induced oxidation of GAPDH in cell systems often results in S-glutathionylation, a modification of interest due to its role in protein thiol regulation [Schuppe-Kostinen et al., 1994; Dalle-Donne et al., 2009]. Lind et al. reported that direct oxidation of GAPDH inhibits enzyme activity but that S-glutathionylation via exchange with GSSG does not inhibit enzyme activity. Their findings suggest a complex mechanism of GAPDH regulation via thiol modification [Lind et al., 1998; Cotgreave et al., 2002]. Effect of GAPDH on Microtubule Polymerization

To assess the effect of GAPDH on microtubule polymerization, rabbit muscle GAPDH was added to tubulin in PME buffer and GTP was used to induce assembly. Separate Evidence for Thiol/Disulfide Exchange Reactions

715 䊏

microtubule and unpolymerized tubulin fractions were analyzed by SDS-PAGE. LDH and lysozyme were also assayed for microtubule binding because they are positively charged like GAPDH. GAPDH, LDH and lysozyme were present in the microtubule fraction as well as the supernatant (Supporting Information Fig. 7S). This result supports a role for charge only in the association of proteins with microtubules. Oxidation of GAPDH by concentrations of H2O2 or HOCl used in Figs. 2 and 3 did not alter its ability to bind to microtubules. In addition, the extent of microtubule polymerization was not affected by control or oxidized GAPDH. The reaction conditions for polymerization require PME buffer and very high tubulin concentration and approximately fivefold more tubulin was used relative to GAPDH. Under these conditions, even if oxidized GAPDH reacted with tubulin via exchange, we do not expect that sufficient tubulin cysteine oxidation would have occurred to affect polymerization.

Conclusions Our results herein on the reactions of oxidized GAPDH with both tubulin and the TRS are important in the context of a repair mechanism for GAPDH, a central enzyme in cellular metabolism. We postulate that the specific cysteines of tubulin and GAPDH that interact via thiol/disulfide exchange are those that are surface-exposed due to steric limitations. Under conditions where GAPDH cysteines could be oxidized in proximity to high localized concentrations of tubulin and microtubules, thiol/disulfide exchange is possible. The TRS is a ubiquitous cytosolic reductase system. Approximate physiologic concentrations of Trx in mammalian liver and brain are 10–12 and 2 mM, respectively, whereas TrxR concentrations are estimated to be 10– 12 nM in liver and 2 nM in brain [Holmgren and Luthman, 1978; Luthman and Holmgren, 1982]. The interactions we present may protect GAPDH from irreversible aggregate formation and modulate oxidantinduced inhibition of GAPDH activity. This hypothesis is strengthened by our studies showing that disulfides of tubulin and GAPDH are repaired by the TRS.

Materials and Methods Materials

Porcine brains were obtained from Smithfield Packing Company in Smithfield, Virginia. Bicinchoninic acid (BCA) protein assay reagent, West Pico chemiluminescence detection system, tris(2-carboxyethyl)phosphine (TCEP) and 5-iodoacetamido-fluorescein (IAF) were from Thermo Pierce (Rockford, IL). Human thioredoxin was from American Diagnostica (Pfungstadt, Germany). Escherichia coli Trx, mouse anti-b-tubulin antibody (clone TUB 2.1) and the goat anti-mouse secondary antibody HRP conjugate were from Sigma (St. Louis, MO). The mouse

䊏 716

Landino et al.

anti-GAPDH antibody (clone MAB374) was from Millipore (Darmstadt, Germany). Rabbit muscle GAPDH, lactate dehydrogenase (LDH), hen egg white lysozyme and STI were from Sigma. All other chemicals were from Fisher (Waltham, MA) or Sigma. Purification of Porcine Brain Tubulin and MAPs

Tubulin was purified from porcine brain by two cycles of temperature-dependent polymerization and depolymerization and subsequent phosphocellulose (PC) chromatography as described [Landino et al., 2011]. MAPs were eluted from the PC column in PME buffer (0.1 M PIPES pH 6.9, 1 mM MgSO4, 2 mM EGTA) containing 500 mM NaCl (PC-MAPs) and stored at 280 C. Purification of Rat Liver TrxR

TrxR was purified from rat liver as described and stored at 280 C in TE buffer (50 mM Tris-HCl, 1 mM EDTA, pH 7.5) [Arner et al., 1999; Landino et al., 2004a]. Labeling of Cysteines With IAF

Tubulin or GAPDH were diluted in 0.1 M phosphate buffer (PB) pH 7.4 and treated with H2O2 for 30 min or with HOCl for 5 min at 22 C in a total reaction volume of 15–20 ml. Catalase (500 U) was added to remove excess H2O2 whereas methionine (200 mM–1 mM) scavenged HOCl. For exchange reactions, the second protein was added and reactions proceeded for 10–30 min at 22 or 37 C. IAF in DMF was added to achieve a 10-fold molar excess relative to protein cys and samples were incubated at 37 C for an additional 30 min. The addition of loading buffer containing b-mercaptoethanol (b-ME) quenches excess IAF and reduces disulfides formed by oxidants. Proteins were resolved by SDS-PAGE on 7.5 or 10% gels and images were captured using a Bio-rad Chemi-doc XRS imaging system. The intensity of the fluorescein-labeled protein bands was measured using Bio-rad Image Lab software. Detection of Tubulin and GAPDH Inter-Chain Disulfides by Western Blot

After treatment with oxidants and exchange as described above, tubulin or GAPDH species (10–15 mg protein per lane) were separated by SDS-PAGE on polyacrylamide gels (7.5% for tubulin, 10% for GAPDH) under nonreducing conditions. Proteins were transferred to PVDF membranes, blocked with 3% milk (tubulin) or 5% BSA (GAPDH) for 30 min and probed with either mouse monoclonal anti-b-tubulin antibody (1:2000) for 2 h or mouse monoclonal anti-GAPDH antibody (MAB374) (1:1500) for 2 h. The antibody complexes were detected using a goat anti-mouse HRP conjugate (1 h, 1:10,000) and Pierce West Pico chemiluminescent substrate. Chemiluminescence was captured using the Bio-rad Chemi-doc XRS imaging system. CYTOSKELETON

GAPDH Activity Assays 1

The GAPDH-catalyzed reduction of NAD in the presence of D,L-glyceraldehyde-3-phosphate (GAP) was monitored at 340 nm in a 96-well plate at 25 C. Typical reactions (200 ml) contained 2.0 mM DL-GAP, 1.0 mM NAD1 and 0.25 mM GAPDH in 0.1 M PB pH 8.0. Reaction rates were calculated from the initial 5 min linear portion. Microtubule Polymerization Assays

GAPDH, LDH or lysozyme (20 mg each) were combined with tubulin (50 ml rxn, 25 mM tubulin, 500 mM cys), for 10 min at 22 C. GTP (1 mM final) was added to induce polymerization and the samples were incubated at 37 C for 20–25 min. Microtubule polymer was collected by centrifugation at 16,000 3 g for 20 min. Control polymerization activity was set at 100% for those samples containing GTP but no oxidant. Controls without GTP were used to establish 0% activity. Supernatant protein concentrations were determined by the BCA protein assay. Protein supernatants and pellets were analyzed by SDS-PAGE with Coomassie Blue staining. Protein pellets were dissolved in 6 M guanidine-HCl and the absorbance was measured at 275 nm [Landino et al., 2004a]. Colorimetric Detection of H2O2

Protein samples were assayed for residual H2O2 using HRP and 2,20 -azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) (ABTS). Typical reactions (200 ml) contained 1.0 mM ABTS and 0.02 mg HRP in 0.1 M PB pH 7.4. Oxidation of ABTS was detected at 405 nm in a 96-well plate reader. Acknowledgments

The authors acknowledge support from the National Institute of Neurological Disorders and Stroke (R15-NS38885 to LML). References

S-nitrosothiols and NO in cellular and cell-free conditions. Am J Physiol Heart Circ Physiol 299:H1212–H1219. Butterfield DA, Bader Lange ML. 2009. Multifunctional roles of enolase in Alzheimer’s disease brain: beyond altered glucose metabolism. J Neurochem 111:915–933. Butterfield DA, Hardas SS, Bader Lange ML. 2010. Oxidatively modified glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and Alzheimer’s disease: many pathways to neurodegeneration. J Alzheimer’s Dis 20:369–393. Clark HM, Hagedorn TD, Landino LM. 2014. Hypothiocyanous acid oxidation of tubulin cysteines inhibits microtubule polymerization. Arch Biochem Biophys 541:67–73. Cotgreave IA, Gerdes R, Schuppe-Kostinen I, Lind C. 2002. S-glutathionylation of glyceraldehyde-3-phosphate dehydrogenase: role of thiol oxidation and catalysis by glutaredoxin. Methods Enzymol 348:175–182. Cumming RC, Andon NL, Haynes PA, Park M, Fischer WH, Schubert D. 2004. Protein disulfide bond formation in the cytoplasm during oxidative stress. J Biol Chem 279:21749–21758. Dalle-Donne I, Rossi R, Colombo G, Giustarini D, Milzani A. 2009. Protein S-glutathionylation: a regulatory device from bacteria to humans. Trends Biochem Sci 34:85–96. Durrieu C, Bernier-Valentin F, Rousset B. 1987. Microtubules bind glyceraldehyde-3-phosphate dehydrogenase and modulate its enzyme activity and quaternary structure. Arch Biochem Biophys 252:32– 40. Glaser PE, Han X, Gross RW. 2002. Tubulin is the endogenous inhibitor of glyceraldehyde 3-phosphate dehydrogenase isoform that catalyzes membrane fusion: implications for the coordinated regulation of glycolysis and membrane fusion. Proc Natl Acad Sci USA 99:14104–14109. Holmgren A, Lu J. 2010. Thioredoxin and thioredoxin reductase: current research with special reference to human disease. Biochem Biophys Res Commun 396:120–124. Holmgren A, Luthman M. 1978. Tissue distribution and subcellular localization of bovine thioredoxin determined by radioimmunoassay. Biochemistry 17:4071–4077. Hwang NR, Yim S-H, Kim YM, Jeong J, Song EJ, Lee Y, Lee JH, Choi S, Lee K-J. 2009. Oxidative modifications of glyceraldehyde3-phosphate dehydrogenase play a key role in its multiple cellular functions. Biochem J 423:253–264.

Anderson PJ. 1979. The structure and amount of tubulin in cells and tissues. J Biol Chem 254:2168–2171.

Landino LM, Hasan R, McGaw A, Cooley S, Smith AW, Masselam K, Kim G. 2002. Peroxynitrite oxidation of tubulin sulfhydryls inhibits microtubule polymerization. Arch Biochem Biophys 398:213–220.

Arner ESJ, Zhong L, Holmgren A. 1999. Preparation and assay of mammalian thioredoxin and thioredoxin reductase. Methods Enzymol 300:226–239.

Landino LM, Iwig JS, Kennett KL, Moynihan KL. 2004a. Repair of peroxynitrite damage to tubulin by the thioredoxin reductase system. Free Radic Biol Med 36:497–506.

Aslund F, Berndt KD, Holmgren A. 1997. Redox potentials of glutaredoxins and other thiol-disulfide oxidoreductases of the thioredoxin superfamily determined by direct protein-protein redox equilibria. J Biol Chem 272:30780–30786.

Landino LM, Moynihan KL, Todd JV, Kennett KL. 2004b. Modulation of the redox state of tubulin by the glutathione/glutaredoxin reductase system. Biochem Biophys Res Commun 314:555–560.

Bedhomme M, Adamo M, Marchand CH, Couturier J, Rouhier N, Lemaire SD, Zaffagnini M, Trost P. 2012. Glutathionylation of cytosolic glyceraldehyde-3-phosphate dehydrogenase from the model plant Arabidosis thaliana is reversed by both glutaredoxins and thioredoxins in vitro. Biochem J 445:337–347. Britto PJ, Knipling L, Wolff J. 2002. The local electrostatic environment determines cysteine reactivity of tubulin. J Biol Chem 277:29018–29027. Broniowska K, Hogg N. 2010. Differential mechanisms of inhibition of glyceraldehyde-3-phosphate dehydrogenase by

CYTOSKELETON

Landino LM, Robinson SH, Skreslet TE, Cabral DM. 2004c. Redox modulation of tau and microtubule-associated protein-2 by the glutathione/glutaredoxin reductase system. Biochem Biophys Res Commun 323:112–117. Landino LM, Skreslet TE, Alston JA. 2004d. Cysteine oxidation of tau and microtubule-associated protein-2 by peroxynitrite: modulation of microtubule assembly kinetics by the thioredoxin reductase system. J Biol Chem 279:35101–35105. Landino LM, Koumas MT, Mason CE, Alston JA. 2007. Modification of tubulin cysteines by nitric oxide and nitroxyl donors alters tubulin polymerization activity. Chem Res Toxicol 20:1693–1700.

Evidence for Thiol/Disulfide Exchange Reactions

717 䊏

Landino LM, Hagedorn TD, Kim SB, Hogan KM. 2011. Inhibition of tubulin polymerization by hypochlorous acid and chloramines. Free Radic Biol Med 50:1000–1008.

totic protein glyceraldehyde-3-phosphate dehydrogenase is essential in oxidative stress-induced aggregation and cell death. J Biol Chem 282:26562–26574.

Lewis SA, Wang D, Cowan N. 1988. Microtubule-associated protein MAP2 shares a microtubule binding motif with tau protein. Science 242:936–939.

Nogales E, Wolf SG, Downing KH. 1998. Structure of the abtubulin dimer by electron crystallography. Nature 391:199–203.

Lind C, Gerdes R, Schuppe-Kostinen I, Cotgreave IA. 1998. Studies on the mechanism of oxidative modification of human glyceraldehyde-3-phosphate dehydrogenase by glutathione:catalysis by glutaredoxin. Biochem Biophys Res Commun 247:481–486. Little C, O’Brien PJ. 1969. Mechanism of peroxide-inactivation of the sulphydryl enzyme glyceraldehyde-3-phosphate dehydrogenase. Eur J Biochem 10:533–538.

Nordberg J, Arner ESJ. 2001. Reactive oxygen species, antioxidants, and the mammalian thioredoxin system. Free Radic Biol Med 31: 1287–1312. Peskin AV, Winterbourn CC. 2001. Kinetics of the reactions of hypochlorous acid and amino acid chloramines with thiols, methionine and ascorbate. Free Radic Biol Med 30:572–579.

Luduena RF. 1998. Multiple forms of tubulin: different gene products and covalent modifications. Int Rev Cytol 178:207–275.

Peskin AV, Winterbourn CC. 2006. Taurine chloramine is more selective than hypochlorous acid at targeting critical cysteines and inactivating creatine kinase and glyceraldehyde-3-phosphate dehydrogenase. Free Radic Biol Med 40:45–53.

Luduena RF. 2013. A hypothesis on the origin and evolution of tubulin. Int Rev Cell Mol Biol 302:41–185.

Radi R. 1996. Kinetic analysis of reactivity of peroxynitrite with biomolecules. Methods Enzymol 269:354–366.

Luduena RF, Roach MC. 1991. Tubulin sulfhydryl groups as probes and targets for antimitotic and antimicrotubule agents. Pharmacol Ther 49:133–152.

Schuppe-Kostinen I, Moldeus P, Bergman T, Cotgreave IA. 1994. S-thiolation of human endothelial cell glyceraldehyde-3-phosphate dehydrogenase after hydrogen peroxide treatment. Eur J Biochem 221:1033–1037.

Luthman M, Holmgren A. 1982. Rat liver thioredoxin and thioredoxin reductase: purification and characterization. Biochemistry 21: 6628–6633. Martyniuk CJ, Fang B, Koomen JM, Gavin T, Zhang L, Barber DS, LoPachin RM. 2011. Molecular mechanism of glyceraldehyde3-phosphate dehydrogenase inactivation by a,b-unsaturated carbonyl derivatives. Chem Res Toxicol 24:2302–2311. McKean PG, Vaughan S, Gull K. 2001. The extended tubulin superfamily. J Cell Sci 114:2723–2733. Mohr S, Hallak H, de Boitte A, Lapetina EG, Brune B. 1999. Nitric oxide-induced S-glutathionylation and inactivation of glyceraldehyde-3-phosphate dehydrogenase. J Biol Chem 274:9427– 9430. Muronetz VI, Wang Z-X, Keith TJ, Knull HR, Strivastava DK. 1994. Binding constants and stoichiometries of glyceraldehyde-3phosphate dehydrogenase-tubulin complexes. Arch Biochem Biophys 313:253–260. Nakajima H, Amano W, Fujita A, Fukuhara A, Azuma Y-T, Hata F, Inui T, Takeuchi T. 2007. The active site cysteine of the proapop-

䊏 718

Landino et al.

Sirover MA. 2011. On the functional diversity of glyceraldehyde-3phosphate dehydrogenase: biochemical mechanisms and regulatory control. Biochim Biophys Acta 1810:741–751. Sparla F, Pupillo P, Trost P. 2002. The C-terminal extension of glyceraldehyde-3-phosphate dehydrogenase subunit B acts as an autoinhibitory domain regulated by thioredoxins and nicotinamide adenine dinucleotide. J Biol Chem 277:44946–44952. Vallee RB. 1986. Reversible assembly purification of microtubules without assembly-promoting agents and further purification of tubulin, microtubule-associated proteins, and MAP fragments. Methods Enzymol 134:89–104. Volker KW, Knull HR. 1997. A glycolytic enzyme binding domain on tubulin. Arch Biochem Biophys 338:237–243. Williams RC, Lee JC. 1982. Preparation of tubulin from brain. Methods Enzymol 85:376–385. Williams RC, Shah C, Sackett D. 1999. Separation of tubulin isoforms by isoelectric focusing in immobilized pH gradient gels. Anal Biochem 275:265–267.

CYTOSKELETON