Corrections

BIOPHYSICS AND COMPUTATIONAL BIOLOGYCorrection for ‘‘A robust method for searching the smallest setof smallest rings with a path-included distance matrix,’’ by ChangJoon Lee, Young-Mook Kang, Kwang-Hwi Cho, and Kyoung TaiNo, which appeared in issue 41, October 13, 2009, of Proc NatlAcad Sci USA (106:17355–17358; first published September 24,2009; 10.1073/pnas.0813040106).

The authors note that, due to a printer’s error, the SupportingInformation materials did not appear online along with the mainarticle. The materials have since been made available online. Theonline version of the main article has been corrected to includea link to the Supporting Information.

www.pnas.org/cgi/doi/10.1073/pnas.0911801106

CELL BIOLOGYCorrection for ‘‘CIB1 functions as a Ca2�-sensitive modulator ofstress-induced signaling by targeting ASK1,’’ by Kyoung WanYoon, Jun-Ho Cho, Jae Keun Lee, Young-Hee Kang, Ji SooChae, Young Mok Kim, Jeehyun Kim, Eun Kyung Kim, SungEun Kim, Ja-Hyun Baik, Ulhas P. Naik, Ssang-Goo Cho, andEui-Ju Choi, which appeared in issue 41, October 13, 2009, ofProc Natl Acad Sci USA (106:17389–17394; first publishedSeptember 29, 2009; 10.1073/pnas.0812259106).

The authors note that during the publication process, the rawdata shown in Fig. 5C were found to be inaccurately labeled andcould not be used to confirm the results. The experiment wasrerun and new data from two repeating experiments were notsignificantly different from those originally shown in Fig. 5C,confirming the results as originally reported. The authors havesubmitted a revised figure, replacing the data originally pub-lished in Fig. 5C with those of the two new experiments to ensurethe raw data are available to readers. The corrected figure andits legend appear below.

Fig. 5. KCl-elicited Ca2� influx prevents the inhibitory action of CIB1 on6-OHDA-induced ASK1 activation and apoptosis. (A) SH-SY5Y cells were incu-bated first with 6-OHDA (100 �M) for 50 min and then without or with KCl (40mM) either alone or together with BAPTA-AM (2.5 �M) for 10 min, as indi-cated. Cell lysates were subjected to immunoprecipitation with anti-ASK1antibody or rabbit preimmune IgG (P.I.), and the resulting precipitates wereexamined by immunoblotting with antibody to CIB1 (Left) or to TRAF2 (Right).(B) SH-SY5Y cells expressing a control siRNA or CIB1 siRNA were treated with6-OHDA, KCl, BAPTA-AM as in A. Cell lysates were assayed for ASK1 activity byimmune complex kinase assay. (C) SH-SY5Y cells expressing a control siRNA orCIB1 siRNA were incubated consecutively with 6-OHDA (100 �M) for 50 min,with KCl (40 mM) either alone or together with BAPTA-AM (2.5 �M) for 30min, and in DMEM containing 2% FBS for 12 h, as indicated. Cells wereanalyzed for apoptosis by flow cytometry. The percentages of apoptotic cellsin a representative experiment are indicated.

www.pnas.org/cgi/doi/10.1073/pnas.0911831106

www.pnas.org PNAS � November 17, 2009 � vol. 106 � no. 46 � 19629

CORR

ECTI

ON

S

Dow

nloa

ded

by g

uest

on

Janu

ary

24, 2

021

Dow

nloa

ded

by g

uest

on

Janu

ary

24, 2

021

Dow

nloa

ded

by g

uest

on

Janu

ary

24, 2

021

Dow

nloa

ded

by g

uest

on

Janu

ary

24, 2

021

Dow

nloa

ded

by g

uest

on

Janu

ary

24, 2

021

Dow

nloa

ded

by g

uest

on

Janu

ary

24, 2

021

Dow

nloa

ded

by g

uest

on

Janu

ary

24, 2

021

Dow

nloa

ded

by g

uest

on

Janu

ary

24, 2

021

CIB1 functions as a Ca2�-sensitive modulatorof stress-induced signaling by targeting ASK1Kyoung Wan Yoona, Jun-Ho Choa, Jae Keun Leea, Young-Hee Kanga, Ji Soo Chaea, Young Mok Kima, Jeehyun Kima,Eun Kyung Kima, Sung Eun Kima, Ja-Hyun Baikb, Ulhas P. Naikc, Ssang-Goo Chod, and Eui-Ju Choia,1

Laboratories of aCell Death and Human Diseases and bMolecular Neurobiology, School of Life Sciences and Biotechnology, Korea University, Seoul 136-701,Korea; cDepartment of Biological Sciences, University of Delaware, Newark, DE 19716; and dDepartment of Animal Biotechnology, Konkuk University,Seoul 143-702, Korea

Edited by Michael Karin, University of California, San Diego, School of Medicine, La Jolla, CA, and approved July 23, 2009 (received for reviewDecember 3, 2008)

Calcium and integrin binding protein 1 (CIB1) is a Ca2�-bindingprotein of 22 kDa that was initially identified as a protein thatinteracts with integrin �IIb. Although it interacts with variousproteins and has been implicated in diverse cellular functions, themolecular mechanism by which CIB1 regulates intracellular signal-ing networks has remained unclear. We now show that, by tar-geting apoptosis signal-regulating kinase 1 (ASK1), CIB1 negativelyregulates stress-activated MAPK signaling pathways. CIB1 wasthus shown to bind to ASK1, to interfere with the recruitment ofTRAF2 to ASK1, and to inhibit the autophosphorylation of ASK1 onthreonine-838, thereby blocking ASK1 activation. Furthermore,CIB1 mitigated apoptotic cell death initiated either by TNF-� inbreast cancer MCF7 cells or by 6-hydroxydopamine (6-OHDA) indopaminergic cells. Ca2� influx induced by membrane depolariza-tion reversed the inhibitory effect of CIB1 on 6-OHDA-inducedASK1 activation and cell death in dopaminergic neurons. Theseobservations thus suggest that CIB1 functions as a Ca2�-sensitivenegative regulator of ASK1-mediated signaling events.

apoptosis � calcium � MAPK

c -Jun NH2-terminal kinase (JNK) and p38 are the mitogen-activated protein kinases (MAPKs) that can be stimulated in

response to a variety of extrinsic and intrinsic cellular stressorsand proinflammatory cytokines (1, 2). MAPK signaling path-ways consist of a MAPK, a MAPK kinase (MAP2K), and aMAPK kinase kinase (MAP3K). Apoptosis signal-regulatingkinase 1 (ASK1) is a serine-threonine kinase that was initiallyidentified as a MAP3K in both the SEK1/MKK4 (or MKK7)-JNK and MKK3 (or MKK6)-p38 signaling pathways (3, 4). ASK1plays a role in apoptosis induced by a variety of cellular stressorsincluding oxidative stress, tumor necrosis factor (TNF)-�, en-doplasmic reticulum stress, and anticancer drugs (4). ASK1-mediated signaling is modulated either positively or negatively byvarious ASK1 binding proteins that include thioredoxin andtumor necrosis factor receptor-associated factors 2 (TRAF2)(5–10).

Calcium and integrin binding protein 1 (CIB1, also known ascalmyrin) is a 22-kDa protein that shares substantial sequencesimilarity with calmodulin and several other calcium-bindingproteins (11). CIB1 contains two canonical Ca2�-binding EF-hand motifs in the COOH-terminal globular domain (12, 13).Although it was initially discovered in platelets as a protein thatinteracts with integrin �IIb (11), CIB1 also binds many otherproteins including presenilin 2, DNA-dependent protein kinase,the polo-like kinases Fnk and Snk (14–16). Furthermore, thebinding of Ca2� to CIB1 induces conformational changes thatmodulate the interaction with at least some of its bindingproteins (17). CIB1 is widely expressed among tissues (18) andis likely to play a role in various cellular processes. However, thebiological functions of CIB1 have remained relatively undefinedto date.

To obtain insights into the biological functions of CIB1, wehave investigated its possible role in regulation of stress-activatedMAPK pathways. We now show that CIB1 negatively regulatessuch pathways by targeting ASK1. CIB1 physically associateswith ASK1 and thereby inhibits the activation of the ASK1-JNKand ASK1-p38 signaling pathways by reactive oxygen species(ROS) and by TNF-�. Ca2� influx reverses the inhibitory effectof CIB1 on ASK1-mediated signaling. These results thus suggestthat CIB1 has a function working as a Ca2�-sensitive antagonistof stress-induced signaling events.

ResultsCIB1 Physically Associates with ASK1. From yeast two-hybridscreening assays with using ASK1 as a bait, we identified severalASK1-binding proteins including CIIA and GSTM1 (8, 10) andCIB1. Subsequent yeast two-hybrid experiments revealed thatCIB1 bound to the ASK1 mutants ASK1-NT and ASK1-�C, butnot to ASK1-�N (Fig. S1 A). ASK1-NT, ASK1-�C, andASK1-�N comprise amino acids 1 to 656, 1 to 936, and 649 to1374 of human ASK1, respectively. To investigate further theinteraction between ASK1 and CIB1, we performed in vitrobinding assays with 35S-labled ASK1 variants and a recombinantGST-CIB1 fusion protein. GST-CIB1 was found to bind to ASK1mutant comprising amino acids 378 to 648 [ASK1 (378–648)]and full-length ASK1 (Fig. 1A). These results thus indicated thatCIB1 binds directly to the portion of ASK1 known as theNH2-terminal regulatory domain (4). Reciprocal in vitro bindingassays with 35S-labled ASK1 and CIB1 variants revealed thatASK1 bound primarily to the NH2-terminal half region of CIB1,which contains two nonfunctional EF-hand-like motifs (12, 13)(Fig. S1B). Coimmunoprecipitation analysis data also revealed aphysical association between the two endogenous proteins inMCF7 cells and that approximately one fourth of total ASK1existed in a complex with CIB1 in the cells (Fig. S1C). Exposureof the cells to H2O2 or TNF-�, both of which activate ASK1, didnot affect the extent of the interaction between CIB1 and ASK1(Fig. 1B).

CIB1 Inhibits ASK1 Activity In Vitro and in Intact Cells. Given thatCIB1 physically associates with ASK1, we examined whetherCIB1 might affect the kinase activity of ASK1. An in vitro kinaseassay showed that GST- CIB1 inhibited ASK1 activity in ASK1immunoprecipitates prepared from 293T cells treated with H2O2

Author contributions: K.W.Y., J.-H.C., J.K.L., E.K.K., and E.-J.C. designed research; K.W.Y.,J.-H.C., J.K.L., Y.-H.K., J.S.C., Y.M.K., J.K., and S.E.K. performed research; J.-H.B., U.P.N., andS.-G.C. contributed new reagents/analytic tools; K.W.Y. and E.-J.C. analyzed data; andK.W.Y. and E.-J.C. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

1To whom correspondence should be addressed. E-mail: ejchoi@korea.ac.kr.

This article contains supporting information online at www.pnas.org/cgi/content/full/0812259106/DCSupplemental.

www.pnas.org�cgi�doi�10.1073�pnas.0812259106 PNAS � October 13, 2009 � vol. 106 � no. 41 � 17389–17394

CELL

BIO

LOG

Y

(Fig. 2A). In contrast, separate in vitro kinase assays revealedthat CIB1 did not inhibit the activities of downstream kinases ofASK1 including MAP2Ks (SEK1/MKK4 and MKK3) andMAPKs (JNK1 and p38) (Fig. S2 A and B). Next, we investigatedthe effect of CIB1 on ASK1 activity in intact cells after the cellswere exposed to ROS-generating stimuli such as H2O2 andTNF-� (7, 19). Forced expression of ectopic CIB1 resulted in adecrease in the H2O2- or TNF-�-stimulated activity of ASK1(Fig. 2B). In contrast, CIB1 did not affect the activation ofMLK3 (Fig. S2C), another MAP3K in the JNK pathway (1).

To examine whether endogenous CIB1 also inhibits ASK1-mediated signaling, we transfected HeLa cells with a vectorexpressing siRNA for CIB1 (CIB1 siRNA) or with a vector fora control siRNA. Immunoblot analysis confirmed a knockdownof CIB1 expression in cells expressing CIB1 siRNA (Fig. 2 C andD). Depletion of endogenous CIB1 by RNA interference(RNAi) potentiated the activity of ASK1 in H2O2- or TNF-�-treated HeLa cells, and this potentiation was abolished byexpression of Flag3-CIB1* in the same cells (Fig. 2C). Flag3-CIB1* is a 3xFlag epitope-tagged CIB1 construct that wasresistant to CIB1 siRNA. Furthermore, The RNAi-mediatedknockdown of CIB1 also enhanced the activities of JNK1 (Fig.S3 A and B) and p38 MAPK (Fig. S3 C and D) in H2O2- orTNF-�-treated cells.

To determine whether CIB1 might block ASK1 activationthrough constitutive association with ASK1, we examined theeffect of H2O2 on the interaction between CIB1 and ASK1 inHeLa cells expressing either control or CIB1 siRNA. Consistentwith the data obtained with 293T cells (Fig. 1B), coimmunopre-cipitation analysis showed that CIB1 was present in ASK1

immunoprecipitates prepared from cells expressing the controlsiRNA and that the extent of the CIB1-ASK1 interaction was notaffected by H2O2 (Fig. 2D). In contrast, H2O2 abolished the

Fig. 1. Physical association of CIB1 with ASK1. (A) A schematic representa-tion of ASK1 indicating the thioredoxin binding region (gray box), the TRAF2binding region (dotted box), and the kinase domain (hatched box) is shown inUpper. The indicated 35S-labled ASK1 variants were reacted with GST-CIB1immobilized on glutathione-agarose beads. The bead-bound 35S-labled pro-teins were eluted and detected by SDS/PAGE and autoradiography. The gelwas also stained with Coomassie brilliant blue. A portion (25%) of the 35S-labled protein input to the binding reaction was also subjected to SDS/PAGEand autoradiography. (B) The 293T cells were left untreated or treated with 1mM H2O2 or 20 ng/mL TNF-� for 20 min, lysed, and subjected to immunopre-cipitation (IP) with antibodies to ASK1 or with rabbit preimmune IgG (control).The resulting precipitates and cell lysates were subjected to immunoblot (IB)analysis with antibodies to CIB1 or to ASK1.

Fig. 2. CIB1 inhibits ASK1 activity. (A) The 293T cells were transfected for 48 hwith a plasmid vector for Flag-ASK1 and were treated with 1 mM H2O2 for 20min where indicated. Cell lysates were subjected to immunoprecipitation withanti-Flag antibody. The resulting precipitates were incubated for 30 min at20 °C with the indicated amounts of GST-CIB1 or GST in 50 �L of Hepes buffer(pH 7.4), and were then assayed for ASK1 activity. The relative activity of ASK1is shown below each lane. (B) The 293T cells were transfected for 48 h with avector encoding HA-ASK1 alone or together with a vector for Flag-CIB1, andwere then incubated in the absence or presence of H2O2 (1 mM) or TNF-� (20ng/mL) for 20 min. Cell lysates were assayed for ASK1 activity by immunecomplex kinase assay. (C) HeLa cells expressing a control siRNA or CIB1 siRNAwere left untreated or treated with H2O2 (1 mM) or TNF-� (20 ng/mL) for 20min. Where indicated, HeLa cells expressing CIB1 siRNA were transientlytransfected for 48 h with a vector encoding Flag3-CIB1* before the chemicaltreatment. Flag3-CIB1* containing three silent third-codon point mutations(GAC 3 GAT, CTT 3 CTC, AGC 3 AGT) within the region targeted by CIB1siRNA was generated by site-directed mutagenesis of human CIB1 gene withthe use of a Quickchange kit (Stratagene). Cell lysates were subjected toimmunoprecipitation with anti-ASK1 antibody, and the resulting precipitateswere assayed for ASK1 activity. The precipitates were also subjected to im-munoblotting with anti-ASK1 antibody. Cell lysates were also examined di-rectly by immunoblot analysis with antibodies to CIB1 or to ASK1. eCIB1,endogenous CIB1. (D) HeLa cells expressing a control siRNA or CIB1 siRNA wereuntreated or treated with H2O2 (1 mM) for 20 min. Cell lysates were subjectedto immunoprecipitation with antibodies to ASK1, and the resulting precipi-tates were subjected to immunoblot analysis with antibodies to CIB1 or tothioredoxin-1 (Trx1).

17390 � www.pnas.org�cgi�doi�10.1073�pnas.0812259106 Yoon et al.

binding between ASK1 and thioredoxin in cells expressingcontrol or CIB1 siRNA. The reduced form of thioredoxin bindsto and inhibits ASK1, but its oxidation in cells exposed toROS-generating stimuli such as H2O2 and TNF-� results in itsdissociation from ASK1 (7, 19).

CIB1 Inhibits TRAF2-ASK1 Interaction and ASK1 Phosphorylation onThr838. Activation of ASK1 requires ASK1 homo-oligomeriza-tion, recruitment of TRAF family proteins to ASK1, and thesubsequent autophosphorylation of a threonine residue in thekinase domain of ASK1 (Thr838 in the human ASK1) (19–23).We therefore examined a possible effect of CIB1 on thoseprocesses required for ASK1 activation. Coimmunoprecipitationanalysis indicated that CIB1 did not affect ASK1 homo-oligomerization (Fig. S4A). We next examined the effect of CIB1on the binding of TRAF2 to ASK1 after transfecting 293T cellswith vectors encoding Flag-TRAF2, HA-ASK1, and Flag-CIB1.Flag-TRAF2 was found to be associated with HA-ASK1, andthis interaction was inhibited by Flag-CIB1 (Fig. 3A). To deter-mine whether endogenous CIB1 also inhibits TRAF2-ASK1interaction, we examined the effect of RNAi-mediated depletionof CIB1 on the interaction. H2O2 induced the association ofTRAF2 with ASK1 in HeLa cells expressing a control siRNA,and this interaction was facilitated in cells expressing CIB1siRNA (Fig. 3B). These results thus suggested that CIB1 inter-feres with the recruitment of TRAF2 to ASK1. Interestingly,CIB1 directly bound to an ASK1 deletion mutant comprisingamino acids 378 to 648 (Fig. 1 A), which is the same region ofASK1 that binds TRAF2 (23). Indeed, CIB1 inhibited thebinding between TRAF2 and ASK1 (1–936) in vitro, whereasmyelin basic protein did not (Fig. S4B).

To investigate the possible effect of CIB1 on ASK1 phosphor-ylation at Thr838, we transfected 293T cells with a vector forHA-ASK1 alone or together with a vector for Flag-CIB1.Immunoblot analysis with antibodies specific for the phospho-Thr838 form of ASK1 revealed that H2O2 treatment increased theextent of ASK1 phosphorylation on Thr838, and this effect wasinhibited by coexpression of CIB1 (Fig. 3C). Furthermore, theH2O2–induced phosphorylation of endogenous ASK1 on Thr838

was higher in HeLa cells expressing CIB1 siRNA than in thoseexpressing a control siRNA (Fig. 3D). We next examinedwhether CIB1 might affect the interaction between ASK1 and itsMAP2K substrate, after transfecting 293T cells with vectors forFlag-CIB1, Myc-ASK1, and HA-MKK3. Exposure of cells toH2O2 promoted the binding of ASK1 to MKK3 and this effectwas abolished by coexpression of CIB1 (Fig. 3E). Moreover,RNAi-mediated depletion of CIB1 in HeLa cells resulted in apotentiation of the H2O2–induced interaction between ASK1and MKK3 (Fig. 3F). Taken together, these results suggestedthat CIB1 inhibits the TRAF2-ASK1 interaction, the phosphor-ylation of ASK1 on Thr838, and the binding of ASK1 to itssubstrate.

RNAi-Mediated Depletion of CIB1 Potentiates ASK1-Mediated Apo-ptosis. The neurotoxin 6-hydroxydopamine (6-OHDA) generatesROS (24), and ROS-initiated stress mediates 6-OHDA-inducedapoptosis in dopaminergic neurons (25). ASK1 has also beenshown to mediate 6-OHDA-induced cell death (26). Indeed,ASK1(K709R), a dominant-negative form of ASK1, blocked6-OHDA-induced apoptosis in dopaminergic neuroblastomaSH-SY5Y cells (Fig. S5A). We therefore investigated whetherCIB1 might inhibit ASK1 activation and apoptosis in SH-SY5Ycells exposed to 6-OHDA. 6-OHDA increased both ASK1activity and apoptosis in cells expressing a control siRNA, andthese effects were potentiated in cells expressing CIB1 siRNA(Fig. 4A). Depletion of CIB1 by RNAi also enhanced H2O2-induced apoptosis in HeLa cells (Fig. S5B).

We also examined the effect of CIB1 on TNF-�-induced

apoptosis in human breast cancer MCF7 cells. TNF-�-inducedapoptosis in MCF7 cells was inhibited by expression ofASK1(K709R) (Fig. S5C), implicating ASK1 as a mediator ofthis process. The RNAi-mediated depletion of CIB1 also po-tentiated ASK1 activation and apoptosis induced by TNF-� inMCF7 cells, compared with those in cells transfected with acontrol siRNA (Fig. 4B).

Ca2� Influx Reverses the Inhibitory Action of CIB1 on ASK1 Signaling.CIB1 is a Ca2�-binding protein (11). We therefore investigatedwhether KCl-induced Ca2� influx might modulate the action ofCIB1 on ASK1 signaling in SH-SY5Y cells exposed to 6-OHDA.KCl-induced membrane depolarization inhibited the physicalassociation between CIB1 and ASK1, and this inhibition was

Fig. 3. CIB1 inhibits the recruitment of TRAF2 to ASK1, the phosphorylationof ASK1 on Thr838, and the interaction between ASK1 and MKK3. (A) The 293Tcells were transfected for 48 h with vectors for Flag-CIB1, HA-ASK1, andFlag-TRAF2, as indicated. The cell lysates were subjected to immunoprecipi-tation with anti-HA antibody, and the resulting precipitates were immuno-blotted with anti-Flag antibody. (B and D) HeLa cells expressing a controlsiRNA or CIB1 siRNA were untreated or treated with 1 mM H2O2 for 20 min,lysed, and subjected to immunoprecipitation with anti-ASK1 antibody. Theresulting precipitates were subjected to immunoblot analysis with antibodiesto TRAF2 (B) or to the phospho-Thr838 form of ASK1 (pASK1) (D). (C) The 293Tcells were transfected for 48 h with plasmid vectors for HA-ASK1 and Flag-CIB1as indicated. The cells were left untreated or treated with 1 mM H2O2 for 20min, lysed, and subjected to immunoprecipitation with anti-HA antibody. Theprecipitates were subjected to immunoblot analysis with antibody to thephospho-Thr838 form of ASK1 (pASK1). (E) The 293T cells were transfected for48 h with vectors for Flag-CIB1, HA-MKK3, and Myc-ASK1 as indicated, incu-bated in the absence or presence of 1 mM H2O2 for 20 min, lysed, and subjectedto immunoprecipitation with anti-Myc antibody. The resulting precipitateswere immunobloted with anti-HA antibody. (F) HeLa cells expressing a controlsiRNA or CIB1 siRNA were untreated or treated with 1 mM H2O2 for 20 min,lysed, and subjected to immunoprecipitation with anti-MKK3 antibody. Theresulting precipitates were examined by immunoblotting with anti-ASK1antibody.

Yoon et al. PNAS � October 13, 2009 � vol. 106 � no. 41 � 17391

CELL

BIO

LOG

Y

reversed by a Ca2� chelator BAPTA-AM (Fig. 5A Left). Fur-thermore, KCl treatment promoted the interaction betweenTRAF2 and ASK1 in the cells exposed to 6-OHDA, and that thiseffect was blocked by BAPTA-AM (Fig. 5A Right). To examinefurther whether Ca2� regulates the ability of CIB1 to bind ASK1,we transfected SH-SY5Y cells with a vector for HA-ASK1together with a vector either for Flag-CIB1 or for Flag-CIB1(D127N/E172Q). CIB1(D127N/E172Q) contains a pointmutation in each of the two functional EF-hand motifs of CIB1and is greatly impaired in the ability to bind Ca2� (27). KCltreatment inhibited the binding between Flag-CIB1 and HA-ASK1, and this inhibition was reversed by BAPTA-AM (Fig.S6A). In contrast, the binding between Flag-CIB1(D127N/E172Q) and HA-ASK1 was not affected by KCl treatment.Together, these results suggested that KCl-induced Ca2� influxdisrupted the interaction between CIB1 and ASK1 in cells. TheCa2� ionophore ionomycin also abolished the interaction be-tween CIB1 and ASK1 in cells (Fig. S6B).

We next examined the effect of KCl-induced Ca2� influx onASK1 activation and apoptosis induced by 6-OHDA in SH-SY5Y cells expressing either control or CIB1 siRNA. KCltreatment potentiated 6-OHDA-elicited ASK1 activation (Fig.5B) and apoptosis (Fig. 5C) in cells expressing the controlsiRNA, and these effects of KCl were abolished by BAPTA-AM.Compared with control cells, cells expressing the CIB1 siRNAshowed higher levels of ASK1 activity and apoptosis induced by6-OHDA. Furthermore, KCl treatment did not affect 6-OHDA-induced ASK1 activation and apoptosis in the CIB1-knockdown

cells. Together, these results suggested that KCl-induced Ca2�

influx abolishes the interaction between CIB1 and ASK1 and theinhibitory action of CIB1 on ASK1 signaling in 6-OHDA-treatedcells.

Finally, we performed similar experiments with primary cul-tures of rat mesencephalic dopaminergic neurons. 6-OHDAinduced activation of ASK1 in neurons transfected with a controlsiRNA, and this effect was potentiated by KCl treatment in amanner sensitive to BAPTA-AM (Fig. 6A). The 6-OHDA-induced activation of ASK1 in cells transfected with CIB1 siRNAwas more pronounced than that in cells transfected with thecontrol siRNA and was not affected by KCl. We also examined6-OHDA-induced cell death in tyrosine hydroxylase (TH)-positive neurons of the cultures transfected with control or CIB1siRNA (Fig. 6B). 6-OHDA reduced the number of TH-positivecells in cultures transfected with the control siRNA, indicative ofthe cytotoxic effect of 6-OHDA on TH-positive dopaminergicneurons. RNAi-mediated depletion of CIB1 potentiated thecytotoxic action of 6-OHDA in TH-positive cells. Whereas KCltreatment further increased the cytotoxicity of 6-OHDA forTH-positive cells and this effect was reversed by BAMTA-AMin cultures transfected with the control siRNA, it had no sucheffect in those transfected with the CIB1 siRNA.

Fig. 4. Depletion of CIB1 by RNAi potentiates ASK1 activation and apoptosisinduced by 6-OHDA or by TNF-�. (A) SH-SY5Y cells expressing a control siRNAor CIB1 siRNA were incubated in the absence or presence of 6-OHDA (100 �M)for 1 h, lysed, and examined for ASK1 activity by immune complex kinase assay(Left). Alternatively, the cells were incubated first in the absence or presenceof 6-OHDA (100 �M) for 2 h and then without 6-OHDA for an additional 12 h.The cells were stained with annexin V and PI, and the percentage of apoptoticcells was determined by flow cytometry (Right). (B) MCF7 cells expressing acontrol siRNA or CIB1 siRNA were incubated in the absence or presence ofTNF� (20 ng/mL) for 20 min, lysed, and assayed for ASK1 activity (Left) byimmune complex kinase assay. Alternatively, the cells were incubated in theabsence or presence of TNF-� (20 ng/mL) plus cycloheximide (3 ng/mL) for 20 hand then examined for apoptosis (Right) as in A. Data from apoptosis assaysin A and B are means � SD of values from three independent experiments.

Fig. 5. KCl-elicited Ca2� influx prevents the inhibitory action of CIB1 on6-OHDA-induced ASK1 activation and apoptosis. (A) SH-SY5Y cells were incu-bated first with 6-OHDA (100 �M) for 50 min and then without or with KCl (40mM) either alone or together with BAPTA-AM (2.5 �M) for 10 min, as indi-cated. Cell lysates were subjected to immunoprecipitation with anti-ASK1antibody or rabbit preimmune IgG (P.I.), and the resulting precipitates wereexamined by immunoblotting with antibody to CIB1 (Left) or to TRAF2 (Right).(B) SH-SY5Y cells expressing a control siRNA or CIB1 siRNA were treated with6-OHDA, KCl, BAPTA-AM as in A. Cell lysates were assayed for ASK1 activity byimmune complex kinase assay. (C) SH-SY5Y cells expressing a control siRNA orCIB1 siRNA were incubated consecutively with 6-OHDA (100 �M) for 50 min,with KCl (40 mM) either alone or together with BAPTA-AM (2.5 �M) for 30min, and in DMEM containing 2% FBS for 12 h, as indicated. Cells wereanalyzed for apoptosis by flow cytometry. The percentages of apoptotic cellsin a representative experiment are indicated.

17392 � www.pnas.org�cgi�doi�10.1073�pnas.0812259106 Yoon et al.

DiscussionWe have demonstrated a function of CIB1, the negative regu-lation of stress-induced signaling through targeting ASK1. CIB1mitigated ASK1 activation and ASK1-mediated cell death, andthis effect was reversed by Ca2� influx. These observations thussuggest that CIB1 functions as a Ca2�-sensitive inhibitor ofASK1-JNK/p38 MAPK signaling axes.

Activation of ASK1 in cellular response to various stimuliincluding ROS requires ASK1 homo-oligomerization, recruit-ment of TRAF2 or TRAF6 to ASK1, and the autophosphory-lation of a threonine residue in ASK1 (Thr838 or Thr845 of humanor mouse ASK1, respectively) (19–23). These events may bepotentially susceptible to modulation by ASK1-interacting pro-teins, among which thioredoxin was the first identified ASK1-inhibitory protein (4, 7). Thioredoxin, a thiol redox-sensitiveprotein, binds and thereby inhibits ASK1 only in its reduced form(7). The generation of ROS in cells converts thioredoxin to theoxidative form, resulting in the dissociation of ASK1 fromthioredoxin. Our results now reveal that CIB1 negatively regu-lates ASK1 through a mechanism distinct from that of thiore-doxin-mediated ASK1 inhibition. CIB1 binds ASK1 in cells, andthis binding is not reduced even by high levels of H2O2. Cellularredox conditions thus do not affect the ability of CIB1 to bindand inhibit ASK1. Instead, the binding of CIB1 to ASK1 wasabolished by the Ca2� ionophore ionomycin (Fig. S6B). CIB1appears to compete with TRAF2 for complex formation withASK1, and inhibition of the TRAF2-ASK1 interaction may bethe primary mechanism by which CIB1 suppresses ASK1-mediated signaling. Given that the recruitment of TRAF2 orTRAF6 to ASK1 is associated with activation of the JNK and p38MAPK pathways in the cellular response to diverse stimuli,including not only oxidative stressand TNF-� but also ER stressand anti-microtubule drugs (19, 22, 28, 29), CIB1 may modulatea variety of cellular functions mediated by the TRAF2/6-ASK1signaling axis. Indeed, siRNA-mediated depletion of CIB1 po-tentiates ASK1 activation induced by an ER stressor tunicamycinand an antimicrotubule agent paclitaxel (Fig. S7).

CIB1 possesses two functional EF-hand motifs in its carboxy-terminal region (12, 13). Ca2� induces a conformational changein CIB1 (17) and modulates the binding and regulatory actionsof CIB1 on several CIB1-interacting proteins (18, 30). In thepresent study, the binding of CIB1 to ASK1 was blocked eitherby ionomycin (Fig. S6B) or by Ca2� influx induced by membranedepolarization (Fig. 5). Furthermore, Ca2� influx potentiated6-OHDA-induced TRAF2-ASK1 signaling events. Our findingsin this study are consistent with a model in which ASK1 exists asa complex with an inhibitory binding partner such as CIB1 andthioredoxin under the resting conditions. In this model, Ca2� andROS accelerate the release of ASK1 from the CIB1-ASK1 andthioredoxin-ASK1 complexes, respectively (Fig. S8). ASK1 lib-erated from the inhibitory complexes becomes activated throughthe assembly with TRAF2. An intriguing scenario is that astimulus that evokes both ROS generation and Ca2� influx isable to induce the activation of ASK1 dissociated from thiore-doxin and from CIB1, thereby promoting TRAF2-ASK1 signal-ing with efficacy higher than that of a stimulus inducing ROSgeneration only (Fig. S8). In relation to this scenario, our studyis under way to clarify a possible regulatory role of CIB1 in theactivation of ASK1-JNK signaling processes initiated by NMDAsignaling, which triggers both Ca2� influx and ROS productionin neurons in the brain (31).

Several Ca2�/calmodulin-binding proteins have been associ-ated with the regulation of ASKl activity (32–34). Ca2�/calmod-ulin-dependent protein kinase II activates ASK1 by phosphor-ylation in neurons (32, 33). In our study, however, KN-93, aninhibitor of Ca2�/calmodulin-dependent protein kinase II, didnot affect the potentiating effect of Ca2� influx on 6-OHDA-induced ASK1 activation (Fig. S9). The Ca2�/calmodulin-activated protein phosphatase calcineurin also activates ASK1 bypromoting its release from 14–3-3 proteins in cardiomyocytes(34). It is thus possible that Ca2� regulates the activation ofASK1 by means of distinct Ca2� sensors, including CIB1, in amanner dependent on cell type and conditions.

On the basis of our observations, we propose that CIB1functions as a switch that antagonizes the TRAF2-ASK1-JNK/

Fig. 6. KCl-elicited Ca2� influx reverses the inhibitory action of CIB1 on6-OHDA-induced ASK1 activation and cell death in rat mesencephalic dopa-minergic neurons. Primary mesencephalic neurons were transfected for 48 hwith control or rat CIB1 siRNA oligonucleotides. The neurons were untreatedor treated with 50 �M 6-OHDA for 50 min and then incubated for 10 min in theabsence or presence of KCl (40 mM) alone or together with BAPTA-AM (2.5�M), as indicated. The cells were lysed and then assayed for ASK1 activity (A)or incubated for an additional 24 h in Neurobasal medium, and subjected toimmunocytochemistry with anti-TH antibody (B). (B) Shown are representa-tive images of the stained cells (Lower) (scale bar, 100 �m), and the viability ofTH-positive neurons (Upper) as a percentage of that for untreated controlsiRNA-transfected cells. Data are means � SD of values from three indepen-dent experiments. *, P � 0.01 represents the result of Student’s t test forcontrol siRNA-transfected cells with no treatment versus those with 6-OHDAtreatment. **, P � 0.01 for differences between control siRNA-transfectedcells treated with 6-OHDA and those treated with 6-OHDA � KCl. ***, P � 0.01for differences between control siRNA-transfected cells treated with 6-OHDA� KCl and those treated with 6-OHDA � KCl � BAPTA-AM. *, P � 0.01 fordifferences between CIB1 siRNA-transfected cells with no treatment and thosewith 6-OHDA treatment. F, P � 0.01 for differences between control siRNA-transfected cells treated with 6-OHDA and CIB1 siRNA-transfected cellstreated with 6-OHDA. NS, no significance between the indicated paired data.

Yoon et al. PNAS � October 13, 2009 � vol. 106 � no. 41 � 17393

CELL

BIO

LOG

Y

p38 signaling axes and thereby regulates a variety of both normaland disease-associated cellular events in a Ca2�-sensitive man-ner. This role of CIB1 in the regulation of stress signalingsuggests CIB1 as a potential molecular target for the develop-ment of new therapeutics specific for human disorders whosepathogenesis is linked to ASK1-mediated signaling events.

Materials and MethodsPrimary Mesencephalic Neuronal Cell Culture. Primary cultures of ventralmesencephalic dopaminergic neurons were prepared from rat embryos atembryonic day 14 and cultured as described in ref. 35.

Immune Complex Kinase Assays. Immune complex kinase assays were per-formed as described in refs. 36 and 37. GST fusion proteins of MKK6(K82A),c-Jun (1–79), and ATF2 were used as substrates for ASK1, JNK, and p38 MAPK,respectively

In Vitro Binding Assay. ASK1 variants were translated in vitro in the presenceof [35S]methionine with the use of a TNT reticulocyte lysate system (Promega).The 35S-labled proteins were incubated at 4 °C for 4 h in a binding buffer (10)with GST-fused proteins immobilized on glutathione-agarose beads. Thebound 35S-labeled proteins were eluted from the beads and analyzed bySDS/PAGE and with a Fuji BAS 2500 phosphoimager.

Coimmunoprecipitation Analysis. Soluble fractions of cell lysates were incu-bated at 4 °C first for 2 h with appropriate antibodies and then for 1 h in theadditional presence of protein G-coupled Sepharose beads (Amersham Bio-sciences). The resulting precipitates were subjected to SDS/PAGE followed byimmunoblot analysis with the use of ECL reagents (Amersham Biosciences).

Apoptotic Cell Death. The cells were fixed, and stained with fluoresceinisothiocyanate (FITC)-conjugated annexin V and propidium iodide (PI). Apo-ptotic cells [annexin V-(�)/PI-(�)] were analyzed by flow cytometry with theuse of CellQuest software (BD Biosciences).

RNA Interference, Immunocytochemistry, and Assay of Cytotoxicity for Mesen-cephalic Dopaminergic Neurons. See SI Materials and Methods.

ACKNOWLEDGMENTS. We thank Drs. H. Ichijo (University of Tokyo, Tokyo) forASK1, R. J. Davis (University of Massachusetts Medical School, Worcester) forJNK1, MKK7, L. I. Zon (Harvard Medical School, Boston) for SEK1, K. Yoshioka(Kanazawa University, Kanazawa, Japan) for Flag-SEK1, J. Woodgett (SamuelLunenfeld Research Institute, Toronto, ON, Canada) for MKK3, J. Han (ScrippsResearch Institute, La Jolla, CA) for MKK6, R. J. Ulevitch (Scripps ResearchInstitute, La Jolla, CA) for p38, and J. S. Gutkind (National Institutes of Health,Bethesda) for MLK3. This work was supported by a BK21 postdoctoral fellow-ship (to J.K.L.); Brain Research Center of the 21st Century Frontier ResearchProgram Grant M103KV010004-08K2201-00410; Korea Research FoundationGrant KRF-2006-341-C00023; and Basic Science Research Program Grant 2009-0080895 through the National Research Foundation of Korea (NRF) funded bythe Ministry of Education, Science and Technology, South Korea (to E.-J.C.).

1. Chang L, Karin M (2001) Mammalian MAPK signalling cascades. Nature 410:37–40.2. Morrison DK, Davis RJ (2003) Regulation of MAP kinase signaling modules by scaffold

proteins in mammals. Annu Rev Cell Dev Biol 19:91–118.3. Ichijo H, et al. (1997) Induction of apoptosis by ASK1, a mammalian MAPKKK that

activates SAPK/JNK and p38 signaling pathways. Science 275:90–94.4. Takeda K, Noguchi T, Naguro I, Ichijo H (2008) Apoptosis signal-regulating kinase 1 in

stress and immune response. Annu Rev Pharmacol Toxicol 48:199–225.5. Chang HY, Nishitoh H, Yang X, Ichijo H, Baltimore D (1998) Activation of apoptosis

signal-regulating kinase 1 (ASK1) by the adapter protein Daxx. Science 281:1860–1863.6. Nishitoh H, et al. (1998) ASK1 is essential for JNK/SAPK activation by TRAF2. Mol Cell

2:389–395.7. Saitoh M, et al. (1998) Mammalian thioredoxin is a direct inhibitor of apoptosis

signal-regulating kinase (ASK) 1. EMBO J 17:2596–2606.8. Cho SG, et al. (2001) Glutathione S-transferase mu modulates the stress-activated

signals by suppressing apoptosis signal-regulating kinase 1. J Biol Chem 276:12749–12755.

9. Park HS, et al. (2002) Heat shock protein hsp72 is a negative regulator of apoptosissignal-regulating kinase 1. Mol Cell Biol 22:7721–7730.

10. Cho SG, et al. (2003) Identification of a novel antiapoptotic protein that antagonizesASK1 and CAD activities. J Cell Biol 163:71–81.

11. Naik UP, Patel PM, Parise LV (1997) Identification of a novel calcium-binding proteinthat interacts with the integrin alphaIIb cytoplasmic domain. J Biol Chem 272:4651–4654.

12. Gentry HR, et al. (2005) Structural and biochemical characterization of CIB1 delineatesa new family of EF-hand-containing proteins. J Biol Chem 280:8407–8415.

13. Hwang PM, Vogel HJ (2000) Structures of the platelet calcium- and integrin-bindingprotein and the alphaIIb-integrin cytoplasmic domain suggest a mechanism for calci-um-regulated recognition; homology modelling and NMR studies. J Mol Recognit13:83–92.

14. Wu X, Lieber MR (1997) Interaction between DNA-dependent protein kinase and anovel protein, KIP. Mutat Res 385:13–20.

15. Kauselmann G, et al. (1999) The polo-like protein kinases Fnk and Snk associate witha Ca(2�)- and integrin-binding protein and are regulated dynamically with synapticplasticity. EMBO J 18:5528–5539.

16. Stabler SM, Ostrowski LL, Janicki SM, Monteiro MJ (1999) A myristoylated calcium-binding protein that preferentially interacts with the Alzheimer’s disease presenilin 2protein. J Cell Biol 145:1277–1292.

17. Yamniuk AP, Nguyen LT, Hoang TT, Vogel HJ (2004) Metal ion binding properties andconformational states of calcium- and integrin-binding protein. Biochemistry43:2558–2568.

18. Shock DD, et al. (1999) Calcium-dependent properties of CIB binding to the integrinalphaIIb cytoplasmic domain and translocation to the platelet cytoskeleton. BiochemJ 342:729–735.

19. Liu H, Nishitoh H, Ichijo H, Kyriakis JM (2000) Activation of apoptosis signal-regulatingkinase 1 (ASK1) by tumor necrosis factor receptor-associated factor 2 requires priordissociation of the ASK1 inhibitor thioredoxin. Mol Cell Biol 20:2198–2208.

20. Gotoh Y, Cooper JA (1998) Reactive oxygen species- and dimerization-induced activa-tion of apoptosis signal-regulating kinase 1 in tumor necrosis factor-alpha signaltransduction. J Biol Chem 273:17477–17482.

21. Tobiume K, Saitoh M, Ichijo H (2002) Activation of apoptosis signal-regulating kinase1 by the stress-induced activating phosphorylation of pre-formed oligomer. J CellPhysiol 191:95–104.

22. Noguchi T, et al. (2005) Recruitment of tumor necrosis factor receptor-associated factorfamily proteins to apoptosis signal-regulating kinase 1 signalosome is essential foroxidative stress-induced cell death. J Biol Chem 280:37033–37040.

23. Fujino G, et al. (2007) Thioredoxin and TRAF family proteins regulates ROS-dependentactivation of ASK1 through reciprocal modulation of the N-terminal homophilicinteraction of ASK1. Mol Cell Biol 27:8152–8163.

24. Heikkila R, Cohen G (1973) 6-Hydroxydopamine: Evidence for superoxide radical as anoxidative intermediate. Science 181:456–457.

25. Blum D, et al. (2001) Molecular pathways involved in the neurotoxicity of 6-OHDA,dopamine and MPTP: Contribution to the apoptotic theory in Parkinson’s disease. ProgNeurobiol 65:135–172.

26. Ouyang M, Shen X (2006) Critical role of ASK1 in the 6-hydroxydopamine-inducedapoptosis in human neuroblastoma SH-SY5Y cells. J Neurochem 97:234–244.

27. Zhu J, Stabler SM, Ames JB, Baskakov I, Monteiro MJ (2004) Calcium binding sequencesin calmyrin regulates interaction with presenilin-2. Exp Cell Res 300:440–454.

28. Nishitoh H, et al. (2002) ASK1 is essential for endoplasmic reticulum stress-inducedneuronal cell death triggered by expanded polyglutamine repeats. Genes Dev16:1345–1355.

29. Mhaidat NM, Thorne R, Zhang XD, Hersey P (2008) Involvement of endoplasmicreticulum stress in docetaxel-induced JNK-dependent apoptosis of human melanoma.Apoptosis 13:1505–1512.

30. Leisner TM, Liu M, Jaffer ZM, Chernoff J, Parise LV (2005) Essential role of CIB1 inregulating PAK1 activation and cell migration. J Cell Biol 170:465–476.

31. Nakamura T, Lipton SA (2009) Cell death: Protein misfolding and neurodegenerativediseases. Apoptosis 14:455–468.

32. Sagasti A, et al. (2001) The CaMKII UNC-43 activates the MAPKKK NSY-1 to execute alateral signaling decision required for asymmetric olfactory neuron fates. Cell 105:221–232.

33. Takeda K, et al. (2004) Involvement of ASK1 in Ca2�-induced p38 MAP kinase activa-tion. EMBO Rep 5:161–166.

34. Liu Q, Wilkins BJ, Lee YJ, Ichijo H, Molkentin JD (2006) Direct interaction and reciprocalregulation between ASK1 and calcineurin-NFAT control cardiomyocyte death andgrowth. Mol Cell Biol 26:3785–3797.

35. Kim SY, et al. (2006) The dopamine D2 receptor regulates the development of dopa-minergic neurons via extracellular signal-regulated kinase and Nurr1 activation. J Neu-rosci 26:4567–4576.

36. Park HS, Lee JS, Huh SH, Seo JS, Choi EJ (2001) Hsp72 functions as a natural inhibitoryprotein of c-Jun N-terminal kinase. EMBO J 20:446–456.

37. Ryoo K, et al. (2004) Negative regulation of MEKK1-induced signaling by glutathioneS-transferase Mu. J Biol Chem 279:43589–43594.

17394 � www.pnas.org�cgi�doi�10.1073�pnas.0812259106 Yoon et al.