21
Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37 Chaperone Complex in Cardiomyocyte Asuka Ota, Jun Zhang, Peipei Ping, Jiahuai Han, Yibin Wang Rationale: p38 is an important stress activated protein kinase involved in gene regulation, proliferation, differentiation, and cell death regulation in heart. p38 kinase activity can be induced through canonical pathway via upstream kinases or by noncanonical autophosphorylation. The intracellular p38 kinase activity is tightly regulated and maintained at low level under basal condition. The underlying regulatory mechanism for canonical p38 kinase activation is well-studied, but the regulation of noncanonical p38 autophosphorylation remains poorly understood. Objective: We investigated the molecular basis for the regulation of noncanonical p38 autophosphorylation and its potential functional impact in cardiomyocytes. Methods and Results: Using both proteomic and biochemical tools, we established that heat shock protein (Hsp)90- Cdc37 chaperones are part of the p38 signaling complex in mammalian cells both in vitro and in vivo. The Hsp90-Cdc37 chaperone complex interacts with p38 via direct binding between p38 and Cdc37. Cdc37 expression is both sufficient and necessary to suppress noncanonical p38 activation via autophosphorylation at either basal state or under TAB1 (TAK1 binding protein-1) induction. In contrast, Cdc37 expression has no impact on p38 activation by canonical upstream kinase MKK3 or oxidative stress. Furthermore, Hsp90 inhibition results in p38 activation via autophosphorylation, and p38 activity contribute to apoptotic cell death induced by Hsp90 inhibition. Conclusion: Our study has revealed a so far uncharacterized function of Hsp90-Cdc37 as an endogenous regulator of noncanonical p38 activity. (Circ Res. 2010;106:1404-1412.) Key Words: p38 mitogen-activated protein kinase noncanonical pathway Hsp90 Cdc37 autophosphorylation O xidative stress, proinflammatory cytokines and many other extracellular stimuli activate intracellular stress responses through stress signaling pathways, including SAPK (stress-activated protein kinase) p38 as a member of the mitogen-activated protein kinase (MAPK) superfamily. 1–5 Although the role of p38 in heart remains controversial, activation of p38 in response to ischemia/reperfusion has been demonstrated repetitively. Several findings suggest p38 contributes to myocyte cell death by promoting produc- tion of proinflammatory cytokines including tumor necrosis factor (TNF), interleukin (IL)-1, and IL-8. This is also supported by the protective effect of p38-specific inhibitor SB203580 in ischemia. On the other hand, protective effect of p38 through activation of heat shock protein (Hsp)27 has also been suggested. There are 2 modes of p38 activation in ischemia. One is the canonical p38 pathway, which is activated by a cascade of phosphorylation through upstream MEKKs (MAPK/extracellular-activated kinase kinases) such as Ask1, 6 and MAPK kinase (MKK)3, -4, or -6, leading to phosphorylation of the TGY motif within the p38 activation loop. 3,4,7 In addition, noncanonical MKK-independent activa- tion of p38 has been demonstrated through intrinsic autophos- phorylation 8,9 that can be facilitated by direct interaction with TAB1 or ZAP-70. It has been speculated that these 2 different pathways are activated by different stimuli, and contribute differently to regulate p38 functions in response to ischemia/reperfusion. The canonical p38 activation is a highly regulated process involving well-studied upstream kinases (MKKs) and protein phosphatases (MKPs) (PP2C/) as its positive and negative regulators respectively. 10 –12 In contrast, the regulatory mech- anism for noncanonical p38 pathway remains poorly under- stood. Autophosphorylation is an intrinsic property of p38 kinase, 8,9 and although TAB1 and ZAP-70 are reported to function as positive inducers of p38 autophosphorylation, 13–16 the molecular mechanism in its negative regulation remains unclear. It is important to note that p38 activity is maintained at low level in cells in the absence of stress stimulus, suggesting that autophosphorylation activity of p38 is nor- mally suppressed at the basal state. In this study, we identified Hsp90-Cdc37 chaperone com- plex as a specific component of p38 kinase signaling com- Original received November 24, 2009; revision received March 2, 2010; accepted March 4, 2010. From the Division of Molecular Medicine (A.O., Y.W.), Departments of Anesthesiology and Physiology (J.Z., P.P.), and Molecular Biology Institute (Y.W.), David Geffen School of Medicine, University of California, Los Angeles; and Key Laboratory of the Ministry of Education for Cell Biology and Tumor Cell Engineering (J.H.), School of Life Sciences, Xiamen University, China. Correspondence to Yibin Wang, PhD, Department of Anesthesiology, Medicine and Physiology, BH 569, CHS, 650 Charles E. Young Dr, David Geffen School of Medicine at UCLA, Los Angeles, CA 90095. E-mail [email protected] © 2010 American Heart Association, Inc. Circulation Research is available at http://circres.ahajournals.org DOI: 10.1161/CIRCRESAHA.109.213769 1404 by guest on June 18, 2018 http://circres.ahajournals.org/ Downloaded from by guest on June 18, 2018 http://circres.ahajournals.org/ Downloaded from by guest on June 18, 2018 http://circres.ahajournals.org/ Downloaded from by guest on June 18, 2018 http://circres.ahajournals.org/ Downloaded from by guest on June 18, 2018 http://circres.ahajournals.org/ Downloaded from by guest on June 18, 2018 http://circres.ahajournals.org/ Downloaded from by guest on June 18, 2018 http://circres.ahajournals.org/ Downloaded from by guest on June 18, 2018 http://circres.ahajournals.org/ Downloaded from by guest on June 18, 2018 http://circres.ahajournals.org/ Downloaded from by guest on June 18, 2018 http://circres.ahajournals.org/ Downloaded from by guest on June 18, 2018 http://circres.ahajournals.org/ Downloaded from by guest on June 18, 2018 http://circres.ahajournals.org/ Downloaded from by guest on June 18, 2018 http://circres.ahajournals.org/ Downloaded from

Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

  • Upload
    lamanh

  • View
    223

  • Download
    1

Embed Size (px)

Citation preview

Page 1: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

Specific Regulation of Noncanonical p38� Activation byHsp90-Cdc37 Chaperone Complex in Cardiomyocyte

Asuka Ota, Jun Zhang, Peipei Ping, Jiahuai Han, Yibin Wang

Rationale: p38 is an important stress activated protein kinase involved in gene regulation, proliferation, differentiation,and cell death regulation in heart. p38 kinase activity can be induced through canonical pathway via upstream kinasesor by noncanonical autophosphorylation. The intracellular p38 kinase activity is tightly regulated and maintained atlow level under basal condition. The underlying regulatory mechanism for canonical p38 kinase activation iswell-studied, but the regulation of noncanonical p38 autophosphorylation remains poorly understood.

Objective: We investigated the molecular basis for the regulation of noncanonical p38 autophosphorylation and itspotential functional impact in cardiomyocytes.

Methods and Results: Using both proteomic and biochemical tools, we established that heat shock protein (Hsp)90-Cdc37 chaperones are part of the p38� signaling complex in mammalian cells both in vitro and in vivo. TheHsp90-Cdc37 chaperone complex interacts with p38 via direct binding between p38 and Cdc37. Cdc37 expression isboth sufficient and necessary to suppress noncanonical p38 activation via autophosphorylation at either basal state orunder TAB1 (TAK1 binding protein-1) induction. In contrast, Cdc37 expression has no impact on p38 activation bycanonical upstream kinase MKK3 or oxidative stress. Furthermore, Hsp90 inhibition results in p38 activation viaautophosphorylation, and p38 activity contribute to apoptotic cell death induced by Hsp90 inhibition.

Conclusion: Our study has revealed a so far uncharacterized function of Hsp90-Cdc37 as an endogenous regulatorof noncanonical p38 activity. (Circ Res. 2010;106:1404-1412.)

Key Words: p38 mitogen-activated protein kinase � noncanonical pathway � Hsp90 � Cdc37 � autophosphorylation

Oxidative stress, proinflammatory cytokines and manyother extracellular stimuli activate intracellular stress

responses through stress signaling pathways, including SAPK(stress-activated protein kinase) p38 as a member of themitogen-activated protein kinase (MAPK) superfamily.1–5

Although the role of p38 in heart remains controversial,activation of p38 in response to ischemia/reperfusion hasbeen demonstrated repetitively. Several findings suggestp38� contributes to myocyte cell death by promoting produc-tion of proinflammatory cytokines including tumor necrosisfactor (TNF)�, interleukin (IL)-1, and IL-8. This is alsosupported by the protective effect of p38-specific inhibitorSB203580 in ischemia. On the other hand, protective effect ofp38 through activation of heat shock protein (Hsp)27 has alsobeen suggested. There are 2 modes of p38 activation inischemia. One is the canonical p38 pathway, which isactivated by a cascade of phosphorylation through upstreamMEKKs (MAPK/extracellular-activated kinase kinases) suchas Ask1,6 and MAPK kinase (MKK)3, -4, or -6, leading tophosphorylation of the TGY motif within the p38 activationloop.3,4,7 In addition, noncanonical MKK-independent activa-

tion of p38 has been demonstrated through intrinsic autophos-phorylation8,9 that can be facilitated by direct interaction withTAB1 or ZAP-70. It has been speculated that these 2 differentpathways are activated by different stimuli, and contributedifferently to regulate p38 functions in response toischemia/reperfusion.

The canonical p38 activation is a highly regulated processinvolving well-studied upstream kinases (MKKs) and proteinphosphatases (MKPs) (PP2C�/�) as its positive and negativeregulators respectively.10–12 In contrast, the regulatory mech-anism for noncanonical p38 pathway remains poorly under-stood. Autophosphorylation is an intrinsic property of p38kinase,8,9 and although TAB1 and ZAP-70 are reported tofunction as positive inducers of p38 autophosphorylation,13–16

the molecular mechanism in its negative regulation remainsunclear. It is important to note that p38 activity is maintainedat low level in cells in the absence of stress stimulus,suggesting that autophosphorylation activity of p38 is nor-mally suppressed at the basal state.

In this study, we identified Hsp90-Cdc37 chaperone com-plex as a specific component of p38 kinase signaling com-

Original received November 24, 2009; revision received March 2, 2010; accepted March 4, 2010.From the Division of Molecular Medicine (A.O., Y.W.), Departments of Anesthesiology and Physiology (J.Z., P.P.), and Molecular Biology Institute

(Y.W.), David Geffen School of Medicine, University of California, Los Angeles; and Key Laboratory of the Ministry of Education for Cell Biology andTumor Cell Engineering (J.H.), School of Life Sciences, Xiamen University, China.

Correspondence to Yibin Wang, PhD, Department of Anesthesiology, Medicine and Physiology, BH 569, CHS, 650 Charles E. Young Dr, DavidGeffen School of Medicine at UCLA, Los Angeles, CA 90095. E-mail [email protected]

© 2010 American Heart Association, Inc.

Circulation Research is available at http://circres.ahajournals.org DOI: 10.1161/CIRCRESAHA.109.213769

1404

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

Page 2: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

plex. Hsp90 constitutes 1% to 2% of cellular proteins17 toregulate the stability and the activity of its client proteins,18

including protein kinases.19–23 In most cases, Hsp90 functionsto stabilize its client protein kinases and promotes kinaseactivation. The loss of prosurvival kinases at both protein andactivity levels are implicated as a major mechanism underly-ing the anticancer effects of Hsp90 inhibitors. Based on theevidence that p38 protein level is not affected by Hsp90inhibition, earlier reports have classified p38 as a nonclientprotein of Hsp90.24–26

Using biochemical and molecular approaches, we haveshown both in vitro and in vivo that p38 specifically interactswith Hsp90 chaperone through direct binding to its cochap-erone Cdc37. Recombinant Hsp90-Cdc37 is sufficient toattenuate autophosphorylation of recombinant p38 kinase inan Hsp90 activity dependent manner. Furthermore, Cdc37 isboth sufficient and necessary for suppressing basal p38activity, and exogenous Cdc37 expression suppresses nonca-nonical p38 activation via TAB1 but has no impact oncanonical p38 activation via MKK3. In cultured cardiomyo-cytes, inhibition of Hsp90 leads to rapid p38 autophosphor-ylation, which contributes to apoptotic cell death resultedfrom Hsp90 inhibition. In conclusion, we establish, for thefirst time, that p38 is a bona fide client protein of Hsp90-Cdc37 chaperone complex. Hsp90-Cdc37 is essential tospecifically suppress noncanonical autophosphorylation ofp38, although it has no impact on canonical p38 activation.Therefore, Hsp90-Cdc37 represents the first identified non-canonical pathway–specific negative regulator for p38. Thisfinding is a potentially important addition to the mechanisticrepertoire for the diverse function of Hsp90 as a mastersignaling modulator.

MethodsImmunoprecipitationRecombinant proteins were cloned from ventricular cDNA libraryprepared using SSII RT and oligo dT primer according to theprotocol of the manufacturer (Invitrogen). Precipitated proteins (0.1to 0.3 �g) were mixed into binding buffer, and incubated at 4°C inthe presence of anti-Cdc37 (Abcam) and Protein A (GE Healthcare)for 3 hours. Anti-FLAG M2 affinity gel (Sigma) was used for FLAGco-IP, and the experiment was conducted according to the protocolof the manufacturer.

Kinase AssayKinase assay was carried out following a protocol provided by CellSignaling. Briefly, 0.3 �g of GST recombinant proteins were mixedin kinase assay buffer, with or without 5 �mol/L SB203580 or10 �mol/L GA. For a detection of de novo phosphorylation, 5 �Ciof �-P32-ATP was included. Mixtures were incubated for 15 minutesat 30°C, and iced immediately. Phosphorylation was detected byautoradiography.

Short Hairpin RNA–Mediated Cdc37 KnockdownshCdc37 was cloned into pSuper. Its efficacy was verified, andpSuper-shCdc37 was digested with PacI and cloned into pAdEasythrough homologous recombination to produce Adv-shCdc37.

TUNEL StainingCells were grown on laminin coated glass coverslips, fixed in 2%paraformaldehyde, and permeabilized in 1% Triton X-100, 1%sodium citrate solution. TUNEL staining was carried out usingApopTag Peroxidase in situ Apoptosis Detection Kit (Chemicon

International) according to the protocol of the manufacturer. Apo-ptotic occurrence was determined by counting TUNEL-positivecells, which colocalize with DAPI (4�,6-diamidino-2-phenylindole)staining, and dividing by the total number of nucleus. Statisticalsignificance was determined by Student t test.

An expanded Methods section is available in the Online DataSupplement at http://circres.ahajournals.org.

ResultsIdentification of HSP90-Cdc37 in Cardiacp38� ComplexWe and others have reported that p38 activation in heart hasa number of pathological consequences.1,14,27–29 To study themolecular mechanisms of p38 regulation and function inheart, p38� protein complex was isolated through coimmu-noprecipitation (co-IP) from transgenic mouse heart express-ing N-terminally FLAG-tagged dominant negative p38�(FLAG-DNp38�). We used DNp38� instead of wild-typep38� (WTp38�) because p38 phosphorylation, as well as theinteraction between WTp38 and many of its binding partnersare relatively transient.13 The molecular identities of the p38signaling complex was established by liquid chromatogra-phy–tandem mass spectrometric analysis with �95% confi-dence and are summarized in Online Table I. In addition toknown p38 interacting proteins, Hsp90�, Hsp90�, and co-chaperone Cdc37 were identified with significant peptidecoverage (Figure 1A). The presence of Hsp90�/� and Cdc37in the p38� complex was confirmed by Western immunoblots(Figure 1B). As a control for specificity, another knownHsp90 cochaperone, Hsc70 was not detected.30

To determine whether p38� and Hsp90-Cdc37 complexinteract directly, we performed co-IP using recombinantproteins. As shown in Figure 1C, Hsp90 interacts muchstronger with FLAG-DNp38� in the presence of Cdc37, withHsp90� showing a higher affinity than Hsp90�. In contrast,Cdc37 was able to interact with either WT or DN of p38� inthe absence of Hsp90 (Figure 1D). Therefore, p38� appears

Non-standard Abbreviations and Acronyms

17-AAG 17-allylamino-17-demethoxygeldanamycin

DN dominant negative

ERK extracellular signal-regulated kinase

GA geldanamycin

GST glutathione S-transferase

Hsp heat shock protein

IL interleukin

IP immunoprecipitation

JNK c-Jun N-terminal kinase

LacZ gene coding for �-galactosidase

MAPK mitogen-activated protein kinase

MKK mitogen-activated protein kinase kinase

NRVM neonatal rat ventricular myocyte

sh short hairpin RNA

TAB TAK1 binding protein

TNF tumor necrosis factor

WT wild type

Ota et al Noncanonical Regulation of p38 Kinase 1405

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

Page 3: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

to bind directly to Cdc37, which can subsequently recruitp38� into Hsp90-Cdc37 complex, a scheme previously dem-onstrated for other Hsp90-Cdc37 client protein kinases.17,31

Consistently, p38 and Hsp90-Cdc37 in total protein extractprepared from heart tissue had broad distribution pattern on asucrose gradient with significant overlap (Online Figure I, A).However, when the p38 immunocomplex was isolated fromheart and separated on the same sucrose gradient, Hsp90 wasdetected only in the same fractions when both Cdc37 andp38 were present, whereas a separate pool of p38 andCdc37 comigrated together without Hsp90 (Online FigureI, B). All of these findings suggest that p38 specificallyinteracts with Hsp90-Cdc37 complex via direct bindingwith Cdc37.

Molecular Basis of Interaction Between p38�and Cdc37To determine the Cdc37 binding domain on p38�, we createda number of GST-tagged truncation mutants of p38� and

performed co-IP assays with Cdc37 (Figure 2A). The resultsindicated that the N terminus of p38� contains a putativeCdc37 binding motif. Based on the sequence analysis, p38MAPK isoforms share a common GXGXYG motif at their Nterminus, which is reported to function as an ATP bindingsite,32 whereas c-Jun N-terminal kinase (JNK) and extracel-lular signal-regulated kinase (ERK) have GXGXQG andGXGXPG, respectively. Interestingly, GXGXF/YG is re-ported as a conserved Cdc37 binding site.23 However, Y35Asingle mutation or Y35A and G31A double mutations onp38� did not abolish its binding to Cdc37; instead, the GYdouble mutant appeared to bind Cdc37 stronger (Figure 2B).The VIWCI residues within the Hsp90-binding motif ofCdc37 are also reported to be critical for binding of anotherestablished client protein kinase Raf-1.33 However, Cdc37protein with mutation at V192A or both V192A and I193Ahad similar p38� binding capacity as the Cdc37WT (Figure2C). Therefore, the direct interaction between N-terminal

Figure 1. Interaction between Hsp90-Cdc37 and p38� protein. A, Hsp90�,Hsp90�, and Cdc37 are identified inDNp38� immunocomplex by liquid chro-matography–tandem mass spectrometry.The total numbers of identified peptidesequences (Queries Matched), total num-ber of identified peptides (PeptideMatched), the percentage of sequencecovered by the identified peptides (Cov-erage %), and peptide mass spectra fin-gerprinting score based on MASCOTdatabase (MOWSE, �40 indicates a sig-nificant match) are shown. B, Immuno-blots on immunocomplex isolated withanti-FLAG antibody (IP: FLAG) fromDNp38� transgenic (TG) and nontrans-genic (NTG) control hearts using antibod-ies as labeled. C, Co-IP using purifiedrecombinant proteins with anti-FLAGafter incubating FLAG-DNp38� proteinwith GST-Cdc37, GST-Hsp90�, andGST-HSP90� as indicated, followed byimmunoblots with anti-GST, antip38�,anti-Cdc37, and anti-Hsp90�/�. Onlysmall-molecular-mass section containing

GST (28KD) was shown for the GST blots. Note the GST signal from lane 2 of the input panel is from the breakdown product of GST-Hsp90�. D, IP with anti-Cdc37 after incubating recombinant GST-Cdc37 with GST-DNp38�, GST-p38�, or GST alone as a control, followedby immunoblot using anti-Cdc37, antip38�, and anti-GST as indicated. Only small-molecular-mass section containing GST (28 kDa) wasshown for the GST blots. Note the GST signals from lanes 1 and 2 of the input panel are from the breakdown product of GST-p38.

Figure 2. Specific interaction betweenCdc37 and p38�. A, IP with anti-Cdc37after incubating Cdc37 with GST-taggedtruncated p38� as labeled, followed byimmunoblots with anti-GST and anti-Cdc37 as indicated. All the experimentswere repeated at least 3 times. B, Puri-fied recombinant p38� (0.1 �g) withY35A (Y) or Y35A and G31A (GY) muta-tions were subjected to co-IP withCdc37 using anti-Cdc37 and analyzedby immunoblots as indicated. C, Cdc37with single mutation at V131A (Cdc37 V)or double mutations at V192A and I193A(Cdc37 VI) were mixed with recombinantWT- or DN-p38� and immunoprecipi-tated using anti-Cdc37, followed by im-munoblots as indicated.

1406 Circulation Research April 30, 2010

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

Page 4: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

domain of p38� and Cdc37 appears to be mediated by so faruncharacterized motifs.

Hsp90-Cdc37 Chaperone Directly Inhibits p38Autophosphorylation In VitroTo investigate the functional implication of Hsp90-Cdc37interaction with p38, we first performed in vitro phosphory-lation assay using recombinant p38� and the isolated p38�immunocomplex, but we did not detect de novo phosphory-lation at the molecular weight of Hsp90 or Cdc37 (OnlineFigure II), suggesting that Hsp90 and Cdc37 are unlikely thedirect substrates of p38�. To examine whether Hsp90-Cdc37

chaperone complex is able to directly regulate p38� kinaseactivity, in vitro phosphorylation assay was performed usingpurified recombinant proteins (Figure 3). As reported previ-ously,34,35 we observed SB203580-sensitive autophosphory-lation of recombinant p38� before coincubation with Cdc37-Hsp90�/� (�incubation). However, overnight incubationwith recombinant Cdc37 and Hsp90�/� before the kinaseassay resulted in a marked decrease in SB203580-sensitiveautophosphorylation of p38� (�incubation). Such inhibitionof p38� autophosphorylation by Hsp90-Cdc37 was blockedby the addition of Hsp90 inhibitor geldanamycin (GA) duringthe kinase assay. Therefore, Hsp90-Cdc37 chaperone com-plex appears to be responsible for inhibiting p38� autophos-phorylation without other cellular proteins and this inhibitoryfunction is chaperone activity dependent.

Hsp90-Cdc37 Inhibits p38 AutophosphorylationIn VivoOur data suggest that Hsp90-Cdc37 may function to suppressthe autophosphorylation of endogenous p38 kinase in cells.However, it is well established that Hsp90 also regulates thestability and the activity of many of its client kinases via ATPdependent chaperone activity.36–39 As shown in Figure 4A,treatment with GA resulted in a robust induction of p38activity within 5 minutes, which lasted for the duration of theexperiment without significant change in the total proteinlevel of p38 (Figure 4A). In contrast, at 5 minutes after GAtreatment, activities of upstream MKK3/6 and other MAPKpathways including ERK and JNK were not affected. At latertime points, JNK activity was also induced to a modest level,along with phospho-Hsp27, a downstream target of p38,whereas ERK activity remained unaffected. Activation of p38phosphorylation was also observed using 17-AAG (17-allylamino-17-demethoxygeldanamycin) as an alternativeHsp90 inhibitor (Online Figure III, A).

Figure 3. p38� activity assay in vitro. Recombinant Hsp90� orHsp90� plus Cdc37 were mixed and incubated in kinase bufferwith [�-32P]-ATP in the presence or absence of 5 �mol/LSB203580 or 10 �mol/L GA for 30 minutes at 30°C. The top 3panels show immunoblots of the total reaction mixtures usingantibodies as indicated. The bottom 2 panels show autoradio-grams of 32P-labeled p38. Quantification for the 32P-labeled p38is indicated at the bottom of each autoradiogram. This experi-ment was repeated twice with similar results. The “– Incubation”samples were identical to the “� Incubation” samples, exceptthe overnight preincubation at 4°C before incubation at 30°C.

Figure 4. Impact of Hsp90 inhibition onp38 activity. A, NRVMs were treated with1.8 �mol/L GA for different time periodsin minutes as indicated. Immunoblots fortotal protein lysates collected from dupli-cate samples at each time point areshown using specific antibodies aslabeled. NRVMs treated with equal vol-ume of DMSO and IL-1� were includedas basal and positive controls for p38activation respectively. B, NRVMs werepreincubated with 1 �mol/L SB203580 orDMSO for 20 minutes before 1.8 �mol/LGA treatment for different time periods.Immunoblots were performed from thetotal protein lysates using antibodies aslabeled. C, p-p38 level normalized tobackground level was quantified forNRVMs treated with DMSO, GA, andGA�SB for 5 minutes (n�5; *P�0.05,**P�0.005). D, MKK3/6 DKO MEFs(DKO) were pretreated with 1 �mol/LSB203580 for 30 minutes, followed by1.8 �mol/L GA or 0.01% DMSO foranother 30 minutes. Immunoblots on thetotal cell lysates was performed with

antibodies as indicated. Nontreated WT MEF (�) was run on the side as a positive control for the immunoblots. E, p-p38 level normal-ized to background level was quantified for DKO MEF treated with DMSO, GA, and GA�SB for 30 minutes (n�4; *P�0.05, **P�0.001).

Ota et al Noncanonical Regulation of p38 Kinase 1407

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

Page 5: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

Addition of p38 inhibitor SB203580 along with GA effec-tively diminished early induction of p38 phosphorylation, acharacteristic feature of p38 kinase autophosphorylation.After longer GA treatment, an increase in phospho-MKK4was observed, and the level of p38 phosphorylation becameless sensitive to SB203580, suggesting an increased contri-bution of canonical p38 activation at later time points,perhaps attributable to secondary and nonspecific effects ofHsp90 inhibition. Phospho-p38 level in the presence orabsence of GA and SB203580 were also quantified todemonstrate the SB-sensitive activity of p38 in neonatal ratventricular myocytes (NRVMs) (Figure 4C). Furthermore,GA-induced autophosphorylation was observed in the

MKK3/6 DKO MEFs (Figure 4D and 4E), again supportingMKK-independent activation of p38. In agreement with thisscheme, T180 but not Y182 of p38 was selectively phosphor-ylated at 5 minutes after GA treatment, whereas only slightincrease in phospho-Tyr-p38 was observed 15 minutes afterGA treatment (Online Figure IV, A). Furthermore, HeLa cellsexpressing p38�T180A showed blunted response to GA,further supporting the notion that T180 is the site for GAinduced p38 autophosphorylation (Online Figure IV, B).

To investigate the impact of p38 activation status on itsinteraction with Hsp90/Cdc37, we studied endogenous p38kinase interaction with Cdc37 in HeLa cells which alsodemonstrated GA induced p38 autophosphorylation (Figure5A). On Hsp90 inhibition, total p38 protein in the Cdc37immunocomplex was reduced relative to the untreated cells,whereas phospho-p38 was not detected in Cdc37 immuno-complex under either basal or stimulated conditions (Figure5B). This observation suggests that Hsp90-Cdc37 preferen-tially interacts with nonphosphorylated p38 and Hsp90 inhi-bition promotes p38 autophosphorylation associated with itsrelease from the Hsp90-Cdc37 complex.

Cdc37 Expression Is Both Sufficient and Necessaryfor Regulating Basal p38 ActivityFrom these biochemical studies, we can speculate that Hsp90-Cdc37 chaperone complex is a negative regulator for basalp38 activity. To demonstrate the functional significance ofHsp90-Cdc37–mediated regulation of basal p38 activity, weused 2 different Adv-shCdc37 constructs to achieve selectiveknockdown of Cdc37 in HeLa cells. As shown in Figure 6A,Cdc37 short hairpin RNA (shCdc37) expression resulted in adosage dependent decrease in Cdc37 protein level in parallelwith an increase in phospho-p38. Knockdown of Cdc37protein also reduced p38 interaction with Hsp90 based onco-IP assay (Online Figure V). On the other hand, overex-pression of Cdc37 in HeLa cells significantly reduced the

Figure 5. Impact of Hsp90 inhibition on p38 activity and interac-tion with Cdc37 in HeLa cells. A, Immunoblots for phospho-and total p38 on HeLa cells treated with 1.8 �mol/L GA for timeperiods, as indicated in the presence of absence of pretreat-ment with 1�mol/L SB203580. B, Co-IP assay on HeLa cellextract with or without 1.8 �mol/L GA treatment for 15 minutes.Immunoprecipitation with anti-Cdc37 was followed by immuno-blots with antibodies as indicated. Relative intensity of the sig-nals of total p38, Cdc37, and Hsp90 in the elution blots werequantified and shown in a bar graph.

Figure 6. Functional impact of Cdc37expression on canonical and noncanonicalp38 activation. A, HeLa cells were infectedwith different multiplicities of infection (0 to500) of adenoviral vectors expressingscramble or 2 different shCdc37 constructs(shCdc37-1 and shCdc37-2) for 48 hours.Immunoblots for Cdc37, phospho-p38, andtotal p38 were performed using specific anti-bodies as indicated. �-Actin was used as aloading control. B, HeLa cells were trans-fected with vector or Myc-Cdc37 for 24hours, followed by treatment with1.8 �mol/L GA for 15 minutes with or with-out 1 �mol/L SB203580 added 20 minutesbefore the GA treatment. Immunoblots forCdc37, phospho-, and total p38, as well asphospho- and total JNK, were performed asindicated. C, HeLa cells were transfectedwith constitutively active MKK3 (MKK3bE)with or without Cdc37 and treated with1 �mol/L SB203580 for 30 minutes. Immu-noblots for p38 and its downstream sub-strates are shown as indicated. D, Cdc37was expressed in HeLa cells with TAB1�.Immunoblots are shown with �-actin as acontrol.

1408 Circulation Research April 30, 2010

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

Page 6: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

basal p38 activity and blocked p38 autophosphorylationinduced by GA (Figure 6B). Therefore, it appears thatHsp90-Cdc37 chaperone complex is both necessary andsufficient to regulate the basal activity of endogenous p38 bynegatively suppressing its autophosphorylation. However, theelevated basal p38 activity in Cdc37 knock-down cells wasno longer very sensitive to p38 inhibition (Figure 4B). Theunderlying reason is unclear but may be partially attributableto toxicity inflicted by Lipofectamine treatment.

Hsp90-Cdc37 Specifically Regulates Noncanonicalp38 Pathway Without Significant Impact onCanonical p38 ActivationUsing H2O2 as an established p38 inducer via canonicalpathway, we observed that overexpression of Cdc37 did nothave a significant effect on p38 activation and its downstreamsignaling induced by H2O2, suggesting that Hsp90-Cdc37complex has no impact on canonical p38 pathway (OnlineFigure VI). To study this further, HeLa cells were cotrans-fected with constitutively active MKK3 (MKK3bE) andCdc37. As expected, MKK3bE led to a significant inductionof phospho-p38 in the presence or absence of p38 inhibitor(Figure 6C). However, coexpression of Cdc37 did not affectp38 induction by MKK3bE, suggesting that canonical p38activation pathway is not affected by Cdc37 (Figure 6C).

It has been shown that TAK1 binding protein-1 (TAB1)promotes autophosphorylation of p38� through direct inter-action.13 Consistently, expression of TAB1�, which lacks theTAK1 binding domain,15 was sufficient to induce p38 acti-vation in HeLa cells (Figure 6D). To determine whetherCdc37 can regulate TAB1� mediated autophosphorylation ofp38, Cdc37 and TAB1� were coexpressed in HeLa cells(Figure 6D). The results showed that the p38 activation byTAB1� was reduced by the exogenous Cdc37, suggestingthat Cdc37 overexpression can negatively regulate p38 auto-phosphorylation induced by TAB1�. Together with the pre-vious results, our data indicate that Hsp90-Cdc37 chaperone

complex selectively regulates the noncanonical p38 activa-tion pathway, but not the MKK-dependent canonical activa-tion pathway.

Role of p38 in Hsp90-Cdc37 Inhibition InducedCardiomyocyte ApoptosisHsp90 is reported to have a potent cytoprotective function inthe heart.40–42 As expected, treating NRVMs with Hsp90inhibitor GA resulted in massive cell death within 6 hours(Figure 7A). Consistent with the increased cell death, theexpression levels of both IL-6 and TNF� were increased inNRVMs treated with GA (Online Figure VII). Similarly,COS7 cells treated with GA showed a significant increase inTUNEL-positive apoptotic cells (Figure 7B). However, therelative low frequency of TUNEL� cells (1.1%) suggest thatapoptosis may not be the only cause to trigger massive celldeath induced by GA. Consistent with its effect on p38autophosphorylation, 17-AAG treatment also led to similarcell death though to a lesser extent (Online Figure III, B).Expression of DN38� was sufficient to block p38 dependentdownstream signaling as demonstrated by the reducedp-Hsp27 and p-MK2 (Figure 7C). DNp38� expression inNRVMs significantly delayed GA induced cell death up to 12hours (Figure 7D). The reduction in cell death by DNp38�correlated with a decrease in the level of TUNEL-positivecells, suggesting that p38 activation contributed to GAinduced apoptosis (Figure 7E). The loss of efficacy ofDNp38� in suppressing apoptosis at higher dose of GAtreatment (9 �mol/L) also suggests that the contribution ofp38 activation in HSP90-mediated cell death regulation ispartial. This is consistent with the earlier data which showsdelayed, but not complete inhibition of cell death in thepresence of DNp38�.

DiscussionIn this report, we have demonstrated that Hsp90-Cdc37chaperone complex is a part of the cardiac p38 signaling

Figure 7. Hsp90 and p38 activ-ity in myocyte survival. A,NRVMs was treated with9 �mol/L GA or DMSO as acontrol. Representative imagesof cell viability were recordedat 0, 6, and 24 hours aftertreatment as indicated. B,TUNEL staining on COS7treated with 9 �mol/L GA for 6hours. Quantification ofTUNEL-positive cells from 3different experiments. *P�0.01,Student’s t test. C, NRVMswere infected with adv-LacZ orAdv-DNp38� (DN) for 48 hours,followed by 9 �mol/L GA treat-ment for specific periods asindicated. Immunoblots forphospho- and total p38 wereperformed. Phospho-MK2 andphospho-Hsp27 were assessedas p38 downstream activities.D, COS7 cells were infected

with Adv-LacZ or Adv-DNp38� as indicated for 48 hours before treatment with 9 �mol/L GA for 0, 12, and 24 hours. Representativeimages of cell viability are shown at 0, 12, and 24 hours after treatment as indicated. E, Quantification of TUNEL-positive apoptoticCOS7 cells among different treatment groups (LacZ vs DNp38�) at 2 doses of GA (3.6 �mol/L vs 9 �mol/L) as indicated. *P�0.05.

Ota et al Noncanonical Regulation of p38 Kinase 1409

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

Page 7: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

complex via direct interaction between p38 and Cdc37.Recombinant Hsp90 and Cdc37 are sufficient to attenuate theintrinsic autophosphorylation activity of the recombinant p38kinase in test tubes. In intact myocytes, Hsp90 inhibition orCdc37 knockdown leads to significant induction of p38autophosphorylation. Cdc37 is both sufficient and necessaryto regulate the basal p38 activity, as well as Hsp90 inhibition–or TAB1�-mediated p38 autophosphorylation, whereas p38activation induced by H2O2 or MKK3 is not hindered.Finally, p38� activity contributes to cell death induced byHsp90 inhibition in cultured cardiomyocytes or COS cells.These data lead us to conclude that Hsp90-Cdc37 chaperonecomplex is a specific regulator of noncanonical pathway forp38� by suppressing basal autophosphorylation.

Among its vast variety of client proteins, Hsp90 is knownto regulate both protein stability and activity of its targetkinases.43,44 Early studies attempting to identify Hsp90 clientkinases by assessing protein degradation in response toHsp90 inhibition have excluded p38 MAPK as an Hsp90client kinase based on the observation that p38 kinase proteinlevel was not affected. Yet, increases in p38 phosphorylationin response to Hsp90 inhibition has been reported,45 andHsp90 has also been shown to promote autophosphorylationof some of the client protein kinases, such as GSK-3�.46 p38homolog in Schizosaccharomyces pombe, Spc1, is reported todirectly bind to yeast Cdc37,47 and Cdc37 has been suggestedto modulate kinase activity independent of Hsp90.48 The datareported here from both in vitro and cellular experimentsclearly demonstrate that mammalian Hsp90 also interactswith p38 through Cdc37. Hsp90 activity is not required forp38 protein stability as no change in the total p38 proteinlevel was observed up to 6 hours of Hsp90 inhibition. Rather,Hsp90-Cdc37 complex functions to suppress p38 autophos-phorylation. This finding not only adds p38 kinase to the listof Hsp90 client proteins but also reveals a novel function ofthe Hsp90-Cdc37 chaperone complex as a negative regulatorof kinase autophosphorylation to the established large reper-toire of Hsp90-Cdc37 functions. This newly identified mech-anism may have general implications in other unidentifiedHsp90-Cdc37 client protein kinases that also possess intrinsicautophosphorylation properties.

Although we have demonstrated that Cdc37 directly bindsto p38� both in vitro and in vivo, the underlying molecularbasis is unclear. It appears that the N-terminal domain ofp38� contains weak Cdc37 binding activity when it isexpressed in its truncated form, but mutations within aputative Cdc37 binding motif (GXGXYG) failed to disruptfull-length p38� interaction with Cdc37. Indeed, GY doublemutant appeared to bind stronger to Cdc37 (Figure 2B),suggesting that protein conformation in the N-terminal do-main, which is known to go through conformational changeon autophosphorylation, may be involved in this interac-tion.8,35 Likewise, although Cdc37 is reported to require V192and I193 residues33 to interact with a client protein kinaseRaf-1, mutations of these sites did not affect its p38�interaction. Clearly, more efforts are needed to identify theresponsible motifs involved in p38�-Cdc37 interaction.

Autophosphorylation is an intrinsic property of p38 kinase,although its physiological role remains uncertain.8,9 TAB1

and ZAP-70 are reported to positively promote p38 autophos-phorylation thus represent a noncanonical pathway for p38activation under different upstream stimuli with differentfunctional outcome.13–16 Previous studies have revealed muchof the regulatory components for canonical p38 pathway,including upstream MKKs, downstream targets and negativemodulating phosphatases.10–12 However, the regulatory com-ponents specific for the noncanonical pathway remain elu-sive. Our report here has provided the first evidence thatCdc37 expression has no impact on p38 activation induced byH2O2 or MKK3, but significantly attenuates TAB1-mediatedautophosphorylation of p38, suggesting that Hsp90-Cdc37chaperone complex is a noncanonical pathway–specific reg-ulator for p38. The molecular basis for this specificity isunclear. We showed that phosphorylated p38 was not detect-able in the Hsp90-Cdc37 immunocomplex and the total p38protein present in the Hsp90-Cdc37 complex was reduced onHsp90 inhibition. This observation suggests that phosphory-lated p38 loses the ability to bind to Cdc37. Thus, we canspeculate that p38 autophosphorylation is inhibited byHsp90-Cdc37 by conferring a change in p38 protein confor-mation. Although this change does not affect its recognitionand phosphorylation by upstream kinases, it hinders p38autophosphorylation induced by noncanonical activator, suchas TAB1.

Hsp90 inhibition induced cell death is an emerging thera-peutic strategy to treat cancer and the underlying mechanismsmainly involve the loss of prosurvival proteins, such as AKT,Raf-1, Cdk4, and p53.49,50 Our data suggest that activation ofp38 via autophosphorylation may also play a significant rolein Hsp90 inhibition induced cell death accompanied byincreased expression of inflammatory cytokines includingIL-6 and TNF�, thus providing yet another possible down-stream mechanism in Hsp90-mediated cell death regulation.Although our data suggest Hsp90 inhibition induced celldeath involves apoptosis, other forms of cell death, includingnecrosis and autophagy, cannot be excluded. The full scopeof both physiological and pathological role of Hsp90-Cdc37complex in p38-mediated stress signaling remains to bedetermined. However, autophosphorylation of p38 in re-sponse to ischemia and its contribution to cardiac injury havebeen reported. Ischemia/reperfusion on MKK3�/� heart dem-onstrated TAB1-mediated autophosphorylation of p38, whichappeared to play a role in necrosis of cardiomyocytes.14

Autophosphorylated p38 was also reported to mediate glu-cose uptake through GLUT4 induced by ischemia51 andTAB1-p38 interaction was implicated in myocyte apoptosisregulation by cGMP-depending kinase I.52 On the other hand,Hsp90 inhibition exacerbated cardiac injury induced bydoxorubicin, a known p38 activator in heart53 and overex-pression of Hsp90 protected ischemia/reperfusion injury inheart.54 Therefore, it can be speculated that Hsp90-Cdc37–mediated p38 autophosphorylation may play a role in ische-mia/reperfusion injury in heart. Our data in Online Figure VIIshow a significant induction of inflammatory cytokines byHsp90 inhibition in NRVMs, suggesting that Hsp90-Cdc37–mediated noncanonical p38 activity may have differentdownstream effect versus TAB1-mediated activation in heartwhich does not appear to induce inflammatory gene induc-

1410 Circulation Research April 30, 2010

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

Page 8: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

tion.15 Our data also suggest the presence of numeroussarcomere-associated proteins in cardiac p38 complex (On-line Table I), yet their functional relevance to Hsp90-Cdc37–mediated p38 activity remains unknown. Further studies withtargeted manipulation of Hsp90-Cdc37-p38 interaction willbe needed to fully establish the functional relevance of thisnew stress signaling pathway.

Other than chaperone function, Hsp90 can also serve as ascaffold protein to coordinate/integrate multiple signal trans-duction activities in different cellular processes.55 We andothers have demonstrated in previous studies that cardiac p38kinase also functions to regulate diverse cellular processesincluding gene transcription,7 proliferation,56 apoptosis,6,57

and cardiac contractility.58,59 Canonical and noncanonicalpathways for p38 activation can have major differences inintracellular localization, downstream targets, and ultimatefunctional outcome.15 In short, identification of Hsp90-Cdc37as an additional component in noncanonical p38 signalingpathway will help to elucidate the molecular network andintricate regulatory mechanisms for this important stresssignal pathway.

AcknowledgmentsWe thank the Cardiovascular Research Laboratory Cell Core (Uni-versity of California, Los Angeles) for providing NRVMs, andProteomic Core for mass spectrometric analysis.

Sources of FundingThis work was supported by NIH grants HL070079, HL062311, andHL080111 (to Y.W. and P.P.) and NIH predoctoral fellowshipF31AG032163 (to A.O.). Y.W. is an Established Investigator ofAmerican Heart Association.

DisclosuresNone.

References1. Bogoyevitch MA. Signalling via stress-activated mitogen-activated

protein kinases in the cardiovascular system. Cardiovasc Res. 2000;45:826–842.

2. New L, Han J. The p38 MAP kinase pathway and its biological function.Trends Cardiovasc Med. 1998;8:220–228.

3. Obata T, Brown GE, Yaffe MB. Map kinase pathways activated by stress:The p38 MAPK pathway. Crit Care Med. 2000;28:N67–77.

4. Widmann C, Gibson S, Jarpe MB, Johnson GL. Mitogen-activatedprotein kinase: conservation of a three-kinase module from yeast tohuman. Physiol Rev. 1999;79:143–180.

5. Raman M, Chen W, Cobb MH. Differential regulation and properties ofMAPKs. Oncogene. 2007;26:3100–3112.

6. Ichijo H, Nishida E, Irie K, ten Dijke P, Saitoh M, Moriguchi T, TakagiM, Matsumoto K, Miyazono K, Gotoh Y. Induction of apoptosis byASK1, a mammalian MAPKKK that activates SAPK/JNK and p38 sig-naling pathways. Science. 1997;275:90–94.

7. Ono K, Han J. The p38 signal transduction pathway: activation andfunction. Cell Signal. 2000;12:1–13.

8. Avitzour M, Diskin R, Raboy B, Askari N, Engelberg D, Livnah O.Intrinsically active variants of all human p38 isoforms. FEBS J. 2007;274:963–975.

9. Levin-Salomon V, Maayan I, Avrahami-Moyal L, Marbach I, Livnah O,Engelberg D. When expressed in yeast, mammalian mitogen-activatedprotein kinases lose proper regulation and become spontaneously phos-phorylated. Biochem J. 2009;417:331–340.

10. Nguyen AN, Shiozaki K. Heat-shock-induced activation of stress mapkinase is regulated by threonine- and tyrosine-specific phosphatases.Genes Dev. 1999;13:1653–1663.

11. Takekawa M, Maeda T, Saito H. Protein phosphatase 2calpha inhibits thehuman stress-responsive p38 and JNK MAPK pathways. EMBO J. 1998;17:4744–4752.

12. Meskiene I, Baudouin E, Schweighofer A, Liwosz A, Jonak C, RodriguezPL, Jelinek H, Hirt H. Stress-induced protein phosphatase 2c is a negativeregulator of a mitogen-activated protein kinase. J Biol Chem. 2003;278:18945–18952.

13. Ge B, Gram H, Di Padova F, Huang B, New L, Ulevitch RJ, Luo Y, HanJ. MAPKK-independent activation of p38alpha mediated by TAB1-dependent autophosphorylation of p38alpha. Science. 2002;295:1291–1294.

14. Tanno M, Bassi R, Gorog DA, Saurin AT, Jiang J, Heads RJ, Martin JL,Davis RJ, Flavell RA, Marber MS. Diverse mechanisms of myocardialp38 mitogen-activated protein kinase activation: evidence for MKK-independent activation by a TAB1-associated mechanism contributing toinjury during myocardial ischemia. Circ Res. 2003;93:254–261.

15. Lu G, Kang YJ, Han J, Herschman HR, Stefani E, Wang Y. TAB-1modulates intracellular localization of p38 MAP kinase and downstreamsignaling. J Biol Chem. 2006;281:6087–6095.

16. Salvador JM, Mittelstadt PR, Guszczynski T, Copeland TD, YamaguchiH, Appella E, Fornace AJ Jr, Ashwell JD. Alternative p38 activationpathway mediated by t cell receptor-proximal tyrosine kinases. NatImmunol. 2005;6:390–395.

17. Riggs DL, Cox MB, Cheung-Flynn J, Prapapanich V, Carrigan PE, SmithDF. Functional specificity of co-chaperone interactions with hsp90 clientproteins. Crit Rev Biochem Mol Biol. 2004;39:279–295.

18. Sreedhar AS, Soti C, Csermely P. Inhibition of hsp90: a new strategy forinhibiting protein kinases. Biochim Biophys Acta. 2004;1697:233–242.

19. Prince T, Sun L, Matts RL. Cdk2: A genuine protein kinase client ofhsp90 and cdc37. Biochemistry. 2005;44:15287–15295.

20. MacLean M, Picard D. Cdc37 goes beyond hsp90 and kinases. Cell StressChaperones. 2003;8:114–119.

21. Sato S, Fujita N, Tsuruo T. Modulation of akt kinase activity by bindingto hsp90. Proc Natl Acad Sci U S A. 2000;97:10832–10837.

22. Pearl LH. Hsp90 and cdc37 – a chaperone cancer conspiracy. Curr OpinGenet Dev. 2005;15:55–61.

23. Terasawa K, Yoshimatsu K, Iemura S, Natsume T, Tanaka K, Minami Y.Cdc37 interacts with the glycine-rich loop of hsp90 client kinases. MolCell Biol. 2006;26:3378–3389.

24. Citri A, Harari D, Shohat G, Ramakrishnan P, Gan J, Lavi S, EisensteinM, Kimchi A, Wallach D, Pietrokovski S, Yarden Y. Hsp90 recognizes acommon surface on client kinases. J Biol Chem. 2006;281:14361–14369.

25. Zhang H, Wu W, Du Y, Santos SJ, Conrad SE, Watson JT, Gramma-tikakis N, Gallo KA. Hsp90/p50cdc37 is required for mixed-lineagekinase (MLK) 3 signaling. J Biol Chem. 2004;279:19457–19463.

26. Miyata Y, Ikawa Y, Shibuya M, Nishida E. Specific association of a setof molecular chaperones including hsp90 and cdc37 with MOK, amember of the mitogen-activated protein kinase superfamily. J BiolChem. 2001;276:21841–21848.

27. Mackay K, Mochly-Rosen D. An inhibitor of p38 mitogen-activatedprotein kinase protects neonatal cardiac myocytes from ischemia. J BiolChem. 1999;274:6272–6279.

28. Wang Y, Huang S, Sah VP, Ross J Jr, Brown JH, Han J, Chien KR.Cardiac muscle cell hypertrophy and apoptosis induced by distinctmembers of the p38 mitogen-activated protein kinase family. J BiolChem. 1998;273:2161–2168.

29. Ren J, Zhang S, Kovacs A, Wang Y, Muslin AJ. Role of p38alpha MAPKin cardiac apoptosis and remodeling after myocardial infarction. J MolCell Cardiol. 2005;38:617–623.

30. Rajapandi T, Greene LE, Eisenberg E. The molecular chaperones hsp90and hsc70 are both necessary and sufficient to activate hormone bindingby glucocorticoid receptor. J Biol Chem. 2000;275:22597–22604.

31. Karnitz LM, Felts SJ Cdc37 regulation of the kinome: When to hold ’emand when to fold ’em. Sci STKE. 2007;2007:pe22.

32. Hemmer W, McGlone M, Tsigelny I, Taylor SS. Role of the glycine triadin the ATP-binding site of camp-dependent protein kinase. J Biol Chem.1997;272:16946–16954.

33. Terasawa K, Minami Y. A client-binding site of cdc37. FEBS J. 2005;272:4684–4690.

34. Jiang Y, Li Z, Schwarz EM, Lin A, Guan K, Ulevitch RJ, Han J.Structure-function studies of p38 mitogen-activated protein kinase. Loop12 influences substrate specificity and autophosphorylation, but notupstream kinase selection. J Biol Chem. 1997;272:11096–11102.

Ota et al Noncanonical Regulation of p38 Kinase 1411

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

Page 9: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

35. Diskin R, Lebendiker M, Engelberg D, Livnah O. Structures of p38alphaactive mutants reveal conformational changes in l16 loop that induceautophosphorylation and activation. J Mol Biol. 2007;365:66–76.

36. Csermely P, Kajtar J, Hollosi M, Jalsovszky G, Holly S, Kahn CR,Gergely P Jr, Soti C, Mihaly K, Somogyi J. ATP induces a conforma-tional change of the 90-kda heat shock protein (hsp90). J Biol Chem.1993;268:1901–1907.

37. Prodromou C, Panaretou B, Chohan S, Siligardi G, O’Brien R, LadburyJE, Roe SM, Piper PW, Pearl LH. The ATPase cycle of hsp90 drives amolecular ‘clamp’ via transient dimerization of the n-terminal domains.EMBO J. 2000;19:4383–4392.

38. Picard D. Heat-shock protein 90, a chaperone for folding and regulation.Cell Mol Life Sci. 2002;59:1640–1648.

39. Nemoto T, Sato N. Oligomeric forms of the 90-kda heat shock protein.Biochem J. 1998;330(Pt 2):989–995.

40. Gray CC, Amrani M, Yacoub MH. Heat stress proteins and myocardialprotection: experimental model or potential clinical tool? Int J BiochemCell Biol. 1999;31:559–573.

41. Nishizawa J, Nakai A, Matsuda K, Komeda M, Ban T, Nagata K.Reactive oxygen species play an important role in the activation of heatshock factor 1 in ischemic-reperfused heart. Circulation. 1999;99:934–941.

42. Proctor CJ, Soti C, Boys RJ, Gillespie CS, Shanley DP, Wilkinson DJ,Kirkwood TB. Modelling the actions of chaperones and their role inageing. Mech Ageing Dev. 2005;126:119–131.

43. Wandinger SK, Richter K, Buchner J. The hsp90 chaperone machinery.J Biol Chem. 2008;283:18473–18477.

44. Dezwaan DC, Freeman BC. Hsp90: the rosetta stone for cellular proteindynamics? Cell Cycle. 2008;7:1006–1012.

45. Dey A, Cederbaum AI. Geldanamycin, an inhibitor of hsp90 increasescytochrome p450 2e1 mediated toxicity in hepg2 cells through sustainedactivation of the p38mapk pathway. Arch Biochem Biophys. 2007;461:275–286.

46. Lochhead PA, Kinstrie R, Sibbet G, Rawjee T, Morrice N, Cleghon V. Achaperone-dependent gsk3beta transitional intermediate mediatesactivation-loop autophosphorylation. Mol Cell. 2006;24:627–633.

47. Tatebe H, Shiozaki K. Identification of cdc37 as a novel regulator of thestress-responsive mitogen-activated protein kinase. Mol Cell Biol. 2003;23:5132–5142.

48. Lee P, Rao J, Fliss A, Yang E, Garrett S, Caplan AJ. The cdc37 proteinkinase-binding domain is sufficient for protein kinase activity and cellviability. J Cell Biol. 2002;159:1051–1059.

49. Garcia-Echeverria C, Sellers WR. Drug discovery approaches targetingthe pi3k/akt pathway in cancer. Oncogene. 2008;27:5511–5526.

50. Messaoudi S, Peyrat JF, Brion JD, Alami M. Recent advances in hsp90inhibitors as antitumor agents. Anticancer Agents Med Chem. 2008;8:761–782.

51. Li J, Miller EJ, Ninomiya-Tsuji J, Russell RR III, Young LH. AMP-ac-tivated protein kinase activates p38 mitogen-activated protein kinase byincreasing recruitment of p38 MAPK to TAB1 in the ischemic heart. CircRes. 2005;97:872–879.

52. Fiedler B, Feil R, Hofmann F, Willenbockel C, Drexler H, Smolenski A,Lohmann SM, Wollert KC. cGMP-dependent protein kinase type Iinhibits TAB1–p38 mitogen-activated protein kinase apoptosis signalingin cardiac myocytes. J Biol Chem. 2006;281:32831–32840.

53. Gabrielson K, Bedja D, Pin S, Tsao A, Gama L, Yuan B, Muratore N.Heat shock protein 90 and erbb2 in the cardiac response to doxorubicininjury. Cancer Res. 2007;67:1436–1441.

54. Kupatt C, Dessy C, Hinkel R, Raake P, Daneau G, Bouzin C, BoekstegersP, Feron O. Heat shock protein 90 transfection reduces ischemia-reperfusion-induced myocardial dysfunction via reciprocal endothelial nosynthase serine 1177 phosphorylation and threonine 495 dephosphoryla-tion. Arterioscler Thromb Vasc Biol. 2004;24:1435–1441.

55. Richter K, Hendershot LM, Freeman BC. The cellular world according tohsp90. Nat Struct Mol Biol. 2007;14:90–94.

56. Engel FB. Cardiomyocyte proliferation: a platform for mammaliancardiac repair. Cell Cycle. 2005;4:1360–1363.

57. Xia Z, Dickens M, Raingeaud J, Davis RJ, Greenberg ME. Opposingeffects of ERK and JNK-p38 map kinases on apoptosis. Science. 1995;270:1326–1331.

58. Liao P, Wang SQ, Wang S, Zheng M, Zhang SJ, Cheng H, Wang Y, XiaoRP. p38 mitogen-activated protein kinase mediates a negative inotropiceffect in cardiac myocytes. Circ Res. 2002;90:190–196.

59. Vahebi S, Ota A, Li M, Warren CM, de Tombe PP, Wang Y, Solaro RJ.p38-MAPK induced dephosphorylation of alpha-tropomyosin is asso-ciated with depression of myocardial sarcomeric tension and ATPaseactivity. Circ Res. 2007;100:408–415.

Novelty and Significance

What Is Known?

● p38 mitogen-activated protein kinase (MAPK) is a stress-inducedsignaling pathway that is involved in regulation of various cellularactivities including proliferation, apoptosis, cell growth, genetranscription, differentiation, and cell motility during developmentand pathological processes, including heart failure.

● p38 MAPK is activated by 2 mechanisms: a canonical pathway mediatedby upstream kinases and a noncanonical pathway mediated byinteracting proteins, such as TAK1 binding protein (TAB)1.

● Canonical p38 activity is known to be highly regulated by the balanceof upstream kinases and phosphatases; however, the functionalsignificance and the regulatory mechanisms of noncanonical p38activity remain poorly understood.

What New Information Does This Article Contribute?

● The heat shock protein (Hsp)90-Cdc37 chaperone complex interacts withp38 and negatively regulates p38 autophosphorylation, which isessential to maintain low p38 activity in the absence of stress stimuli.

● Hsp90-Cdc37 specifically regulates TAB1-mediated noncanonicalactivation of p38, but has no effect on MAPK kinase (MKK)-dependent canonical activation of p38.

● Cell death induced by Hsp90-Cdc37 inhibition is contributed by p38activation in cardiomyocytes.

As a stress-response pathway, p38 MAPK is under tightregulation in basal states; on activation, it can lead to manydownstream effects via canonical versus noncanonical path-ways. Previous work has focused mostly on the regulatorymechanism of the canonical p38 pathway; the underlyingmechanisms for the regulation of noncanonical p38 pathway areunknown. In this report, we identified that Hsp90-Cdc37 is partof the p38 complex and functions as a negative regulator of p38autophosphorylation. Although MKK-dependent p38 activationthrough the canonical pathway is not affected, the noncanonicalactivation of p38 through autophosphorylation is greatly en-hanced on inhibition or loss of Hsp90-Cdc37. We also show thatp38 activation induced by Hsp90 inhibition plays a role in apoptosis.Therefore, we have established that Hsp90/Cdc37-mediated sup-pression of p38 autophosphorylation is a critical mechanism fornoncanonical p38 kinase regulation and that p38 is a new clientprotein and regulatory target for Hsp90/Cdc37 in stress signaling.This is the first report to demonstrate the molecular and functionallink between these 2 important cell signaling molecules, therebyrevealing a new connection between 2 major nodes in stresssignaling network. The mechanism discovered herein may play asignificant role in stress-induced cardiac injury and other patholog-ical conditions.

1412 Circulation Research April 30, 2010

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

Page 10: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

Asuka Ota, Jun Zhang, Peipei Ping, Jiahuai Han and Yibin WangComplex in Cardiomyocyte

Activation by Hsp90-Cdc37 ChaperoneαSpecific Regulation of Noncanonical p38

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 2010 American Heart Association, Inc. All rights reserved.is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation Research

doi: 10.1161/CIRCRESAHA.109.2137692010;106:1404-1412; originally published online March 18, 2010;Circ Res. 

http://circres.ahajournals.org/content/106/8/1404World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://circres.ahajournals.org/content/suppl/2010/03/18/CIRCRESAHA.109.213769.DC1Data Supplement (unedited) at:

  http://circres.ahajournals.org//subscriptions/

is online at: Circulation Research Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further informationEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculation Researchin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:

by guest on June 18, 2018http://circres.ahajournals.org/

Dow

nloaded from

Page 11: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

CIRCRESAHA/2009/213769/R2-1

Supplementary Information Materials and Methods Production and isolation of recombinant proteins. Mouse ventricular myocytes were homogenized with Trizol (Invitrogen) for RNA isolation following company’s protocol, and cDNA was prepared using oligo dT primer and SSII RT (Invitrogen) following company’s protocol. The cDNAs encoding Cdc37, Hsp90α, and Hsp90β were subcloned into pGEX 5x-3 vector (GE healthcare) for GST tag, and pET28a (Novagen) for His- or FLAG-tagged proteins. BL21 strain E. coli was transformed with pGEX 5X-3 or pET28a vectors containing gene of interest, and expanded in 3 mL culture for overnight in 37oC shaker in 2xYTA growth media with appropriate antibiotics. 30 mL of 2xYTA media with appropriate antibiotics were inoculated with overnight culture at 1:100, and incubated in 37oC shaker until absorbance at OD600 is between 0.5 – 0.6. Culture was then induced with 0.1 mmol/L IPTG for 2 hours at 37oC. Bacterial pellet was collected through centrifugation at 77,000g for 10 minutes 4oC. Pellet was resupend in ice-cold 1xPBS, and rinsed twice. Once the rinse is complete, pellet was resuspended in 1.5 mL 1xPBS with Complete Mini protease inhibitor cocktail tablet (Roche Applied Science), and 10% glycerol, and sonicated. Lysate was centrifuged at 16,000g for 30 minutes at 4oC, and supernatant was kept for further purification. GST-tagged proteins were purified using Glutathione Sepharose 4B (GE Healthcare) using company’s protocol. His-tagged proteins were purified using TALON Metal Affinity Resins (Clontech Laboratories, Inc.) following company’s protocol. The eluates were analyzed by 1D Western immunoblot and coomassie staining to verify the expression and purity of the isolated recombinant proteins. Immunoprecipitation. 0.1 - 0.3 μg of recombinant proteins were resuspended in 200 μL binding buffer (150 mM NaCl, 20 mM Tris-HCl, pH 7.3, 1% Triton X-100, 1 mM EDTA, Complete Mini protease inhibitor cocktail tablet (Roche Applied Science), 1 mM PMSF, and 10% glycerol) and 2 μL of appropriate antibody overnight at 4oC. nProtein A Sepharose 4 Fast Flow (GE Healthcare) was rinsed three times in 10x volume of cold 1xPBS, and resuspended in 1x volume of 1x PBS to generate 50% slurry nProtein A Sepharose 4 Fast Flow. 10 μL of 50% slurry was added to each tube and incubated for 2 hours at 4oC. Beads were washed in binding buffer for 5 minutes, and centrifuged at 500g for 5 minutes. Wash was repeated 6 times. Finally, the beads were boiled in loading buffer containing 0.2% 2-Mercaptoethanol at 95oC for 10 minutes. Anti-FLAG M2 affinity gel (Sigma) was used for FLAG co-IP, by incubating with protein mixtures at 4oC for 2 hours. Beads were washed with the binding buffer 6 times, and boiled in 1x NuPAGE LDS sample buffer (Invitrogen) with 0.2% β-mercaptoethanol. Cell culture. NRVM were prepared by enzymatic digestion of 48 hours old neonatal rat ventricular tissues with collagenase (Worthington) and pancreatin (Sgima) in 1xADS buffer at 37oC as described previously. NRVM were culture in DMEM supplemented with 100U/mL Pen/Strep (Invitrogen) and 5% ITS (w/v) (BD Biosciences). HeLa cells were cultured in DMEM supplemented with 10% FBS, and 100U/mL Pen/Strep at 37oC with 5% CO2. HeLa cells were cultured to 90% confluency in 6-well dish (Corning) and transfected with Cdc37 expressing vector using Lipofectamine 2000 (Invitrogen) and Opti-MEM (Gibco) with half the amount of suggested amount of lipofectamine to minimize the toxicity. Cells were incubated for 24-48 hours to allow expression of proteins, and lysed in 1% Triton X-100 lysis buffer (150 mmol/L NaCl, 20 mmol/L Tris-HCl, pH 7.3, 1% Triton X-100, 1 mmol/L EDTA, 1 mmol/L EGTA, 0.1 mmol/L Na3VO4, 2.5 mmol/L Na4P2O7, Complete Mini protease inhibitor cocktail tablet (Roche Applied Science), 1 mmol/L PMSF, and 10% glycerol) for further analyses. HeLa cells were infected with Adv-MEK3bE or Adv-TAB1b in serum-free DMEM supplemented with 100U/mL Pen/Strep for 2 hours. The media was replaced with fresh DMEM supplemented with 10% FBS and 100U/mL Pen/Strep, and cultured for additional 24 hours at 37oC. GA (Sigma-Aldrich) was resuspended in DMSO at a concentration of 1mg/mL for a stock solution. SB203580 was also resuspended in DMSO at 1mg/mL.

Page 12: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

CIRCRESAHA/2009/213769/R2-2

TUNEL staining. Coverslips were rinsed in acetone, and kept in 70% ethanol prior to use. Coverslips were flamed to evaporate ethanol, and placed into 24-well plate (Corning). 80 μL of Laminin (Gibco) was placed on each coverslip, and incubated at 37oC for 2 hours. Coverslips were rinsed in 1xPBS twice, and cells were seeded at 50% confluency in complete culture media, and incubated overnight. Cells were then infected with Adv for 24 hours, and treated with GA at a given concentration for the duration of time as specified. Cells were rinsed in 1xPBS twice, and fixed in 2% paraformaldehyde for 30 mimutes at room temperature. Cells were rinsed in 1xPBS twice, 5 minutes each at room temperature, and permeabilized in 1% Triton X-100, 1% sodium citrate solution for 30 minutes. TUNEL staining was carried out using ApopTag Peroxidase in situ Apoptosis Detection Kit (Chemicon International) following company’s protocol. Briefly, cells were incubated in equilibration buffer for 10 minutes, followed by TdT enzyme for 1 hour at 37oC. Reactions were terminated by incubating coverslips in stop wash buffer for 10 minutes at room temperature, and rinsed in 1xPBS three times. Apoptotic cells were then detected by incubating in anti-Digox-Fluorescein in 5% BSA/5% Goat serum. Nuclei were stained with DAPI for 5 minutes at room temperature before mounting coverslips on glass slides. Apoptotic occurrence was determined by counting TUNEL positive cells which co-localize with DAPI staining, and dividing by the total number of nucleus. Statistical significance was determined by student’s t-test. Kinase assay. 0.3 μg of recombinant proteins were incubated in 30 μL total of incomplete kinase assay buffer (25 mmol/L Tris-HCl, pH 7.5, 5 mmol/L β-glycerophosphate, 2 mmol/L DTT, 0.1 mmol/L Na3VO4, 10 mmol/L MgCl2, 100 μmol/L ATP ) at 4 oC as described earlier. 32P-γ-ATP, GA and SB203580 were added prior to the assay, and incubated on ice for 15 minutes. Samples were then incubated at 30 oC for 30 minutes, and run on the 4-12% SDS PAGE. Gel was fixed in 50% methanol and 10% acetic acid for 10 minutes at room temperature, and rinsed in double distilled water three times for 5 minutes each. Gel was dried using Gel Dryer (Bio-Rad) and exposed on phospho-screen for 24 hours. The cassette was scanned using Typhoon imager (GE Healthcare) using ImageQuant software (GE Healthcare). Western blot. The concentrations of protein lysates were measured using BCA protein assay reagent following company’s protocol (Thermo Scientific). 40 μg of proteins were added to 1x NuPAGE LDS sample buffer (Invitrogen) with 0.2% 2-mercaptoethanol (Sigma), and boiled for 5 minutes at 95oC. Proteins were separated on 4-12% or 10% Bis/Tris SDS-PAGE (Invitrogen) in 1x MOPS buffer at 150V. Proteins were transferred onto nitrocellulose membrane (GE Healthcare) at 30V for 120 minutes. Membrane was blocked in 5% non-fat milk for 30 minutes at room temperature and washed twice in 1xTBST for 5 minutes. Primary antibodies were diluted either in 3% BSA (Sigma) or 5% non-fat milk overnight at 4 oC. Membrane was washed 5 times in 1xTBST for 5 minutes, and incubated in secondary antibody diluted in 5% non-fat milk for 1 hour at room temperature. Membrane was exposed on film using SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientific) following company’s protocol. Antibodies used are p-p38 MAPK, p38 MAPK, p38α MAPK, p-JNK, JNK, p-ERK, ERK, p-MK2, MK2, p-MEK3/6, MEK3/6, p-MEK4, MEK4, p-Hsp27, p-Akt T308, p-Akt S473, Akt (Cell Signaling), Hsp90α/β Cdc37, Hsc70 (Abcam), Cdc37, MEK3/6, p-p38 Y182, β-Actin, Hsp27, and GST (Santa Cruz). shRNA construct. Oligos encoding shRNA sense and anti-sense sequences were ordered from Invitrogen and resupended at a final concentration of 25 μmol/L. The sequences of Cdc37 shRNA tested are: Cdc37 shRNA 1 (sense) – GCCCACCAGACAATCGTCATGCAAT, Cdc37 shRNA 2 (sense) – ACAGATCAAGC ACTTTGGCATGCTT, Cdc37 shRNA 3 (sense) – GAGGAACTCCAGAAGTGCTTCGATG, and Cdc37 shRNA 4 (sense) – CAGAAACACAAGACCTTCGTGGAAA. Cdc37 shRNA 2 was used for further knockdown experiments. 5 μL each of sense and anti-sense strands were annealed in 50 mmol/L NaCl, 1 mmol/L EDTA, and 10 mmol/L Tris-HCl at pH 7.5 by gradually cooling temperature from 95oC to 25 oC over a period of 59 minutes using GeneAmp PCR Systems 9700 (Applied Bioscience). The oligos were annealed into pSuper through BamHI/HindIII restriction sites. Adv was produced through homologous recombination between pSuper and pAdEasy using BJ strain E. coli. pAdEasy was digested with PacI (New England

Page 13: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

CIRCRESAHA/2009/213769/R2-3

Biolabs) followed by phenol/chloroform purification. Purified DNA was transfected into 293 cells using Lipofectamine 2000 to produce Adv. Knockdown efficiency for each construct was determined by Western blot analysis prior to Adv amplification. qRT-PCR. RNA was isolated from cells using 1 mL of Trizol (Invitrogen) following company’s protocol. cDNA was synthesized from RNA using Oligo-dT primers and SuperScriptII RT (Invitrogen) following company’s protocol. 1 μL of cDNA in duplicate was analyzed for qRT-PCR using SYBR Green Supermix (Bio-Rad) in a reaction volume of 50 μL with primers specific for IL-6 or TNFα as listed below.

IL-6: GCCAGAGTCATTCAGAGCAA, GGTTTGCCGAGTAGACCTCA TNFα: CTCTTCAAGGGACAAGGCTG, TGGAAGACTCCTCCCAGGTA

The reaction was carried out for 40 cycles with annealing temperature at 60oC using Thermocycler (Bio-Rad). p-value was determined by student’s t-test (n=3).

Page 14: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

CIRCRESAHA/2009/213769/R2-4

Supplementary Figures and Tables Online Table I: Proteins Identified in p38α complex by LC-MS/MS

Protein Peptides Matched

MOWSEScore

PRDX2 3 66

PRDX5 8 111

PRDX6 3 75

GPX3 3 109

GAPDH 21 358

MYOM2 76 1,350

ORACLE 1 9 169

FHL1 3 67

MLP 9 199

β-Tubulin 12 522

α-Tubulin 8 381

*HSP90α 8 216

*HSP90β 20 401

*CDC37 7 160

PP2Cα 6 228

PP2Cβ 19 391

MNK1 7 363

MKK3 6 135

MK2 12 437

MSK1 3 52

Proteins known to interact with p38α are indicated in blue.

Page 15: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

CIRCRESAHA/2009/213769/R2-5

Online Figure I. Sucrose gradient distribution of p38, Hsp90 and Cdc37. A) Total lysates from wildtype mouse cardiac tissue and B) DNp38α protein complex isolated from mouse heart by anti-FLAG immunoprecipitation were separated to a sucrose gradient ranging from 10 – 50%. Immunoblots for every other fraction are shown as indicated.

A

B

Page 16: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

CIRCRESAHA/2009/213769/R2-6

Online Figure II. Kinase assay of p38α complex in vitro. Flag-p38 immunocomplex isolated from DNp38α expressing mouse heart and NTG mouse heart (as a control) were incubated with recombinant GST-p38α or GST and subjected to kinase assay with 5 μCi 32P-γ-ATP for indicated amount of time to determine the presence of p38 substrates in the immunocomplex. The total reaction mixture was separated on a SDS-PAGE and analyzed by autoradiography. Arrowheads indicate the molecular weight of Hsp90 (90kD) and Cdc37 (50kD). Asterisks indicate possible p38 substrates.

Page 17: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

CIRCRESAHA/2009/213769/R2-7

Online Figure III. Effect of 17-AAG on cardiomyocytes. A) NRVM was pre-treated with or without 1 μmol/L SB203580 for 25 minutes, and cultured in the presence of 3.6 μmol/L 17-AAG for indicated amount of time. The total lysates were analyzed by immunoblot as indicated. B) NRVM was treated with 9 μmol/L of 17-AAG or DMSO for 1, 20, or 48 hours and observed under light microscope at 200x.

A B

Page 18: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

CIRCRESAHA/2009/213769/R2-8

Online Figure IV. Autophosphorylation of p38 in response to GA. A) NRVM were treated with 1.8 μmol/L GA for 5 min or 15 min as indicated. Total phospho-p38 and phospho-Tyr-p38a were analyzed by IP using anti-p38 antibody followed by immunoblots using antibodies as indicated. B) HeLa cells were transfected with either vector alone (left panel) or Myc-p38α T180A expressing vector, and treated with 1μmol/L SB203580 or 0.01% DMSO (w/v) for 20 minutes prior to 1.8 μmol/L GA treatment for 5 minutes. Total cell lysates were analyzed by immunoblot using antibodies as indicated.

A

B

Page 19: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

CIRCRESAHA/2009/213769/R2-9

Online Figure V. Cdc37 dependent interaction of p38 and Hsp90. HeLa cells were infected with increasing dosage of shCdc37 or scramble shRNA ranging from 0 to 500 moi for 24 hours. Total cell lysates were analyzed by immunoprecipitation with anti-Hsp90α/β followed by immunoblots for Cdc37 and p38 as indicated.

Page 20: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

CIRCRESAHA/2009/213769/R2-10

Online Figure VI. Impact of Cdc37 overexpression on p38 activation in response to H2O2 vs GA. HeLa cells were transfected with vector alone or vector expressing Myc-Cdc37 for 48 hours. Cells were treated with 0.01% (w/v) DMSO for 30 minutes, or 100 μmol/L H2O2 for 30 minutes, or 3.6 μmol/L GA for 15 minutes. Immunoblots of the total lysates were performed using antibodies as indicated.

Page 21: Specific Regulation of Noncanonical p38 Activation by …circres.ahajournals.org/content/circresaha/106/8/1404.full.pdf · Specific Regulation of Noncanonical p38 Activation by Hsp90-Cdc37

CIRCRESAHA/2009/213769/R2-11

Online Figure VII. Increased expression of inflammatory cytokines in response to GA. NRVM were treated with 0.01% DMSO or 1.8 μmol/L of GA for either 1 hour or 6 hours. cDNA was prepared from triplicate samples, and qRT-PCR was determined to determine the expression level of IL-6 and TNF-α at mRNA level. p-value was determined by student’s t-test compared to the DMSO treated samples. * p < 0.05.