Article

Meetali Singh,

0022-2836/$ - see front m

A Novel C-Terminal Homologue of Aha1Co-Chaperone Binds to Heat Shock Protein90 and Stimulates Its ATPase Activity inEntamoeba histolytica

Varun Shah and Utpal Tat

u

Department of Biochemistry, Indian Institute of Science, Bangalore 560 012, India

Correspondence to Utpal Tatu: tatu@biochem.iisc.ernet.inhttp://dx.doi.org/10.1016/j.jmb.2014.01.008Edited by J. Buchner

Abstract

Cytosolic heat shock protein 90 (Hsp90) has been shown to be essential for many infectious pathogens and isconsidered a potential target for drug development. In this study, we have carried out biochemicalcharacterization of Hsp90 from a poorly studied protozoan parasite of clinical importance, Entamoebahistolytica. We have shown that Entamoeba Hsp90 can bind to both ATP and its pharmacological inhibitor,17-AAG (17-allylamino-17-demethoxygeldanamycin), with Kd values of 365.2 and 10.77 μM, respectively,and it has a weak ATPase activity with a catalytic efficiency of 4.12 × 10−4 min−1 μM−1. Using inhibitor17-AAG, we have shown dependence of Entamoeba on Hsp90 for its growth and survival. Hsp90 function isregulated by various co-chaperones. Previous studies suggest a lack of several important co-chaperones inE. histolytica. In this study, we describe the presence of a novel homologue of co-chaperone Aha1 (activator ofHsp90 ATPase), EhAha1c, lacking a canonical Aha1 N-terminal domain. We also show that EhAha1c iscapable of binding and stimulating ATPase activity of EhHsp90. In addition to highlighting the potential ofHsp90 inhibitors as drugs against amoebiasis, our study highlights the importance of E. histolytica inunderstanding the evolution of Hsp90 and its co-chaperone repertoire.

© 2014 Elsevier Ltd. All rights reserved.

Introduction

Heat shock protein 90 (Hsp90) is an essentialmolecular chaperone in all eukaryotic organisms. Itsfunction as an activator and regulator of importantsignaling and cellular homeostatic processes, suchas heat shock response, steroid signaling, kinasematuration, protein trafficking, immunity, and manyothers, has been very well elucidated over the lastfew decades by many research groups [1,2]. Hsp90plays a critical role in manifestation of infection byvirulent human and animal pathogens. Hsp90inhibition by inhibitors like geldanamycin (GA)inhibits stage transition in the malarial parasitePlasmodium falciparum, and previous studies haveshown complete clearance of the parasite fromrodent malaria model upon treatment with GAderivative 17-allylamino-17-demethoxygeldanamy-cin (17-AAG) [3–10]. Similar studies have shownimportant roles of Hsp90 in the survival of otherprotozoan parasites like Trypanosoma evansi, Giar-dia lamblia, Toxoplasma gondii, Leishmania dono-vani, and Eimeria tenella [3,8,11–14].

atter © 2014 Elsevier Ltd. All rights reserve

One important question in the Hsp90 field is howthe activity and specificity of Hsp90 were regulatedconsidering its vast clientele. Efforts from variousresearch groups have led to the identification ofmany co-chaperones that associate with Hsp90.These proteins can regulate Hsp90 in multiple wayseither by regulating the ATPase activity, by priming itto interact with a certain client group, by assisting information of a multichaperone complex, or by aidingin client maturation [1,15]. Many distinct functionalco-chaperones are known to be associated withHsp90. These include HOP, p23, Aha1, Hch1,Cdc37, FKBP, cyclophilin Cyp40, Sgt1, Pih1, Tah1,Cns1, and protein phosphatase PP5 [15]. HOP(Hsp70–Hsp90 organizing protein), as the namesuggests, enables the transfer of metastable clientprotein associated with the Hsp70–Hsp90 complexto Hsp90. Aha1 (activator of Hsp90 ATPase) and itsN-terminal homologue, Hch1, are the activators ofHsp90 ATPase activity while p23 inhibits the same[16,17]. Aha1 protein is characterized by thepresence of two domains joined by a small linker.Yeast Aha1 N-terminal domain is responsible for

d. J. Mol. Biol. (2014) 426, 1786–1798

1787EhAha1c — A Novel Co-Chaperone of EhHsp90

activating Hsp90 ATPase activity and shares ho-mology with Hch1. The C-terminal domain of yeastAha1 has no ATPase-stimulating activity [16].Cdc37 also represses Hsp90 ATPase but primesHsp90 to interact with specific kinase substrates.FKBP and Cyp40 are prolyl isomerases and are partof the Hsp90 multichaperone complex. Sgt1 partic-ipates in kinetochore assembly and PP5 helpsregulates posttranslational modifications of Hsp90and its co-chaperone Cdc37 [18]. The role of Pih1and Tah1 is implicated in chromatin remodelingcomplexes and small nucleolar RNP maturation[19]. The role of Cns1 is not yet clearly understood;however, it is known to be essential for cell survivaland its temperature sensitive mutant shows re-duced Hsp90 function and its overexpressionsuppresses Cyp40 mutation [20,21]. Most of theseco-chaperones are conserved across species butcertain protozoan parasites show a completeabsence of a few of these co-chaperones. Cdc37,along with Aha1, p23, Cyp40, and Pih1, features inthis list [15].Almost nothing is known about the role of Hsp90 in

Entamoeba histolytica. Entamoeba is the causativeagent of amoebiasis with 50 million cases ofinvasive disease reported per year along with100,000 deaths per year. The treatment regime forEntamoeba has often shown dose-limiting sideeffects or emergence of drug resistance.In this study, we have carried out biochemical and

functional characterization of Entamoeba Hsp90(EhHsp90) and have shown sensitivity of Entamoe-ba growth to Hsp90 inhibition, suggesting thepotential of Hsp90 inhibitors as anti-amoebic drugs.Previous bioinformatics analysis suggests that Ent-amoeba exhibits a depleted co-chaperone repertoirelacking four important co-chaperones, namely, Cdc37,p23, Aha1, and Cyp40 [15]. We describe a novelco-chaperone, namely, EhAha1c in Entamoeba,which shows a weak homology to Aha1 C-terminaldomain. Unlike previous reports that show completeabsence of stimulation with the C-terminal domainconstruct of yeast Aha1 [16], we show for the first timein this study that a naturally occurring C-terminalhomologue of Aha1 is capable of binding andstimulating the ATPase activity of EhHsp90. Thesestudies have important implications on our under-standing of the evolution of Hsp90 and its co-chaperone repertoire.

Results

EhHsp90 is a functional ATPase

Hsp90s usually depend on their ATPase activityfor their role as chaperones. Hsp90 N-terminus hasa Bergerat-type ATP-binding fold that is involved

in ATP binding, and catalysis is through a criticalarginine contributed by the middle domain [1].Despite general conservation at the level of itsprimary structure, Hsp90s from different speciesdiffer in their biochemical activities. Therefore, todetermine the biochemical activity of EhHsp90, itwas cloned from E. histolytica genomic DNA inpRSET-A vector as a 6×-His tagged fusion proteinand confirmed by insert release by double digestionby BamHI and XhoI (Fig. 1a). This fusion protein wasexpressed in Escherichia coli BL21 pLysS and waspurified to homogeneity using Ni-NTA chromatogra-phy (Fig. 1b and c). Purified EhHsp90 was used foranalyzing ATP binding using the method of fluores-cence quenching. Hsp90 displays measurable dif-ferences in tryptophan fluorescence intensity uponconformational change induced by ligand binding.EhHsp90 was incubated with varying concentrationsof ATP and difference in fluorescence intensity wasplotted against ATP concentration. The dissociationconstant (Kd) for ATP binding was determined to be365.2 μM (Fig. 1d, Fig. S1a, and Table 1). Further,ATPase activity of EhHsp90 was analyzed byincubating it with varying concentrations of ATP.γ-32P-ATP was used as a tracer and ATP hydrolysiswas monitored on a TLC plate. Fractional cleavageof ATP by EhHsp90 was used to calculate ATPaseactivity, which was plotted against corresponding ATPconcentrations. Km and Kcat for ATP hydrolysis werefound to be 432.5 μM and 178.2 × 10−3 min−1,respectively (Fig. 1e). The catalytic efficiency ofEhHsp90, 4.12 × 10−4 min−1 μM−1, is higher thanthat of human Hsp90.

17-AAG binds EhHsp90 with high affinity andinhibits its ATPase activity

17-AAG is a derivative of antibiotic GA. There aremany reports from studies in humans, modelorganisms, and parasitic protozoa that establishGA and 17-AAG as potent Hsp90 inhibitors[3,5,8,10–12,28,31]. GA is a competitive inhibitorof Hsp90 and binds to the ATP-binding pocket in theN-terminal domain of Hsp90 with a higher affinitythan ATP [32]. Inhibition of Hsp90 by GA disruptssubstrate maturation, leading to derailment of manycellular pathways and, ultimately, cell death. In thecase of E. histolytica, there are no reports on eitheressentiality of Hsp90 or the effect of Hsp90 inhibitionon cellular homeostasis. Therefore, the first courseof action was to check the efficacy of 17-AAGbinding to EhHsp90. Varying concentrations of17-AAG were incubated with EhHsp90 and changein intrinsic fluorescence of Hsp90 was plottedagainst 17-AAG concentrations. Dissociation con-stant (Kd) was observed to be 10.77 μM (Fig. 2a, Fig.S1b, and Table 1).

Fig. 1. EhHsp90 cloning and biochemical characterization. (a) Cloning of Entamoeba Hsp90 in pRSET-A vector. Lane1, undigested EhHsp90 clone; lane 2, insert (2.1 kb EhHsp90) released on double digestion with BamHI and XhoIenzymes. (b) Immunoblot using mouse anti-his antibody to validate expression of recombinant His-tagged EhHsp90. Lane1, uninduced EhHsp90 gene harboring E. coli lysate; lane 2, IPTG-induced EhHsp90 expressing E. coli lysate.(c) EhHsp90 purification profile. Coomassie-stained gel shows purified fraction of recombinant His-tagged EhHsp90protein using Ni-NTA chromatography (molecular mass, ~86 kDa). (d) Binding affinity of natural ligand ATP to EhHsp90using tryptophan fluorescence. Change in intrinsic fluorescence intensity upon ligand binding was plotted against ligandconcentration. Dissociation constant, Kd, for ATP binding was found to be 365.2 μM. (e) Rate of ATP hydrolysis wasmeasured by the hydrolysis of radiolabeled ATP to ADP. A Michaelis–Menten plot shows the fractional cleavage ofγ-32P-labeled ATP plotted against ATP concentration; Km was observed to be 432.5 μM, Kcat and Kcat/Km were found to be178.2 × 10−3 (min−1) and 4.12 × 10−4 (min−1 μM−1), respectively. Inset graph is the representative Lineweaver–Burkeplot for EhHsp90 ATPase.

1788 EhAha1c — A Novel Co-Chaperone of EhHsp90

Inhibition of Hsp90 ATPase activity compromisesHsp90 function. Therefore, the effect of 17-AAG onEhHsp90 ATPase activity was checked. EhHsp90was incubated with saturating concentrations ofATP, and 17-AAG concentration was varied, rangingfrom 5 μM to 300 μM. Percent remaining activity foreach 17-AAG concentration was calculated and wasplotted against log10 [17-AAG] concentration. IC50for 17-AAG-mediated EhHsp90 ATPase activity wasobserved to be 30.90 μM (Fig. 2b and Table 1).

E. histolytica growth is sensitive toHsp90 inhibition

Hsp90 expression in E. histolytica was validatedby an immunoblot using a specific antibody raisedagainst full-length EhHsp90 (Fig. 2c and d). Toexamine the effect of Hsp90 inhibition on growth,Entamoeba trophozoites were grown to log phaseand were treated with varying concentrations of17-AAG in the range of 10 nM to 100 μM for 24 h.

Table 1. Biochemical properties of Hsp90 from E. histolytica and comparison with Hsp90 from other organisms[8,22–25]

Organism Molecularweight

pI GI50 growth(nM)

Km ATP(μM)

Kcat/Km

(min−1 μM−1)Kd ATP(μM)

Kd GA(μM)

IC50 ATPase

E. histolytica 82.99 4.97 546 (17-AAG) 432.5 4.12 × 10−4 365.2 10.77 (17-AAG) 30.90 μM (17-AAG)H. sapiens [8,22–24] 84.6 5.01 700 (GA) 324 4.6 × 10−5 240 ± 14 4.4 702 nM (GA)S. cerevisiae [22,25] 81.4 4.84 — 511 ± 54 15 × 10−5 132 ± 47 1.2 —

1789EhAha1c — A Novel Co-Chaperone of EhHsp90

Cell survival was measured by counting viable cellsusing trypan blue dye exclusion methodology.Percent survivability was plotted against log10[17-AAG] concentration. GI50 for 17-AAG treatmentwas observed to be 546 nM (Fig. 2e). At higherconcentrations of drug, complete cell death wasobserved, thereby establishing a functionally impor-tant role for Hsp90 in E. histolytica. GI50 observed forEntamoeba is lower than that reported for human(Table 1).

Fig. 2. Binding of Hsp90 inhibitor 17-AAG and its effect oncompetitive inhibitor 17-AAG toEhHsp90using tryptophan fluoresbinding was plotted against ligand concentration. Dissociationinhibition of ATPase activity of EhHsp90 is 30.90 μM. IC50 waconcentration of 17-AAG in logarithmic scale. (c) SDS-PAGE prEhHsp90 to show expression of Hsp90 in Entamoeba using spEntamoeba growth for 17-AAG was determined by counting via17-AAG using the trypan blue dye exclusion method. IC50 grostandard deviation between two independent experiments.

Identification of a novel Aha1 homologue inE. histolytica

Previously, it has been shown that protozoanparasites lack multiple co-chaperones of Hsp90and Entamoeba, in particular 4 out of about 12 co-chaperones described above. These include Cdc37,Aha1, p23, and Cyp40 (Table 2). In order to verify thereport, genome-wide search was done for homo-logues of Cdc37, p23, and Aha1. Our search

Hsp90 activity and parasite growth. (a) Binding affinity ofcence:Change in intrinsic fluorescence intensity upon ligandconstant, Kd, for 17-AAG binding is 10.77 μM. (b) IC50 fors determined by plotting percent activity remaining versusofile of whole cell lysate of E. histolytica. (d) Immunoblot forecific antibody raised against EhHsp90 in rabbit. (e) IC50 forble cells after 24 h treatment with varying concentrations ofwth was determined to be 546.4 nM. Error bars represent

Table 2. Distribution of co-chaperones in E. histolytica

H. sapiens S. cerevisiae E. histolytica

Hop/Sti1 + + +PP5 + + +Cns1 + + +FKBP52 + − +Cyclophilins + + +Cyp40 + + −p23/Sba1 + + −Aha1 + + − (C-terminal

homologue present)Sgt1 + + +Pih1 + − +Cdc37 + + −

1790 EhAha1c — A Novel Co-Chaperone of EhHsp90

revealed the presence of an Aha1 homologue,EHI_065800, sharing 35% identity with the C-terminalhalf of yeast and human Aha1 and which would bereferred to as EhAha1c (Fig. 3a). Sequence alignmentof EhAha1c with yeast Aha1 showed conservation ofcritical residues of Aha1 C-terminal domain in-volved in binding to Hsp90 (Fig. 3a) [16]. Similarly,it was observed that residues involved in Aha1binding in Hsp90 [8] are also conserved inEhHsp90 (data not shown). For understanding theevolution of Aha1 in the eukaryotic lineage, weperformed phylogenetic analysis as shown inFig. 3b. Hsp90 sequences were chosen to makethe phylogenetic tree as it is known to be wellconserved and Aha1 has evolved to regulate thefunction of Hsp90. Simultaneously, using yeastAha1 sequence as template, BLASTp analysis wasperformed to look for Aha1 homologues in the sameorganisms chosen for tree building. Figure 3bshows that Aha1 originated in early eukaryotes astwo different proteins that represent the twodomains of the full-length Aha1 present in highereukaryotes. However, these proteins originated intwo different organisms Giardia and Entamoebahaving Aha1_N domain and Aha1_C domain,respectively. With increasing complexity and associ-ated demand on Hsp90 function, these two domainsfused and full-length Aha1 originated in highereukaryotes and yeast. Among higher eukaryotes,most have completely lost the proteins with individualdomain and have retained only the full-length Aha1.However, in some organisms like Saccharomycesand Candida, in addition to full-length protein, anN-terminal homologue, Hch1, is retained. In manyapicomplexan organisms, Aha1 is characterized bythe presence of two Aha1_N domains and theseorganisms have additionally retained proteins withonly Aha1_C domain alone. While it is well estab-lished that the function of Aha1 ismediated through itsN-terminal domain (Aha1_N) [16], it is interesting tosee that Entamoeba has only the Aha1_C domainprotein.

EhAha1c binds EhHsp90

For examining whether EhAha1c is indeedexpressed as a protein in Entamoeba, immunoblotanalysis of Entamoeba lysate was carried out usingan antibody specific to EhAha1c. A clear single bandcorresponding to EhAha1c was observed on theblot, confirming its expression in Entamoeba cells(Fig. 4a). Immunofluorescence using EhHsp90 andEhAha1c antibodies shows cytosolic co-localizationof both the proteins (Fig. 4b).In order to evaluate the functionality of EhAha1c,

we cloned the gene from Entamoeba cDNA intopRSET-A vector with an N-terminal His tag and weexpressed the protein in E. coli pLysS and purified itusing Ni-NTA affinity chromatography (Fig. 4c andd). Purified protein was used to assess binding ofEhAha1c and EhHsp90 using Far Western blotting.EhAha1c was immobilized on nitrocellulose mem-brane and EhHsp90 was used as bait protein.Interaction was probed using EhHsp90-specificantibody. Bovine serum albumin (BSA) was used asnegative control.Only if there is an interaction betweenbait protein (i.e., Hsp90) and immobilized proteins, asignal specific to Hsp90 will be observed. As can beseen in Fig. 4f, signal was observed on the spot whereEhAha1c was immobilized onto the membrane alongwith the Hsp90 spot, which was used as a positivecontrol, and no signal was observed for BSA,suggesting that the interaction between EhAha1cand EhHsp90 is a specific interaction.We also analyzed the interaction between

EhHsp90 and EhAha1c using circular dichroism(CD) measurements. CD spectra for EhHsp90 showan ordered secondary structure with the majority ofthe structural component being α-helices; however,upon addition of increasing concentration of EhA-ha1c, a dose-dependent change in EhHsp90 con-formation was observed with a decrease in α-helicalcontent as can be seen in Fig. 4e. α-Helical contentof EhHsp90 alone is 55.3% and β-sheet content is8.2%. However, upon addition of four molar excessof EhAha1c, α-helical content is reduced to 48.3%and β-sheet content increases to 10.5%.The interaction between two proteins was validated

using surface plasmon resonance (SPR), whereinEhHsp90 was immobilized on a CM5 chip by aminecoupling and EhAha1c was used as the analyte.Analyte was injected over the protein-coated surfaceand a blank channel at a constant flow rate. Responseunit showed a dose-dependent increase with increas-ing concentrationofEhAha1cas canbeseen inFig. 4g.Evaluation of sensorgrams for EhHsp90 and EhAha1cshowed the dissociation constant, Kd, to be 8 μM.

EhAha1c stimulates EhHsp90 ATPase activity

To examine the function of EhAha1c, we incubat-ed the purified protein with EhHsp90 in different

Fig. 3. A novel truncated C-terminal homologue of Aha1 in Entamoeba. (a) Sequence alignment for Entamoeba Aha1.In Entamoeba, only the C-terminal half of Aha1 is present and shares 50% similarity with yeast Aha1. ▲ shows residuesknown to be critical in yeast Aha1 C-terminal domain for binding to Hsp90. (b) Phylogenetic tree for Hsp90s in protozoaand higher eukaryotes showing distribution of Aha1 homologues. Percentages represent identity to human Aha1. Thereare four different classes of Aha1 homologues—typical full-length Aha1 with Aha1_N and C-terminal Aha1_C domain,Hch1 present in yeast andGiardia and contains only Aha1_N domain, Aha1c that has only C-terminal domain Aha1_C andpresent only in protozoa, and Aha1 homologue containing two Aha1_N domains and no Aha1_C domain and is restrictedto apicomplexan protozoa. All protozoa harboring Aha1c also have a corresponding Aha1_N domain-containing Aha1homologue; however, Entamoeba exclusively harbors Aha1c and has no cognate Aha1_N domain-containing homologue.

1791EhAha1c — A Novel Co-Chaperone of EhHsp90

Fig. 4. EhAha1c expresses in Entamoeba and interacts with EhHsp90. (a) Immunoblot showing expression of EhAha1cin Entamoeba using specific antibody raised against EhAha1c in mice. (b) Immunofluorescence showing co-localization ofEhHsp90 and EhAha1c in Entamoeba cytosol. (c) Cloning of Entamoeba Aha1c in pRSET-A vector. Lane 1, DNA ladder;lane 2, insert release for positive clone upon double digestion. (d) EhAha1c purification profile. SDS-PAGE gel showspurified fraction of recombinant His-tagged EhAha1 protein using Ni-NTA chromatography (molecular mass, ~18.3 kDa).(e) CD spectrum showing the conformational change induced in EhHsp90 secondary structure upon binding withincreasing concentration of EhAha1c. (f) Far Western blot showing binding of EhAha1c to EhHsp90. EhAha1c was spottedon activated nitrocellulose membrane, BSA was used as negative control, and EhHsp90 was used as positive control.Membrane was probed with EhHsp90 followed by rabbit anti-EhHsp90 antibody. (g) Interaction of EhAha1c tosurface-bound EhHsp90 was measured by SPR at 25 °C. Sensorgram shows dose-dependent increase in binding ofEhAha1c to EhHsp90.

1792 EhAha1c — A Novel Co-Chaperone of EhHsp90

1793EhAha1c — A Novel Co-Chaperone of EhHsp90

molar ratios, and its effect on EhHsp90 ATPaseactivity was observed. The activity of EhHsp90 wascalculated in the absence and presence of EhAha1cas shown in Fig. 5a. At a molar concentration of 2:1for EhHsp90 and EhAha1c (one dimer of Hsp90 andone EhAha1c monomer), stimulation of ATPaseactivity was observed. This ATPase stimulationincreases in a dose-dependent manner and amaximal three fold stimulation of ATPase activitywas observed at a 1:2 molar ratio of EhHsp90 andEhAha1c.Another protozoan parasite,Giardia, also does not

have full-length Aha1, but it has an N-terminal Hch1homologue and, therefore, it was of interest toassess the effect of the C-terminal Aha1 homologueon GlHsp90, which has probably never encounteredthis protein endogenously. Also, we assessed theeffect of EhAha1c on Hsp90 from a system where acanonical full-length Aha1 is present and for thesame we chose human Hsp90. We incubated allthree Hsp90s—EhHsp90, GlHsp90 and HsHsp90—at 3 μM concentration with four times molar excess

Fig. 5. EhAha1c is a functional co-chaperone of Hsp90. (a) Ehfold stimulation inATPase activity of EhHsp90 (2 μM) by increasinstimulate ATPase activity of other Hsp90s, namely, human Hsp9EhAha1c protects inhibition of EhHsp90ATPaseactivity by its phaIC50 for EhHsp90 ATPase increases by 10-fold.

of EhAha1c and calculated the fold change inATPase activity of all three proteins in the presenceand absence of EhAha1c. Our results showed thatEhAha1c was capable of stimulating the ATPaseactivity of both Giardia Hsp90 and human Hsp90along with EhHsp90 by three- and two fold,respectively (Fig. 5b). These results collectivelysuggest that EhAha1c is a bona fide co-chaperoneof Hsp90, capable of binding and stimulatingATPase activity even though it lacks the canonicalN-terminal domain of Aha1.We also examined whether binding of EhAha1c to

Hsp90 has any effect on inhibition of Hsp90 ATPaseactivity by its pharmacological inhibitor, 17-AAG. Weincubated 3 μM EhHsp90 protein with four timesmolar excess of EhAha1c (12 μM) and performedATPase assay with varying concentrations of17-AAG. As mentioned above, EhHsp90 shows anIC50 (ATPase) of 30.90 μM; however, in the pres-ence of EhAha1c, sensitivity of EhHsp90 ATPaseactivity towards 17-AAG-mediated inhibition de-creases with a higher IC50 value of 115 μM (Fig. 5c).

Aha1c stimulates EhHsp90 ATPase activity. The plot showsg themolar concentration of EhAha1c. (b) EhAha1c can also0 and Giardia Hsp90, by two- to three fold, respectively. (c)rmacological inhibitor, 17-AAG. In the presenceof EhAha1c,

1794 EhAha1c — A Novel Co-Chaperone of EhHsp90

Discussion

Hsp90 shows a high level of sequence conserva-tion across eukaryotic species. Its function has beenshown to be essential for cellular growth and survivalin all the organisms studied so far. Hsp90 has gainedpopularity as a candidate drug target for cancertherapy and now its potential as a drug target forinfectious organisms is being explored by manyresearchers. E. histolytica is considered one of themost infectious disease-causing organisms respon-sible for the third highest morbidity and mortalityworldwide [33]. Hsp90 is a very dynamic moleculeand substrate maturation requires ATPase activity.Hsp90 undergoes transition from an open ADPbound state to a closed ATP bound state. ATPhydrolysis restores the open state [1,32]. In thisstudy, we have carried out biochemical characteri-zation of EhHsp90. We have determined bindingaffinities for ATP and pharmacological inhibitor17-AAG binding to EhHsp90. 17-AAG exhibitsmuch stronger affinity for Hsp90 binding comparedto ATP. EhHsp90 is a functional ATPase with acatalytic efficiency of 4.12 × 10−4 μM−1 min−1.While Hsp90 functions in a multi chaperone

complex and most of its co-chaperones are con-served across various species, many protozoa lackone or a few of them. Entamoeba serves as anexcellent model not only to understand the evolu-tionary and functional aspects of these co-chaper-ones in terms of their direct effect on Hsp90 activityand client maturation but also to help understand ifthese functions are exclusive and are dispensable inprimitive organisms or are taken up by othermolecular players. Our search in the genomedatabase showed the presence of a novel Aha1homologue having only the Aha1_C domain withoutany cognate Aha1_N domain. Additionally, anothercharacteristic form of Aha1 was observed to bepresent in a few apicomplexan protozoa where Aha1protein possessed two Aha_N domains. This sug-gests that Hsp90 evolution alone does not dictateAha1 evolution and different organisms have mod-ified Aha1 with requirements pertaining to theirbiological niche.Yeast Aha1 has been shown to bind and stimulate

the ATPase activity of Hsp90 by 12-fold while Hch1(Aha1_N alone) can stimulate activity by nearly4-fold. When recombinant forms of N- and C-termi-nal domains of Aha1 were used separately, theN-terminal domain showed ATPase stimulationequivalent to Hch1; however, the C-terminal domainalone did not show any effect on Hsp90 ATPaseactivity [16,17]. There is one report where it wasshown that a construct with human Aha1 C-terminaldomain along with the linker region had some abilityto weakly stimulate Hsp90 activity [34]. In contrast tothese reports, we found that a naturally presentEhAha1c not only binds but also stimulates Hsp90

ATPase activity. We find that EhAha1c does notshow any bias and possesses the ability to stimulateATPase activity of Giardia and human Hsp90 aswell, suggesting its role as a bona fide co-chaperoneof Hsp90 like Aha1 and Hch1.Over the last decade, multiple Hsp90 inhibitors

have been explored for their anti-infective potential.Like other protozoa, we find Entamoeba to be highlysensitive to Hsp90 inhibition compared to mamma-lian cell lines. Hsp90 inhibition led to arrest in growthand eventually cell death in a dose-dependentmanner, thereby suggesting EhHsp90 as a drugtarget. EhAha1c showed a protective role in phar-macological inhibition of Hsp90 ATPase activity.Therefore, it will be interesting to explore if EhHsp90inhibition can be potentiated by blocking interactionof EhAha1c with EhHsp90. It would be interesting toevaluate if EhAha1c is essential. Previously, adouble knockout of Aha1 and Hch1 in yeast hasbeen shown to result in hypersensitivity to heatshock and a compromised function in terms offacilitating client maturation [17].To conclude, our study for the first time describes a

novel co-chaperone of Hsp90, EhAha1c, whichrepresents only the C-terminal domain of full-lengthAha1. Our study suggests that the two domains ofAha1—Aha1_N and Aha1_C—have evolved inde-pendently in the primitive protozoa as functionalactivators of Hsp90 ATPase and during the course ofevolution became part of a single protein for anenhanced function, as in the case of highereukaryotes. It will be of immense importance tounderstand the interplay of co-chaperones withHsp90 in primitive organisms like Entamoeba tohelp our understanding about the evolution andfunction of Hsp90 co-chaperone repertoire. A studyof Hsp90 and its co-chaperone in primitive parasitessuch as E. histolytica is likely to produce importantinsights into the evolution of Hsp90 co-chaperonerepertoire in eukaryotes.

Experimental Procedures

E. histolytica strain

E. histolytica strain HM-1: IMSS trophozoites wereaxenically cultured in TY1-S-33 medium at 33.5 °Cin screw cap glass tubes.

Cloning and purification

EhHsp90 (Gene id: EHI_196940) was amplifiedfrom Entamoeba genomic DNA using the followingprimers: Eh_90f BamHI: GCCGCGGGATCCATGGGAAATAGAAAAGATCAATC (sense) and Eh_90rXhoI: GCCGCGCTCGAGTTAATCAACTTCTTCCATC (antisense). The amplified product of

1795EhAha1c — A Novel Co-Chaperone of EhHsp90

2157 bp was cloned in pRSET-A vector as a 6×-Histag fusion protein. Similarly, EhAha1c (Gene id:EHI_065800) was amplified from Entamoeba cDNAusing the following primers: Eh_Aha1f BamHI:GCCGCGGGATCCATGGCACAAATTGAAATTACTTC (sense) and Eh_Aha1R XhoI: GCCGCGCTCGAGTTAATATCCAAACATTTTCTTCATTC(antisense). The amplified product of 396 bp wascloned in pRSET-A vector as His tag fusion protein.Giardia Hsp90 clone in pRSET-A vector wasavailable in the laboratory (unpublished) and humanHsp90 clone is a kind gift from Prof. David O. Toft(Mayo Clinic, Rochester, MN). All proteins wereexpressed in E. coli BL21 pLysS expression strainand proteins were purified to homogeneity usingNi-NTA affinity chromatography. In brief, E. coli strainsof clone EhHsp90 and EhAha1c were grown in LBbroth and induced with 0.5 mM IPTG at 16 °C for 8 h.Cells were pelleted down and lysed in buffer containingTris–Cl, pH 7.5, 1 M NaCl, and 10 mM imidazole andprotease inhibitor. Lysate supernatant was allowed tobind to Ni-NTA beads. Beads bound to protein werethen washed with buffer containing a gradient of imid-azole concentration ranging from 10 mM to 60 mM.Protein was finally eluted using 200 mM imidazolecontaining buffer. Protein was then dialyzed in suitablebuffers described in the following sections for differentexperiments. Protein concentration was estimatedusing Bradford's reagent using BSA as standard.

Binding studies for ATP and 17-AAG

ATP binding was determined using the method offluorescence quenching upon ligand binding asdescribed previously [8]. Briefly, 50 μg of protein inbinding buffer (40 mM Hepes–KOH buffer, pH 7.4,5 mM MgCl2, and 100 mM KCl) was incubated withdifferent concentrations of ATP (0.05–10 mM).Intrinsic tryptophan fluorescence was measured byscanning the emission spectrum in the wavelengthrange of 300–400 nm and excitation at 280 nm. Thefluorescence intensity at λmax = 340 nm was select-ed for calculations. Difference in intrinsic fluores-cence of protein alone and in the presence of ligandwas plotted against ligand concentration. Data wereanalyzed using GraphPad Prism 5.0 using nonlinearregression analysis with single site-specific binding.A similar procedure was carried out for 17-AAGbinding with 50 μg protein in binding buffer (50 mMTris and 1 mM ethylenediaminetetraacetic acid) withconcentrations of 17-AAG ranging from 500 nM to50 μM). The final concentration of dimethyl sulfoxidein the assay was 1%.

ATPase assay, ATPase inhibition assay, andATPase stimulation assay

EhHsp90 (1.5 μM) in 40 mM Tris–Cl buffer,pH 7.4, containing 100 mM KCl and 5 mM MgCl2

was incubated with varying concentrations of ATP(50 to 4000 μM) as previously described [8].γ-32P-ATP with a specific activity of 0.55 Ci/mmolwas used as a tracer. 17-AAG (300 μM) was used incontrol reaction to negate out nonspecific or back-ground activity. Control activity was subtracted fromtotal activity. ATPase activity was plotted againstATPase concentration. Data were analyzed viaGraphPad Prism 5.0 using Michaelis–Mentenkinetics.Similarly, ATPase inhibition assay was carried out

except that purified EhHsp90 was incubated with asaturating concentration of ATP (2 mM) and17-AAG concentration was varied from 2.5 μM to150 μM. 300 μM 17-AAG was used in controlreaction. Percentage residual ATPase activitywas plotted against log of concentration of inhibitorand the result was analyzed using GraphPad Prism5.0.Assay for ATPase stimulation by EhAha1c was

carried out by incubating 2 μM EhHsp90 with 8 mMATP and varying molar ratios of EhAha1c (0.25 to 4).EhAha1c alone for corresponding concentrationswas taken as control in addition to control reactionswith 300 μM 17-AAG.

Immunoblot for EhHsp90 and EhAha1c

Entamoeba trophozoites grown to log phase wereharvested by chilling on ice for 5 min and centrifug-ing at 300g. Cells were lysed by hypotonic lysis inlysis buffer containing 20 mM Tris–Cl buffer, pH 7.4,containing 2 mM ethylenediaminetetraacetic acid,2 mM ethylene glycol bis(β-aminoethyl ether) N,N′-tetraacetic acid, 2 mM DTT, 2 mM PHMB, 0.5%Triton X-100, and protease inhibitors. Lysate wasresolved on 12.5% polyacrylamide SDS gel underreducing conditions and subjected to immunoblot-ting with EhHsp90- and EhAha1c-specific antibod-ies. Antibodies against full-length EhHsp90 andEhAha1c were raised in rabbit and mouse,respectively.

Assay for Entamoeba sensitivity to 17-AAG

Entamoeba trophozoites grown to log scale wereharvested and 15,000 cells were seeded per wellin TYI-S-33 medium in a 96-well plate. Cells wereallowed to adhere. Medium was replaced withmedium containing varying concentrations of17-AAG from 10 nM to 100 μM. Dimethyl sulfoxide(0.2%) was used as control. Cells were treatedwith drug for 24 h. After treatment, viable Ent-amoeba cells were counted under a microscopeusing the trypan blue dye exclusion method.Percent survival above control was plotted againstlog concentration of 17-AAG and data wereanalyzed using GraphPad Prism 5 and GI50 wasdetermined.

1796 EhAha1c — A Novel Co-Chaperone of EhHsp90

Aha1 sequence alignment and phylogenetic tree

Hsp90 sequences were retrieved from NCBIdatabase and analyzed using MEGA 5.1 software.Sequences used for Hsp90 were as follows:Saccharomyces cerevisiae NP_013911.1, Homosapiens NP_005339.3, E. histolytica XP_653132.1/EHI_196940, Cryptosporidium parva XP_626924.1,Babesia bovis XP_001611554.1, P. falciparumXP_001348998.1, T. gondii XP_002368278.1, Try-panosoma brucei A44983, G. lamblia BAJ33526.1,Neospora caninum XP_003881046.1, Mus muscu-lus NP_032328.2, Gallus gallus NP_996842.1,Danio rer io NP_571403.1, Brugia malayiEDP29326.1, E. coli Htpg EDV65681.1, and Klebsi-ella Htpg CCI78437.1. A multiple sequence align-ment was carried out and phylogenetic tree wasconstructed using the Neighbor-Joining method insoftware Mega 5.2 [26,35]. The optimal tree with thesum of branch length = 2.21793029 is shown. Thepercentage of replicate trees in which the associatedtaxa clustered together in the bootstrap test (500replicates) is shown next to the branches [27]. Thetree is drawn to scale, with branch lengths in thesame units as those of the evolutionary distancesused to infer the phylogenetic tree. The evolutionarydistances were computed using the Poisson correc-tion method and are in the units of the number ofamino acid substitutions per site. All positionscontaining gaps and missing data were eliminated.Sequences for Aha1 homologues in protozoan

parasites were retrieved from EupathDB usingmanual BLASTp searches against sequences ofhuman and yeast Aha1 and yeast Hch1. Sequencesfor Aha1 for other organisms were retrieved from theNCBI database. The sequences used for Aha1 wereas follows: E. histolytica EHI_065800, Entamoebadispar EDI_198720, Acanthamoeba castellaniiACA1_074410, G. lambliaGLP15_885, P. falciparumPF3D7_0306200, P. falciparum PF3D7_0308500,C. parva cgd7_2050, C. parva cgd2_320, B. bovisBBM_II00940, N. caninum NCLIV_018620, N. cani-num NCLIV_012030, T. brucei Tb927.10.13710, T.gondii TGGT1_045940, T. gondii TGGT1_124070,SaccharomycesHch1 P53834,SaccharomycesAha1NP_010500.3, B. malayi XP_001901650.1, D. rerioNP_997767.1,G. gallus XP_004941893.1,M.muscu-lus NP_666148.1, Pan troglodytes XP_001165345.1,and H. sapiens AAH00321.1.

Indirect immunofluorescence analysis

Entamoeba cells grown to log phase were chilledand harvested. Cells were washed and were allowedto adhere on coverslips for 15 min at 33.5 °C. Cellswere then fixed with acetone chilled at −20 °C. Thecoverslips were blocked with 3% BSA in phospha-te-buffered saline for 1 h. Primary antibodies againstEhHsp90 and EhAha1c were diluted in phosphate-

buffered saline and were added onto the cover-slips and incubated for 1 h. Three washes weregiven to the coverslips maintaining gentle agita-tion. Coverslips were then incubated with fluores-cein isothiocyanate-conjugated anti-rabbitsecondary antibody and tetramethylrhodamineisothiocyanate-conjugated anti-mouse secondaryantibody. Threewasheswere given to the coverslipsmaintaining gentle agitation. The coverslips weremounted on glass slides with 90% glycerol contain-ing 2% DABCO and 1 μg/ml DAPI and visualizedunder a confocal laser scanning microscope (LeicaTCS SP8).

EhAha1c binding to Hsp90

Far Western blotting was carried out to identifyinteraction of EhAha1c and EhHsp90 in vitro.EhAha1c was immobilized on nitrocellulose alongwith BSA (negative control) and EhHsp90 aspositive control. Membrane was blocked usingskimmed milk for 45 min. Blot was probed bypurified bait protein EhHsp90 and incubated for1 h followed by four washes in TBST. Interactionof EhAha1c and EhHsp90 was then probedusing antibody specific to EhHsp90, which wasincubated for 8 h. Primary antibody incubationwas followed by four washes. Membrane was thenincubated for 1 h in buffer with horseradishperoxidase-conjugated secondary anti-rabbit anti-body. Following washes, signal was developedusing chemiluminescence.CD spectra for EhHsp90 were recorded in the

presence and absence of EhAha1c on a nitrogen-flushed JASCO spectropolarimeter. EhHsp90 andEhAha1c were allowed to interact for 30 min at roomtemperature (25 °C). Spectra for EhHsp90 andEhHsp90 complex with different EhAha1c concen-trations were obtained. Also, spectra for eachEhAha1c concentration was taken as control.Three scans for each sample were taken at ascan speed of 20 nm min−1. Change in secondary-structure conformation was observed.SPR analysis was carried out on a Biacore T200

instrument. The surface of the CM5 chip wasactivated for amine coupling using 1-ethyl-3(3-dimethylaminopropyl)-carbodiimide hydrochloride/N-hydroxysuccinimide. EhHsp90 was immobilizedon the surface in 10 mM acetate buffer, pH 5, andthereafter the surface was blocked with 1 M etha-nolamine at pH 8.5 for 6 min. Various concentra-tions of EhAha1c (in HBS-EP buffer, GE Healthcare)were injected at a flow rate of 30 μL/min, and 10 mMglycine–HCl buffer, pH 3 (GE Healthcare), was usedto regenerate the surface. Similar concentrations ofEhAha1c were also injected in flow channel with noimmobilized protein to serve as blank. Bindingkinetics were evaluated using Biacore T200 evalu-ation software (GE Healthcare).

1797EhAha1c — A Novel Co-Chaperone of EhHsp90

Supplementary data to this article can be foundonline at http://dx.doi.org/10.1016/j.jmb.2014.01.008.

Acknowledgments

The authors would like to acknowledge Prof. AlokBhattacharya, Prof. Sudha Bhattacharya, andDr. Swati Tiwari, JNU, for helping with Entamoebaculture. We thank Dr. Bob Kennedy Dass and Dr.Uma Sinha Dutta from GE Healthcare for help withSPR experiment. We acknowledge funding from theDBT-IISc partnership grant. Funding from CSIR forMeetali Singh and from DBT for Varun Shah isacknowledged.

Received 11 November 2013;Received in revised form 19 January 2014;

Accepted 23 January 2014Available online 29 January 2014

Keywords:Hsp90;

17-AAG;Entamoeba;

co-chaperones;Aha1

Abbreviations used:Hsp90, heat shock protein 90; 17-AAG, 17-allylamino-17-demethoxygeldanamycin; Aha1, activator of Hsp90 AT-Pase; SPR, surface plasmon resonance; GA, geldana-

mycin; BSA, bovine serum albumin.

References

[1] Taipale M, Jarosz DF, Lindquist S. HSP90 at the hub ofprotein homeostasis: emerging mechanistic insights. NatRev Mol Cell Biol 2010;11:515–28.

[2] Neckers L, Ivy SP. Heat shock protein 90. Curr Opin Oncol2003;15:419–24.

[3] Acharya P, Kumar R, Tatu U. Chaperoning a cellular upheavalin malaria: heat shock proteins in Plasmodium falciparum. MolBiochem Parasitol 2007;153:85–94.

[4] BanumathyG,SinghV,PavithraSR, TatuU.Heat shockprotein90 function is essential for Plasmodium falciparum growth inhuman erythrocytes. J Biol Chem 2003;278:18336–45.

[5] Pavithra SR, Banumathy G, Joy O, Singh V, Tatu U. Recurrentfever promotes Plasmodium falciparum development in humanerythrocytes. J Biol Chem 2004;279:46692–9.

[6] Corbett KD, Berger JM. Structure of the ATP-binding domain ofPlasmodium falciparum Hsp90. Proteins 2010;78:2738–44.

[7] Kumar R, Pavithra SR, Tatu U. Three-dimensional structureof heat shock protein 90 from Plasmodium falciparum:molecular modelling approach to rational drug design againstmalaria. J Biosci 2007;32:531–6.

[8] Pallavi R, Roy N, Nageshan RK, Talukdar P, Pavithra SR,Reddy R, et al. Heat shock protein 90 as a drug target againstprotozoan infections: biochemical characterization of HSP90from Plasmodium falciparum and Trypanosoma evansi andevaluation of its inhibitor as a candidate drug. J Biol Chem2010;285:37964–75.

[9] Shahinas D, Liang M, Datti A, Pillai DR. A repurposing strategyidentifies novel synergistic inhibitors of Plasmodium falciparumheat shock protein 90. J Med Chem 2010;53:3552–7.

[10] Wider D, Peli-Gulli MP, Briand PA, Tatu U, Picard D. Thecomplementation of yeast with human or Plasmodiumfalciparum Hsp90 confers differential inhibitor sensitivities.Mol Biochem Parasitol 2009;164:147–52.

[11] Graefe SE, Wiesgigl M, Gaworski I, Macdonald A, Clos J.Inhibition of HSP90 in Trypanosoma cruzi induces a stressresponse but no stage differentiation. Eukaryot Cell2002;1:936–43.

[12] Echeverria PC, Matrajt M, Harb OS, Zappia MP, Costas MA,Roos DS, et al. Toxoplasma gondii Hsp90 is a potential drugtarget whose expression and subcellular localization aredevelopmentally regulated. J Mol Biol 2005;350:723–34.

[13] Wiesgigl M, Clos J. Heat shock protein 90 homeostasiscontrols stage differentiation in Leishmania donovani. MolBiol Cell 2001;12:3307–16.

[14] Rochani AK, Singh M, Tatu U. Heat shock protein 90inhibitors as broad spectrum anti-infectives. Curr Pharm Des2013;19:377–86.

[15] Johnson JL, Brown C. Plasticity of the Hsp90 chaperonemachine in divergent eukaryotic organisms. Cell StressChaperones 2009;14:83–94.

[16] Retzlaff M, Hagn F, Mitschke L, Hessling M, Gugel F, KesslerH, et al. Asymmetric activation of the hsp90 dimer by itscochaperone aha1. Mol Cell 2010;37:344–54.

[17] Panaretou B, Siligardi G, Meyer P, Maloney A, Sullivan JK,Singh S, et al. Activation of the ATPase activity of hsp90 by thestress-regulated cochaperone aha1.Mol Cell 2002;10:1307–18.

[18] Vaughan CK, Mollapour M, Smith JR, Truman A, Hu B, GoodVM, et al. Hsp90-dependent activation of protein kinases isregulated by chaperone-targeted dephosphorylation ofCdc37. Mol Cell 2008;31:886–95.

[19] Eckert K, Saliou JM, Monlezun L, Vigouroux A, Atmane N,Caillat C, et al. The Pih1–Tah1 cochaperone complex inhibitsHsp90 molecular chaperone ATPase activity. J Biol Chem2010;285:31304–12.

[20] Marsh JA, Kalton HM, Gaber RF. Cns1 is an essential proteinassociated with the hsp90 chaperone complex in Saccharo-myces cerevisiae that can restore cyclophilin 40-dependentfunctions in cpr7Delta cells. Mol Cell Biol 1998;18:7353–9.

[21] Dolinski KJ, Cardenas ME, Heitman J. CNS1 encodes anessential p60/Sti1 homolog in Saccharomyces cerevisiaethat suppresses cyclophilin 40 mutations and interacts withHsp90. Mol Cell Biol 1998;18:7344–52.

[22] Prodromou C, Roe SM, O'Brien R, Ladbury JE, Piper PW,Pearl LH. Identification and structural characterization of theATP/ADP-binding site in the Hsp90 molecular chaperone.Cell 1997;90:65–75.

[23] Roe SM, Prodromou C, O'Brien R, Ladbury JE, Piper PW,Pearl LH. Structural basis for inhibition of the Hsp90molecular chaperone by the antitumor antibiotics radicicoland geldanamycin. J Med Chem 1999;42:260–6.

[24] McLaughlin SH, Ventouras LA, Lobbezoo B, Jackson SE.Independent ATPase activity of Hsp90 subunits creates aflexible assembly platform. J Mol Biol 2004;344:813–26.

1798 EhAha1c — A Novel Co-Chaperone of EhHsp90

[25] Owen BA, Sullivan WP, Felts SJ, Toft DO. Regulation of heatshock protein 90 ATPase activity by sequences in thecarboxyl terminus. J Biol Chem 2002;277:7086–91.

[26] Tamura K, Peterson D, Peterson N, Stecher G, Nei M,Kumar S. MEGA5: molecular evolutionary geneticsanalysis using maximum likelihood, evolutionary distance,and maximum parsimony methods. Mol Biol Evol2011;28:2731–9.

[27] Zharkikh A, Li WH. Estimation of confidence in phylogeny:the complete-and-partial bootstrap technique. Mol Phylo-genet Evol 1995;4:44–63.

[28] Cowen LE, Singh SD, Kohler JR, Collins C, Zaas AK, SchellWA, et al. Harnessing Hsp90 function as a powerful, broadlyeffective therapeutic strategy for fungal infectious disease.Proc Natl Acad Sci U S A 2009;106:2818–23.

[29] Chase G, Deng T, Fodor E, Leung BW, Mayer D,Schwemmle M, et al. Hsp90 inhibitors reduce influenzavirus replication in cell culture. Virology 2008;377:431–9.

[30] Dolgin E, Motluk A. Heat shock and awe. Nat Med2011;17:646–9.

[31] Neckers L, Tatu U. Molecular chaperones in pathogenvirulence: emerging new targets for therapy. Cell HostMicrobe 2008;4:519–27.

[32] Pearl LH, Prodromou C. Structure and mechanism of theHsp90 molecular chaperone machinery. Annu Rev Biochem2006;75:271–94.

[33] Morf L, Singh U. Entamoeba histolytica: a snapshot of currentresearch and methods for genetic analysis. Curr OpinMicrobiol 2012;15:469–75.

[34] Koulov AV, LaPointe P, Lu B, Razvi A, Coppinger J, DongMQ, et al. Biological and structural basis for Aha1 regulationof Hsp90 ATPase activity in maintaining proteostasis in thehuman disease cystic fibrosis. Mol Biol Cell 2010;21:871–84.

[35] Saitou N, Nei M. The neighbor-joining method: a newmethodfor reconstructing phylogenetic trees. Mol Biol Evol1987;4:406–25.