Phytochemistry 94 (2013) 36–44

Contents lists available at SciVerse ScienceDirect

Phytochemistry

journal homepage: www.elsevier .com/locate /phytochem

Sequence analysis and structure prediction of enoyl-CoA hydratase fromAvicennia marina: Implication of various amino acid residueson substrate–enzyme interactions

0031-9422/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.phytochem.2013.05.018

Abbreviations: ECH, enoyl-CoA hydratase; AMECH, Avicennia marina enoyl-CoAhydratase; DAC-CoA, dimethyl amino cinnamoyl-CoA; SPDV, swiss PDB viewer.⇑ Corresponding author. Tel.: +92 21 99261701/2; fax: +92 21 4819018 19.

E-mail address: asmat.salim@iccs.edu (A. Salim).

Uzma Jabeen a, Asmat Salim b,⇑a H.E.J. Research Institute of Chemistry, International Center for Chemical and Biological Sciences, University of Karachi, Karachi 75270, Pakistanb Dr. Panjwani Center for Molecular Medicine and Drug Research, International Center for Chemical and Biological Sciences, University of Karachi, Karachi 75270, Pakistan

a r t i c l e i n f o a b s t r a c t

Article history:Received 16 November 2012Received in revised form 22 May 2013Available online 26 June 2013

Keywords:Homology modelingStructure predictionSubstrate interactionsSalt stress

Enoyl-CoA hydratase catalyzes the hydration of 2-trans-enoyl-CoA into 3-hydroxyacyl-CoA. The presentstudy focuses on the correlation between the functional and structural aspects of enoyl-CoA hydratasefrom Avicennia marina. We have used bioinformatics tools to construct and analyze 3D homology modelsof A. marina enoyl-CoA hydratase (AMECH) bound to different substrates and inhibitors and studied theresidues involved in the ligand–enzyme interaction. Structural information obtained from the modelswas compared with those of the reported crystal structures. We observed that the overall folds were sim-ilar; however, AMECH showed few distinct structural changes which include structural variation in themobile loop, formation and loss of certain interactions between the active site residues and substrates.Some changes were also observed within specific regions of the enzyme. Glu106 is almost completelyconserved in sequences of the isomerases/hydratases including AMECH while Glu86 which is the othercatalytic residue in most of the isomerases/hydratases is replaced by Gly and shows no interaction withthe substrate. Asp114 is located within 4 Å distance of the catalytic water which makes it a probable can-didate for the second catalytic residue in AMECH. Another prominent feature of AMECH is the presence ofstructurally distinct mobile loop having a completely different coordination with the hydrophobic bind-ing pocket of acyl portion of the substrate.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

In higher plants, fatty acids are stored in large amounts as tria-cylglycerols that are degraded by b-oxidation and sequential re-moval of carbon units resulting in the formation of acetyl-CoA(Parani et al., 1999). The peroxisome is the site of numerous impor-tant biochemical reactions in plants, including b-oxidation cycle(Reumann et al., 2004). b-Oxidation also plays a significant roleduring the vegetative and reproductive growth phases, and is in-volved in plant responses to stress mainly in the synthesis of jas-monic acid (Goepfert and Poirier, 2007). The complete b-oxidation process consists of the cyclic repetition of four basicreactions catalyzed by a dehydrogenase, a hydratase, a seconddehydrogenase and a thiolase (Modis et al., 1998). Enoyl-CoAhydratase (ECH) (EC 4.2.1.17) catalyzes one of the reactions in fattyacid metabolism i.e. the hydration of 2-trans-enoyl-CoA into

3-hydroxyacyl-CoA (Engel et al., 1996). This reaction involves addi-tion of a hydroxyl group from an activated water molecule to theb-carbon (C3) and protonation of the a-carbon (Müller-Newenand Stoffel, 1993).

Enoyl-CoA hydratase has also been identified as one of the pro-teins related to various salt stress responses (Goepfert and Poirier,2007). Being the first member of crotonase superfamily, it consistsof repeated bba units that assemble into two approximately per-pendicular b-sheets surrounded by a-helices (Hamed et al.,2008). This enzyme has broad substrate specificity. It is active withsubstrates of varying chain lengths (C4–C20) (Kim and Battaile,2002). Crotonyl-CoA, hexadienoyl-CoA, and b-hydroxyacyl-CoAderivatives containing, 4, 6, 8, 9, and 12 carbon atoms are all potentsubstrates (Engel et al., 1996).

The structural studies done so far have revealed that two glu-tamic acid residues act as catalytic acid for providing the a-protonduring the hydratase reaction and as the catalytic base for the acti-vation of a water molecule (Muller et al., 1995; Gerlt and Gassman,1993; Bahnson et al., 2002). However, one of the glutamic acids isnot conserved in all isomerases/hydratases suggesting that someother residue may be involved in water activation reaction. The

U. Jabeen, A. Salim / Phytochemistry 94 (2013) 36–44 37

primary conserved feature of the enzyme is a common protein foldthat produces an oxyanion hole formed by alanine and glycine aswell as side chain atoms of the two glutamic acid residues that sta-bilize carbanion transition states of the enzyme (Holden et al.,2001; Bell et al., 2002).

In the present study, we have analyzed the primary sequence ofAvicennia marina enoyl-CoA hydratase (AMECH) and constructed3D homology models of AMECH bound to different ligands usingthe known structural coordinates of enoyl-CoA hydratases. A. mar-ina (grey mangrove) is a salt tolerant plant. As ECH has been shownto have some role in stress related conditions (Goepfert and Poirier,2007), we wanted to observe whether there is any significant cor-relation with the structural aspects of the enzyme as well. Thestructural information obtained from the models is compared withthose of the reported crystal structures. This includes the study ofthe active site residues with substrate and inhibitors which playimportant roles in catalysis, and mutation predictions at the cru-cial sites to see if the change in the amino acids resulted in differ-ent interactions with the substrates as well as with theneighboring residues.

2. Results and discussion

2.1. Multiple sequence alignment and phylogenetic analysis

Multiple sequence alignment of A. marina enoyl-CoA hydratase(AMECH) was performed with the members of enoyl-CoA hydra-tase/isomerase family whose structures have been solved anddeposited in the protein data bank (Supplemental material;Table S1). These sequences are part of family HD2 (abbreviatedfor hyhydroxyacyl dehydratases that catalyze the forward reac-tion) as classified in the ThYme (Thioester-active EnzYmes) data-

Fig. 1. Pairwise sequence alignment of Avicennia marina enoyl-CoA hydratase (AMECH) aCoA hydratase/isomerase (PDB id: 3JU1) showing 30% identity, while rat enoyl-CoA hydrhexadienoyl-CoA (PDB id: 1MJ3), (d) acetoacetyl-CoA (PDB id: 1DUB) and (e) octanoyl-C3JU1, 301 out of 321 residues of AMECH were aligned while in case of 1EY3, 1MJ3, 1DUB anecessary. These sequences and corresponding crystal structures were used for model b

base at http://www.enzyme.cbirc.iastate.edu (Cantu et al., 2011,2012). Of the 22 amino acid residues involved in the interactionwith ligand atoms, none of them is entirely conserved. Two gluta-matic acid residues are conserved in most of the enoyl-CoA hydra-tases and are involved in the catalytic reaction of hydration (Bellet al., 2002; Bahnson et al., 2002). In AMECH, one of the two cata-lytic residues is Glu106 which is better conserved in ECH memberswhile the other glutamic acid is replaced by Gly at position 86(AMECH numbering). The probable candidate for the second cata-lytic residue, Asp114 is only weakly conserved. Analysis of themultiple sequence alignment shows that the second glutamic acidresidue is not conserved in 9 species; out of which 5 possess Asp attheir respective equivalent positions. Multiple sequence alignmentshows that Gly34, Asp36, Gly78 and Gly84 are completely con-served in ECH sequences while 22 other residues are conservedin more than 70% sequences. Aligned residues were analyzed inthe homology models and crystal structures of the templates. Itwas observed that Glu106 hydrogen bonds with the substratewhile Asp114 is located within the hydrogen bonding distancewith the catalytic water. Other amino acid residues that showinteraction with the ligands are Ala32 and Val36 that form H-bonds with the carbonyl oxygen of hexadienoyl-CoA and Gly34and Gly83 that form H-bonds with the thiocarbonyl group ofDAC-CoA, hexadienoyl-CoA and acetoacetyl-CoA. One of thenoticeable differences between AMECH sequence and other mem-bers of the ECH family is the presence of Arg28 which is involved inthe formation of two ionic interactions with the phosphates of thesubstrate, hexadienoyl-CoA. Arg28 is very weakly conserved and isreplaced by Lys in most of the members of this family. Unrelatedamino acids are also present in other sequences at this position.

Multiple sequence alignment of A. marina enoyl-CoA hydratasewith that of two salt sensitive (Zea mays and Glycine max) and two

gainst protein data bank entries using BLAST server: (a) Shewanella oneidensis enoyl-atases bound with (b) dimethylamino cinnamoyl-CoA (DAC-CoA) (PDB id: 1EY3), (c)oA (PDB id: 2DUB) each showing 28% identities with the target sequence. In case ofnd 2DUB, 157 residues were aligned. The alignments were manually modified whereuilding of AMECH.

38 U. Jabeen, A. Salim / Phytochemistry 94 (2013) 36–44

salt tolerant (Arabidopsis thaliana and Shewanella oneidensis) spe-cies show no significant changes at the substrate binding residues.In some places where changes were observed, the amino acidswere similar in nature (Supplemental material; Table S2).

A phylogenetic tree was generated by using the multiple se-quence alignment of A. marina enoyl-CoA hydratase (AMECH) with26 amino acid sequences of enoyl-CoA hydratase/isomerase familymembers whose 3D structures are known. AMECH is closely re-lated to ECH from Shewanella oneidensis and Rhodobacter sphaero-ides (Supplemental material; Fig. S1).

2.2. Model building

Shewanella oneidensis enoyl-CoA hydratase/isomerase (PDB id:3JU1) gave highest homology (30%) with the target sequence(Fig. 1a). Four rat enoyl-CoA hydratases having sequence identitiesof 28% each were also used as templates as they have bound sub-strates (DAC-CoA, hexadienoyl-CoA) and inhibitors (acetocetyl-CoA and octanoyl-CoA). Details of pairwise alignment and tem-plates are outlined in Fig. 1b–e.

2.3. Predicted 3D structures and model validation

Homology model of enoyl-CoA hydratase was built from 321amino acid residues. Homology model of AMECH shows that ithas crotonase-like fold (Haataja et al., 2011) similar to that of en-oyl-CoA hydratases and isomerases belonging to other species.The model consists mainly of a-helical regions (Fig. 2). However,the catalytic region consists of b-strands which are localized nearthe N-terminal region. There are twelve a-helices and five b-sheetsthat are well-defined. Sub-domain 12 a-helical region is sur-rounded by five b-sheets and constitutes typical hot-dog fold. Fiveb-strands form two mixed b-sheets lying almost perpendicular to

Fig. 2. Homology model of Avicennia marina enoyl-CoA hydratase (AMECH): (a) The mohelical regions. Glu106, a conserved catalytic residue and Asp114 which is predicted as cthe crystal coordinates of Shewanella oneidensis enoyl-CoA hydratase, (b) a portion of thecatalytic residues Glu106 and Asp114 along with Gly34, Val36 and Gly83 which partici

one another. The enzyme has a low overall sequence identity buthas been shown to contain a common sequence pattern, the en-oyl-CoA hydratase/isomerase signature. This observation is madewhen the sequences are aligned and folds are compared from thesame species as well as from variable species. In this study,although sequence homology between AMECH and that of thetemplates is low i.e. 28–30%, yet there is little difference in theoverall folds.

The constructed models were examined for validation using dif-ferent criteria. The Ca RMSD values for each homology model aftersuperposition with the respective crystal structures of templatesare shown in Table 1. The stereochemical quality of the predictedmodels was evaluated using PROCHECK through the Ramachan-dran plots that show less than 2% of the residues in disallowedand more than 80% in the most favored region (Table 1). The ProSAenergy plot shows the local model quality by plotting energies as afunction of amino acid sequence position. In case of the homologymodels of AMECH, ProSA analysis reveals Z-scores which lie in thelow energy conformation states (Supplemental material; Fig. S2).These results suggest that the models are of good quality.

2.4. Catalytic mechanism

Various studies have been carried out including the crystallo-graphic analyses to elucidate the mechanism of this family of en-zymes. It has been demonstrated that Glu164 (Glu106 inAMECH) is the only protic amino acid residue completely con-served among the characterized hydratase/isomerase superfamilyand is confirmed as an essential catalytic residue by site-directedmutagenesis, suggesting that it is probably involved in the proton-ation/deprotonation step (Müller-Newen and Stoffel, 1993; Mulleret al., 1995). Crystal structure of rat hydratase showed anotherprotic residue, Glu144, to be also involved in catalysis (Engel

del showing secondary structural elements. The structure mainly consists of alphaatalytic residue are shown in ball and stick configuration. The model was built usingAMECH model comprising residues, Cys31-Phe120. This portion contains predicted

pate in ligand binding.

Table 1Details of the homology models of Avicennia marina enoyl-CoA hydratase (AMECHa).

Models Templates Root mean square deviation (rmsd)

(ÅA0

)

PROCHECK statistics

AMECH without ligand Shewanella oneidensis enoyl-CoA hydratase (3JU1) 1 86.0% most favored11.6% allowed1.8% generouslyallowed0.7% disallowed

AMECH bound with bDAC-CoA (a) Shewanella oneidensis enoyl-CoA hydratase (3JU1) 1 83.9.% most favored12.3% allowed

(b) Rat enoyl-CoA hydratase bound with aDAC-CoA (1EY3) 0.9 3.2% generouslyallowed0.7% disallowed

AMECH bound with hexadienoyl-CoA

(a) Shewanella oneidensis enoyl CoA-hydratase (3JU1) 1 85.3% most favored11.9% allowed

(b) Rat enoyl-CoA hydratase bound with hexadienoyl-CoA(1MJ3)

0.9 1.8% generouslyallowed1.1% disallowed

AMECH bound with acetoacetyl-CoA

(a) Shewanella oneidensis enoyl-CoA hydratase (3JU1) 1 84.9% most favored11.6% allowed

(b) Rat enoyl-CoA hydratase bound with acetoacetyl-CoA(1DUB)

0.9 2.8% generouslyallowed0.7% disallowed

AMECH bound with octanoyl-CoA (a) Shewanella oneidensis enoyl CoA-hydratase (3JU1) 1 80.7% most favored14.0% allowed

(b) Rat enoyl-CoA hydratase bound with octanoyl-CoA (2DUB) 0.9 3.9% generouslyallowed1.4% disallowed

a Target sequence is Avicennia marina enoyl-CoA hydratase (Q9SE41_AVIMR).b DAC-CoA, dimethyl amino cinnamoyl-CoA.

U. Jabeen, A. Salim / Phytochemistry 94 (2013) 36–44 39

et al., 1996). This residue is also completely conserved in the mul-tifunctional proteins that catalyze the hydratase reaction but not inmonofunctional enoyl-CoA isomerases. Glu144 acts as a catalyticbase and activates the water molecule for the attack at C3, whereasGlu164 is the catalytic acid protonating C2 (Engel et al., 1996). Thewater molecule needed for activation is bound between Glu144and Glu164 in the crystal structure of rat enoyl-CoA hydratase (En-gel et al., 1996). It has been shown that mutation of Glu164 totallyabolishes isomerase activity, while the activity in the Glu144 var-iant is lowered only 10-fold. These findings suggest that Glu164is needed for both the hydratase and isomerase activities whereasGlu144 is only essential for the water activation reaction of hydra-tase (Kiema et al., 1999). Bahnson and co-workers (2002) has laterrevised this model suggesting that Glu164 acts both as acid andbase while Gu144 is involved in positioning the bound water mol-ecule. The amino acid residue responsible for the protonation ofthe c-carbon in the isomerase reaction is however not known (Kie-ma et al., 1999). AMECH is a monofunctional enzyme and has onecatalytic glutamic acid (Glu106). The other glutamic acid residue isnot present in AMECH. This position in AMECH is occupied byGly86 which is not involved in catalysis. Interestingly, we observedanother acidic residue, Asp114 which is located at an appropriateposition for hydration of the substrate to occur. This residue wasseen conserved in some ECHs in which Gly is present in place ofGlu at position 86 (Supplemental material; Table S1). In ECHswhere both glutamates are conserved, the Asp is not present at thisposition. Superposition of part of AMECH with Arabidopsis thalianamultifunctional enzyme (PDB id; 2WTB) and Shewanella oneidensisenoyl-CoA hydratase/isomerase (PDB id: 3JU1) shows that Asp114and Glu106 are located within the 4 Å distance of the catalyticwater as in case of Glu157 and Asp165 of 3JU1 and Glu119 andGlu139 of 2WTB (Fig. 3). Asp114 is the only acidic residue otherthan Glu106 which is present within this distance and seems tobe the only candidate for the second catalytic residue (Fig. 3). How-ever, further structural and kinetics data are required to confirmthe exact mechanism.

2.5. Substrate-binding

2.5.1. DAC-CoAThe location of the catalytic site and mode of binding of the

DAC-CoA can be predicted from the sequence and structural sim-ilarities among the members of ECH family. Multiple sequencealignment of members of ECH revealed that Glu86 is not presentin AMECH. In AMECH, Gly is present instead of Glu at position 86while Asp is present in dienoyl-CoA isomerase. Gly86 does notshow any coordination with the substrate either directly or viathe catalytic water presumably because of the small and un-charged side chain. It therefore does not seem to play any rolein catalysis. Glu106 forms hydrogen bond with the catalyticwater which in turn forms hydrogen bond with the substrate(Fig. 4a). Analysis of substrate–enzyme interactions revealedinteractions with Ala32, Val36, Asn62 and Asp114 (Fig. 4b). Likeother ligand bound structures of ECH, the carbonyl group of thesubstrate form H-bonds with Ala32, Gly34, Val36 and Gly83 (En-gel et al., 1996; Kim and Battaile, 2002; Gerlt and Gassman,1993).

We have observed that the thiocarbonyl group of substrate inAMECH has three hydrogen bonding partners (Fig. 4); it acceptshydrogen atoms from Gly34 and Gly83 as in case of rat ECH (Bahn-son et al., 2002). These residues comprise the oxyanion hole for thesubstrate. From the multiple sequence alignment it is evident thatin most of the hydratases, Gly83 (equivalent residue of Gly141 inrat ECH) and Gly34 (equivalent residue of Ala98 in rat ECH) arestrongly and nearly conserved, respectively. Gly141 is an impor-tant residue in catalysis and could potentially contribute in stabi-lizing the transition state formed during the catalytic mechanismof ECH (Bell et al., 2001). We have observed similar interactionsof Gly34 and Gly83 with the substrate. The amide nitrogen ofGly34 is aligned in an optimal geometry to donate its hydrogenbond to the thiocarbonyl group. It has been predicted in earlierstudies that the flanking residues, Gly33 and Asp35 that arestrongly conserved insert the peptide plane in a precise orientation

Fig. 3. Superposition of part of AMECH model with (a) Shewanella oneidensis enoyl-CoA hydratase/isomerase (Pdb id: 3JU1) and (b) Arabidopsis thaliana multifunctionalenzyme (Pdb id; 2WTB). Asp114 and Glu106 are located within the 4 Å distance of the catalytic water as in case of Glu157 and Asp165 of 3JU1 and Glu119 and Glu139 of2WTB.

40 U. Jabeen, A. Salim / Phytochemistry 94 (2013) 36–44

as in case of the template structure (Modis et al., 1998; Bahnsonet al., 2002).

Gly83Pro mutation was carried out to see if any change in theinteraction with substrate takes place. This glycine is conservedin the hydratase family and is one of the residues that interact withthe substrate through hydrogen bonding (Bell et al., 2001). Gly83together with Ala40 comprise the oxyanion hole for stabilizingnegative charge accumulation on the thioester carbonyl group(Feng et al., 2002). In our study, Gly83Pro mutation resulted inthe loss of coordination with the thiocarbonyl group of the sub-strate With Pro83 being relatively exposed at the surface, thehydrogen bonding is lost between Pro83 and the substrate(Fig. 4c and d). However hydrogen bonding interaction withGly34 remains the same. Also the hydrogen bond between

Glu106 and water molecule remains the same in the original mod-el, mutant model as well as in rat ECH. Experimental study carriedout by Feng et al. (2002) showed that in rat enoyl-CoA hydratasethis mutation eliminates the hydrogen bond with the substrateand there was a decrease the kcat for substrate hydration by3400-fold.

2.5.2. Hexadienoyl-CoAIn our predicted model, the carbonyl group of hexadienoyl-CoA

forms H-bonds with Ala32, Gly34, Val36 and Gly83. Phosphategroups of the substrate form two ionic bonds with Arg28(Fig. 5a). This interaction is not found between the rat ECH andthe substrate (Bell et al., 2002). As a result of this interaction, thesurface accessibility of Arg28 is decreased from 61.98 to 50.11 Å

Fig. 4. Homology model of AMECH bound with DAC-CoA (AMECH residues are in bold text): superposition of target and the template structure, 1EY3 showing interaction ofamino acid residues with the substrate involving (a) active site residues and (b) amino acid residues other than the active site residues; G83P mutation: thiolester carbonyl ofDAC-CoA is shown coordinated to amide nitrogen of Gly83 in the (c) original model while this coordination is not observed in case of (d) mutated Pro83.

Fig. 5. AMECH bound with hexadienoyl-CoA (AMECH residues are in bold text): (a) The carbonyl group of hexadienoyl-CoA showing hydrogen bonding interactions and twoionic bonds between phosphate groups and Arg28. (b) Original and (c) mutant (Gly86?Asp) AMECH model superposed with rat ECH.

U. Jabeen, A. Salim / Phytochemistry 94 (2013) 36–44 41

in the homology model. It has been shown that charged aminoacids are optimally positioned so as to maximize electrostatic

interactions in hyperthermophilic proteins (Xiao and Honig,1999). However, the catalytic properties of thermostable enzymes

42 U. Jabeen, A. Salim / Phytochemistry 94 (2013) 36–44

seem to be unaffected by the mutations disrupting the ion pairs.Therefore, extensive ion-pair networks may provide a generalstrategy for manipulating enzyme thermostability of multisubunitenzymes (Vetriani et al., 1998). The involvement of Arg28 inAMECH in the substrate interaction may also provide additionalstability in the enzyme–substrate reaction in case of this salt toler-ant plant.

We have modeled Gly86?Asp mutation in this model. In theoriginal model, Gly86 did not interact with the substrate and thesurrounding water. Therefore, first mutation is done replacingGly with Glu because Glu is better conserved in the ECH familyas a catalytic residue (Bell et al., 2002; Bahnson et al., 2002). Asno interaction is observed between mutated Glu and water (resultsnot shown), another mutation with Asp is carried out. In this mu-tant model, Asp86 shows hydrogen bonding interaction withwater. The same interaction is also observed between Glu144(equivalent to Gly86 in AMECH) and water in case of rat ECH(Fig. 5b and c) (Engel et al., 1996). Asp86 could therefore be a bet-ter replacement for Gly86.

2.6. Inhibitor binding

2.6.1. Acetoacetyl-CoAFig. 6a shows possible mode of binding of acetoacetyl-CoA in

the active site pocket. We observe that the organization of the ac-tive site of AMECH is similar to that of rat liver ECH (Engel et al.,1996). Gly34 and Ala98 in AMECH and rat ECH, respectively, areH-bonded to O1 of the acetoacetyl group (Engel et al., 1996).Gly83 which is strongly conserved in the superfamily of ECH alsocoordinates in the same manner with the acetoacetyl group (Engelet al., 1996). Crystal structure of rat ECH shows that Gly172 makesH-bond with the water molecule (Bahnson et al., 2002). In AMECH,Asp114 (equivalent residue of Gly172) although chemically differ-ent, also show the same interaction. Conserved H-bonded interac-tions are also observed between Gly34, Val36 and Ala32 and AN1and between Ala32 and AN6 of the adenine ring, respectively. InAMECH, some ionic interactions are lost that were observed in caseof rat ECH (Bahnson et al., 2002). In rat ECH, Lys92 (equivalent res-idue of Arg28 in AMECH) and Lys101 (equivalent residue of Ser37in AMECH) form ionic interactions with the phosphate group of theinhibitor (Bahnson et al., 2002). No such interaction is predictedbetween AMECH and the inhibitor.

Fig. 6. Superposed models of AMECH and rat ECH bound with acetoacetyl-CoA and oclocated near the (a) C2, C3 and O3 of the acetoacetyl-CoA and (b) C2 of the octanoyl-Co

2.6.2. Octanoyl-CoAOctanoyl-CoA is a competitive inhibitor that lacks the b-hydro-

xy group. It irreversibly inactivates the enzyme by means of cova-lent linkage with the catalytic active residues of ECH (Wu et al.,2008). Superposition of the catalytic site of the AMECH bound withrat ECH revealed some important features (Fig. 6b). H-bondinginteraction is observed between the octanoyl-CoA and Glu106 asin case of rat ECH (Engel et al., 1998). A hydrogen bond observedbetween Glu144 (equivalent residue at position 86 in AMECH)and water in rat ECH (Engel et al., 1998) is lost in AMECH due tothe replacement of Glu with Gly at this position. The catalytic res-idue closest to the octanoyl-CoA is Glu106. In AMECH, the carbonyloxygen of octanoyl-CoA forms two hydrogen bonds with Gly34 andGly83 as in rat ECH (Engel et al., 1998). AN1 and AN6 atoms of theadenine ring also form two H-bonds with Val36 and Gly34, respec-tively, that are structurally equivalent to Ileu100 and Ala98 of ratECH (Engel et al., 1998). In case of AMECH at position 37, Ser ispresent instead of Lys; there is a loss of H-bond between Ser37and AN3 atom of the adenine ring of the inhibitor.

2.7. Mobile loop

The ‘‘mobile loops’’ of AMECH and rat ECH are structurally dif-ferent (Fig. 7). Mobile loop constitutes residues 55–61 in AMECHequivalent to 113–119 in rat and are different in all sequences ana-lyzed (Supplemental material; Table S1). In AMECH, it comprises,55Phe, Arg, Lys, Glu, Phe, Thr and Met61. Phe55 and Phe59 showcoordination with the hydrophobic binding pocket of the substrate.Mobile loop in ECH plays an important role in controlling the com-parative reaction rate of varying chain lengths. The magnitude andflexibility of the mobile loop reflect an ability to adapt the acyl-binding pocket structure to the substrate specificity profiles ofthe enzymes with regards to acyl chain length and the presenceof conjugated systems (Bahnson et al., 2002; Arent et al., 2010).

A comparison of the position of this loop in the predicted mod-els bound to acetoacetyl-CoA, DAC-CoA and hexadienoyl-CoA showthat it did not superimpose well with an average RMSD 1.98 Å.While the protein folds of predicted structures are similar, differ-ences are observed near the fatty acid binding pocket due to thedifferent nature of the ligand. Phe59 which is structurally equiva-lent to Leu in rat ECH moves into the inter trimer space while thearomatic ring of Phe55 is positioned away from the acyl portion ofsubstrate and therefore shows no interaction. Presence of phenyl

tanoyl-CoA (AMECH residues are in bold text): in AMECH Glu106 and Asp114 areA as in case of rat ECH.

Fig. 7. Mobile Loop (AMECH residues are in bold text): residues in the superposedmodels of AMECH and rat ECH bound with DAC-CoA. Tyr54, Phe55 and Phe59 of themobile loop of AMECH are shown coordinating with the hydrophobic acyl portionof DAC-CoA.

U. Jabeen, A. Salim / Phytochemistry 94 (2013) 36–44 43

ring of Tyr54 at the beginning of the mobile loop increases the sizeof the hydrophobic coordinating pocket to accommodate larger Rgroups of the substrate. ND2 amide group of Asn62 which is pres-ent at the end of the mobile loop is coordinated with ND1 and CDBof substrate at a distance of 3.54 Å and 2.85 Å, respectively. Thisloop is therefore responsible for different specificities for differentsubstrates.

3. Conclusion

A. marina and other enoyl-CoA hydratases share similar foldsdespite low sequence identity (28–30%) between the sequences.The analyses based on substrate/inhibitor–enzyme interactionssupport the significance of amino acid residues implicated in catal-ysis. Although in most cases the interaction between active siteresidues of AMECH and their respective ligands are similar, somenew interactions were predicted while some were lost. Glu106equivalent to Glu164 in rat ECH is proposed as prompting active-site acid in the hydration reaction and is almost completely con-served in sequences of isomerases/hydratases. The other conservedresidue in most ECH sequences is Glu/Asp144 which is equivalentto Gly86 in AMECH. This amino acid is responsible for activatingwater molecule. Gly86 in AMECH show no interaction with thesubstrate or with water. Asp114 however is located within thehydrogen bonding distance of catalytic water molecule. It is there-fore predicted that the Asp114 could be the other catalytic residuein the hydration reaction. The prominent feature of the AMECH isthe presence of structurally different mobile loop having a com-pletely different coordination with the hydrophobic binding pocketof the acyl portion of the substrate.

A. marina is a salt tolerant plant. The regulation mechanism thatwithstands various stresses determines salt tolerance and sensitiv-

ity. The plant cell membrane itself is a basic and potential barrierto a number of external factors. Some proteins involved in lipidmetabolism protect cells against salinity stress by changing the li-pid composition of the plasma membrane. Enoyl-CoA hydratase isone of the enzymes involved in the peroxisomal b-oxidation cycle.With only one structural study of ECH being available from a salttolerant plant (Arent et al., 2010), the homology modeling studyof AMECH has given a structural outline of the interactions be-tween catalytic residues and various ligands. Further study involv-ing the crystal structure of enoyl-CoA hydratase from salt tolerantspecies is needed to demonstrate the significance of amino acids incatalysis as well as their roles in the mechanism of salt tolerance.

4. Experimental

4.1. Multiple sequence alignment and phylogenetic analysis

Primary amino acid sequences of members of the enoyl-CoAhydratase/isomerase family were retrieved from UniProt database(Apweiler et al., 2004). Multiple sequence alignment was doneusing either online BLAST server at www.ncbi.nlm.nih.gov/BLAST(Altschul et al., 1997) or CLUSTAL W Version 2 at www.ebi.ac.uk/Tools/msa/clustalw2/ (Larkin et al., 2007; Goujon et al., 2010). Phy-logenetic trees were constructed for multiple sequence alignmentby cluster algorithm by the online server at www.gene-bee.msu.su/services/hlp/phtree-hlp.html (Brodsky et al., 1992;Chumakov and Yushmanov, 1988; Yushmanov and Chumakov,1988).

4.2. Pairwise sequence analysis and model building

Primary amino acid sequence of AMECH (Q9SE41_AVIMR) wasretrieved from UniProt database (Apweiler et al., 2004). The aminoacid sequence of AMECH was submitted to the BLAST server usingPSI-BLAST algorithms (Altschul et al., 1997) against the sequencesin the PDB (Berman et al., 2000).

All the models were built using MODELLER 9.10 (http://sali-lab.org/modeller/). The program executes with ‘‘automodel’’ envi-ronment having built-in optimized parameters i.e. a combinationof spatial restraints and Charmm energy terms enforcing properstereochemistry to obtain an objective function. The final modelis obtained by optimizing the objective function in Cartesian space.The optimization is carried out by the use of the variable targetfunction method employing methods of conjugate gradients andmolecular dynamics with simulated annealing. Several slightly dif-ferent models can be calculated by varying the initial structure.The variability among these models can be used to estimate the er-rors in the corresponding regions of the fold. Protein structureswere visualized and analyzed using Swiss PDB Viewer (SPDV) 4.1(Gues et al., 1995–2000) and WebLab viewer 4.0 (http://accel-rys.com/products/discovery-studio/visualization-download.php).

4.3. Model assessment

Assessment of the predicted homology model is based on theanalyses of geometry, stereochemistry and energy distributionsin the model. The consistency of the predicted homology modelis assessed using the ENERGY command of the MODELLER. Assess-ment was further conducted by programs PROCHECK (Laskowskiet al., 1993) and ProSa (Sippl, 1993; Wiederstein and Sippl,2007). The variability among the models was compared by super-position of the Ca traces of the model onto the template fromwhich the RMSD value for positional differences between theequivalent atoms are determined.

44 U. Jabeen, A. Salim / Phytochemistry 94 (2013) 36–44

4.4. Prediction of mutation

Mutation prediction was carried out at the catalytic site in theAMECH homology models by using MODELLER 9.10 (http://sali-lab.org/modeller/). Three mutant models were predicted: (a)Gly86 to Asp and (b) Gly83 to Pro mutations in AMECH bound withhexadienoyl-CoA while (c) Gly83 to Pro mutations in AMECHbound with DAC-CoA.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.phytochem.2013.05.018.

References

Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, Z., Miller, W., Lipman, D.J.,1997. Gapped BLAST and PSI-BLAST: a new generation of protein databasesearch programs. Nucleic Acids Res. 25, 3389–3402.

Apweiler, R., Bairoch, A., Wu, C.H., Barker, W.C., Boeckmann, B., Ferro, S., Gasteiger,E., Huang, H., Lopez, R., Magrane, M., Martin, M.J., Natale, D.A., O’Donovan, C.,Redaschi, N., Yeh, L.S., 2004. UniProt: the Universal Protein knowledgebase.Nucleic Acids Res. 32, D115–119.

Arent, S., Christensen, C.E., Pye, V.E., Nørgaard, A., Henriksen, A., 2010. Themultifunctional protein in peroxisomal beta-oxidation: structure andsubstrate specificity of the Arabidopsis thaliana protein MFP2. J. Biol. Chem.285, 24066–24077.

Bahnson, B.J., Anderson, V.E., Petsko, G., 2002. Structural mechanism of enoyl-CoAhydratase: three atoms from single water are added in either an E1cb stepwiseor concerted fashion. Biochemistry 41, 2621–2629.

Bell, A.F., Wu, J., Feng, Y., Tonge, P.J., 2001. Involvement of glycine 141 in substrateactivation by enoyl-CoA hydratase. Biochemistry 40, 1725–1733.

Bell, A.F., Feng, Y., Hofstein, H.A., Parikh, S., Wu, J., Rudolph, M.J., Kisker, C., Whitty,A., Tonge, P.J., 2002. Stereoselectivity of enoyl-CoA hydratase results frompreferential activation of one of two bound substrate conformers. Chem. Biol. 9,1247–1255.

Berman, H.M., Westbrook, J., Feng, Z., Gilliland, G., Bhat, T.N., Weissing, H.,Shindyalov, I.N., Bourne, P.E., 2000. The Protein Data Bank. Nucleic Acids Res.28, 235–242.

Brodsky, L.I., Vasiliev, A.V., Kalaidzidis, Y.L., Osipov, Y.S., Tatuzov, R.L., Feranchuk,S.I., 1992. GeneBee: the program package for biopolymer structure analysis,1992. DIMACS 8, 127–139.

Cantu, D.C., Chen, Y., Lemons, M.L., Reilly, P.J., 2011. ThYme: a database forthioester-active enzymes. Nucleic Acids Res. 39, D342–D346.

Cantu, D.C., Dai, T., Beversdorf, Z.S., Reilly, P.J., 2012. Structural classification andproperties of ketoacyl reductases, hydroxyacyl dehydratases and enoylreductases. Prot. Eng. Des. Select. 25, 803–811.

Chumakov, K.M., Yushmanov, S.V., 1988. The maximum topological similarityprinciple in molecular systematics. Mol. Genet. Microbiol. Virusol. 3, 3–9.

Engel, C.K., Mathieu, M., Zeelen, J.P., Hiltunen, J.K., Wierenga, R.K., 1996. CrystalStructure of Enoyl-coenzyme A (CoA) hydratase at 2.5 resolution: a spiral folddefines the CoA-binding pocket. EMBO J. 15, 5135–5145.

Engel, C.K., Kiema, T.R., Hiltunen, J.K., Wierengar, R.K., 1998. The crystal structure ofenoyl-CoA hydratase complexed with octanoyl-CoA reveals the structuraladaptations required for binding of a long chain fatty acid-CoA molecule. J.Mol. Biol. 275, 847–859.

Feng, Y., Hofstein, H.A., Zwahlen, J., Tonge, P.J., 2002. Effect of mutagenesis on thestereochemistry of enoyl-CoA hydratase. Biochemistry 41, 12883–12890.

Gerlt, J.A., Gassman, P.G., 1993. An explanation for rapid enzyme catalyzed protonabstraction from carbon acids: importance of late trasition states in concertedmechanisms. J. Am. Chem. Soc. 115, 11552–11568.

Goepfert, S., Poirier, Y., 2007. b-Oxidation in fatty acid degradation and beyond.Curr. Opin. Plant Biol. 10, 245–251.

Goujon, M., McWilliam, H., Li, W., Valentin, F., Squizzato, S., Paern, J., Lopez, R., 2010.A new bioinformatics analysis tools framework at EMBL-EBI. Nucleic Acids Res.38, W695–W699.

Gues, N., Peitsch, M., Schwede, T., Diemand, A., 1995–2000. Deep View/Swiss-PdbViewer. V. 3.7.

Haataja, T.J., Koski, M.K., Hiltunen, J.K., Glumoff, T., 2011. Peroxisomalmultifunctional enzyme type 2 from the fruitfly: dehydrogenase andhydratase act as separate entities, as revealed by structure and kinetics.Biochem. J. 435, 771–781.

Hamed, R.B., Batchelar, E.T., Clifton, I.J., Schofield, C.J., 2008. Mechanisms andstructures of crotonase superfamily enzymes. How nature controls enolate andoxyanion reactivity. Cell. Mol. Life Sci. 65, 2507–2527.

Holden, H.M., Benning, M.M., Haller, T., Gerlt, J.A., 2001. The crotonase superfamily:divergently related different reactions involving acyl coenzyme a thioesters.Acc. Chem. Res. 34, 145–157.

Kiema, T.R., Engel, C.K., Schmitz, W., Filppula, S.A., Wierenga, R.K., Hiltunen, J.K.,1999. Mutagenic and enzymological studies of the hydratase and isomeraseactivities of 2-enoyl-CoA hydratase-1. Biochemistry 38, 2991–2999.

Kim, J.-J.P., Battaile, K.P., 2002. Burning fat: the structural basis of fatty acid b-oxidation. Curr. Opin. Struct. Biol. 12, 721–728.

Larkin, M.A., Blackshields, G., Brown, N.P., Chenna, R., McGettigan, P.A., McWilliam,H., Valentin, F., Wallace, I.M., Wilm, A., Lopez, R., Thompson, J.D., Gibson, T.J.,Higgins, D.G., 2007. ClustalW and ClustalX version 2. Bioinformatics 23, 2947–2948.

Laskowski, R.A., McAurthur, M.W., Moss, D.S., Thornton, J.M., 1993. PROCHECK: aprogram to check the stereochemical quality of protein structures. J. Appl.Crystallogr. 26, 283–291.

Modis, Y., Filppula, S.A., Novikov, D.K., Norledge, B., Hiltunen, J.K., Wierenga, R.K.,1998. The crystal structure of dienoyl-CoA isomerase at 1.5 A resolution revealsthe importance of aspartate and glutamate sidechains for catalysis. Structure 6,957–970.

Muller, N.G., Janssen, U., Stoffel, W., 1995. Enoyl-CoA hydratase and isomerase forma superfamily with a common active-site glutamate residue. Eur. J. Biochem.228, 68–73.

Müller-Newen, G., Stoffel, W., 1993. Site-directed mutagenesis of putative active-site amino acid residues of 3,2-trans-enoyl-CoA isomerase, conserved withinthe low-homology isomerase/hydratase enzyme family. Biochemistry 32,11405–11412.

Parani, M., Lakshmi, M., Senthikumar, P., Parida, A., 1999. Molecular cloning andnucleotide sequence of 2-enoyl-COA hydratase from the mangrove speciesAvicenia marina. The Electronic Plant Gene Register. Plant Physiol. 121, 1053–1055.

Reumann, S., Ma, C., Lemke, S., Babujee, L., 2004. AraPerox. A database of putativeArabidopsis proteins from plant peroxisomes. Plant Physiol. 136, 2587–2608.

Sippl, M.J., 1993. Recognition of errors in three-dimensional structures of proteins.Proteins 17, 355–362.

The UniProt Consortium, 2011. Ongoing and future developments at the UniversalProtein Resource. Nucleic Acids Res. 39, D214–D219.

Vetriani, C., Maeder, D.L., Tolliday, N., Yip, K.S., Stillman, T.J., Britton, K.L., Ricem,D.W., Klumpm, H.H., Robb, F.T., 1998. Protein thermostability above 100 �C: akey role for ionic interactions. Proc. Natl. Acad. Sci. USA 95, 12300–12305.

Wiederstein, M., Sippl, M.J., 2007. ProSA-web: interactive web service for therecognition of errors in three-dimensional structures of proteins. Nucleic AcidsRes. 35 (Web Server issue), W407–W410.

Wu, L., Lin, S., Li, D., 2008. Comparative inhibition studies of enoyl-CoA hydratase 1and enoyl-CoA hydratase 2 in long-chain fatty acid oxidation. Org. Lett. 10,3355–3358.

Xiao, L., Honig, B., 1999. Electrostatic contributions to the stability ofhyperthermophilic proteins. J. Mol. Biol. 289, 1435–1444.

Yushmanov, S.V., Chumakov, K.M., 1988. Algorithms of the maximum topologicalsimilarity phylogenetic trees construction. Mol. Genet. Microbiol. Virusol. 3, 9–15.