Enhanced antifungal and insect α-amylase inhibitory activities of Alpha-TvD1, a peptide variant of Tephrosia villosa defensin (TvD1) generated through in vitro mutagenesis

Page 1

EAt

Sa

b

c

a

ARR2AA

KTWAT�P

1

pwTdflTbti[

Kp

0d

Peptides 33 (2012) 220–229

Contents lists available at SciVerse ScienceDirect

Peptides

j our na l ho me p age : www.elsev ier .com/ locate /pept ides

nhanced antifungal and insect �-amylase inhibitory activities oflpha-TvD1, a peptide variant of Tephrosia villosa defensin (TvD1) generated

hrough in vitro mutagenesis

. Vijayana,1, J. Imanib,1, K. Tanneeruc, L. Guruprasadc,∗, K.H. Kogelb,∗, P.B. Kirti a,∗

Department of Plant Sciences, University of Hyderabad, Hyderabad 500046, IndiaResearch Center for BioSystems, Land Use and Nutrition (IFZ Giessen), Justus Liebig University, Giessen, GermanySchool of Chemistry, University of Hyderabad, Hyderabad 500046, India

r t i c l e i n f o

rticle history:eceived 11 November 2011eceived in revised form7 December 2011ccepted 29 December 2011vailable online 8 January 2012

eywords:ephrosia villosaild type TvD1

lpha-TvD1

a b s t r a c t

TvD1 is a small, cationic, and highly stable defensin from the weedy legume, Tephrosia villosa with demon-strated in vitro antifungal activity. We show here peptide modifications in TvD1 that lead to enhancedantifungal activities. Three peptide variants, S32R, D37R, and Alpha-TvD1 (-G-M-T-R-T-) with variationsin and around the �2–�3 loop region that imposes the two �-strands, �2 and �3 were generated throughin vitro mutagenesis. Alpha-TvD1 exhibited enhanced antifungal activity against the fungal pathogens,Fusarium culmorum and Fusarium oxysporum with respective IC50 values of 2.5 �M and 3.0 �M, whencompared to S32R (<5.0 �M and >5.0 �M), D37R (5.5 �M and 4.5 �M), and the wild type TvD1 (6.5 �M).Because of the enhanced antifungal activity, this variant peptide was characterized further. Growth ofF. culmorum in the presence of Alpha-TvD1 showed deformities in hyphal walls and nuclear damage.With respect to the plant pathogenic bacterium, Pseudomonas syringae pv. tomato strain DC3000, both

enebrio molitor-amylase inhibitionrotein–protein docking

Alpha-TvD1 and the wild type TvD1 showed comparable antibacterial activity. Both wild type TvD1 andAlpha-TvD1 displayed inhibitory activity against the �-amylase of the mealworm beetle, Tenebrio molitor(TMA) with the latter showing enhanced activity. The human salivary as well as barley �-amylase activ-ities were not inhibited even at concentrations of up to 50 �M, which has been predicted to be due todifferences in the pocket size and the size of the interacting loops. Present study shows that the variant

anced

Alpha-TvD1 exhibits enh

. Introduction

Plant defensins are small (5–10 kDa), highly stable, cysteine-richeptides, which can act as broad-spectrum antimicrobial peptidesith potent antifungal activity against certain fungal species [44].

he characteristic feature of plant defensin is the presence of fourisulfide bridges between cysteine residues in the mature peptideorming the cysteine stabilized motif, but some plant defensinsike PhD1 from Petunia hybrida exhibit five disulfide brides [18].hey carry one �-helix and three anti-parallel �-strands stabilizedy four disulfide bridges folding into a CS�� configuration with

hree extra loops and net positive charge [5]. They were implicatedn host defense and detrimental to fungal pathogens and bacteria29,30,39,42].

∗ Corresponding authors.E-mail addresses: [email protected] (L. Guruprasad),

[email protected] (K.H. Kogel),[email protected] (P.B. Kirti).1 Contributed equally to the investigation.

196-9781/$ – see front matter © 2012 Elsevier Inc. All rights reserved.oi:10.1016/j.peptides.2011.12.020

antifungal as well as insect �-amylase inhibitory activity.© 2012 Elsevier Inc. All rights reserved.

The mode of action of plant defensins is still not preciselyunderstood. They have been found to interact with componentsin the plasma membrane of higher fungi resulting in membranepermeabilization [41]. Their antifungal activities could be due tothe presence of glycosyl ceramide residues in the fungal cell wall,whose interaction with the defensin peptides results in membranedamage followed by pore formation, ion leakage, and cell deatheventually due to ionic imbalance [40,43].

In addition to antimicrobial activities, plant defensins arealso reported to exhibit plant developmental [1,44], antiprolifer-ative [45], HIV-1 reverse transcriptase inhibitor [28], proteinaseinhibitory [46] and �-amylase inhibitory [3,22,31] activities andalso inhibit plant parasites [47]. Plant defensins, such as VrD1(Vigna radiata defensin) and VuD1 (V. unguiculata defensin), exhibitinsect �-amylase inhibitory activity, which was detrimental toinsect pests like Tenebrio molitor and Z. subfasciatus, respectively[21,22,31]. The defensins with �-amylase inhibitory activity inter-

fere in insect digestion leading to deprivation of energy fromstarch [6]. In VrD1, the amino acids such as -G-M-T-R-T- presentin the loop-3 region have been shown to be associated with thegut �-amylase inhibitory activity in T. molitor, whereas VrD2 has

Page 2

tides

taa�ibi

hiedTfiiBat[

wd(lsdptg([fiGomhtwbaalta

vapfaawtbwaT

2

2

vt

S. Vijayan et al. / Pep

he amino acid sequence -D-D-F-R- in the corresponding regionnd lacked the inhibitory activity [21]. However, VuD1 has samemino acid sequence as in VrD2, but exhibited Z. subfasciatus gut-amylase inhibitory activity and the interaction between the pos-

tively charged amino acids outside the loop-3 of the peptide haseen reportedly associated with the gut �-amylase inhibitory activ-

ty [31].In an earlier study, we characterized the defensin TvD1 from the

oary peas Tephrosia villosa, which showed constitutive expressionn all the tissues, such as seed, stem, young and older leaves, flow-rs and roots; this is in contrast to some constitutively expressedefensins of Arabidopsis, which did not show expression in roots.vD1 has been shown to be detrimental to several plant pathogeniclamentous fungal species, such as Fusarium oxysporum, F. monil-

forme, Rhizoctonia solani, Phytophthora parasitica, Curvularia sp.,otrytis cinerea, and Alternaria helianthi. The homology modelings well as pair-wise alignment of the amino acid sequence showedhat TvD1 shared 91% homology with V. radiata defensin 2 (VrD2)44].

In general, plant defensins possess three anti-parallel �-strandsith an �-helix connected with different extra loops. In someefensins like VrD1 (V. radiata, mungbean defensin 1) and VrD2V. radiata defensin 2), extra loops were notified as loop-1 or L1,oop-2 or L2, and loop-3 or L3 [21]. Previous reports showed thatubstitution of non-polar or uncharged amino acids through siteirected mutagenesis in the �2−�3 loop (or loop-3) region of theeptide greatly affected its antifungal activity [38]. Consistent withhis, changing the positive charge potential at �2−�3 loop by sin-le amino acid substitution in MsDef1 (Medicago sativa defensin)R38Q) [38] or Raphanus sativus antifungal peptide, RsAFP2 (V39R)14] strongly affected their activity against some phytopathogenicungi. Moreover, substitution of the four amino acids (-D-D-F-R-)n the loop-3 of VrD2 with the sequence of five amino acids (--M-T-R-T-) present in the loop-3 of VrD1 resulted in inhibitionf �-amylase of the stored grain pest, T. molitor [21]. MsDef1 is aorphogenic defensin, which perturbs hyphal growth by inducing

yperbranching, whereas MtDef4 is a non-morphogenic defensinhat does not induce hyperbranching. The MsDef1-�4 variant, inhich the �-core of MsDef1 was replaced by that of MtDef4,

ehaved in a way similar to MtDef4 by exhibiting non-morphogenicntifungal activity [35]. Hence, substitution of uncharged aminocids with positively charged or hydrophobic ones at the �2−�3oop of TvD1 through site directed mutagenesis appears to offerhe possibility of significantly increasing its antimicrobial as wells anti-insect activity.

With this information in the background, we generated threeariants of the peptide, TvD1 through in vitro mutagenesis in andround �2−�3 loop, and studied their inhibitory activity on somelant fungal pathogens. Subsequently, the most effective peptiderom the three variants, Alpha-TvD1 was studied for its inhibitoryctivity against the plant bacterial pathogen Pseudomonas syringaend �-amylase inhibitory activity in the insect pest, T. molitor alongith the wild type TvD1. By structure modeling, the interaction of

he wild type TvD1 and Alpha-TvD1 with T. molitor �-amylase haseen analyzed using protein–protein docking studies. Furthermore,e conducted a comparative study of binding of Alpha-TvD1 to �-

mylases of T. molitor, H. sapiens, and barley in this investigation.hese observations are reported in this communication.

. Materials and methods

.1. Construction of expression vectors

To produce the different recombinant peptides of wild type andariant TvD1 peptides, an expression vector was constructed. Ini-ially, a PCR based-two step DNA synthesis was used to generate

33 (2012) 220–229 221

different TvD1 DNA sequences [20]. The DNA sequence that wassubmitted earlier to the NCBI GenBank under the accession numberAY907349 [44] was used in the amplification of different fragments.The gene was synthesized without the coding sequence for the sig-nal peptide. In the first step, the wild type TvD1 gene was amplifiedby using the following primers: TvD1-A-F-5′-AAG ACA TGC GAGAAC CTG GCA GAT ACG TAT AGG GGT CCA TGC TTC ACC ACT GGAAGC TGT GAC GAT CAT TGC AAG-3′ and TvD1-B-R-5′-ACA TCT TTTAGT ACA CCA GCA GCG AAA ATC GTC CCT GCA CCT TCC ACT CAG TAAGTG CTC CTT ATT CTT GCA ATG ATC G-3′ by using Pfu-DNA poly-merase. Similarly, the Alpha-TvD1 (mutant) was synthesized usingthe following primers: TvD1-A-F-5′-AAG ACA TGC GAG AAC CTGGCA GAT ACG TAT AGG GGT CCA TGC TTC ACC ACT GGA AGC TGTGAC GAT CAT TGC AAG-3′ and TvD1-C-R-5′-ACA TCT TTT AGT ACACCA GCA AGT ACG AGT CAT ACC CCT GCA CCT TCC ACT CAG TAAGTG CTC CTT ATT CTT GCA ATG ATC G-3′. To produce the completefragment, a second step DNA synthesis was performed using thefollowing primers: forward (BamHI-DP-TvD1-F-5′-C GGA TCC GACCCG AAG ACA TGC GAG AAC CTG GCA-3′) and reverse (XhoI – Stop-TvD1-R-5′-GTG CTC GAG TTA ACA TCT TTT AGT ACA CCA-3′). Thetwo restriction sites for BamHI and XhoI are underlined; the twocodons for the acid-sensitive dipeptide (Asp–Pro) that lies imme-diately upstream of the TvD1 coding sequence are represented inbold capital letters, and the stop codon is represented in italics. ThePCR-amplified DNA fragment was cloned into the expression vectorpET32a(+) (Novagen, USA) using BamHI and XhoI restriction sites.For the second PCR, all the primers used were the same as in the caseof wild type TvD1 and Alpha-TvD1. Likewise, the other two mutantswere generated by using the wild type fragment as template. Themutants S32R and D37R were generated by using the same strategyas mentioned above. For the first PCR, the forward primers, (S32R)F-5′-CAC TTA CTG AGA GGA AGG TGC A-3′ and (D37R)-F-5′-TGC AGGGAC GGT TTT CGC TGC T-3′ and reverse primer XhoI – Stop-TvD1-R-5′-GTG CTC GAG TTA ACA TCT TTT AGT ACA CCA-3′ were used togenerate the two variants. All the amplified fragments were doubledigested with the respective restriction enzymes and cloned at thecorresponding sites in the expression vector, pET32a(+). The clonedfragments were confirmed by sequencing.

2.2. Expression and purification of recombinant proteins

E. coli strain-BL21(DE3) cells were transformed withthe above mentioned expression plasmids through electro-transformation and the recombinant colonies were confirmed byPCR. For peptide production, cells were grown at 37 ◦C in Luriabroth (LB) with 100 �g/ml ampicillin. When the cultures reachedan optical density (OD600) of 0.6–0.8, the heterologous expres-sion was induced by the addition of 1 mM IPTG. Six hours afterinduction, cells were harvested by centrifugation at 5000 × g for5 min at 4 ◦C. Cell disruption was performed by high pressurehomogenization (French press, at a pressure of 40 kpsi; 35 ml) inbinding buffer A (8 M urea, 10 mM imidazole, 0.1 M monobasicsodium phosphate, 0.01 M Tris/HCl, pH 8.0), 10 ml buffer pergram cell wet weight. Insoluble contaminants were eliminated bycentrifugation at 12,000 × g for 15 min at 4 ◦C. The use of pET32a(+)allowed the expression of the fusion protein with the N-terminalfusion partners, Thioredoxin-His-S-tag linked to the target peptideby an acid-sensitive dipeptide. Purification was done by IMAC(immobilized metal ion affinity chromatography) using the His-Tagof the fusion protein. The cell lysate was loaded on an equilibratedNi-NTA column and washed with 10 column volumes (CV) ofbuffer B (8 M urea, 50 mM imidazole, 0.1 M monobasic sodium

phosphate, and 0.01 Tris/HCl, pH 6.3) and the fusion protein elutedby the addition of 1 ml of elution buffer C (8 M urea, 250 mMimidazole, 0.1 M sodium hydrogen phosphate, 0.01 M Tris/HCl,pH 4.5). The elution has been repeated till the entire quantity of

Page 3

2 tides

tAbtsAnwflNc3fw

2

am([eittcftaCoc

aaoat

2

tgprwwo(ts

2

APkplc5tm

22 S. Vijayan et al. / Pep

he recombinant peptide was recovered from the fraction. Thesp–Pro bond of the acid-sensitive dipeptide has been shown toe extremely labile when incubated at pH ∼1.5. For release of thearget peptides from the fusion partners, the pH in the proteinolution was adjusted to 1.4 by drop-wise addition of 6 M HCl.fter incubation for 16–18 h at 55 ◦C at pH 1.4, the solution waseutralized by the addition of 10 M NaOH, and the fusion partneras purified from the target peptide by a second IMAC-step. Theow-through containing the target peptides with an additional-terminal proline was concentrated approximately 5× to a finaloncentration of 2 �g/ml by using centrifugal filter devices (2 ml,

kDa cut off) using the Vivaspin columns (Vivascience Ltd, UK)ollowing manufacturer’s instructions. The recombinant peptidesere stored in buffer 10 mM Tris/HCl at 4 ◦C until use.

.3. Antifungal activity assays

Inhibitory activity of recombinant wild type TvD1, S32R, D37R,nd Alpha-TvD1 peptides on conidial germination was tested byicro-spectrophotometry with the fungal pathogens F. oxysporum

DSM62059) and Fusarium culmorum, kindly gifted by J. Chellowski7,8]. Briefly, 10 �l of the purified peptides were diluted to differ-nt concentrations, such as 0.5 �M, 5.0 �M, and 10 �M, pipettednto the wells of a 96-well microtiter plate containing 90 �l of theest fungal spore suspension (∼2.5 × 104 spores/ml) in potato dex-rose broth (PDB) and incubated at 28 ◦C. The effect of the peptideoncentration was tested for its antifungal activity in triplicatesor each pathogen. Conidial germination and growth were moni-ored microscopically, and the OD was measured 30 min and 48 hfter inoculation, respectively, using a Microplate Reader at 595 nm.ontrols were tested identically except that the test peptides weremitted. Value of the growth inhibition lower than 10% was notonsidered as significant.

Growth inhibition is defined as the ratio of the correctedbsorbance at 595 nm of the control minus the correctedbsorbance of the test sample, divided by the corrected absorbancef the control. Corrected absorbance is defined as the absorbancet 48 h minus that at 30 min. IC50 is defined as the protein concen-ration at which 50% inhibition was reached [36].

.4. Assay for chitin deposition at hyphal tips

This assay was conducted to observe chitin deposition at hyphalips of F. culmorum [27]. Pre-germination and incubation of fun-al cultures with the wild type and Alpha-TvD1 peptide wereerformed in 96-well microplates. Following incubation with theecombinant peptides, a final concentration 1 mM of Congo Red dyeas added to distinguish between growing hyphal tips and thoseith inhibited growth. Fluorescence was examined under the flu-

rescence microscope with wavelength between 543 and 560 nmAxioplan-2 Imaging Microscope, Zeiss, Germany). Growing hyphalips interact poorly with the dye, while those with inhibited growthtain deeply.

.5. Antibacterial activity

Antibacterial activity of recombinant peptides, wild type andlpha-TvD1 was examined with the plant pathogenic bacterium. syringae pv. tomato strain DC3000 carrying the vector pVSP61,indly gifted by Jane Parker, Germany [16] in sterile 96-welllates (microtiter plates) in a final volume of 100 �l as fol-

ows. Aliquots (50 �l) of a suspension containing bacteria at

oncentrations of 106 CFU/ml in culture medium were added to0 �l of double-distilled water containing the peptides in serialwo-fold dilution in LB medium. Inhibition in growth was deter-

ined by measuring the absorbance at 595 nm with BioTek EL808

33 (2012) 220–229

Microplate Reader (BioTek Instruments Inc.) after incubation for48 h at 28 ◦C. The antibacterial activity was expressed as IC50[24].

2.6. Membrane permeabilization assay

The conidia of F. culmorum at a concentration of ∼4 × 104 wereincubated in 1/4-strength potato dextrose broth for 14–18 h at28 ◦C. Subsequently, the germinated conidia were incubated withthe recombinant peptides, wild type and Alpha-TvD1 at a concen-tration of 2.5 and 5 �M, respectively, for 3 h with gentle agitation.Fluorescence in conidia and hyphae was visualized 10 min afteradding Sytox Green dye (0.5 �M) using a fluorescence microscope(Axioplan-2 Imaging, Zeiss, Germany) with excitation and emissionwavelengths of 488 and 538 nm.

2.7. Insect larval gut preparation

Larvae of T. molitor in third instar were obtained commercially(Fauna Topics, http://www.futtertiere.de) and larvae of similar sizewere kept on ice. The entire larval gut region was dissected out andhomogenized in 0.9% (w/v) sodium chloride (NaCl) solution. Sam-ples were then centrifuged for 10 min at 12,000 × g at 4 ◦C and theprecipitate was removed. The supernatant was saved as enzymefrom the insect and used in the assay with the recombinant pep-tides.

The insect �-amylase inhibitory activity of the recombinantpeptides was determined using the Bernfeld method [2]. Differentconcentrations of wild type TvD1 and Alpha-TvD1 were assayedagainst the purified enzyme (25 �g/ml) in a buffer containing50 mM sodium acetate (pH 6.5), 5 mM calcium chloride, and 10 mMNaCl. Soluble starch (1%, w/v) was used as the substrate. The recom-binant peptides and the enzymes were pre-incubated for 20 min at37 ◦C before the addition of 300 �l of 3,5-dinitrosalicylic acid (DNS).The reaction was stopped by heat inactivation of the enzyme for10 min at 100 ◦C and the absorbance was measured at 530 nm. One�-amylase unit was defined as the quantity of the enzyme thatincreased the absorbance at 530 nm by 0.1 OD during 20 min assay.Each assay was performed in triplicate and distilled water was usedas negative control.

2.8. Protein structure modeling

The wild type TvD1 comprising the sequence motif “37D-D-F-R40” was mutated to “G-M-T-R-T”. This 48 amino acid mutantAlpha-TvD1 sequence was subjected to NCBI BLAST search againstprotein structure data bank (PDB) to identify homologues as tem-plate for comparative structure modeling. The homology modelof the mutant peptide was constructed using MODELER softwareimplemented in the Discovery studio-2.1. MODELER is a homol-ogy/comparative modeling program for constructing the 3D modelof a protein structure from its amino acid sequence [34]. It is basedon the sequence alignment between the query protein and tem-plate structure. The program automatically constructs a model forall non-hydrogen atoms in the protein by the satisfaction of spatialrestraints that include non-homologous loops and energy opti-mization of the final model.

The defensin homology model built was validated using theRamachandran plot and 3D profiles [23,33]. The 3D structuresuperposition of �-amylases from T. molitor (PDB ID: 1TMQ),human (PDB ID: 3OLD), and barley (PDB ID: 1HT6) was per-

formed using MAPSCI server [19]. The structural overlay identifiescommon substructures in a set of proteins that assess the relat-edness between structures with the evolutionary history andfunction.

Page 4

S. Vijayan et al. / Peptides 33 (2012) 220–229 223

Table 1Inhibitory effect of the wild type TvD1, Alpha-TvD1, S32R, and D37R peptides on the fungal pathogens F. culmorum and F. oxysporum with IC50 values (given are mean valuesand standard deviation from three replicates of two independent experiments).

Peptide Percentage growth inhibition

F. culmorum F. oxysporum

0.5 �M 5 �M 10 �M IC50 0.5 �M 5 �M 10 �M IC50

Wild type TvD1 2.7 ± 1.1 34.7 ± 3.1 93.9 ± 0.7 6.5 4.3 ± 1.2 37.5 ± 1.1 90.4 ± 0.8 6.5Alpha-TvD1 17.6 ± 0.6 75.0 ± 1.1 95.3 ± 0.7 2.5 21.3 ± 4.1 88.3 ± 1.1 94.7 ± 1.1 3.0S32R 5.4 ± 1.8 50.7 ± 2.0 92.2 ± 2.4 <5.0 14.9 ± 1.1 51.1 ± 2.1 90.7 ± 0.8 >5.0

C at thet

2

csTesirbtTetcrtFusoeFtTv

3

3

wpatwTdht

TIo

C

D37R 4.1 ± 1.1 56.3 ± 4.1 93.2 ± 2.1

omparison of statistical significance of mean values using Student’s t-test shows threatments are significantly different at 5% level.

.9. Protein–protein docking

The 3D model structure of Alpha-TvD1 was docked into therystal structure of T. molitor larval �-amylase (TMA) crystaltructure (PDB ID: 1TMQ) using the ZDOCK software [10,11,12,25].he PDB ID: 1TMQ corresponds to the crystal structure of a het-rodimer complex of TMA with RBI. For protein–protein dockingtudies, we removed the RBI (Ragi bifunctional �-amylase/trypsinnhibitor) from the crystal structure complex and used TMA as aeceptor and the mutant Alpha-TvD1 as a ligand. ZDOCK is a rigid-ody docking program that requires minimal information abouthe binding site and is targeted at initial-stage unbound docking.he program uses individual protein structures determined byxperimental or computational methods as inputs and predictshe structure of a number of protein complexes (i.e., the top 2000omplexes). The ZDOCK search scheme exhaustively searches allotational and translational space for the ligand protein relativeo the receptor protein, which is fixed at its starting orientation.or the rotational search, evenly distributed Euler angles weresed [10]. This angle set is equivalent to a uniformly distributedet of points on a projective sphere, which ensures that minimalrientations are required to cover the entire rotational space. Forach rotation, the algorithm rapidly scans translational space usingFT based on a 3D grid with step size of 1.2 A. The conformation ofhe complex with high ZDOCK score was used in further analysis.he docking results of the protein–protein interactions wereisualized on Discovery studio-2.1.

. Results

.1. Generation of mutants and prokaryotic expression

Wild type TvD1 and the mutants Alpha-TvD1, S32R, and D37Rere generated by PCR and cloned in the expression vectorET32a(+). In all the mutants, the signal peptide was removed andn acid sensitive dipeptide (Asp–Pro) was introduced betweenhe ∼16 kDa Trx-His-S-tag and the target peptide. Hence, the tagould be removed from the mature peptide after purification.

he generated clones were sequenced and confirmed for theesired mutations in the TvD1 peptide. The mutant Alpha-TvD1as the amino acids -G-M-T-R-T- in the �2–�3/loop-3 region ofhe peptide removing the original -D-D-F-R- sequence in wild type

able 2nhibition of the growth of the bacterium P. syringae by wild type TvD1 and Alpha-TvD1 wf two independent experiments).

Peptide Percentage growth inhibition

P. syringae

1 �M 2 �M 3 �M

Wild type TvD1 16.4 ± 2.1 23.3 ± 1.5 31.5

Alpha-TvD1 16.5 ± 4.2 26.4 ± 3.6 34.8

omparison of statistical significance of mean values using Student’s t-test shows that th

5.5 14.5 ± 0.1 48.9 ± 0.6 91.5 ± 0.1 4.5

differences between the wild type TvD1 and the peptide variants in 0.5 and 5.0 �M

TvD1. The variants S32R and D37R were obtained by site-directedmutagenesis to generate variants with enhanced positive netcharge in this region. The vectors with the gene constructs werethen transformed into E. coli BL21 (DE3) cells for sequence analysesand expression. The recombinant peptides were expressed in E. coliand purified by IMAC for use in subsequent antifungal bioassays(Supplementary Fig. S1).

3.2. Alpha-TvD1 showed highest antimicrobial activity inbioassays

Alpha-TvD1, at 5 �M concentration, inhibited the growth ofconidia of F. culmorum and F. oxysporum by 75% and 88.3%, respec-tively (Fig. 1A and B; Table 1), while S32R, D37R, and the wildtype TvD1 showed comparatively reduced antifungal activity with50.7%, 56.3%, and 34.7% inhibition against F. culmorum, and 51.1%,48.9%, and 37.5% inhibition against F. oxysporum, respectively.When tested at lower concentrations (0.5 �M), Alpha-TvD1 dis-played 17.6% and 21.3% inhibition, respectively, against F. culmorumand F. oxysporum, while S32R, D37R, and wild type TvD1 showedonly 5.4%, 4.1%, and 2.7% inhibition against F. culmorum and 14.9%,14.5%, and 4.3%, respectively, against F. oxysporum. At 10 �M con-centration, all the peptides were inhibitory to both the fungalpathogens with inhibition as high as 90%.

The calculated IC50 values for the peptides, such as wild typeTvD1, Alpha-TvD1, S32R, and D37R, were 6.5 �M, 2.5 �M, <5.0 �M,and 5.5 �M, respectively, against the test fungus, F. culmorum and6.5 �M, 3.0 �M, <5.0 �M, and 4.5 �M, respectively, against F. oxys-porum. Statistical analyses between the wild type and differentvariants of the peptide using Student’s t-test (significance at 5%level) showed that the antifungal activity of Alpha-TvD1, S32R, andD37R was significantly higher than the wild type TvD1 (P > 0.05)at 0.5 �M and 5 �M against both the test fungal pathogens. At10 �M concentration (P < 0.05), there was no significant differencebetween the wild type and peptide variants. Similarly, there wasno significant difference between Alpha-TvD1 and wild type TvD1in controlling the growth of the bacterial pathogen, P. syringae pv.

tomato strain DC3000. Overall, the inhibitory concentration washigher than that was required for fungal inhibition (Table 2). At1 �M, the assay showed around 16% inhibition for both the wildtype TvD1 and Alpha-TvD1, and more than 50% inhibition was

ith IC50 values (given are mean values and standard deviation from three replicates

5 �M 10 �M IC50

± 1.7 52.7 ± 1.6 73.1 ± 2.0 <5± 3.1 54.3 ± 3.1 76.2 ± 3.5 <5

ere is no significant difference between the wild type TvD1 and Alpha-TvD1.

Page 5

224 S. Vijayan et al. / Peptides 33 (2012) 220–229

F , S32Rw

rb

3

eti

ig. 1. Conidial germination assay with the peptides of wild type TvD1, Alpha-TvD1ere taken 48 h post treatment. Bars represent 30 �m.

eached at 5 �M concentration with calculated IC50 < 5.0 �M foroth the peptides.

.3. TvD1 peptides inhibited fungal cell wall synthesis

In the Congo Red staining assay, the tips of the hyphae thatmerged out of the germinating conidia of F. culmorum exposedo wild type TvD1 and Alpha-TvD1 at 5 �M concentration stainedntensely with the dye elucidating bleb formation that suggested

, and D37R using Fusarium culmorum (A) and Fusarium oxysporum (B). Photographs

weakening of the fungal cell wall synthesis in treatment with boththe peptides (Fig. 2A).

3.4. Fungal membrane showed permeabilization in the presenceof TvD1 peptides

Membrane permeabilization assays were performed usingthe filamentous fungus, F. culmorum with the wild type TvD1and Alpha-TvD1 peptides and observations were made using

Page 6

S. Vijayan et al. / Peptides 33 (2012) 220–229 225

Fig. 2. (A) Microscopic analysis of chitin deposition in hyphae of germinating F. culmorum spores in the presence/absence of wild type TvD1 and Alpha-TvD1 (5 �M peptideeach) after 18 hpi in PDB (Potato Dextrose Broth). (B) Microscopic analysis of membrane permeabilization with the fungus F. culmorum in the presence of wild type TvD1a ‘a’ repi

flnflTgft

3o

aawa5ftSsg

3.6. Protein structure modeling

The multiple sequence alignment between the target protein(Alpha-TvD1), the template structure (PDB ID: 2GL1) and several

Table 3Percent inhibition of the insect Tenebrio molitor gut �-amylase by wild type andAlpha-TvD1 (given are mean values and standard deviation from three replicates).

Protein concentration(�M)

Percentage inhibition

Alpha-TvD1(IC50 < 17 �M)

Wild type TvD1(IC50 < 23 �M)

50 85.5 ± 2.3 78.1 ± 2.725 74.8 ± 3.1 60.0 ± 1.5

nd Alpha-TvD1 peptide at 2.5 �M and 5 �M concentrations using SYTOX-Green. (ndicates the nuclei. Bars represent 30 �m.)

uorescence microscopy. At peptide concentration of 2.5 �M,uclei of the germinating spores got stained and were visible underuorescence microscope for both the wild type TvD1 and Alpha-vD1 (Fig. 2B). At a higher concentration (5 �M), the nucleus of theerminating spores got disorganized completely and got disruptedor both the peptides as evidenced by the fluorescence throughouthe cell (Fig. 2B).

.5. ˛-Amylase inhibitory activity of the peptides using the larvaef the mealworm beetle, T. molitor

Wild type and Alpha-TvD1 peptides were tested for the gut �-mylase inhibitory activity of T. molitor. At 10 �M, there was 16%nd 32.5% inhibition of the �-amylase activity, respectively, forild type and Alpha-TvD1 (Table 3). The inhibition of �-amylase

ctivity was found to be dose dependent. At the concentration of0 �M, nearly complete inhibition of amylase activity was reachedor both the peptides. The IC50 was <23 �M and <17 �M for wild

ype and Alpha-TvD1, respectively (Supplementary Fig. S2, Table 3).tatistical analysis between the wild type TvD1 and Alpha-TvD1howed that Alpha-TvD1 exhibited significant (P > 0.05) T. molitorut �-amylase inhibitory activity at the tested concentrations of 10

resents bright field microscopy, ‘b’ represents fluorescent field microscopy. Arrow

and 25 �M. Both the peptides did not exhibit any inhibitory effectwhen tested against human salivary �-amylase as well as barley�-amylase even at concentrations of 50 �M recombinant peptide(data not shown).

10 32.5 ± 1.2 16.0 ± 1.7

Comparison of statistical significance of mean values using Student’s t-test showsthat the differences between the wild type TvD1 and the peptide variant, Alpha-TvD1in 10 and 25 �M treatments are significantly different at 5% level.

Page 7

226 S. Vijayan et al. / Peptides 33 (2012) 220–229

F ervedw Alpha

ltBchairrddiogs

3

aAurtoc

ir1c�tb

ig. 3. The multiple sequence alignment of defensins from legume plants. The consith connected lines. The region of mutation in TvD1 (AAX86993.1) can be seen in

egume defensins is shown in Fig. 3. The PDB ID: 2GL1 correspond tohe NMR solution structure of VrD2 [21] that was identified by theLAST searches. This family of proteins comprise eight conservedysteines. The modeled structure of Alpha-TvD1 comprised one �-elix and three anti-parallel �-strands connected by three loopsnd stabilized by four disulfide bonds (Supplementary Fig. S3). Val-dation of the protein structure using PROCHECK showed that 92.7%esidues were in the most favored regions and the remaining 7.3%esidues were in the additionally allowed regions of Ramachan-ran plot. Also dihedral angles, volume, and accessible surface areaata were within the expected values (data not shown). The val-

dation of protein structure using 3D profiles indicated that 100%f the residues had an averaged 3D–1D score > 0.2 indicating veryood correlation between the sequence and its corresponding 3Dtructure.

.7. Protein–protein docking

In order to understand the mechanism of �-amylase inhibitiont molecular level, we carried out docking studies; 3D structure oflpha-TvD1 was docked into the structure of T. molitor �-amylasesing ZDOCK. The active site of �-amylase comprised three acidicesidues (Asp-185, Glu-222, and Asp-287). In the crystal struc-ure of PDB ID: 1TMQ, RBI fitted into the substrate-binding sitef TMA with extensive nonbonding interactions stabilizing theomplex.

To validate ZDOCK methodology for this protein system, wenitially docked RBI into the TMA. The score was 1508.240 andeproduced the exact crystal structure conformation as in PDB ID:TMQ (data not shown). The docking of Alpha-TvD1 with TMA indi-

ated that in the pose with the highest ZDOCK score (1498.527),2−�3 loop between the �-strands 2 and 3 was positioned in

he substrate binding site of TMA (Fig. 4A). The Alpha-TvD1 haseen stabilized by the formation of hydrogen bonds, ionic and

cysteines are indicated in yellow background and the disulfide pattern is indicated-TvD1. 2GL1 is the NMR solution structure of Vigna radiata Defensin 2 (VrD2).

hydrophobic interactions with TMA protein. In the loop-1 of Alpha-TvD1, the side-chain OH of Thr-16 formed hydrogen bonds with theside-chain oxygen of the Asn-331 of TMA (2.52 A). The side chainOH of Thr-17 formed a hydrogen bond with main chain oxygenof the Thr-291 (2.173 A). The main chain NH of Gly-12 formed ahydrogen bond with the main-chain carbonyl oxygen of Val-151(2.72 A). The Phe-15 of the loop-1 showed the pi–pi stacking inter-actions with the Trp-57 of the TMA strengthening the binding. Themutated residue, Arg-40 with side-chain NH groups in the loop-3of Alpha-TvD1 formed hydrogen bonds with the side-chain oxy-gen atoms of the catalytic triad residue Asp-185 (2.59 A). The otherresidue, Met-38 in the mutated motif was surrounded by Ala-186,Val-223, and Ile-224 side-chains of the TMA. The ionic interactionsformed between the Alpha-TvD1 and TMA further strengthen thebinding. The Arg-40 of Alpha-TvD1 showed ionic interactions withAsp-185 and Asp-287. Similarly, Arg-11 had ionic interactions withthe Glu-135 of TMA (Supplementary Fig.S4).

We also observed in the modeling that the side-chain NH of theTrp-43 on the �-strand-1 forms a hydrogen bond with the side-chain OH of the Tyr-139 (2.79 A) and the side-chain OH of Ser-19 onthe helix forms another hydrogen bond with the side-chain oxygenof the Asp-332 (3.21 A). The above hydrophilic and hydrophobicinteractions as shown in Fig. 4B appear to be responsible for thebinding of Alpha-TvD1 to the TMA.

We compared the 3D structures of TMA (PDB ID: 1TMQ) withthe 3D structures of human �-amylase (PDB ID: 3OLD) and barley�-amylase (PDB ID: 1HT6), respectively, by structure superposi-tion using MAPSCI. The structure superposition indicated that thethree structures superimpose with a main-chain RMSD (root meansquare deviation) of 0.85 A (Fig. 4C). The side-chains of the three

catalytic acidic residues exactly superimposed in all structures. Thecomparison of the substrate binding site indicated major differ-ences in the pocket size around the active site, due to the variationsin their loops size (Supplementary Fig. S5).

Page 8

S. Vijayan et al. / Peptides

Fig. 4. (A) The docking of mutant Alpha-TvD1 with T. molitor �-amylase after remov-ing the RBI (Ragi bifunctional �-amylase/trypsin inhibitor). The side chains of threecatalytic acidic residues in �-amylase are indicated. (B) The interactions betweenAlpha-TvD1 (secondary structure with ribbon representation and side chains in balland stick model are shown in orange) and T. molitor �-amylase (side chains areshown in green stick model). The hydrogen bonds between side-chains of AlphaTvD1 and T. molitor �-amylase are shown in dashed lines (pink). (C) Structural super-position of �-amylases from T. molitor (blue), barley (green) and human (magenta).The location of the docked mutant Alpha-TvD1 (orange) is shown. The interactionbetween Alpha-TvD1 and the TMA shows that it is greater compared to the othertwo amylases from barley and humans. (For interpretation of the references to colorin this figure legend, the reader is referred to the web version of the article.)

33 (2012) 220–229 227

4. Discussion

TvD1 is a potent antifungal peptide [44] and hence, an appro-priate candidate peptide for studies on mutagenesis aimed atimproving its potency against various soil borne filamentous fungalpathogens. In order to enhance its antifungal property, mutationalanalysis was carried out in the present study by modifying someamino acid residues in the �2−�3 loop that connect the �-strand-2and �-strand-3 of TvD1. Possible enhancement in the antimicrobialactivity would help in generating transgenic crops with enhancedresistance against microbes and other pests. Mutational analysisof the plant defensin was first carried out in RsAFP2 [14] and thishas led to the determination of the specific amino acids responsiblefor its antifungal activity. Similarly, some amino residues responsi-ble for antifungal activity in MsDef1 and MsDef2 were determinedthrough mutational analysis [38].

In the present investigation, we substituted the amino acidresidue at 32 by replacing the serine with arginine (S32R), asparticacid with arginine at 37 (D37R) and looked for enhanced antifun-gal activity. Similarly, it was also shown that changing the positivecharge potential by replacing the existing amino acids at 9 and39 into arginine in RsAFP2 has increased its antifungal activityby two fold [14]. In VrD2, insertion of five amino acids (-G-M-T-R-T-) by replacing the existing 4 amino acid residues (-D-D-F-R-)enhanced the insect �-amylase inhibitory activity, when comparedto VrD1 [21]. The substitution of the �-core motif of MsDef1 withthat of MtDef4 enhanced the antifungal activity of MsDef1 morethan two-fold [35]. Hence, some of these replacements/insertionswere carried out in the wild type TvD1 to develop mutants withpossible enhancement in antimicrobial and insecticidal activity.

Through PCR based mutagenesis, mutants such as S32R andD37R were synthesized [20], whereas Alpha-TvD1 and wild typeTvD1 were synthesized by using overlapping primers. In all thesepeptides, the signal peptide was removed. Since the peptide ispH stable, an acid sensitive dipeptide D-P (Aspartic acid–Proline)was introduced between the Thioredoxin-His-S-tag and the maturepeptide in the bacterial expression vector pET32a. The inclusion ofdipeptide did not show any drawbacks in the activity of the pep-tide [21,48]. And, also after acid digestion, proline will be the firstamino acid residue in the mature peptide and it did not have anysuppressive role in the peptide [48].

From the spore germination assays, it could be clearly seen thatAlpha-TvD1 displayed enhanced activity against F. culmorum and F.oxysporum, when compared to the other mutants, S32R and D37R,which possessed comparatively better activity than the wild typeTvD1. Hence, the relative antifungal activity of TvD1 by argininesubstitution was reportedly dependent on the nature of the testfungus. The same was observed in the case of RsAFP2 confirmingthat the composition and structure of the putative receptor on thehyphal membrane between different fungal species controlled theantifungal activity [14]. Similarly, VrD1 also possessed the sameamino acids as in Alpha-TvD1 with reported inhibitory activityagainst the fungal pathogens, F. oxysporum, Pyricularia oryzae, R.solani, and Trichophyton rubrum along with bruchid larvae [9,13].The antifungal activity of Alpha-TvD1 is comparable with that ofVrD1 against the fungal pathogen F. oxysporum (IC50 = 5.4 �g/ml)and it was slightly more than the value for the wild type TvD1(IC50 = 6.5 �M). Hence, �2−�3 loop may be the major determinantfor antifungal activity particularly against the fungus F. oxysporumfor both the peptides. The IC50 was 90.7 �g/ml against the fungusR. solani for VrD1, but the corresponding value was much lower forTvD1 (IC50 = 38 �g/ml) [9,44]. Hence, it can be concluded that the

active site and inhibitory activities of the peptides vary with thefungal pathogens. From the antifungal assays, it could be seen thatAlpha-TvD1 was more potent against the fungal pathogens, F. cul-morum and F. oxysporum and hence, it was characterized further.

Page 9

2 tides

Aatar

bnRbfasoii

euotctmmcTcmcpfldbpabebonspp[

pdipswhaiosptr3Tted3

28 S. Vijayan et al. / Pep

lpha-TvD1 did not exhibit any enhancement in the antibacterialctivity when compared to the activity that is already present inhe wild type peptide. In addition, the required concentration forntibacterial activity of the mutant peptide was higher than thatequired against fungal pathogens.

Congo Red displays a strong affinity for �-glucans and hence, itinds to chitin in the fungal cell wall. Actively growing hyphae doot have chitin deposits at the tips and hence, show little Congoed staining. Sometimes, the growing hyphal tips would appear asalloon-like structures due to the impeded growth and thinning ofungal cell walls. It was shown that in the presence of wild typend Alpha-TvD1, the growing hyphal tips appeared as balloon-liketructures with hyperbranching in the fungus F. culmorum. Similarbservations were made on the fungus Magnoporthe grisea withntense staining at the growing tip due to the deposition of chitinn the presence of the Aspergillus giganteus antifungal peptide [26].

To determine the mechanism of action, we have studied theffect of TvD1 on the plasma membranes of F. culmorum hyphaesing the dye, SYTOX Green. In general, microbial cell death canccur in the presence of antimicrobial peptides in two ways: (i)hrough disruption of the plasma membrane leading to leakage ofytoplasmic contents or (ii) through interaction with intracellularargets [4]. It was observed that TvD1 permeabilized the plasma

embrane of susceptible fungal hyphae in a dose-dependentanner that correlated with growth inhibition. At sub-inhibitory

oncentrations (2.5 �M) of wild type TvD1 as well as Alpha-vD1, membrane permeabilization was detected. However, theytoplasm of permeabilized hyphae appeared normal under theicroscope with very little DNA fragmentation. At higher inhibitory

oncentration (5 �M) of both peptides, permeabilized hyphae dis-layed significant cytoplasmic granulation with the SYTOX Greenuorescence pattern getting diffused across the cell indicating theisruption of nuclei. This suggests that peptide-induced mem-rane permeabilization is required for growth inhibition. Althoughermeabilization of membranes has been reported for manyntimicrobial peptides, the mechanisms of permeabilization coulde very different and the mechanism in many instances remainedlusive [43]. Various models have been suggested including thearrel-stave model, which involves the formation of pores byligomerization of amphipathic peptides to form hydrophilic chan-els. In the toroidal pore model, the pore includes lipid head groupstabilizing the high positive charge of the peptides; and the car-et model hypothesizes layering of the plasma membrane withositively charged protein causing detergent-like destabilization4].

Protein structure modeling studies have shown that the TvD1ossessed one �-helix and three �-strands similar to other plantefensins [44]. Validation of the protein structure using PROCHECK

ndicated a good stereochemical quality and stability and the 3Drofiles indicated a good correlation between 1D sequence and 3Dtructure of the model built. The docking studies of Alpha-TvD1ith T. molitor �-amylase indicated that in the structure with theighest ZDOCK score, several nonbonding interactions are medi-ted via some loops in TMA. The highest scoring docking posendicates that the location of Alpha-TvD1 is very similar to thatf RBI in the complex and that these inhibitors are located on theubstrate binding site of TMA acting as a lid to the active site. Therotein–protein interaction studies between TMA and test pep-ides (wild type and Alpha-TvD1) are in accordance with a previouseport, which suggested an interaction between the L3 or loop-

of V. radiata defensin VrD1 with the substrate binding site ofMA [22]. The addition of five amino acids, -G-M-T-R-T- from VrD1

o Alpha-TvD1 replacing the existing four amino acids -D-D-F-R-nhanced insect T. molitor �-amylase inhibitory activity as in theefensin VrD2. It was also reported that the amino acids in the loop-

of VrD2 determine the insect �-amylase inhibitory activity [21].

33 (2012) 220–229

Similarly, the mutant Alpha-TvD1 with the five amino acid inser-tion in the �2−�3 loop exhibited enhanced �-amylase inhibitoryactivity against the insect T. molitor, when compared to the wildtype TvD1. In contrast to VrD2, which did not exhibit any insect�-amylase inhibitory effect due to the presence of -D-D-F-R- in the�2−�3 loop, the wild type TvD1 has interestingly same amino acidresidues in the �2−�3 loop and exhibited �-amylase inhibitoryactivity. Hence, from these results, it could be clearly envisagedthat the amino acids in the �2−�3 loop are not exclusively respon-sible for the insect �-amylase inhibitory activity, but other aminoacids in the peptides are also involved in its activity as in VuD1, adefensin from V. ungiculata [31].

From the experimental results, we also observed that Alpha-TvD1 has very low or no inhibitory activity against human andbarley �-amylases even at 50 �M concentration. In order to find outthe reason for the specificity of Alpha-TvD1 for TMA activity inhibi-tion, we have compared the 3D structures of TMA (PDB ID: 1TMQ)with the 3D structures of human salivary �-amylase (PDB ID: 3OLD)and barley �-amylase (PDB ID: 1HT6), respectively, by structuresuperposition using MAPSCI. The structure superposition indicatedthat the three structures superimpose with low RMSD and the side-chains of the three catalytic acidic residues exactly superimpose inall the structures. The comparison of the substrate binding site indi-cated major differences in the pocket size, due to the variations intheir loop size. These major structural differences are, therefore,responsible for the differential activity of Alpha-TvD1 toward TMAat a molecular level identified through in silico studies.

TvD1 did not possess any inhibitory effect on human as well asbarley enzymes as efficiently as in the case of the insect �-amylase.Pelegrini et al. [31] used molecular modeling for identifying themechanism that VuD1 uses to present such a characteristic fea-ture. TvD1 has an asparagine residue at position 40 in the �2−�3loop; the presence of a positively charged residue in this positionseemed to be important for the inhibition of mammal �-amylasesin VuD1. The fact that VuD1 does not possess a charged residuein this region might explain why it did not inhibit PPA (porcinepancreas �-amylase), while it inhibited insect amylases. Further-more, it was observed that the longer loops of PPA could impede thebinding of VuD1 [31]. Interestingly, VuD1 did not show any inhi-bition against the filamentous fungal pathogens, F. oxysporum andAspergillus flavus [31].

A similar observation was also made in a study carried outwith wheat �-amylase inhibitors against the PPA, which showedthat steric impediment might also be responsible for the speci-ficity pattern [15]. Hence, it appears that the wild type and mutantAlpha-TvD1 appear to be safe for human consumption as well asplant transformation and over-expression studies. Previous stud-ies have reported the production of transgenic plants expressing�-amylase inhibitors isolated from the common bean Phaseolus vul-garis in pea [32], chickpea [17], and V. radiata [37] to enhance plantresistance toward the insect pests. Hence, the variant Alpha-TvD1could be useful in pest management programs as an alternativestrategy against the storage pest particularly the insect larvaeT. molitor along with significant antifungal activity and appearto be promising candidate genes for deployment in transgeniccrops.

5. Conclusion

Through in vitro mutagenesis in the �2–�3 loop, different pep-tide variants of T. villosa defensin (TvD1) were generated, such as

S32R, D37R, and Alpha-TvD1. Alpha-TvD1 showed enhanced anti-fungal activity against the plant fungal pathogens, F. oxysporumand F. culmorum and hence, we proceeded with further charac-terization of this peptide. Also, Alpha-TvD1 displayed enhanced

Page 10

tides

ipap

A

vm(

A

t

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[Ha-DEF1, a sunflower defensin, induces cell death in Orobanche parasitic plants.

S. Vijayan et al. / Pep

nsect �-amylase activity. Hence, Alpha-TvD1 appears to be aotent candidate gene for engineering crop plants particularlygainst the soil borne filamentous fungi as well as some insectests.

cknowledgments

The current investigations have been supported by the exchangeisit grant of the Department of Science and Technology, Govern-ent of India, New Delhi and German Academic Exchange Program

DAAD) given to PBK and KHK.

ppendix A. Supplementary data

Supplementary data associated with this article can be found, inhe online version, at doi:10.1016/j.peptides.2011.12.020.

eferences

[1] Allen A, Snyder AK, Preuss M, Nielsen EE, Shah DM, Smith TJ. Plant defensinsand virally encoded fungal toxin KP4 inhibit plant root growth. Planta2007;227:331–9.

[2] Bernfeld P. Amylases a and b. Methods in enzymology, vol. 1. New York: Aca-demic Press, Inc.; 1995. pp. 149–58.

[3] Bloch C, Richardson M. A new family of small (5 kDa) protein inhibitors ofinsect alpha-amylases from seeds of sorghum (Sorghum bicolor (L.) Moench)have sequence homologies with wheat gamma-purothionins. FEBS Lett1991;279:101–4.

[4] Brogden KA. Antimicrobial peptides: pore formers or metabolic inhibitors inbacteria. Nat Rev Microbiol 2005;3:238–50.

[5] Broekaert WF, Terras FRG, Cammune BPA, Osborn RW. Plant defensins: novelantimicrobial peptides components of the host defense systems. Plant Physiol1995;108:1353–8.

[6] Carvalho ADO, Gomes VM. Plant defensins-prospects for the biological func-tions and biotechnological properties. Peptides 2009;30:1007–20.

[7] Chelkowski Perkowski J. Mycotoxins in cereal grain. Part 15. Distribu-tion of deoxynivalenol in naturally contaminated kernels. Mycotoxin Res1992;8:27–30.

[8] Chelkowski J, Zawadzki M, Zajkowski P, Logrieco A, Bottalico A. Moniliforminproduction by Fusarium species. Mycotoxin Res 1990;6:41–5.

[9] Chen JJ, Chen GH, Hsu HC, Li SS, Chen CS. Cloning and functional expres-sion of a mungbean defensin VrD1 in Pichia pastoris. J Agric Food Chem2004;52:2256–61.

10] Chen R, Weng Z. Docking unbound proteins using shape complementarity,desolvation, and electrostatics. Proteins 2002;47:281–94.

11] Chen R, Li L, Weng Z. ZDock: an initial-stage protein docking algorithm. Proteins2003;52:80–7.

12] Chen R, Weng Z. A novel shape complementarity scoring function for protein-protein docking. Proteins 2003;51:397–408.

13] Chen KC, Lin CY, Kuan CC, Sung HY, Chen CS. A novel defensin encoded bya mungbean cDNA exhibits insecticidal activity against bruchid. J Agric FoodChem 2002;50:7258–63.

14] De Samblanx GW, Goderis IJ, Thevissen K, Raemaekers R, Fant F, BorremansF, et al. Mutational analysis of a plant defensin from Radish (Raphanus sativusL.) reveals two adjacent sites important for antifungal activity. J Biol Chem1997;272:1171–9.

15] Franco OL, Rigden DJ, Melo FR, Bloch Jr C, Silva CP, Grossi-de-Saı MF.Activity of wheat �-amylase inhibitors towards bruchid �-amylases andstructural explanation of observed specificities. Eur J Biochem 2000;267:1466–73.

16] Garcıa AV, Blanvillain-Baufumeı S, Huibers RP, Wiermer M, Li G, Gobbato E,et al. Balanced nuclear and cytoplasmic activities of EDS1 are required for acomplete plant innate immune response. PLoS Pathogens 2010;6:e1000970.

17] Ignacimuthu S, Prakash S. Agrobacterium-mediated transformation of chickpeawith �-amylase inhibitor gene for insect resistance. J Biosci 2006;31:339–45.

18] Janssen BJC, Schirra HJ, Lay FT, Anderson MA, Craik DJ. Structure of Petuniahybrida Defensin 1, a novel plant defensin with five disulfide bonds. Biochem-

istry 2003;42:8214–22.

19] Ilinkin I, Ye J, Janardan R. Multiple structure alignment and consensus identifi-cation for proteins. BMC Bioinformatics 2010;11:71.

20] Ke SH, Madison EL. Rapid and efficient site-directed mutagenesis by single-tube‘megaprimer’ PCR method. Nucleic Acids Res 1997;25:3371–2.

[

33 (2012) 220–229 229

21] Lin KF, Lee TR, Tsai PH, Hsu MP, Chen CS, Lyu PC. Structure-based pro-tein engineering for �-amylase inhibitory activity of plant defensin. Proteins2007;68:530–40.

22] Liu YJ, Cheng CS, Lai SM, Hsu MP, Chen CS, Lyu PC. Solution structure of theplant defensin VrD1 from mungbean and its possible role in insecticidal activityagainst bruchids. Proteins 2006;63:777–86.

23] Luthy R, Bowie JU, Eisenberg D. Assessment of protein models with three-dimensional profiles. Nature 1992;356:83–5.

24] Makovitzki A, Viterbo A, Brotman Y, Chet I, Shai Y. Inhibition of fungal and bacte-rial plant pathogens in vitro and in plants with ultrashort cationic lipopeptides.Appl Environ Microbiol 2007;73:6629–36.

25] Mintseris J, Pierce B, Wiehe K, Anderson R, Chen R, Weng Z. Integrating statisti-cal pair potentials into protein complex prediction. Proteins 2007;69:511–20.

26] Moreno AB, Martinez Del Pozo A, San Segundo B. Biotechnologically relevantenzymes and proteins. Antifungal mechanism of the Aspergillus giganteus AFPagainst the rice blast fungus Magnaporthe grisea. Appl Microbiol Biotechnol2006;72:883–95.

27] Moreno M, Segura A, Garcia-Olmedo F. Pseudothionin-St1, a potato peptideactive against potato pathogens. Eur J Biochem 1994;223:135–9.

28] Ngai PHK, Ng TB. Phaseococcin, an antifungal protein with antiproliferativeand anti-HIV-1 reverse transcriptase activities from small scarlet runner beans.Biochem Cell Biol 2005;83:212–20.

29] Osborn RW, De Samblanx GW, Thevissen K, Goderis I, Torrekens S, Van Leuven F,et al. Isolation and characterisation of plant defensins from seeds of Asteraceae,Fabaceae, Hippocastanaceae and Saxifragaceae. FEBS Lett 1995;368:257–62.

30] Pelegrini PB, Franco OL. Plant �-thionins: novel insights on the mechanisms ofactions of a multi-functional class of defense proteins. Int J Biochem Cell Biol2005;37:2239–53.

31] Pelegrini PB, Lay FT, Murad AM, Anderson MA, Franco OL. Novel insights on themechanism of action of �-amylase inhibitors from the plant defensin family.Proteins 2008;73:719–29.

32] Prescott VE, Campbell PM, Moore A, Mattes J, Rothenberf ME, Foster PS, et al.Transgenic expression of bean �-amylase inhibitor in peas results in alteredstructure and immunogenicity. J Agric Food Chem 2005;53:9023–30.

33] Ramachandran GN, Ramakrishnan C, Sasisekharan V. Stereochemistry ofpolypeptide chain configurations. J Mol Biol 1963;7:95–9.

34] Sali A, Blundell TL. Comparative protein modelling by satisfaction of spatialrestraints. J Mol Biol 1993;234:779–815.

35] Sagaram US, Pandurangi R, Kaur J, Smith TJ, Shah DM. Structure-activity deter-minants in antifungal plant defensins MsDef1 and MtDef4 with different modesof action against Fusarium graminearum. PLoS One 2011;6:e18550.

36] Song X, Wang J, Wu F, Li X, Teng M, Gong W. cDNA cloning, functional expressionand antifungal activities of a dimeric plant defensin SPE10 from Pachyrrhizuserosus seeds. Plant Mol Biol 2005;57:13–20.

37] Sonia MS, Saini R, Singh RP, Jaiwal PK. Agrobacterium tumefaciens mediatedtransfer of Phaseolus vulgaris �-amylase inhibitor-1 gene into mungbean Vignaradiata. Plant Cell Rep 2007;26:187–98.

38] Spelbrink RG, Dilmac N, Allen A, Smith TJ, Shah DM. Differential antifungal andcalcium channel-blocking activity among structurally related plant defensins.Plant Physiol 2004;135:2055–67.

39] Terras FRG, Eggermont K, Kovaleva V, Raikhel NV, Osborn RW, Kester A, et al.Small cysteine-rich antifungal proteins from radish: their role in host defence.Plant Cell 1995;7:573–88.

40] Thevissen K, Waranecke DC, Francois IE, Leipelt M, Heinz E, Ott C, et al. Defensinsfrom insects and plant interact with fungal glucosyl ceramides. J Biol Chem2004;279:3900–5.

41] Thomma BP, Cammue BP, Thevissen K. Mode of action of plant defensins sug-gests therapeutic potential. Curr Drug Targets Infect Disord 2003;3:1–8.

42] Thomma BPHJ, Camme BPA, Thevissen K. Plant defensins. Planta2002;216:193–202.

43] Van der Weerden NL, Lay FT, Anderson MA. The plant defensin NaD1, enters thecytoplasm of Fusarium oxysporum hyphae. J Biol Chem 2008;283:14445–52.

44] Vijayan S, Guruprasad, Kirti PB. Prokaryotic expression of a constitutivelyexpressed Tephrosia villosa defensin and its potent antifungal activity. ApplMicrobiol Biotechnol 2008;80:1023–32.

45] Wang HX, Ng TB. Isolation and characterization of an antifungal peptide withantiproliferative activity from seeds of Phaseolus vulgaris cv ‘Spotted Bean’. ApplMicrobiol Biotechnol 2007;74:125–30.

46] Wijaya R, Neumann GM, Condron R, Hughes AB, Polya GM. Defense proteinsfrom seed of Cassia fistula include a lipid transfer protein homologue and aprotease inhibitory plant defensin. Plant Sci 2000;159:243–55.

47] Zélicourt AD, Letousey P, Thoiron S, Campion C, Simoneau P, Elmorjani K, et al.

Planta 2007;226:591–600.48] Zorko M, Japelj B, Hafner-Bratkovic I, Jerala R. Expression, purification and

structural studies of a short antimicrobial peptide. Biochim et Biophys Acta2009;1788:314–23.