Aa

AMa

Rb

c

d

a

ARRAA

KPAACPSS

1

cctraoi

CeaMMtG

vH

0h

Peptides 45 (2013) 78–84

Contents lists available at SciVerse ScienceDirect

Peptides

j ourna l ho me pa g e: www.elsev ier .com/ locate /pept ides

ntimicrobial activity of pleurocidin is retained in Plc-2, a C-terminal 12-aminocid fragment

ndre L.A. Souzaa, Paola Díaz-Dellavallec, Andrea Cabrerac, Patricia Larrañagac,arco Dalla-Rizzac, Salvatore G. De-Simonea,b,d,∗

National Institute of Science and Technology on Innovation on Neglected Diseases (INCT-IDN)/Center for Technological Development in Health (CDTS), FIOCRUZ,io de Janeiro, RJ, BrazilDepartment of Molecular and Cell Biology, Federal Fluminense University, Niterói, RJ, BrazilLaboratory of Proteins, Biotechnology Unit, INIA Las Brujas, Canelones, UruguayProtein and Peptides Laboratory Oswaldo Cruz Institute, FIOCRUZ, Rio de Janeiro, RJ, Brazil

r t i c l e i n f o

rticle history:eceived 12 June 2012eceived in revised form 26 March 2013ccepted 27 March 2013vailable online 17 April 2013

eywords:

a b s t r a c t

An analysis of a series of five peptides composed of various portions of the pleurocidin (Plc) sequenceidentified a l2-amino acid fragment from the C-terminus of Plc, designated Plc-2, as the smallest fragmentthat retained a antimicrobial activity comparable to that of the parent compound. MIC tests in vitrowith low-ionic-strength medium showed that Plc-2 has potent activity against Pseudomonas aeruginosa,Escherichia coli and Staphylococcus aureus but not against Enterococcus faecalis. The antifungal activityof the synthetic peptides against phytopathogenic fungi, such as Fusarium oxysporum, Colletotrichum sp.,Aspergillus niger and Alternaria sp., also identified Plc-2 as a biologically active peptide. Microscopy studies

leurocidinntimicrobial peptidesntifungal activityationic peptidesore-formingynthetic peptides

of fluorescently stained fungi treated with Plc-2 demonstrated that cytoplasmic and nuclear membraneswere compromised in all strains of phytopathogenic fungi tested. Together, these results identify Plc-2 asa potential antimicrobial agent with similar properties to its parent compound, pleurocidin. In addition,it demonstrated that the KHVGKAALTHYL residues are critical for the antimicrobial activity described forpleurocidin.

mall active sequence

. Introduction

Eukaryotic antimicrobial peptides (AMPs) are a promisinglass of molecular tools that have tremendous potential to belinically relevant due to their activity against a wide spec-rum of microorganisms. Over the last few decades, intenseesearch has been focused on their biosynthesis and the mech-

nisms of their activity [6,16]. They are important componentsf the natural defenses of most living organisms and have beensolated from a wide variety of animals, plants and bacteria

Abbreviations: AMPs, antimicrobial peptides; BHI, brain heart infusion broth;DI, cellular damage index; CFU, colony-forming units; CFW, calcofluor white; EB,thidium bromide; HPLC, high performance liquid chromatography; MALDI, matrix-ssociated laser desorption ionization; MBC, minimal bactericidal concentration;FC, minimal fungicidal concentration; MIC, minimal inhibitory concentration;HB, Muller-Hillton; PDA, potato dextrose agar; PDB, potato dextrose broth; TBS,

rypcase soy broth; PBS, phosphate buffered saline; Plc, pleurocidin; SG, Sytoxreen; TBS, trypcase soy broth.∗ Corresponding author at: National Institute of Science and Technology on Inno-ation on Neglected Diseases (INCT-IDN)/Center for Technological Development inealth (CDTS), FIOCRUZ, Brazil. Tel.: +55 21 38658183; +55 21 38658240.

E-mail address: dsimone@ioc.fiocruz.br (S.G. De-Simone).

196-9781/$ – see front matter © 2013 Elsevier Inc. All rights reserved.ttp://dx.doi.org/10.1016/j.peptides.2013.03.030

© 2013 Elsevier Inc. All rights reserved.

(http://www.bbem.univ.triest.it/-tossi/search.htm). Many of theseAMPs are genetically encoded, while others are secondary metabo-lites. Yet, both avenues generate peptides that display positive,negative or neutral net charge in physiological pH.

The positive charge of the vast majority of AMPs allows theinitial interaction with the negatively charged lipopolysacharides(LPS) in the outer membrane, and later, with negatively chargedphospholipids in the inner membrane. However, there are excep-tions such as the magainins and cecropins, which do not showany interaction with chiral centers in the membrane since the Land D enantiomers of these peptides have similar antimicrobialactivity. In contrast, apidaecin kills bacteria by a mechanism thatinvolves stereospecificity. The selectivity of these peptides for bac-terial membranes over eukaryotic membranes has been ascribedto the lack of cholesterol and cationic lipids in the bacterial mem-branes and the limited amounts of anionic lipids in the eukaryoticmembranes.

The skin and skin mucus of several fish species have been shownto contain AMPs. A number of those show sequence similarity to

AMPs isolated from other organisms. Pleurocidin (Plc), a �-helicalcationic peptide isolated from the skin-secreted mucous of the win-ter flounder Pleuronectes americanus [6,7] is predicted as a 60 or68 residue precursor prepropolypeptide that undergoes proteolytic

dx.doi.org/10.1016/j.peptides.2013.03.030http://www.sciencedirect.com/science/journal/01969781http://www.elsevier.com/locate/peptidesmailto:dsimone@ioc.fiocruz.brhttp://www.bbem.univ.triest.it/-tossi/search.htmdx.doi.org/10.1016/j.peptides.2013.03.030

A.L.A. Souza et al. / Peptides 45 (2013) 78–84 79

Table 1Peptides, sequences and properties.

Pep Amino acids sequ ence MM r Mma Charge Hn pIPlc GWGSFFKKAAHVGKHVGKAALTHYL 2716.22 0.830 +4 4.43 10.02Plc-1

KLKHVGLHVGKLK 1456.84 0.015 +4 −0.06 10.48Plc-2

KHVGKAALTHYL 1224.43 0.160 +2 0.66 9.70

Plc-3

KKAAHVGKHVGXAAKK 1529.43 1.383 +5 5.68 10.60

Plc-4

KKAAHVGKHVGKAAL 1514.84 1.283 +4 4.37 10.60Plc-5

GWGSFFKKAAHVG 1391.60 0.727 +2 2.78 10.00

a ], as d

cppiehsamomoFdTatIpo

ataoetagtac(saEiocpts

2

2

C

ionization (MALDI) mass spectroscopy (Kratos Kompact MALDI,Manchester, UK). The amino acid sequences of the peptides are

The molecular moment (Mm) was calculated by the method of Eisenberg et al. [11

leavage of its amino-terminal signal and carboxy-terminal anionicropiece to form the active peptide [8,9]. The resulting activeeptide consists of 25 amino acid residues and possesses a net pos-

tive charge at physiological pH. The activity has been extensivelyxplored and has shown a broad spectrum of antimicrobial andemolytic activity [14,17,40,42]. Its codifying gene is similar to thetructural gene of the mammalian antimicrobial peptide PR-39 [15]nd its primary structure showed some homology with the der-aseptins from frogs [28] and ceratotoxins (medflies) [24] classes

f AMPs [31,33]. The presence of common motifs in PR-39, der-aseptins and ceratotoxins, suggests that the antimicrobial activity

f pleurocidin is due to only a portion of the pleurocidin sequence.urthermore, to make Plc useful as a therapeutic drug requireselineating the feature responsible for its activities as an AMP.here are examples of peptide and protein fragments retaining thentimicrobial activity of the parent molecule and, in some instance,heir activities even exceed that of a close relative molecule [5,21].n addition, the N- and C-terminal regions of antimicrobial peptideslay an important role in the organism-specific interaction processr pore formation in plasma membrane [4,23].

AMPs are not only an interest against human pathogens, theyre also excellent candidates for serving as biologically based pes-icides for agriculture. To capitalize on their potential use, studiesre in progress to elucidate the mechanisms of their action with anngoing search to identify the particular residues and structurallements responsible. Such endeavors can lead to modificationsowards more selective compounds with lower intrinsic toxicitynd reduced negative environmental impacts [27]. Towards thisoal, an analysis of the peptide fragments in pleurocidin was ini-iated. In this study, the activity of the peptides was examinedgainst bacteria and filamentous phytopathogenic fungi. We dis-overed that a small sequence of the 12 amino acid C-terminusKHVGKAALTHYL) possed a high percentage (≥80%) of the precur-or’s lytic activity against Pseudomonas aeruginosa, Staphylococcusureus and Escherichia coli and no effect or very small effect againstnterococcus faecalis. In fungi of agronomical interest, a high activ-ty was observed against all the fungi evaluated, except Aspergilluschraceus. The measured MIC values were close to those of commer-ial fungicides. Four other synthetic peptides, spanning the wholeleurocidin sequence, were tested, but a reduced growth inhibi-ion was obtained for P. aeruginosa and for the three other bacteriapecies compared to pleurocidin.

. Materials and methods

.1. Materials

Amino acids for peptide synthesis were acquired fromalbiochem-Novabiochem Corp. (Germany). The sequencing

escribed in Section 2. Plc, pleurocidin peptide.

reagents and HPLC columns were from Shimadzu (Kyoto, Japan).Piperidine, acetonitrile and trifluoroacetic acid were purchasedfrom Fluke. Brain heart infusion broth (BHI), trifluoroethanol andall other analytical reagents were purchased from Merck (Darm-stad, Germany). Sytox green (SG) was acquired from MolecularProbes (Invitrogen Corp, Carlsbad, CA, USA) and calcofluor white(CFW) (Fluorescent Brightener 28) from Sigma–Aldrich (St Louis,MO, USA). Potato dextrose broth (PDB) and potato dextrose agar(PDA) were purchased from HiMedia (Mumbai, India) and Oxoid(Hampshire, England), respectively.

2.2. Microorganisms

The following bacteria were used for testing the AMP activityof the pleurocidin fragments; wild-type bacteria E. faecalis (ATCC-29212), E. coli (ATCC-35218), P. aeruginosa (ATCC-27853) and S.aureus (ATCC 25325). All bacteria were obtained from the NationalInstitute of Health Quality Control (INCQS), Oswaldo Cruz Foun-dation, RJ, Brazil and maintained in tubes with BHI at 37 ◦C untilreaching the exponential log phase. All strains were stored at −80 ◦Cuntil use and cultures were grown in 3% (w/v) Trypticase Soy Broth(TBS) at 37 ◦C.

Six fungal isolates, all known plant pathogens, were used forantifungal activity assays. The fungi Alternaria sp., Fusarium oxyspo-rum, A. niger, A. ochraceus, Cladosporium fulvum and Colletotrichumsp. were supplied by the Department of Protección Vegetal of theInstituto Nacional de Investigación Agropecuaria (INIA, Las Bru-jas, Uruguay). Fungi were cultured on PDA at 27 ◦C. Fungal sporeswere collected as described [2]. The concentration of the sporan-gial suspensions were estimated using a cell counting chamber andadjusted to 2 × 106 spores mL−1 [1], and stored in 20% glycerol at−80 ◦C until use.

2.3. Peptide synthesis

Peptides were synthesized by the solid-phase synthesis methodin a PSS-8 (Shimadzu, Kyoto, Japan) Pep Synthesizer according tothe fluoren-9-methyloxycarbonyl (Fmoc)-polyamide active esterchemistry [26]. The synthesized peptides were purified using aVydac (Altech Associates, Inc., USA) reverse-phase C18 column andthe purity was confirmed by matrix-associated laser desorption

listed in Table 1.Before the biological assays, lyophilized peptides were solubi-

lized in sterile Milli-Q water to a final concentration of 1 mM andfiltered sterilized through a 0.22-�m pore filter.

8 eptid

2

cd

�

wwitr

2

cB0GastM

2d

v2iitmasco

2

dIat7Moccam(2fctatata

0 A.L.A. Souza et al. / P

.4. Computational analysis and molecular moment calculation

The mean hydrophobic moment (�H) values for the pleuro-idin peptide fragments at different angles (ı) were calculated asescribed [10,11] by the equation:

H(ı) = (√

˙Hn sin(ıH))2 + (˙Hn cos(ıH))2Nhere N is the number of residues and n is the specific residueithin the peptide sequence; Hn is the hydrophobic value, accord-

ng to the normalized consensus hydropathy scale [34] assignedo residue n; and � is the angle (in radians) between successiveesidues (e.g., ı is equal to 100◦ for an � helix).

.5. Assay of bactericidal activity

Screening for the bacterial effect was performed at a peptideoncentration of 100 �g mL−1 using the tube dilution method.riefly, 0.9 mL of the bacterial suspension was incubated with.1 mL of peptide solution (1 mg mL−1) at 37 ◦C for 2 h, aerobically.rowth controls were performed with brain heart infusion (BHI)nd saline. Negative growth controls were performed under theame conditions with 10 �g mL−1 of gentamicin. Colony forma-ions units (CFU) were counted by streaking remaining bacteria on

ueller-Hinton Agar.

.6. Antimicrobial colony forming unit (CFU) assay andetermination of MIC

First, 5 × 105 CFU mL−1 was incubated for 18 h at 37 ◦C in a finalolume of 100 �l of MHB with 0.1–100 �g mL−1 of peptides using4-well polypropylene plates. Visual verification of microbial sed-

mentation and the absorbance at 600 nm confirmed the minimalnhibitory concentration (MIC). The minimal bactericidal concen-ration (MBC) was determined by streaking 5-�L aliquots of the

icrotiter plate reaction mixtures used to determine the MIC onto Mueller-Hinton (MHB) agar plate. The wells containing the threeerial dilutions above and below the MIC were analyzed. The lowestoncentration of peptide that ablated the bacterial colony growthn the agar plate was deemed the MBC.

.7. Assay of antifungal activity

The in vitro antifungal activities of peptide solutions wereetermined by a quantitative micro-spectrophotometric assay [2].

nhibition of growth was measured in 96-well microtiter platest 595 nm. Routine tests were performed with 20 �L of a peptideest solution, 10 �L of a spore suspension (2 × 106 spores mL−1) and0 �L of potato dextrose broth (PDB) (HiMedia, Mumbai, India).icrocultures containing 20 �L of sterile distilled water in place

f test solutions were used as negative control. The commer-ial fungicide Captan (0.2 mg mL−1) [35] was used as the positiveontrol. The plates were allowed to stand for 30 min at 27 ◦C tollow the spores to sediment, after which, the absorbance waseasured at 595 nm in a Multiscan Spectrum microplate reader

Thermo Electron Corp., Varta, Finland). After a 48 h incubation at7 ◦C, growth was recorded by measuring absorbance. All assaysor antifungal activity were performed, at a minimum, in tripli-ate. The growth inhibition percentage was determined based onhe equation [(�C − �T)/�C] × 100, where �C was the correctedbsorbance of the control microculture at 595 nm and �T was

he corrected absorbance of the test microculture. The correctedbsorbance values equaled the absorbance at 595 nm of the cul-ure measured after 48 h minus the absorbance at 595 nm measuredfter 30 min.

es 45 (2013) 78–84

2.8. Determination of the fungal MIC and MFC

A microplate method, as previously described [13], was usedwith slight modifications to determine the MIC of peptide test solu-tions. Briefly, 20 �L from a 500 �g mL−1 stock solution was added tothe first column of a microplate. Then, double serial dilutions wereperformed using distilled sterile water for the remaining columns.In each well, 20 �L of peptide dilutions were mixed with 180 �Lof the fungal spore suspension (2 × 106 spores mL−1 in fresh PDB).The microplates were incubated for 48 h at 27 ◦C. All experimentswere performed in triplicate. The MIC readings were measured asabsorbance at 595 nm. MIC was defined as the lowest peptide con-centration that inhibited 90% fungal growth. The in vitro minimalfungicidal concentration (MFC) was determined as described byEspinel-Ingroff et al. [12]. After a 48 h incubation, 20 �L from eachwell was subcultured onto PDA and incubated at 27 ◦C until growthwas observed in the growth control subculture. Cultures with thelowest peptide concentration that displayed no visible growth (rep-resenting an inhibition >98%) in comparison to cultures from thepositive well (Captan) and the growth control well (peptide-freemedium) onto PDA plates were used to determine the MFC.

2.9. Microscopic visualization

The mycelium of the different test fungi was grown for 48 h(2 × 106 spores mL−1 in PDB) before culturing in the presence ofpeptides at different concentrations. After 48 h, specimens werestained with two different fluorescent dyes [32]. First, the SGstock solution (4 �mol L−1) was added to a final concentration of0.2 �mol L−1. After 5 min in the dark, 0.1% (w/v) CFW was addedto a final concentration of 50 �g mL−1 and samples were incubatedagain for 5 min in the dark [29]. Finally, the mycelium was washedwith sterile water, centrifuged and resuspended in 20% glycerol.Wash conditions were determined based on evaluations of themycelium structure. Fluorescence signals were imaged using a flu-orescence microscope with the corresponding filters (Optiphot-2,Nikon, Japan).

2.10. Hemolytic activity

Peptides were screened for hemolytic activity by treating a 1%suspension of freshly isolated and washed human red blood cells(O+ donor) in PBS, pH 7.4. Peptide samples (10 �L) at various con-centrations (5–100 �g mL−1) were added and, after gentle mixing,the tubes were incubated at 37 ◦C for 30 min before centrifugation(4000 × g for 5 min at 25 ◦C). Supernatants (100 �l) were removedand diluted ten fold with PBS. The absorbance at 567 nm was mea-sured and the relative optical density was compared to that of a cellsuspension treated with 0.2% Triton X-100 (100% hemolysis).

3. Results

3.1. Peptide design and identification of Plc-2 as the corepleurocidin fragment retaining full antibacterial activity

Antimicrobial activities of derived pleurocidin peptides wereevaluated using Gram-positive (S. aureus and E. faecalis) and Gramnegative (P. aeruginosa and E. coli) representative bacteria. Detec-tion and quantification of antimicrobial activity were determinedby reduction of Alamar Blue. An evaluation of the Plc-2 fragmentbegan by testing the synthetic peptides Plc-1–5, which represent

the N-terminal, middle, and C-terminal segments of pleurocidin(Table 1). Significant killing activity against E. coli, S. aureus and P.aeruginosa was retained only by Plc-2 and the Plc-4 as compared tothe parent peptide of pleurocidin. The MIC ranged from 4.0 �M to

A.L.A. Souza et al. / Peptides 45 (2013) 78–84 81

Table 2Antibacterial activity (MIC, �M) of the different peptides.

Clinical isolate Plc-1 Plc-2 Plc-3 Plc-4 Plc-5

P. aeruginosa (ATCC-27853) 56.0 9.1 52.0 29.5 52.0E. coli (ATCC-35218) 43.4 4.0 56.1 16.1 43.6

9a

popdttiasPiao

3

ibiioeateb

wTTtWtT

Fca

Table 3Minimum inhibitory concentration (MIC) and minimum fungicidal concentration(MFC) of the different peptides.

Fungal pathogens MIC (�g mL−1) MFC (�g mL−1)

Pleurocidin Plc-2 Pleurocidin Plc-2

Alternaria sp. 3.12 6.25 – 6.25Fusarium oxysporum 3.12 3.12 6.25 6.25Aspergillus niger 1.56 6.25 3.12 12.5Aspergillus ochraceus >25 >50 >25 >50

S. aureus (ATCC-25325) 40.2 4.6 56.3 20.2 58.0E. faecalis (ATCC-29212) 52.3 – 47.2 – 57.0

.1 �M. Peptides Plc-1, Plc-3, Plc-4 and Plc-5 presented an oppositectivity with an MIC of 40.2–58.0 (Table 2).

The relationship between bacterial growths inhibitions versuseptide dose is shown in Fig. 1. Plc-2 and Plc-4 inhibited growthf S. aureus and E. coli at 2.5 (Fig. 1) and 5 �g/ml (data not shown)eptide concentrations, respectively. Plc-1, Plc-3, Plc-4 and Plc-5id not exhibit any growth inhibition. Plc-3 and Plc-5 also failedo show any antimicrobial activities through all the tested concen-rations. Notably, Plc-2 showed the highest antimicrobial activityn the experiment with S. aureus, and E. coli and it functionedt a 2.5 �g mL−1 peptide concentration. Plc-2 and Plc-4 have theequence KHVGKAAL in common. The antimicrobial activity oflc-2 was comparable to the original pleurocidin peptide, whichndicated that the C-terminal duodecapeptide was the minimalctive fragment and that the C-terminal Tre, His, Tyr, Lys residuesf Plc were required together for activity.

.2. Antifungal activity

Antimicrobial peptides target cytoplasmic membranes andntracellular macromolecules. As a general feature, most antimicro-ial peptides are amphipathic and this property serves a key role

n their antimicrobial activity by promoting microbial membranenteractions. However, microbial cell surfaces such as membranesr cell walls are composed of a variety of components, which gen-rate significant differences between the surfaces of prokaryotend eukaryote cells [17,18,42,44]. Previous studies have shown thathe pleurocidin peptide presents a selective membrane-disruptionffect in some fungi [22], but its mechanism of action remains toe determined.

The antifungal activities of the short pleurocidin peptidesere screened in vitro against Alternaria sp. and F. oxysporum.

able 3 shows the MIC and MFC values for the different fungi.he MIC and MFC values of pleurocidin ranged from 0.79 �g mL−1

−1 −1 −1

o >25 �g mL and 3.12 �g mL to >50 �g mL , respectively.hereas the MIC and MFC values of Plc-2 ranged from 3.12 �g mL−1

o >50 �g mL−1 and 6.25 �g mL−1 to >50 �g mL−1, respectively.hese values illustrate the relative antifungal potency of the two

0

20

40

60

80

100

120

P. aeruginosa S. aeureus E. fecalis E. coli

No drug

Plc-1

Plc-2

Plc-3

Plc-4

Plc-5

Plc

ig. 1. Susceptibility of P. aeruginosa, S. aureus, E. fecalis and E. coli to the pleuro-idin peptides (Plc, Plc-1, Plc-2, Plc-3, Plc-4 and Plc-5) measured by the assay ofntimicrobial activity. Screening was performed at 100 �g mL−1.

Cladosporium fulvum 6.25 12.5 25 >50Colletorichum sp. 0.79 3.12 3.12 6.25

peptides, with MIC values quite comparable to the conventionalfungicide captan. The highest inhibitory activity of the two peptideswas observed against Colletotrichum sp., and the lowest inhibitionwas noted against A. ochraceus. Plc-2 was less active than pleu-rocidin, except against F. oxysporum, for which the MIC and theMFC values were the same. Both peptides exhibited fungistatic andfungicidal activity for all the ascomycete fungi tested.

3.3. Morphological alterations

Significant morphological changes were observed when thephytopathogenic fungi were exposed to pleurocidin and Plc-2at concentrations that partially inhibit growth (Fig. 2). Most ofthese fungi exhibited increased branching (hyper-branching) andswelling of the hyphae in the presence of the peptides. Condensedhyphal aggregates were commonly observed when fungi weretreated with peptides followed by staining with CFW. The fluores-cent probe SG was used to assess cell permeation of fungi treatedwith both peptides. All the fungi showed identical fluorescent stain-ing. Cellular membranes were compromised and also disruptedif the fungal structures were incubated with pleurocidin or Plc-2(Fig. 2).

3.4. Structural and compositional studies of Plc-2

The fact that Plc-2 is reduced in size compared to pleurocidinmight alter its structural properties. The Plc-2 peptide presentedthe smallest charge (+2) and highest pI (9.7). Its major molecularmoment (0.16) was at the low end for all of the synthesized peptides(Table 1). Comparing the primary structure of Plc-2 with the struc-ture of antimicrobial peptides with similar activity (dermaseptin-1,ceratotoxin and PR39) together with the results presented here(Fig. 1 and Table 2), it can be determined that the sequence GKAALwas the critical amino acid sequence for the antimicrobial activityassociated with the parental pleurocidin antimicrobial peptide.

The Schiffer–Edmundson helical wheel showed that the amphi-pathic �-helical structure (containing hydrophobic residues on oneface of the helix and hydrophilic residues on the opposite face)present in the pleurocidin is also present in the Plc-2 (Fig. 3).

4. Discussion

Pleurocidin is one of the most noteworthy antimicrobial pep-tides studied in the past few years. It displays a broad spectrum ofactivity against bacteria, fungus and some leukemical and eukary-otic cells [17]. In general, the activity of AMP depends on thecomposition of the membranes with which it interacts togetherwith its structural features; primary structure, conformation, netcharge, net hydrophobicity, hydrophobic moment, amphipathicity,

size, and polar angle [43].

Therefore, to find the minimal active fragment, Plc was furthertruncated and assayed against a representative set of Gram-positive(S. aureus and E. faecalis) and Gram negative (P. aeruginosa and

82 A.L.A. Souza et al. / Peptides 45 (2013) 78–84

Fig. 2. Microscopy visualization of Alternaria sp. (A) and Fusarium oxysporum (B) treated with the peptide Plc-2 in vitro. (A) Hyphal morphology of Alternaria sp. Panel A1–A3shows the same area without Plc-2 treatment. A1: bright field; A2: treatment with CFW; A3: treatment with SG. Panel A4–A6 shows the same area after treatment with Plc-2.A4: bright field; A5: treatment with CFW; A6: treatment with SG. (B) Hyphal morphology of Fusarium oxysporum. Panel B1–B3 shows the same area without Plc-2 treatment.B1: bright field; B2: treatment with CFW; B3: treatment with SG. Panel B4–B6 shows the same area after treatment with pleurocidin. B4: bright field; B5: treatment withCFW; B6: treatment with SG. CFW, calcofluor white; SG, sytox green.

Fig. 3. Alignment of Plc-2 sequence and antimicrobial peptides with similar activity (A) and the Schiffer-Edmundson helical-wheel diagram projection demonstratingprobable amphipathic �-helical conformation of pleurocidin (B, left) and Plc-2 peptide (B, right). Diagrams of left helical-wheel and right helical net of five repeat KAALTsequence. When 2 and 6 positions on the helical-wheel and in the helical net were occupied by cationic residues, the �-helice takes amphiphilic structures, and B stands forK. Diagrams of left B showed one hydrophobic face VAWAGFYGA in pleurocidin.

eptide

EapltPbbi

ptpshethsbsstspm

tttacdiahStasTtcTubTahd

�Astpmabbp

ipei

A.L.A. Souza et al. / P

. coli) bacteria and fungi. The bacterium S. aureus has acquired number of genes that provide antibiotic resistance against allenicillins, including methicillin and other narrow-spectrum �-

actamase-resistant penicillin antibiotics. Recently, it has becomehe major cause of hospital-acquired infections [33]. The strain. aeruginosa typically infects the pulmonary tract, urinary tract,urns, and wounds and also causes other blood infections. It haseen reported to be responsible for one in ten hospital-acquired

nfections are from Pseudomonas [30,39].An important distinction for Gram-positive bacteria is that they

ossess thicker cell wall than Gram-negative organisms. This pep-idoglycan layer of Gram-positive bacteria remains a relativelyorous structure [38]. The essential function of the cell wall is toerve as a selective permeability barrier, protecting bacteria fromarmful agents, such as detergents, drugstoxins and degradativenzymes, yet allow the penetration of nutrients to sustain bac-erial growth. Evidence from genetic and chemical experimentsave proven that the cell wall plays an essential role in providing aelective permeability barrier for S. aureus and other Gram-positiveacteria. The peptide fragments used here caused cell wall effectsuch as breaks, thinning, and disintegration as well as abnormaleptation. Our experimental results using Plc-1–5, which representhe N-terminal, middle, and C-terminal segments of pleurocidinuggested that only the amino acids present in Plc-2 enhanced theermeability of membrane with a similar potency of the parentolecule, which requires passage through the cell wall.Examining the data in Table 1, the hydrophobicity of the pep-

ides was determined not to be a possible element of influence onhe observed activities. Therefore, it was assumed that the struc-ure would be a primary contributing aspect to the mechanism ofction. The C-terminal Plc-2 fragment (KHVGKAALTHYL) of pleuro-idin showed a perfect amphipathic structure in a �-helical wheeliagram (Fig. 3). Most AMPs are an amphipathic and this property

s a key role in antimicrobial activity by microbial membrane inter-ction. In fact, it was previously demonstrated that the pleurocidinad a �-helical structure in the membrane-mimetic condition [36].imilarly, NMR structural studies that covered all of the Plc-2 pep-ide sequence showed that in an aqueous solution the Plc presents

random coil conformation [37]. However, it assumed an �-helicaltructure in TFE and in dodecylphosphocholine (DPC) micelles [22].hus, the Plc-2 �-helical structure described in this work is similaro that for many other AMPs, which cause lysis and release of intra-ellular contents by binding to the surface of bacterial membranes.he �-helical structure produces a significant destabilizing effectpon membranes, which insert themselves into the membrane byinding more efficiently than other structural configurations [19].his conclusion is also in agreement with a report from Yamadand Natori [41], in which the fragment corresponding to the �-elical region of sapecin B, a derived peptide belonging to the insectefensin family, showed broad antibacterial activity.

However, antimicrobial activity is not restricted exclusively to-helical structures. Lee et al. [27] attributed the activity of theMP tenecin to a fragment (amino acids 29–43) located in a �-heet region of the peptide. Similarly other authors founded thathe antimicrobial activity of some peptides was establish in amphi-athic beta-sheet segments [19]. Microbial cell surfaces such asembranes or cell wall are composed of various components,

nd they exhibited significant differences in surface componentsetween bacteria and fungi. Therefore, it may be possible the mem-rane composition influences the activity of an AMP by influencingreferential interactions with �-helical or �-sheet structures.

Some known AMP can induce abnormal morphological changes

n the hyphae structure of phytopathogenic fungi [3] and humanathogenic fungi [20]. In our study, we chose to examine two fungixamined Alternaria sp. and F. oxysporum that are of economicalmportance. The abnormal morphological changes in membrane

s 45 (2013) 78–84 83

structure of hyphae were evaluated in vivo with the fluorescentmembrane probe SG. This probe was used to assess cell permeationof fungi treated with both peptides. All the fungi showed identicalfluorescent staining. Cellular membranes were compromised andalso disrupted if the fungal structures were incubated with pleu-rocidin or Plc-2. The MIC and MFC values measured illustrate therelative antifungal potency of the two peptides with MIC valuesquite comparable to the conventional fungicide captan.

The highest inhibitory activity of the two peptides was observedagainst Colletotrichum sp., and the lowest inhibition was notedagainst A. ochraceus (Table 2). Plc-2 was less active than pleurocidin,except against F. oxysporum, for which the MIC and the MFC valueswere the same. From these studies, we conclude that both peptidesexhibited fungistatic and fungicidal activity for all the ascomycetefungi tested. In order to understand the correlation between anti-fungal activity and cell-surface accumulation, we examined theeffects of the peptides on the cell. The integrity of the cell wall,cellular membrane and the nuclear membrane were compromisedsince nuclear staining was used as an indicator for the degree ofaccess from membrane damage by antifungal activity. These resultsindicate that the antifungal effects of Plc-2 are due to the accumu-lation of Plc-2 within the plasma membrane through interactionbetween peptides and the plasma membrane, rather than with thecell wall. Therefore, membrane damage from the membrane inter-action of Plc-2, is the major cause of cell death. The results are alsosupported by amphipathic �-helical conformation as the structuralfeature of Plc-2, which is presented by Schiffer–Edmundson helicalwheel modeling. Amphipathic �-helical peptides, with antibac-terial activity, exhibit membrane-disrupting activity and Plc alsodisplays �-helical structure in an aqueous solution, as well as inmembrane mimetic environments [25]. Comparing the primarystructure of Plc-2 with the structure of other antimicrobial pep-tides with similar activity (dermaseptin-1, ceratotoxin and PR39)and the results obtained in this work (Fig. 1 and Table 2) it canbe strongly suggested that the sequence GKAAL was the criticalamino acid sequence for the antimicrobial activity of the pleuro-cidin antimicrobial peptide.

In summary, the potential of Plc-2 as a therapeutic peptide agentwas investigated and the results suggested that one possible rea-son for it exerting a similar activity of Plc was because it appearedto maintenance of the native structure of pleurocidin. Actually,Plc-2 itself had strong antimicrobial activity with a much lowerhemolytic effect in comparison with melittin and other well knownantibiotics, such as ampicillin, vancomycin, cefotaxime, chloram-phenicol and kanamycin (data not shown). Thus, when the problemof the structure of Plc-2 is improved by peptide engineering, thispeptide may help form a leading model for developing new andnovel therapeutic agents.

5. Conclusions

In this study, we investigated the antibacterial and antifun-gal activities to determine mechanisms of action for pleurocidinand short Plc derived peptides. Our results from anti-bacterialactivity indicated that Plc-2 a C-terminal 12-amino acid fragmentof pleurocidin contained the critical amino acid for the cytolyticactivity. The data also showed that the strong antibacterial activ-ity against human pathogenic Gram-positive and Gram-negativebacteria was not damaging to human erythrocytes. In addition,Plc-2 also exhibits a potent activity against fungicide-resistantpathogens. Staining experiments indicated that both pleurocidin

and Plc-2 compromised the integrity of the fungi membrane and, inaddition bind to nuclear acid, thus affecting the nuclear membrane.The antifungal effects of Plc-2 are thought to be due to membranedisruption after the accumulation in the plasma membrane of the

8 eptid

ftaa

A

TSoSpTDtDi

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[[

[

[

[

[

[

[

[

[

[

[

[

[[

[

[

4 A.L.A. Souza et al. / P

ungal cell. Thus, the broad activity of this peptide may be helpfulo form a leading model for developing new and novel therapeuticgents for public health and could have applications to commercialgriculture.

cknowledgements

This work received financial assistance from the Science andechnology Ministry of Brazil (MCT), Research Foundation of thetate of Rio de Janeiro (FAPERJ), Coordination of Improvementf Higher Education Personnel (CAPES), the Brazilian Council forcientific Research (CNPq) and DYCIT (Uruguay). We thank thelatform of peptide synthesis of FIOCRUZ (PDTIS) for the facilities.hanks are also due to Dr M.C. Lourenç o from the Microbiologyepartment (Instituto de Pesquisas Evandro Chagas-FIOCRUZ) for

he bacterial assays and to Ms Nora Altier and Carolina Leoni fromepartment of Protección Vegetal (INIA Las Brujas) for the fungal

solates.

eferences

[1] Abril M, Curry KJ, Smith BJ, Wedge DE. Improved microassays used to test natu-ral product-based and conventional fungicides on plant pathogenic fungi. PlantDis 2008;92:106–12.

[2] Broekaert WF, Terras FRG, Cammue BPA, Vanderleyden J. An automated quanti-tative assay for fungal growth inhibition. FEMS Microbiol Lett 1990;69:55–60.

[3] Cavallarin L, Andreu D, San-Segundo B. Cecropin A-derived peptides arepotent inhibitors of fungal plant pathogens. Mol Plant Microbe Interact1998;11:218–27.

[4] Cho J, Choi H, Lee DG. Influence of the N- and C-terminal regions of antimi-crobial peptide pleurocidin on antibacterial activity. J Microbiol Biotechnol2012;22:1367–74.

[5] Chen YH, Yang JT, Chau KH. Determinations of the helix and betaforms of proteins in aqueous solutions by circular dichroism. Biochemistry1974;13:3350–9.

[6] Cole AM, Darouiche RO, Legarda D, Connell N, Diamond G. Characterizationof a fish antimicrobial peptide: gene expression, subcellular localization, andspectrum of activity. Antimicrob Agents Chemother 2000;44:2039–45.

[7] Cole AM, Weis P, Diamond G. Isolation and characterization of pleurocidin, anantimicrobial peptide in the skin secretions of winter flounder. J Biol Chem1997;272:12008–13.

[8] Douglas SE, Gallant JW, Gonz Z, Hew C. Cloning and development expression ofa family of pleurocidin-like antimicrobial peptides from winter flunder, Pseu-dopleuronectes americanus (Walbaum). Dev Comp Immunol 2001;25:137–47.

[9] Douglas SE, Patrzykat A, Pytuck J, Gallant JW. Identification, structure anddifferential expression of novel pleurocidins clustered on the genome ofthe winter flunder, Pseudopleronectes americanus (Walbaum). Eur J Biochem2003;270:3720–30.

10] Eisenberg D. Three-dimensional structure of membrane and surface proteins.Annu Rev Biochem 1984;53:595–623.

11] Eisenberg D, Weiss RM, Terwilliger TC. The hydrophobic moment detects peri-odicity in protein hydrophobicity. Proc Natl Acad Sci U S A 1984;81:140–4.

12] Espinel-Ingroff A, Fothergill A, Peter J, Rinaldi MG, Walsh TJ. Testing conditionsfor determination of minimum fungicidal concentrations of new and estab-lished antifungal agents for Aspergillus spp.: NCCLS Collaborative Study. J ClinMicrobiol 2002;40:3204–8.

13] Fernández-Andreu CM, Pimentel-Turino T, Martínez-Machín G, González-Miranda M. Determinación de la concentración mínima inhibitoria de flucona-zol frente a Cryptococcus neoformans. Rev Cubana Med Trop 1999;51:55–7.

14] Fuchs PC, Barry AL, Brown SD. In vitro antimicrobial actiovity of MSI-78, amagainin analog. Antimicrob Agents Chemother 1998;42:1213–6.

15] Gudmundsson GH, Magnusson KP, Chowdhary BP, Johansson M, Anderson L,Boman HG. Structure of the gene for porcine peptide antibiotic PR-39, a cathelingene family member? Comparative mapping of the locus for the human peptideantibiotic FALL-39. Proc Natl Acad Sci U S A 1995;92:7085–9.

16] Hancoc RE, Sahl HG. Antimicrobial and host-defense peptides as new anti-

infective therapeutic strategic. Nat Biotechnol 2006;24:1551–7.

17] Hilchie AL, Doucette CD, Pinto DM, Patrzykat A, Douglas S, Hoskin DW.Pleurocidin-family cationic antimicrobial peptides are cytolytic for breast car-cinoma cells and prevent growth of tumor xenografts. Breast Cancer Res2011;13:R102.

[

[

es 45 (2013) 78–84

18] Jung HJ, Park Y, Sung WS, Suh BK, Lee J, Hahm KS, Lee DG. Fungicidaleffect of pleurocidin by membrane-active mechanism and design of enan-tiomeric analogue for proteolytic resistance. Biochim Biophys Acta 2007;1768:1400–5.

19] Jin Y, Hammer J, Pate M, Zhang Y, Zhu F, Zmuda E, Blazyk J. Antimicrobialactivities and structures of two linear cationic peptide families with vari-ous amphipathic beta-sheet and alpha-helical potentials. Antimicrob AgentsChemother 2005;49:4957–64.

20] Lee KHS, Hong SY, Oh JE, Kwon M, Yoon JH, Lee J, Lee BL, Moon HM. Identificationand characterization of the antimicrobial peptide corresponding to C-terminalbeta-sheet domain of tenecin 1, an antibacterial protein of larvae of Tenebriomolitor. Biochem J 1998;334:99–105.

21] Lee DG, Park Y, Nam HN, Kim HK, Kim PI, Choi BH, Hahm KS. Antifungalmechanism of an antimicrobial peptide, HP(2–20), derived from N-terminusof Helicobacter pylori ribosomal protein L1 against Candida albicans. BiochemBiophys Res Commun 2002;291:1006–13.

22] Lee J, Lee DG. Structure–antimicrobial activity relationship between pleuro-cidin and its enantiomer. Exp Mol Med 2008;40:370–6.

23] Ma QQ, Shan AS, Dong A, Cao YP, Lv YF, Wang L. The effects of Leu or Valresidues on cell selectivity of �-helical peptides. Protein Pept Lett 2011;18:1112–8.

24] Marchini D, Giordano AG, Amons R, Bernini LF, Dallai R. Purification and primarystructure of ceratotoxin A and B, two antibacterial peptides from the femalereproductive accessory glands of the medfly Ceratitis capitata (insecta: Diptera).Insect Biochem Mol Biol 1993;23:591–8.

25] Mason AJ, Chotimah IN, Bertani P, Bechinger B. A spectroscopic study of themembrane interaction of the antimicrobial peptide pleurocidin. Mol MembrBiol 2006;23:185–94.

26] Merrifield RB. Solid phase synthesis. Science 1986;232:341–7.27] Montesinos E. Antimicrobial peptides and plant disease control. FEMS Micro-

biol Lett 2007;270:1–11.28] Mor A, Nguyen VH, Delfour A, Migliore-Samour D, Nicolas P. Isolation, amino

acid sequence and synthesis of dermaseptin, a novel antimicrobial peptide ofamphibian skin. Biochemistry 1991;30:8824–30.

29] Muñoz A, Marcos JF. Activity and mode of action against fungal phytopathogensof bovine lactoferricin-derived peptides. J Appl Microbiol 2006;101:1199–207.

30] Raymond J. Pseudomonas aeruginosa and hospital-acquired infections. ArchPediatr 2006;13(Suppl. 1):S2–4.

31] Redondo-Lopez V, Cook RL, Sobel JD. Emerging role of lactobacilli in thecontrol and maintenance of the vaginal bacterial microflora. Rev Infect Dis1990;12:856–72.

32] Reed JD, Edwards DL, Gonzalez CF. Synthetic peptide combinatorial libraries:a method for the identification of bioactive peptides against phytopathogenicfungi. Mol Plant Microbe Interact 1997;10:537–49.

33] Romero-Vivas J, Rubio M, Fernandez C, Picazo JJ. Mortality associated withnosocomial bacteremia due to methicillin-resistant Staphylococcus aureus. ClinInfect Dis 1995;21:1417–23.

34] Saint NS, Cadiou H, Bessin Y, Molle G. Antibacterial peptide pleurocidinfrom ion channels in planar lipid bilayers. Biochim Biophys Acta 2002;1564:359–64.

35] Satish S, Mohana DC, Ranhavendra MP, Raveesha KA. Antifungal activity ofsome plant extracts against important seed borne pathogens of Aspergillus sp.J Agric Technol 2007;3:109–19.

36] Syvitski RT, Burton I, Mattatall NR, Douglas SE, Jakeman DL. Structural char-acterization of the antimicrobial peptide pleurocidin from winter flounder.Biochemistry 2005;44:7282–93.

37] Sung WS, Lee DG. Pleurocidin-derived antifungal peptides with selec-tive membrane-disruption effect. Biochem Biophys Res Commun2008;369:858–61.

38] Sutcliffe IC, Russell R. Lipoproteins of Gram-positive bacteria. J Bacteriol1995;175:1123–8.

39] Todar. Online textbook of bacteriology.40] Tao R, Tong Z, Lin Y, Xue Y, Wang W, Kuang R, Wang P, Tian Y, Ni L. Antimicro-

bial and antibiofilm activity of pleurocidin against cariogenic microorganisms.Peptides 2011;32:1748–54.

41] Yamada K, Natori S. Characterization of the antimicrobial peptide derived fromsapecin B, an antibacterial protein of Sarcophaga peregrina (flesh fly). BiochemJ 1994;298:623–8.

42] Yang JY, Shin SY, Lim SS, Hahm KS, Kim Y. Structure and bacterial cell selectiv-ity of a fish-derived antimicrobial peptide, pleurocidin. J Microbiol Biotechnol2006;16:880–8.

43] Yeaman MR, Yount NY. Mechanisms of antimicrobial peptide action and resis-tance. Pharmacol Rev 2003;55:27–53.

44] Yoshida K, Mukai Y, Niidome T, Takashi C, Tokunaga Y, Hatakeyama T, AoyagiH. Interaction of pleurocidin and its analogs with phospholipid membrane andtheir antibacterial activity. J Pept Res 2001;57:119–26.

http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0005http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0010http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0015http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0020http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0025http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0030http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0035http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0040http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0045http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0050http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0050http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0050http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0050http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0050http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0050http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0050http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0050http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0050http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0050http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0050http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0050http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0050http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0050http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0050http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0055http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0060http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0065http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0070http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0075http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0080http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0080http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0080http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0080http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0080http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0080http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0080http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0080http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0080http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0080http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0080http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0080http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0080http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0080http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0080http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0080http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0080http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0080http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0080http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0085http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0090http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0095http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0100http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0105http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0110http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0115http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0120http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0125http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0130http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0130http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0130http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0130http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0130http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0130http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0130http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0130http://refhub.elsevier.com/S0196-9781(13)00141-1/sbref0130http://refhub.elsevier.c