FEBS Letters 589 (2015) 2812–2818

journal homepage: www.FEBSLetters .org

A new active antimicrobial peptide from PD-L4, a type 1 ribosomeinactivating protein of Phytolacca dioica L.: A new function of RIPsfor plant defence?

http://dx.doi.org/10.1016/j.febslet.2015.08.0180014-5793/� 2015 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.

Abbreviations: CD, circular dichroism; CNBr, cyanogen bromide; LPS,lipopolysaccharide; Na/P, sodium phosphate buffer; TFE, trifluoroethanol; Tris�Cl,Tris(hydroxymethyl)aminomethane–HCl buffer

Author contributions: This work was carried out in collaboration between allauthors. Authors Elio Pizzo and Andrea Bosso have performed experiments onhuman cell lines and CD experiments. Authors Anna Zanfardino and MarioVarcamonti have performed experiments on bacterial strains (antimicrobial activ-ity). Authors Antonella MA Di Giuseppe, Nicola Landi and Sara Ragucci have carriedout the purification of the recombinant protein and its chemical fragmentation toobtain PD-L4 fragment. Author Eugenio Notomista analysed the antimicrobialactivity of peptides from Phytolaccaceae RIPs by bioinformatic approach. AuthorADM designed the study and wrote the first draft of the manuscript. All authorsread and approved the final manuscript.⇑ Corresponding authors.

E-mail addresses: eliodoro.pizzo@unina.it (E. Pizzo), antimo.dimaro@unina2.it(A. Di Maro).

1 These authors contributed equally to this work.

Elio Pizzo a,⇑,1, Anna Zanfardino a,1, Antonella M.A. Di Giuseppe b, Andrea Bosso a, Nicola Landi b,Sara Ragucci b, Mario Varcamonti a, Eugenio Notomista a, Antimo Di Maro b,⇑aDepartment of Biology, University of Naples ‘‘Federico II” Via Cintia, I-80126 Napoli, ItalybDepartment of Environmental, Biological and Pharmaceutical Sciences and Technologies, Second University of Naples, I-81100 Caserta, Italy

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

Article history:Received 9 July 2015Revised 30 July 2015Accepted 9 August 2015Available online 20 August 2015

Edited by Michael Ibba

Keywords:Antimicrobial peptideCationic antimicrobial peptideRibosome inactivating proteinPhytolacca dioica

We investigated the antimicrobial activity of PD-L4, a type 1 RIP from Phytolacca dioica. We foundthat this protein is active on different bacterial strains both in a native and denatured/alkylatedform and that this biological activity is related to a cryptic peptide, named PDL440–65, identifiedby chemical fragmentation. This peptide showed the same antimicrobial activity of full-lengthprotein and possessed, similarly to several antimicrobial peptides, an immunomodulatory effecton human cells. It assumes an alpha-helical conformation when interact with mimic membraneagents as TFE and likely bacterial membranes are a target of this peptide. To date PDL440–65 is thefirst antimicrobial peptide identified in a type 1 RIP.� 2015 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.

1. Introduction

Ribosome-inactivating proteins (RIPs) are N-b-glycosylase (EC:3.2.2.22) which remove a single adenine residue (A4324 in rat liverribosomal ribonucleic acid [rRNA] [1]) from the sugar-phosphatebackbone of 28S rRNA. These enzymes, identified mainly in plantsbut also in fungi and bacteria [2,3], are divided into two groups [4]:

type 1 RIPs, consisting of a single enzymatically active chain andtype 2 RIPs, made of two chains, one catalytically active (similarto type 1 RIPs and named A chain) and the other containing a lectindomain (B chain). Type 2 RIPs are generally extremely toxic tomany cell types because the lectin B chain facilitates their entry[5], with exceptions like, for instance, those present in Sambucus[6]; on the contrary type 1 RIPs, devoid of a lectin binding chain,are internalized much less efficiently by cells and consequentlyhave relatively low toxicity [7].

The enzymology of RIPs has been well characterized, whereastheir biological and physiological functions in plants have not yetfully clarified. For this reason, biological role of RIPs in plantsremains open to argumentations [8]. It is proposed that they areinvolved in plant defence in abiotic and biotic stress [8]. RIPs areoften found in large amounts in different plant tissues such asleaves (e.g. saporin S-6 from Saponaria officinalis L. [9,10]) andseeds (e.g. ricin from Ricinus communis L. [11]) therefore someauthors consider RIPs also as storage proteins [12]. Whatevertheir physiological role, the interest for these enzymes is due totheir potential use (i) for the development of immunotoxins fortumor therapy [13], (ii) for their use as natural phyto-pesticides[14,15], and (iii) for the production of transgenic plants endowedwith specific parasite resistance [16,17]. It was found that many

E. Pizzo et al. / FEBS Letters 589 (2015) 2812–2818 2813

RIPs are also able to remove adenine residues from deoxyribonu-cleic acid (DNA) or from other polynucleotides [18]; recentreports revealed that some RIPs possess RNase or DNase activi-ties, superoxide dismutase (SOD) or phospholipase properties[8,19].

In the last decade, many authors reported that antimicrobialactivity of several cationic proteins is due to their structuralcharacteristics and not necessarily to their functional/enzymaticproperties [20]. More particularly, antimicrobial activity dependsessentially on specific stretches of residues or cryptic peptidesdetectable in the primary structure of proteins. Among these,cationic antimicrobial peptides (commonly named CAMPs) areable to interact with the bacterial membrane through electrostaticinteractions, hence they can cause direct cell death or significantlyalter cell survival by damaging membrane or by binding to specificintracellular target crossing membranes. Some of these cationicpeptides, identified in RNases [21], lysozymes [22], lactoferrin[23], thrombin [24], have been thoroughly investigated.

In this framework lies type 1 RIPs that are cationic and so ableto bind strongly negative substrates, as polyanions (e.g. RNA, DNA,etc.) [1,7,8]. Because in general cationic proteins possess intrinsicantimicrobial properties (see above), we thought to verify if alsoa type 1 RIP showed antimicrobial activity on different bacterialstrains and to identify possibly structural determinants that couldconfer this property. In order to test this proof of concept wechoose PD-L4, a type 1 RIP isolated from Phytolacca dioica L. sum-mer leaves, which is enzymatically and structurally well character-ized [25–27].

2. Materials and methods

2.1. Materials and general procedures

All chemicals utilized in this work were of analytical grade and,except where indicated, were purchased from Sigma–Aldrich(Milan, Italy).

Bacterial cultures, plasmid purifications and transformationswere performed according to Sambrook [28]. Protein sequencedeterminations were performed as previously reported [29], usinga Procise sequencer, Model 491C (Applied Biosystems, Foster City,CA).

Bacterial strains used for antimicrobial activity assays were:Pseudomonas aeruginosa PAOI, Staphylococcus aureus ATCC 6538P,Escherichia coli DH5a and Agrobacterium tumefaciens AGL0.

2.2. Peptide synthesis

The peptide PDL440–65 was synthesized by InBios (Naples, Italy)using solid-phase 9-fluorenylmethoxy carbonyl (Fmoc) chem-istry and purified to a purity >95% using reverse-phase high-performance liquid chromatography (HPLC). Peptide mass wasconfirmed by mass spectrometry.

2.3. Protein purification and the its chemical (with CNBr)fragmentation

Expression, refolding and purification of recombinant rPD-L4 inE. coli expression system were performed according to Del VecchioBlanco et al. [30]. Each step of the expression and purification pro-cedure was monitored by SDS–PAGE analyses. The eluted proteinfrom the cation exchange chromatography on SOURCE 15S4.6/100 PE column (1.7 mL) using Akta purifier system (GE Health-care, Milan, Italy) was collected, dialyzed, and kept frozen untiluse. Chemical fragmentation by CNBr method and correspondingresults are reported in Supplemental Material section.

2.4. Anti-microbial activity assay

A single colony of P. aeruginosa, E. coli, S. aureus or A. tumefacienswas suspended in 5 mL of LB (Luria–Bertani medium; Difco,Detroit, MI) and incubated overnight at 37 �C. When the culturereached an OD600 of 1 unit, it was diluted 1:100 in 20 mM Na/P,pH 7.0. Samples were prepared by adding 40 lL of bacterial cellsand the peptide, native protein or denatured protein at differentconcentrations and adjuster to 1 mL final volume was reachedwith 20 mM Na/P at pH 7.0. Negative controls were representedby cells incubated without any compound or with BSA, andassayed at the same concentration of tested molecules; cellsincubated with 0.05 mg/mL ampicillin (for E. coli and S. aureus),0.01 mg/mL colistin (for P. aeruginosa) or 0.05 mg/mL carbenicillin(for A. tumefaciens) were used as positive control. All samples werefirst incubated at 37 �C for 4 h, and then corresponding dilutions(1:100 and 1:1000) were plated on solid LB medium plates andincubated overnight at 37 �C. The following day survived cells wereestimated by colonies counting on each placed and compared withthe controls. All compounds were tested in triplicate experiments,standard deviations were always <5% for each experiment.

2.5. DAPI/PI dual staining and fluorescence microscopy imageacquisition

A working solution (5 lg/mL) of 40,6-diamidino-2-phenylindoledihydrochloride (DAPI; Sigma Aldrich) was prepared by dissolvingthe content of a 25 mg package in sterile Milli-Q water and storedat �20 �C until used. For the dual staining, 10 lL of E. coli culturewere mixed with 5 lL of DAPI solution (1 lg/mL DAPI final concen-tration) and 10 lL of 120 lM propidium iodide (Sigma Aldrich)[31] and incubated in the dark for 30 min at room temperature[32]. Samples were observed using an Olympus BX51 fluorescencemicroscope using a DAPI (Sigma, Aldrich) filter. Standard acquisi-tion time was 1000 ms for DAPI/PI dual staining and the Imageswere captured using an Olympus DP70 digital camera andprocessed [33].

2.6. SDS–PAGE and western immunoblot analysis

CaCo-2 or HaCat cells were harvested in lysis buffer [50 mMTris�Cl, pH 7.5, 5 mM EDTA, 150 mM NaCl, 1% NP-40, 1 mMphenylmethyl-sulfonyl fluoride, 0.5% sodium deoxycholate, andprotease inhibitors (SIGMAFASTTM Tablets, Sigma–Aldrich)]; corre-sponding total protein extracts were prepared as previouslydescribed [34]. Briefly, cell lysates were incubated on ice for40 min and then centrifuged at 15,000g for 15 min to remove celldebris. Protein concentration of resulting supernatants was thendetermined by using Bio-Rad protein assay kit (Bio-Rad, RomeItaly). Finally, after the addition of 2� Laemmli buffer (SigmaAldrich), samples were boiled at 100 �C for 5 min and resolved bySDS–polyacrylamide gel electrophoresis (10% or 12%) and subse-quently proteins were transferred to polyvinylidenedifluoride(PVDF) membranes (Millipore Milan, Italy) as described elsewhere[34]. Membranes were blocked in 5% w/v milk buffer (5% w/v non-fat dried milk in 50 mM Tris�Cl pH 8.0 containing 200 mMNaCl and0.2% Tween 20). Then membranes were incubated with primaryantibody diluted in 5% w/v milk or bovine serum albumin in Tris�ClpH 8.0 200 mM NaCl, 0.2% Tween-20, pH 8.0 for 2 h at room tem-perature or overnight at 4 �C. Primary antibodies were anti-rabbitpErks 42/44 (Cell Signaling, EuroClone, Milan, Italy), anti-goatb-actin (Santa-Cruz Biotechnology DBA Milan, Italy). Data werevisualized by enhanced chemi-luminescence method (ECL,GE-Healthcare Milan, Italy) using HRP-conjugated secondaryantibody (Santa-Cruz Biotechnology DBA) incubated 1 h at room

Fig. 1. Antibacterial activity on E. coli (black bars), P. aeruginosa (dark gray bars), S.aureus (white bars) and A. tumefaciens (light gray bars) of recombinant PD-L4 or itsdenatured and alkylated form. Positive control for each strain was performed usingthe appropriate antibiotic which concentrations are indicated in methods (Sec-tion 2.4). Standard deviations were always <5% for each experiment.

Fig. 2. Antibacterial activity on E. coli (black bars), P. aeruginosa (dark gray bars), S.aureus (white bars) and A. tumefaciens (light gray bars) of PDL440–65 peptidecompared to recombinant PD-L4 and its denatured or alkylated form. Standarddeviations were always <5% for each experiment.

2814 E. Pizzo et al. / FEBS Letters 589 (2015) 2812–2818

temperature and analyzed by Quantity One software of Chemi-DocTM XRS system (Bio-Rad).

3. Results and discussion

3.1. Antimicrobial activity of native and denatured PD-L4 on differentbacteria

PD-L4 was produced as recombinant protein (rPD-L4), asdescribed in Supplementary Materials (see Fig. S1A). Its antibacte-rial activity was analyzed on E. coli, P. aeruginosa, S. aureus and onthe plant pathogen A. tumefaciens (Fig. 1). As for many cationicenzymes, their antimicrobial activity is not necessarily related totheir intrinsic catalytic efficacy, we decided to test also the antimi-crobial activity of rPD-L4 previously subjected to denaturation fol-lowed by S-pyridylethylation. As shown in Fig. 1, both rPD-L4 andits chemically modified form, when assayed at 3 lM, showed a sim-ilar antimicrobial activity. This result demonstrates that rPD-L4cytotoxic propensity is likely due to its intrinsic aminoacidic com-position. A similar antimicrobial trend for denatured rPD-L4 hasbeen obtained also modifying it with different alkylating agents,such as bromopropilamine and iodoacetic acid, (data not shown).

3.2. Analysis of antimicrobial activity of PD-L4 peptides obtained bychemical fragmentation

Since results above described were obtained on full-lengthnative or denatured protein, in order to verify if bactericidal actionof rPD-L4 was due to a specific cluster of residues, we decided tosubject rPD-L4 to a chemical fragmentation by CNBr (see Supple-mental Material). As shown in Fig. S1B, rPD-L4 sequence contains5 methioninyl residues, so that CNBr fragmentation determines atheoretical pattern of fragmentation, consisting of six peptides.

CNBr reaction mixture (see methods) has been subjected toreduction by b-mercaptoethanol and S-pyridylethylation followedby RP-HPLC (Fig. S2A). Fragments, corresponding to main peaks,were so identified by Edman degradation and MALDI-TOF MS(Fig. S2B) and then assayed.

Antimicrobial assays on P. aeruginosa, revealed that only peptide5, corresponding to residues 40–65, showed a significant bacterici-dal action (data not shown). Based on this indication, we tested thispeptide, named PDL440–65, on E. coli, P. aeruginosa, S. aureus and onA. tumefaciens and compared it to the activity of native or dena-tured/alkylated rPD-L4 (Fig. 2). In all considered cases, antimicro-bial action of PDL440–65, was comparable to that of both proteinforms and was dose dependent, as shown in Supplemental Fig. S3.These data strongly support the hypothesis that cytotoxic actionof rPD-L4 is due to structural properties of peptide PDL440–65.

Since CAMPs usually directly damage target membranes, weperformed a fluorescence microscopy experiment to verify theeffect of PDL440–65 on bacterial membranes. To test this action,we used E. coli cells, DAPI as fluorescent stain for DNA and propid-ium iodide, the latter detectable only in cells with damaged mem-branes and therefore it is an useful indicator of cell death. Asshown in Fig. 3 (panel 1), results obtained point out that a signifi-cant amount of cells, after treatment with PDL440–65, developed ared fluorescence light as a consequence of a possible impairmentof membrane integrity. Same trend has been observed on E. colicells treated with rPD-L4 (Fig. 3, panel 2). As expected, fluorescencedeveloped by untreated cells was blue because intact membranesavoid the entry of propidium iodide [35].

3.3. Conformational studies of PDL440–65 peptide by circular dichroism

Since residues 40–65 in PD-L4 (PDB code: 2Z4U) are structuredas b-sheet fold, we performed circular dichroism experiments to

study conformational changes of PDL440–65 in buffer solution andin presence of different mimic membrane agents. Far-UV CDspectra indicate that PDL440–65 peptide is unstructured in phos-phate buffer but adopts a a-helical structure already in 30% TFEand in the presence of 20 mM SDS (Fig. 4). This behavior indicatesthat PDL440–65 peptide is prone to assume a specific conformationwhen interacting with membrane-mimicking agents like TFE(trifluoroethanol) or SDS. Alpha helical content of the peptidewas estimated [36] and reported in Fig. 4. It is worth noting thathelix content did not exhibit a significant change at concentrationsof TFE higher than 50%, thus suggesting a high propensity toacquire an ordered structure. In fact, it has been already shownthat peptides with high helical-propensity reach their maximumhelical content at concentrations of TFE between 30% and 50%[37]. Besides, it should be noted that PDL440–65 assumes a specificfold in presence of TFE or SDS, but this conformation is differentwith respect to that adopted by residues 40–65 in rPD-L4. Thistrend is already described for some peptides that are able to switchbetween different secondary structures (generally from a b-sheetto an a-helical structure [38]) and are known as chameleonpeptides [39].

To further characterize the structural properties of PDL440–65

peptide, we analyzed also its interaction with LPS (lipopolysaccha-ride), the main constituent of the outer membrane of Gram nega-tive bacteria. When PDL440–65 peptide was analyzed in presence

Fig. 3. Fluorescence microscopy by DAPI/PI dual staining on E. coli cells treated withrPD-L4 (Panel 1B) or treated with PDL440–65 peptide (Panel 2B). Panels 1A and 2Arefer to untreated cells. Corresponding 1–2 C and 1–2 D refer to optical analysis.

Fig. 5. Effects of PDL440–65 on CaCo-2 and HaCat cells. Cells were grown for 24, 48and 72 h in presence of 3 or 30 lM of peptide. Cell survival values (means oftriplicates) are expressed as percent of the corresponding values obtained in controlcultures grown in the absence of peptide. Standard deviations were always <5% foreach experiment.

E. Pizzo et al. / FEBS Letters 589 (2015) 2812–2818 2815

of increasing concentrations of LPS from E. coli strain 0111:B4,dichroic spectra indicated that the peptide was random coiled(data not shown) and so likely not able to bind LPS.

Fig. 4. CD spectra of PDL440–65 peptide in phosphate buffer (black line), in thepresence of different concentrations of TFE or in presence of 20 mM SDS. In table arereported secondary structure content (%) obtained by deconvoluted CD data.

3.4. Cytotoxicity assays and inflammatory properties of PDL440–65peptide on human cells

As already underlined above, the promising interest in the useof CAMPs as alternative antibiotics stems from their selectiveaction on bacterial cells with respect to eukaryotic cells. We thusstudied the cytotoxic effect of PDL440–65 peptide towards two dif-ferent human cell lines, HaCat and CaCo-2 cells. Addition ofincreasing concentrations (3 and 30 lM) of PDL440–65 peptide toHaCat or CaCo-2 cells, at three different times of incubation (24,48 or 72 h), did not result in any significant effect on viability ofcells as judged by measuring mitochondrial functionality (Fig. 5).

It is known that several CAMPs have the ability to block the pro-duction of cytokines produced in response to LPS by either directlyup-regulating inhibitory pathways in cells [40] or interfering withthe ability of LPS to bind LPS-binding proteins. In order to investi-gate the hypothesis that PDL440–65 peptide might elicit anti-inflammatory effects and thus immunomodulatory activities onhuman cells, although it is not able to interact with LPS (see CDsection), we analyzed by western blotting its effects on LPS-treated CaCo-2 cells. As shown in Fig. 6, in CaCo-2 cells subjected

Fig. 6. Effect of PDL440–65 peptide on ERK phosphorylation level in CaCo-2 cells bywestern blotting. Protein extracts from CaCo-2 cells, treated as described inSection 2.6, were analyzed as follows: lane C, protein extract from untreated cells;lane 1–4, protein extracts from CaCo-2 cells under different conditions. Lane 1, LPStreatment (1 h); lane 2, peptide alone (1 h); lane 3, PDL440–65 and LPS simultane-ously administrated (1 h); lane 4, pre-treatment (1 h) with LPS and subsequentlyincubation (1 h) with PDL440–65.

2816 E. Pizzo et al. / FEBS Letters 589 (2015) 2812–2818

to LPS treatment, we observed, as expected, an increasing of ERKtyrosine phosphorylation level (lane 1), whereas this cytokinedecreased when PDL440–65 peptide was administrated simultane-ously to LPS (lane 3). A similar trend has been observed also whenCaCo-2 cells were pre-treated with LPS and subsequentlyincubated with PDL440–65 (lane 4). It should be also noted thatincubation of CaCo-2 cells with PDL440–65 alone (lane 2) does notalter significantly ERK tyrosine phosphorylation level. This set ofevidences indicates that PDL440–65 is able to trigger ananti-inflammatory response in LPS treated CaCo-2 cells albeit CDanalysis showed no direct binding to LPS.

3.5. Comparative analysis in family

In order to verify if antimicrobial peptides similar to PDL440–65are present in other type 1 RIPs from Phytolaccaceae [41] a multiplealignment with PD-L4 was carried out (Fig. S4). A regionhomologous to PDL440–65 (position 41–71, numbering according

Fig. 7. (A) Extract of multi alignment of type 1 RIPs from Phytolaccaceae. Asterisks and higon the right side. Identical residues (⁄), conserved substitutions (:) and semi-conserved sSupplement Materials. (B) Antimicrobial activity scores for two different bacterial straiobtained by the bioinformatics tool described in the text.

to the consensus sequence) was detected and the correspondingidentity and similarity values are reported in Fig. S5.

This region, highlighted in Fig. 7A, shows the presence of sixconserved residues (K9, Y10, L12, L15, T28 and L29). Almost allthe examined peptides are cationic (net positive charges from +1to +3) with the exception of peptides from dioicin 2 and PAP-IIwhich show null net charge.

Given the apparent heterogeneity of the sequences, we decidedto analyze them through a recently developed bioinformaticsapproach [42] that allows to assign an ‘‘antimicrobial potencyscore” to any peptide of known sequence. This approach is basedon the finding that antimicrobial potency [defined as the Log (1/MIC)] increases linearly with the product CmHnL where C is thenet charge, H is a measure of the hydrophobicity of the hydropho-bic residues and L is the length of the peptide. Exponents m and nare strain dependent variables that determine the relative weightof charge and hydrophobicity on the score and potency of thepeptide. These two values are strain specific and can only be

hlighting in gray refer to conserved residues. Net charge (D) of peptides is indicatedubstitutions (.) are reported. For the accession number of proteins see paragraph 1.5ns (P. aeruginosa and S. aureus) of peptides reported in panel A. Scores have been

E. Pizzo et al. / FEBS Letters 589 (2015) 2812–2818 2817

determined experimentally. We analyzed the peptides of Fig. 7A byusing exponents determined for a Gram negative strain(P. aeruginosa) and a gram positive strain (S. aureus). As shown inFig. 7B also the bioinformatics approach shows a very heteroge-neous picture with scores that range from 0 (for the two peptideswith net charge = 0) to 6.5–7, a score that corresponds to MIC val-ues in the range 100–50 lM [42]. Very interestingly, the highestscores, indicated with asterisks, were found in the case of threeproteins, PD-L2, PAP-S1aci and PAP-S. This analysis suggests theexistence of a new putative family of antimicrobial peptides inPhytolaccaceae and that PDL440–65 is the first member of this classto be experimentally characterized.

4. Conclusions

Cationic antimicrobial peptides (CAMPs) are ancestral elementsof the innate immune system and are present in all multicellularorganisms. The majority of CAMPs, in spite of their different origin,share common structural parameters, as small size, cationicity, andamphipathicity. A wide variety of cationic proteins includes, insidetheir primary structure, CAMPs that can be defined ‘‘cryptic”.PD-L4 is a type 1 RIP from P. dioica well characterized from a struc-tural, functional and enzymatic point of view. This paper reportsfor the first time the characterization of its antimicrobialproperties. Our data indicate that a cryptic peptide, encompassingresidues 40–65 of full-length protein, showed a significant antimi-crobial activity on different bacterial strains and an intriguingimmunomodulatory effect on human cells. An in silico analysisallowed to identify homologous cryptic peptides in other type 1RIPs from Phytolaccaceae. As some of these cryptic CAMPs belongto seed proteins, it could be speculated that these bioactive mole-cules play a key role during seedling development, a processstrongly susceptible to attack by pathogens. This possibility issupported by some studies describing the degradation of RIPsduring seedling development [43,44]. Fragments released by thedegradation process could participate to defense by exerting directantimicrobial effects as shown for PDL440–65.

Conflict of interest

The authors declare that there are no conflicts of interest.

Acknowledgments

This work was supported by funding of Second University ofNaples (located in Caserta) and Italian Cystic Fibrosis Foundation(FFC). Nonetheless, this study was made possible by care and abne-gation of all participants, despite the absence of dedicated fundsand chronic difficulties afflicting the Italian scientific community.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.febslet.2015.08.018.

References

[1] Endo, Y. and Tsurugi, K. (1987) RNA N-glycosidase activity of ricin A-chain.Mechanism of action of the toxic lectin ricin on eukaryotic ribosomes. J. Biol.Chem. 262, 8128–8130.

[2] Girbes, T., Ferreras, J.M., Arias, F.J. and Stirpe, F. (2004) Description,distribution, activity and phylogenetic relationship of ribosome-inactivatingproteins in plants, fungi and bacteria. Mini Rev. Med. Chem. 4, 461–476.

[3] Di Maro, A., Citores, L., Russo, R., Iglesias, R. and Ferreras, J.M. (2014) Sequencecomparison and phylogenetic analysis by the maximum likelihood method of

ribosome-inactivating proteins from angiosperms. Plant Mol. Biol. 85, 575–588.

[4] Barbieri, L., Battelli, M.G. and Stirpe, F. (1993) Ribosome-inactivating proteinsfrom plants. Biochim. Biophys. Acta 1154, 237–282.

[5] Spooner, R.A. and Lord, J.M. (2015) Ricin trafficking in cells. Toxins (Basel) 7,49–65.

[6] Tejero, J., Jimenez, P., Quinto, E.J., Cordoba-Diaz, D., Garrosa, M., Cordoba-Diaz,M., Gayoso, M.J. and Girbes, T. (2015) Elderberries: a source of ribosome-inactivating proteins with lectin activity. Molecules 20, 2364–2387.

[7] Puri, M., Kaur, I., Perugini, M.A. and Gupta, R.C. (2012) Ribosome-inactivatingproteins: current status and biomedical applications. Drug Discov. Today 17,774–783.

[8] Stirpe, F. (2013) Ribosome-inactivating proteins: from toxins to usefulproteins. Toxicon 67, 12–16.

[9] Stirpe, F., Gasperi-Campani, A., Barbieri, L., Falasca, A., Abbondanza, A. andStevens, W.A. (1983) Ribosome-inactivating proteins from the seeds ofSaponaria officinalis L. (soapwort), of Agrostemma githago L. (corn cockle) andof Asparagus officinalis L. (asparagus), and from the latex of Hura crepitans L.(sandbox tree). Biochem. J. 216, 617–625.

[10] Ferreras, J.M. et al. (1993) Distribution and properties of major ribosome-inactivating proteins (28 S rRNA N-glycosidases) of the plant Saponariaofficinalis L. (Caryophyllaceae). Biochim. Biophys. Acta 1216, 31–42.

[11] Youle, R.J. and Huang, A.H. (1976) Protein bodies from the endosperm of castorbean: subfractionation, protein components, lectins, and changes duringgermination. Plant Physiol. 58, 703–709.

[12] Stirpe, F. (2004) Ribosome-inactivating proteins. Toxicon 44, 371–383.[13] Gilabert-Oriol, R., Weng, A., Mallinckrodt, B., Melzig, M.F., Fuchs, H. and

Thakur, M. (2014) Immunotoxins constructed with ribosome-inactivatingproteins and their enhancers: a lethal cocktail with tumor specific efficacy.Curr. Pharm. Des. 20, 6584–6643.

[14] Corrado, G., Bovi, P.D., Ciliento, R., Gaudio, L., Di Maro, A., Aceto, S., Lorito, M.and Rao, R. (2005) Inducible expression of a Phytolacca heterotepala ribosome-inactivating protein leads to enhanced resistance against major fungalpathogens in tobacco. Phytopathology 95, 206–215.

[15] Lodge, J.K., Kaniewski, W.K. and Tumer, N.E. (1993) Broad-spectrum virusresistance in transgenic plants expressing pokeweed antiviral protein. Proc.Natl. Acad. Sci. U.S.A. 90, 7089–7093.

[16] Carlini, C.R. and Grossi-de-Sa, M.F. (2002) Plant toxic proteins with insecticidalproperties. A review on their potentialities as bioinsecticides. Toxicon 40,1515–1539.

[17] Logemann, J., Jach, G., Tommerup, H., Mundy, J. and Schell, J. (1992) Expressionof a barley ribosome-inactivating protein leads to increased fungal protectionin transgenic tobacco plants. Nat. Biotechnol. 10, 305–308.

[18] Barbieri, L., Gorini, P., Valbonesi, P., Castiglioni, P. and Stirpe, F. (1994)Unexpected activity of saponins. Nature 372, 624.

[19] de Virgilio, M., Lombardi, A., Caliandro, R. and Fabbrini, M.S. (2010) Ribosome-inactivating proteins: from plant defense to tumor attack. Toxins (Basel) 2,2699–2737.

[20] Hancock, R.E. and Scott, M.G. (2000) The role of antimicrobial peptides inanimal defenses. Proc. Natl. Acad. Sci. U.S.A. 97, 8856–8861.

[21] Torrent, M., Pulido, D., Valle, J., Nogues, M.V., Andreu, D. and Boix, E. (2013)Ribonucleases as a host-defence family: evidence of evolutionarily conservedantimicrobial activity at the N-terminus. Biochem. J. 456, 99–108.

[22] Mine, Y., Ma, F. and Lauriau, S. (2004) Antimicrobial peptides released byenzymatic hydrolysis of hen egg white lysozyme. J. Agric. Food Chem. 52,1088–1094.

[23] Tomita, M., Takase, M., Wakabayashi, H. and Bellamy, W. (1994) Antimicrobialpeptides of lactoferrin. Adv. Exp. Med. Biol. 357, 209–218.

[24] Papareddy, P., Rydengard, V., Pasupuleti, M., Walse, B., Morgelin, M., Chalupka,A., Malmsten, M. and Schmidtchen, A. (2010) Proteolysis of human thrombingenerates novel host defense peptides. PLoS Pathog. 6, e1000857.

[25] Di Maro, A. et al. (1999) Isolation and characterization of four type-1ribosome-inactivating proteins, with polynucleotide:adenosine glycosidaseactivity, from leaves of Phytolacca dioica L.. Planta 208, 125–131.

[26] Di Maro, A., Chambery, A., Carafa, V., Costantini, S., Colonna, G. and Parente, A.(2009) Structural characterization and comparative modeling of PD-Ls 1–3,type 1 ribosome-inactivating proteins from summer leaves of Phytolaccadioica L.. Biochimie 91, 352–363.

[27] Chambery, A., Pisante, M., Di Maro, A., Di Zazzo, E., Ruvo, M., Costantini, S.,Colonna, G. and Parente, A. (2007) Invariant Ser211 is involved in the catalysisof PD-L4, type I RIP from Phytolacca dioica leaves. Proteins 67, 209–218.

[28] Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: ALaboratory Manual, 2nd ed, Cold Spring Harbor Laboratory Press, Cold SpringHarbor, NY.

[29] Di Maro, A., Ferranti, P., Mastronicola, M., Polito, L., Bolognesi, A., Stirpe, F.,Malorni, A. and Parente, A. (2001) Reliable sequence determination ofribosome-inactivating proteins by combining electrospray massspectrometry and Edman degradation. J. Mass Spectrom. 36, 38–46.

[30] Del Vecchio Blanco, F., Cafaro, V., Di Maro, A., Scognamiglio, R., Siniscalco, G.,Parente, A. and Di Donato, A. (1998) A recombinant ribosome-inactivatingprotein from the plant Phytolacca dioica L. produced from a synthetic gene.FEBS Lett. 437, 241–245.

[31] Williams, B.C., Murphy, T.D., Goldberg, M.L. and Karpen, G.H. (1998)Neocentromere activity of structurally acentric mini-chromosomes inDrosophila. Nat. Genet. 18, 30–37.

2818 E. Pizzo et al. / FEBS Letters 589 (2015) 2812–2818

[32] Zotta, T., Guidone, A., Tremonte, P., Parente, E. and Ricciardi, A. (2012) Acomparison of fluorescent stains for the assessment of viability and metabolicactivity of lactic acid bacteria. World J. Microbiol. Biotechnol. 28, 919–927.

[33] Manzo, N., Di Luccia, B., Isticato, R., D’Apuzzo, E., De Felice, M. and Ricca, E.(2013) Pigmentation and sporulation are alternative cell fates in Bacilluspumilus SF214. PLoS One 8, e62093.

[34] Vivo, M., Ranieri, M., Sansone, F., Santoriello, C., Calogero, R.A., Calabro, V.,Pollice, A. and La Mantia, G. (2013) Mimicking p14ARF phosphorylationinfluences its ability to restrain cell proliferation. PLoS One 8, e53631.

[35] Galdiero, S., Falanga, A., Berisio, R., Grieco, P., Morelli, G. and Galdiero, M.(2015) Antimicrobial peptides as an opportunity against bacterial diseases.Curr. Med. Chem. 22, 1665–1677.

[36] Scholtz, J.M. and Baldwin, R.L. (1992) The mechanism of a-helix formation bypeptides. Annu. Rev. Biophys. Biomol. Struct. 21, 95–118.

[37] Sonnichsen, F.D., Van Eyk, J.E., Hodges, R.S. and Sykes, B.D. (1992) Effect oftrifluoroethanol on protein secondary structure: an NMR and CD study using asynthetic actin peptide. Biochemistry 31, 8790–8798.

[38] Yang, J.J., Pikeathly, M. and Radford, S.E. (1994) Far-UV circular dichroismreveals a conformational switch in a peptide fragment from the beta-sheet ofhen lysozyme. Biochemistry 33, 7345–7353.

[39] Oldfield, C.J. and Dunker, A.K. (2014) Intrinsically disordered proteinsand intrinsically disordered protein regions. Annu. Rev. Biochem. 83,553–584.

[40] Scott, M.G., Vreugdenhil, A.C., Buurman, W.A., Hancock, R.E. and Gold, M.R.(2000) Cutting edge: cationic antimicrobial peptides block the binding oflipopolysaccharide (LPS) to LPS binding protein. J. Immunol. 164, 549–553.

[41] Parente, A., Chambery, A., Di Maro, A., Russo, R. and Severino, V. (2014)Ribosome-inactivating proteins from Phytolaccaceae (Stirpe, F. and Lappi, D.,Eds.), Ribosome-inactivating Proteins: Ricin and Related Proteins, Wiley-Blackwell, Weinheim, Germany.

[42] Pane, K., et al. (2015). Antimicrobial potency of cationic antimicrobial peptidescan be predicted from their amino acid composition: application to thedetection of ‘‘Cryptic” antimicrobial peptides (submitted for publication).

[43] Di Maro, A., Berisio, R., Ruggiero, A., Tamburino, R., Severino, V., Zacchia, E. andParente, A. (2012) Structural and enzymatic properties of an in vivo proteolyticform of PD-S2, type 1 ribosome-inactivating protein from seeds of Phytolaccadioica L.. Biochem. Biophys. Res. Commun. 421, 514–520.

[44] Barnesa, D.J., Baldwinb, B.S. and Braascha, D.A. (2009) Ricin accumulation anddegradation during castor seed development and late germination. Ind. CropsProd. 30, 254–258.