Vertebrate secreted RNases (ribonucleases) are small proteins that play important roles in RNA metabolism, angiogenesis or host defence. In the present study we describe the antimicrobial properties of the N-terminal domain of the hcRNases (human canonical RNases) and show that their antimicrobial activity is well conserved among their lineage. Furthermore, all domains display a similar antimicrobial mechanism, characterized by bacteria agglutination followed by membrane permeabilization. The results of the present study show that, for all antimicrobial hcRNases, (i) activity is retained at the N-terminus and (ii) the antimicrobial mechanism is conserved. Moreover, using computational analysis we show that antimicrobial propensity may be conserved at the N-terminus for all vertebrate RNases, thereby suggesting that a defence mechanism could be a primary function in vertebrate RNases and that the N-terminus was selected to ensure this property. In a broader context, from the overall comparison of the peptides’ physicochemical and biological properties, general correlation rules could be drawn to assist in the structure-based development of antimicrobial agents.

INTRODUCTION

RNase (ribonuclease) A-homologues comprise a vertebrate-specific superfamily that includes a wide network of diverse gene lineages [1]. Eight human members (known as canonical RNases) belong to the RNase A-like family [2,3]. They are all small secreted proteins (approximately 15 kDa), sharing a typical fold, with six or eight cysteine residues arranged in three or four disulfide bonds [49]. All RNases include a conserved catalytic triad comprising two histidine residues and one lysine residue, the latter located inside the characteristic RNase signature (CKXXNTF) [2] (Supplementary Figure S1 at http://www.biochemj.org/bj/456/bj4560099add.htm).

All hcRNases (human canonical RNases) are catalytically active in variable degrees [10] and some of them share relevant antimicrobial properties [11]. However, antimicrobial and catalytic activities of hcRNases are apparently unrelated [9,11,12]. In particular, studies in zebrafish [13] and chicken RNases [14] show they display antimicrobial properties independent of a catalytic activity. Interestingly, vertebrate RNases have high isoelectric points, a common feature in antimicrobial proteins, required for interaction with the negatively charged membranes of pathogens [11,15].

In summary, the presence of antimicrobial activity in diverse lineages of vertebrate RNases suggests that host defence might be a relevant physiological role of the superfamily [16]. However, a sequential or structural evolutionarily selected pattern that embodies the antimicrobial activity has not yet been reported.

In this context, we previously described that, for RNase 3, a functional antimicrobial domain is located at the N-terminus of the protein [1720]. This domain can be further dissected into two active regions: (i) residues 24–45 are essential for the antimicrobial action, and (ii) residues 8–16 contribute to agglutination and membrane destabilization [17].

Given that many RNases display antimicrobial properties unrelated to their ribonuclease activity, it is appealing to hypothesize that an N-terminal antimicrobial domain could be preserved and/or refined along evolution to embody the antimicrobial activity. To test this hypothesis, we have synthesized the N-terminal domains of all hcRNases and studied their antimicrobial properties. The results of the present study reveal that (i) antimicrobial activity is largely confined at the N-terminus, and that (ii) the mechanism of action of the N-terminal domains is similar to that of the original proteins. Moreover, complementary computational analyses suggest that the N-terminal domain in the protein might have been selected by evolution to provide a host-defence function.

MATERIALS AND METHODS

Materials and strains

DOPC (1,2-dioleoyl-sn-glycero-3-phosphocholine) and DOPG (1,2-dioleoyl-sn-glycero-3-phosphoglycerol) were from Avanti Polar Lipids. LPS (lipopolysaccharide) from Escherichia coli was purchased from Sigma–Aldrich.

ANTS (8-aminonaphthalene-1,3,6-trisulfonic acid), DPX (p-xylenebispyridinium bromide), BC [BODIPY® cadaverine, where BODIPY is boron dipyrromethene (4,4-difluoro-4-bora-3a,4a-diaza-s-indacene)] and SYTOX were purchased from Invitrogen. pET11 expression vector and E. coli BL21(DE3) cells were from Novagen.

Strains used were E. coli (BL2, Novagen), Staphylococcus aureus (A.T.C.C. 502A), Acinetobacter baumannii (A.T.C.C. 15308), Pseudomonas sp (A.T.C.C. 15915), Micrococcus luteus (A.T.C.C. 7468) and Enterococcus faecium (A.T.C.C. 19434).

Peptide synthesis

Fmoc-protected amino acids and HBTU [2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate] were obtained from Iris Biotech. Fmoc-Rink-amide (MBHA) resin was from Novabiochem. HPLC-grade ACN (acetonitrile) and peptide-synthesis-grade DMF (dimethylformamide), DIEA (N,N-di-isopropylethylamine) and TFA (trifluoroacetic acid) were from Carlo Erba-SDS.

Solid-phase peptide synthesis was done by Fmoc-based chemistry on Fmoc-Rink-amide (MBHA) resin (0.1 mmol) in a model 433 synthesizer (Applied Biosystems) running FastMoc protocols. Couplings used an 8-fold molar excess each of Fmoc-amino acid and HBTU and a 16-fold molar excess of DIEA. Side chains of trifunctional residues were protected with t-butyl (aspartate, glutamate, serine, threonine and tyrosine), t-butyloxycarbonyl (lysine and tryptophan), 2,2,4,6,7-pentamethyldihydrobenzofuran-5-sulfonyl (arginine) and trityl (asparagine, glutamine and histidine) groups. After chain assembly, full deprotection and cleavage were carried out with TFA/water/tri-isopropylsilane (95:2.5:2.5, by vol.) for 90 min at room temperature (25°C). Peptides were isolated by precipitation with ice-cold diethyl ether and separated by centrifugation (3000 g for 20 min at 4°C), dissolved in 0.1 M acetic acid, and freeze-dried. Analytical reversed-phase HPLC was performed on a Luna C18 column (4.6 mm×50 mm, 3 μm; Phenomenex). Linear 5–60% gradients of solvent B (0.036% TFA in ACN) into solvent A (0.045% TFA in water) were used for elution at a flow rate of 1 ml/min and with UV detection at 220 nm. Preparative HPLC runs were performed on a Luna C18 column (21.2 mm×250 mm, 10 μm; Phenomenex), using linear gradients of solvent B (0.1% in ACN) into solvent A (0.1% TFA in water), as required, with a flow rate of 25 ml/min. MALDI–TOF mass spectra were recorded in the reflector or linear mode in a Voyager DE-STR workstation (Applied Biosystems) using an α-hydroxycinnamic acid matrix. Fractions of adequate (>90%) HPLC homogeneity and with the expected mass were pooled, freeze-dried and used in subsequent experiments.

MIC (minimal inhibitory concentration) and MAC (minimal agglutination concentration) determinations

Antimicrobial activity was expressed as the MIC100, defined as the lowest protein concentration that completely inhibits microbial growth. The MIC of each peptide was determined from two independent experiments performed in triplicate for each concentration. Bacteria were incubated at 37°C overnight in LB broth and diluted to give approximately 5×105 CFU (colony-forming units)/ml. The bacterial suspension was incubated with peptides at various concentrations (0.1–10 μM) at 37°C for 4 h in phosphate buffer. Samples were plated on to Petri dishes and incubated at 37°C overnight.

For MAC determination, bacterial cells were grown at 37°C to mid-exponential phase (D600=0.6), centrifuged at 5000 g for 2 min, and resuspended in LB to give an absorbance of 0.2 at 600 nm. A 200 μl aliquot of the bacterial suspension was incubated with peptides at various (0.1–10 μM) concentrations at 25°C for 4 h. Aggregation behaviour was observed by visual inspection and the MAC was expressed as described previously [20].

Haemolytic activity

Fresh sheep RBCs (red blood cells) (Oxoid) were washed three times with PBS [35 mM phosphate buffer and 0.15 M NaCl (pH 7.4)] by centrifugation for 5 min at 3000g and resuspended in PBS at 2×107 cells/ml. RBCs were incubated with peptides at 37°C for 4 h and centrifuged at 13000 g for 5 min. The supernatant was separated from the pellet and its absorbance measured at 570 nm. T100% haemolysis was defined as the absorbance obtained by sonicating RBCs for 10 s. HC50 was calculated by fitting the data to a sigmoidal function.

FACS assay

Bacterial cells were grown at 37°C to mid-exponential phase (D600=0.6), centrifuged at 5000 g for 2 min, resuspended in 10 mM sodium phosphate buffer and 100 mM NaCl (pH 7.4) to give a final D600=0.2 and pre-incubated for 20 min. A 500 μl aliquot of the bacterial suspension was incubated with 5 μM peptide for 4 h. After incubation, 25000 cells were subjected to FACS analysis using a FACSCalibur cytometer (BD Biosciences) and a dot-plot was generated by representing the low-angle forward scattering (FSC-H) in the x-axis and the side scattering (SSC-H) in the y-axis to analyse the size and complexity of the cell cultures. Results were analysed using FlowJo (Tree Star).

LUV (large unilamellar vesicle) liposome preparation

LUVs containing DOPC, DOPG or DOPC/DOPG (3:2 molar ratio) of a defined size (approximately 100 nm) were prepared as described previously [21]. LUVs were obtained from a vacuum-drying lipid chloroform solution by extrusion through 100 nm polycarbonate membranes. The lipid suspension was frozen and thawed ten times before extrusion. A 1 mM stock solution of liposome suspension in 10 mM phosphate buffer and 100 mM NaCl (pH 7.4) was prepared.

Liposome leakage assay

The ANTS/DPX liposome leakage fluorescence assay was performed as described previously [21]. Briefly, a unique population of LUVs of DOPC/DOPG (3:2 molar ratio) lipids was obtained containing 12.5 mM ANTS, 45 mM DPX, 20 mM NaCl and 10 mM Tris/HCl (pH 7.4). The ANTS/DPX liposome suspension was diluted to a 30 μM concentration and incubated at 25°C in the presence of peptide. Leakage activity was followed by monitoring the increase in fluorescence at 535 nm [22].

Liposome agglutination assay

Liposome agglutination was analysed by DLS (dynamic light scattering) using a Malvern 4700 photon correlation spectrometer (Malvern Instruments). An argon laser (λ=488 nm) was used to cover the wide size range involved. Hydrodynamic radius measurements were always carried out at a reading scattering angle of 90°. From the intensity measurements recorded, data were processed by the CONTIN software (Malvern), and the hydrodynamic diameter, the polydispersity index and the total number of counts were calculated. The incubation buffer was 10 mM Tris/HCl and 100 mM NaCl (pH 7.4). Measurements were performed at 25°C, a 200 μM final liposome concentration, and a 5 μM peptide concentration.

LPS-binding assay

LPS binding was assessed using the fluorescent probe BC as described previously [23]. Briefly, the displacement assay was performed by adding 1–2 μl aliquots of peptide to 1 ml of a continuously stirred mixture of LPS (10 μg/ml) and BC (10 μM) in 10 mM phosphate buffer and 100 mM NaCl (pH 7.5). Fluorescence measurements were performed on a Cary Eclipse spectrofluorimeter, with 580 and 620 nm as BC excitation and emission wavelengths respectively. The excitation slit was set at 2.5 nm, and the emission slit was set at 20 nm. Final values correspond to an average of four replicates and were the mean of a 0.3 s continuous measurement. Occupancy factor was calculated as described previously [23].

CD spectroscopy

Far-UV CD spectra were recorded in a Jasco J-715 spectropolarimeter as described previously [24]. Mean-residue ellipticity [θ] (deg·cm2·dmol−1) was calculated as

 
formula

where θ is the experimental ellipticity (deg), MRW is the mean residue molecular mass of the peptide, c is the molar concentration of peptide and l is the cell path length. Data from four consecutive scans were averaged. Spectra of peptides [4–8 μM in 5 mM sodium phosphate (pH 7.5)] in the absence and presence (1 mM) of SDS and LPS (1 mM, nominal molecular mass=90 kDa) were recorded. Samples were centrifuged for 5 min at 10000 g before use.

SYTOX Green uptake

For SYTOX Green assays, bacterial E. coli and S. aureus cells were grown to mid-exponential growth phase (D600~0.6) in LB medium and then centrifuged (5000 g for 2 min at 25°C), washed and resuspended in phosphate buffer. Cell suspensions (D600~0.2) were incubated with 1 μM SYTOX Green (Invitrogen) for 15 min in the dark before the influx assay. At 2–4 min after initiating data collection, 5 μM peptide was added to the cell suspension, and the increase in SYTOX Green fluorescence was measured (485 and 520 nm excitation and emission wavelengths respectively) for 40 min in a Cary Eclipse spectrofluorimeter. Maximum fluorescence was that resulting from cell lysis with Triton X-100.

Bacterial cytoplasmic membrane depolarization assay

Membrane depolarization was monitored as described previously [25]. Briefly, E. coli and S. aureus strains were grown at 37°C to an D600 of 0.2, centrifuged at 5000 g for 7 min, washed with 5 mM Hepes (pH 7.2) containing 20 mM glucose and resuspended in 5 mM Hepes-KOH, 20 mM glucose and 100 mM KCl (pH 7.2) to an D600 of 0.05. DiSC3(5) (3,3-dipropylthiadicarbocyanine iodide) was added to a final concentration of 0.4 μM, and changes in the fluorescence were continuously recorded after the addition of peptide (5 μM). The concentration required to achieve half of total membrane depolarization was estimated from non-linear regression analysis.

RESULTS

Design of hcRNase-derived peptides

As the reference 1–45 segment in hcRNase 3 has been found to be the minimal domain retaining full antimicrobial properties [20,24], in a search for similar antimicrobial properties, peptides 18, comprising equivalent regions of other RNase N-termini [residues 1–45 of hRNases (human RNases) 2, 6, 7 and 8; residues 1–48 of hRNases 1 and 4; and residues 1–47 of RNase 5; Table 1] were selected for synthesis. In all cases, the peptides included the first two α-helices as well as the first β-strand from the parental protein. As described previously [12], the two cysteine residues in the N-terminal region, disulfide-linked to a distant cysteine residue in the native protein, were replaced by isosteric serine residues. Peptides of satisfactory quality, with the expected molecular mass, were obtained after Fmoc solid-phase synthesis and HPLC purification (Table 1). Their CD secondary structures in the presence of structuring agents [TFE (trifluoroethanol) and DPC (dodecylphosphocholine) micelles] were found to be similar to those of the cognate proteins [24,26].

Table 1
Physicochemical properties of the hcRNase-derived peptides

In every peptide, the underlined serine residue (S) corresponds to cysteine residues in the native protein. Hydrophobicity was computed using the GRAVY scale.

  [M+H]+ 
Peptide Sequence Measured mass (Da) Theoretical mass (Da) Retention time (min)* Hydrophobicity pI 
—KESRAKKFQRQHMDSDSSPSSSSTYSNQMMRRRNMTQGRSKPVNTFVH 5630.45 5628.72 4.93 −1.575 11.40 
KPPQFTWAQWFETQHINMTSQ——QSTNAMQVINNYQRRSKNQNTFLL 5453.32 5452.67 8.56 −1.069 10.28 
RPPQFTRAQWFAIQHISLNPP——RSTIAMRAINNYRWRSKNQNTFL 5478.34 5478.74† 5.05† −0.764 11.88 
–QDGMY-QRFLRQHVHPEET-GGSDRYSNLMMQRRKMTLYHSKRFNTFIH 5812.31 5824.84 6.69 −1.228 10.15 
–QDNSRYTHFLTQHYDAKPQ-GRDDRYSESIMRRRGLTS-PSKDINTFIH 5638.15 5635.76 6.43 −1.430 9.40 
WPKRLTKAHWFEIQHIQPSPL——QSNRAMSGINNYTQHSKHQNTFLH 5410.85 5404.74 6.88 −1.096 10.45 
KPKGMTSSQWFKIQHMQPSPQ——ASNSAMKNINKHTKRSKDLNTFLH 5206.90 5205.67 6.20 −1.209 10.75 
KPKDMTSSQWFKTQHVQPSPQ——ASNSAMSIINKYTERSKDLNTFLH 5206.95 5204.59 7.48 −1.044 9.70 
  [M+H]+ 
Peptide Sequence Measured mass (Da) Theoretical mass (Da) Retention time (min)* Hydrophobicity pI 
—KESRAKKFQRQHMDSDSSPSSSSTYSNQMMRRRNMTQGRSKPVNTFVH 5630.45 5628.72 4.93 −1.575 11.40 
KPPQFTWAQWFETQHINMTSQ——QSTNAMQVINNYQRRSKNQNTFLL 5453.32 5452.67 8.56 −1.069 10.28 
RPPQFTRAQWFAIQHISLNPP——RSTIAMRAINNYRWRSKNQNTFL 5478.34 5478.74† 5.05† −0.764 11.88 
–QDGMY-QRFLRQHVHPEET-GGSDRYSNLMMQRRKMTLYHSKRFNTFIH 5812.31 5824.84 6.69 −1.228 10.15 
–QDNSRYTHFLTQHYDAKPQ-GRDDRYSESIMRRRGLTS-PSKDINTFIH 5638.15 5635.76 6.43 −1.430 9.40 
WPKRLTKAHWFEIQHIQPSPL——QSNRAMSGINNYTQHSKHQNTFLH 5410.85 5404.74 6.88 −1.096 10.45 
KPKGMTSSQWFKIQHMQPSPQ——ASNSAMKNINKHTKRSKDLNTFLH 5206.90 5205.67 6.20 −1.209 10.75 
KPKDMTSSQWFKTQHVQPSPQ——ASNSAMSIINKYTERSKDLNTFLH 5206.95 5204.59 7.48 −1.044 9.70 
*

All peptides except RNase 3 were analysed as described in the Materials and methods section. For RNase 3, data are from [12].

Data from [12]; HPLC as described in the Materials and methods section, except that the gradient was 20–45% solvent B into solevnt A over 15 min.

Antimicrobial activity of hcRNases is retained at the N-terminus

To assess the antimicrobial activity of peptides 18 we determined their MIC for three representative Gram-negative and Gram-positive bacteria (Table 2). Five peptides (1, 3, 4, 6 and 7) tested active against all strains, with MIC values between 0.2 and 10 μM, a range close to that observed for several hcRNases [9,17,24]. In contrast, peptides 2, 5 and 8 were inactive against all strains. When these results are compared with available literature data on hcRNases, a good agreement between the two sets of data is observed (Table 2), with only peptide 8 differing from previously reported results [27]. As we suspected that the discrepancy observed could be due to the inoculum size (5×105 cells/ml in our case, 5×104 cells/ml in Rudolph et al. [27]), peptide 8 was reassayed using a 5×104 cells/ml inoculum and MIC values in the 1–10 μM range, compatible with the previous report [27], were determined. Thus we conclude that the antimicrobial activity of hcRNases is consistently preserved at the N-terminus.

Table 2
Antimicrobial and haemolytic activities of the hcRNase N-terminal peptides

n.r., not reported (antimicrobial activity has not been assayed for some proteins).

 Antimicrobial activity MIC100 (μM) 
 Gram-negative bacteria Gram-positive bacteria Haemolytic activity HC50 (μM) 
Peptide E. coli P. aeruginosa A. baumanii S. aureus M. luteus E. faecium Protein antimicrobial activity* Sheep RBCs 
n.r. 11.6 
>10 >10 >10 >10 >10 10 − >100 
0.4 0.4 0.3 0.5 0.6 0.6 +++ 9.5 
0.3 0.4 0.5 0.3 1.2 0.8 n.r. 15.2 
>10 >10 >10 >10 >10 >10 − >100 
1.2 1.2 1.2 2.5 2.5 n.r. 17.4 
1.2 1.2 1.2 10 2.5 +++ 10.4 
>10 10 >10 >10 >10 >10 15.1 
 Antimicrobial activity MIC100 (μM) 
 Gram-negative bacteria Gram-positive bacteria Haemolytic activity HC50 (μM) 
Peptide E. coli P. aeruginosa A. baumanii S. aureus M. luteus E. faecium Protein antimicrobial activity* Sheep RBCs 
n.r. 11.6 
>10 >10 >10 >10 >10 10 − >100 
0.4 0.4 0.3 0.5 0.6 0.6 +++ 9.5 
0.3 0.4 0.5 0.3 1.2 0.8 n.r. 15.2 
>10 >10 >10 >10 >10 >10 − >100 
1.2 1.2 1.2 2.5 2.5 n.r. 17.4 
1.2 1.2 1.2 10 2.5 +++ 10.4 
>10 10 >10 >10 >10 >10 15.1 
*

Antimicrobial activity of the parental proteins obtained from [17].

To characterize the cell selectivity of peptides 18, their haemolytic activity was tested on sheep RBCs. For antimicrobially active peptides (1, 3, 4 and 68), haemolytic activities (measured as HC50) between 10 and 20 μM were found (Table 2), whereas no haemolysis (HC50>100 μM) was detected for antimicrobially inert peptides 2 and 5. For the most antimicrobially active peptides (3, 4, 6 and 7) MICs were one or two orders of magnitude lower than HCs, suggesting that hcRNase-derived peptides are good candidates for drug development, as already shown for RNase 3 [20].

The antimicrobial mechanism of hcRNases primarily involves the N-terminus

To characterize the antimicrobial mechanism of hcRNase-derived peptides 18, we first assessed their ability to depolarize and permeabilize bacterial cytoplasmic membranes. Results with representative Gram-negative and Gram-positive organisms (Table 3) show that only the peptides with high bactericidal activity (3, 4, 6 and 7) can cause significant membrane depolarization and permeabilization, in agreement with the antimicrobial values (Table 2).

Table 3
Depolarization and bacteria leakage IC50 values for hcRNase-derived peptides
 Depolarization (nM) Bacteria leakage (nM) 
Peptide E. coli S. aureus E. coli S. aureus 
150±10 350±13 >10 9.1±1 
>1000 >1000 >10 >10 
40±1 50±7 3.8±0.1 4.5±0.5 
40±3 30±2 3.8±0.2 6.1±0.1 
>1000 >1000 >10 >10 
40±4 40±2 5.0±0.3 7.3±0.7 
40±3 70±1 5.7±0.1 7.3±0.2 
>1000 >1000 >10 >10 
 Depolarization (nM) Bacteria leakage (nM) 
Peptide E. coli S. aureus E. coli S. aureus 
150±10 350±13 >10 9.1±1 
>1000 >1000 >10 >10 
40±1 50±7 3.8±0.1 4.5±0.5 
40±3 30±2 3.8±0.2 6.1±0.1 
>1000 >1000 >10 >10 
40±4 40±2 5.0±0.3 7.3±0.7 
40±3 70±1 5.7±0.1 7.3±0.2 
>1000 >1000 >10 >10 

Another specific characteristic of antimicrobial proteins, already demonstrated for the N-terminal domain of hcRNase 3 (peptide 3) [12,15], is the ability to agglutinate Gram-negative bacteria, judged as important in holding pathogens at bay at the infection focus. To test whether other hcRNase-derived peptides shared this property, their MAC was determined (Table 4). Interestingly, all antimicrobially active peptides (1, 3, 4, 6 and 7) showed agglutinating properties to a certain degree, suggesting that this antimicrobial feature is conserved for all hcRNases. Furthermore, MAC values correlated well with LPS-binding affinities (Table 4 and Supplemenatry Table S1 at http://www.biochemj.org/bj/456/bj4560099add.htm), suggesting that binding to the Gram-negative bacterial cell wall would trigger the agglutinating mechanism, as described previously [27a]. In agreement with this, no agglutination was observed for Gram-positive S. aureus, confirming the fact that an LPS external layer is necessary for bacterial agglutination. To further explore these agglutinating properties, we used FACS. After incubation of E. coli cell cultures with the hcRNase-derived peptides, significant agglutination was only observed for active peptides 1, 3, 4, 6 and 7 (Figure 1A), with peptide 3, followed by 6, displaying the highest activity. Additionally, bacterial cells were incubated with SYTO9 and propidium iodide to assess their viability simultaneously with agglutination (Figure 1B). Results showed dead bacteria to be present only when significant agglutinating activity was also detected, suggesting that both activities are part of a sole mechanism. One may therefore generalize the behaviour of all active hcRNase-derived peptides as having the ability to dock on to external LPS layers and promote Gram-negative bacteria agglutination, triggering, in turn, membrane depolarization and lipid bilayer leakage, eventually resulting in cell death. For Gram-positive cells, peptides would probably diffuse across the permeable cell wall and exert their lytic action on to the cell membrane [25].

Table 4
LPS affinity, agglutination activity and liposome leakage of hcRNase-derived peptides

Values are expressed as ED50. n.d., not detected at the assayed concentration range (0.1–10 μM).

   Liposome leakage (nM) 
Peptide LPS binding (μM) MAC* (μM) DOPC DOPC/DOPG DOPG 
5.6±1.3 5.0±0.5 >1000 >1000 206±77 
7.1±1.0 n.d. >1000 >1000 >1000 
2.9±0.3 0.9±0.1 >1000 95±5 99±2 
1.3±0.2 5.0±0.5 >1000 580±10 53±5 
n.d. n.d. >1000 >1000 >1000 
1.9±0.3 1.8±0.1 >1000 210±30 86±10 
3.2±1.2 5±0.5 >1000 180±20 108±4 
n.d. n.d. >1000 >1000 >1000 
   Liposome leakage (nM) 
Peptide LPS binding (μM) MAC* (μM) DOPC DOPC/DOPG DOPG 
5.6±1.3 5.0±0.5 >1000 >1000 206±77 
7.1±1.0 n.d. >1000 >1000 >1000 
2.9±0.3 0.9±0.1 >1000 95±5 99±2 
1.3±0.2 5.0±0.5 >1000 580±10 53±5 
n.d. n.d. >1000 >1000 >1000 
1.9±0.3 1.8±0.1 >1000 210±30 86±10 
3.2±1.2 5±0.5 >1000 180±20 108±4 
n.d. n.d. >1000 >1000 >1000 
*

MAC values obtained with E. coli cultures.

Bacterial agglutination measured by FACS

Figure 1
Bacterial agglutination measured by FACS

E. coli cell cultures were incubated with hcRNase-derived peptides at a 5 μM concentration for 4 h and analysed using a FACSCalibur cytometer. Low-angle forward scattering (FSC-H) is represented on the x-axis and the side scattering (SSC-H) on the y-axis to analyse the size and complexity of the cell cultures. (A) Plots are coloured to show cell density. Low (blue) to high (red) cell densities are represented. The percentages of cells belonging to each quadrant are indicated. (B) Plots are coloured to show cell viability. Live cells (stained with Syto9, green) and dead cells (stained with propidium iodide, red) are coloured in the plot. Control live and dead cells were analysed as a reference.

Figure 1
Bacterial agglutination measured by FACS

E. coli cell cultures were incubated with hcRNase-derived peptides at a 5 μM concentration for 4 h and analysed using a FACSCalibur cytometer. Low-angle forward scattering (FSC-H) is represented on the x-axis and the side scattering (SSC-H) on the y-axis to analyse the size and complexity of the cell cultures. (A) Plots are coloured to show cell density. Low (blue) to high (red) cell densities are represented. The percentages of cells belonging to each quadrant are indicated. (B) Plots are coloured to show cell viability. Live cells (stained with Syto9, green) and dead cells (stained with propidium iodide, red) are coloured in the plot. Control live and dead cells were analysed as a reference.

Previously we showed that the hcRNase 3 N-terminal domain (peptide 3) retains the ability of the cognate protein to bind and disrupt model membranes [20,24]. To determine whether all hcRNase-derived peptides behave similarly, we studied the effect of the peptides on lipid bilayer models. Changes in the tryptophan spectrum upon incubation with liposomes, a usual readout for membrane interaction, could only be assessed for peptides bearing this residue (2, 3, 6, 7 and 8; Supplementary Table S1), but in all cases showed a good agreement with antimicrobial activity. As reported for many other antimicrobial peptides, hcRNase-derived peptides were able to bind anionic liposomes, whereas no significant binding was observed for neutral ones, suggesting that electrostatic interactions are indispensable for peptide binding to lipid vesicles.

Next, the ability of the peptides to perturb bilayer stability was tested in a liposome leakage assay (Table 4). As expected, only active peptides displayed high leakage activity, and values observed for liposome content release were in good agreement with those observed for depolarization activity in bacteria (Table 3).

For a better characterization of the mechanism, the ability of hcRNase-derived peptides to agglutinate liposomes was studied by DLS, which allows monitoring of the vesicle size. All peptides were tested against both neutral and negatively charged vesicles (Figure 2), but only the latter agglutinated upon peptide interaction, again highlighting the role of electrostatic interactions in promoting agglutination. More interestingly, the only peptides capable of promoting agglutination in this assay with model membranes were those previously found to be active in either MAC or FACS experiments (Table 3 and Figure 1).

Liposome aggregation measured by DLS

Figure 2
Liposome aggregation measured by DLS

Liposomes containing DOPC (light grey), DOPG (black) or a 3:2 molar mixture of DOPC/DOPG (dark grey) were incubated with 5 μM hcRNase-derived peptides and the size was measured after 15 min by DLS using a Malvern 4700 spectrometer. The plots represent the intensity against the liposome mean size. For peptides 6 and 7 with DOPG liposomes, large aggregates were observed which made it difficult to determine a defined particle mean size.

Figure 2
Liposome aggregation measured by DLS

Liposomes containing DOPC (light grey), DOPG (black) or a 3:2 molar mixture of DOPC/DOPG (dark grey) were incubated with 5 μM hcRNase-derived peptides and the size was measured after 15 min by DLS using a Malvern 4700 spectrometer. The plots represent the intensity against the liposome mean size. For peptides 6 and 7 with DOPG liposomes, large aggregates were observed which made it difficult to determine a defined particle mean size.

As some hcRNases are rich in lysine residues (e.g. RNase 7) while others are rich in arginine residues (e.g. RNase 3) we committed ourselves to test whether the content of arginine or lysine could modulate the mechanism of action of the hcRNases N-termini and synthesized the lysine and arginine variants of peptides 3 and 7 respectively. In all cases, we did not find any significant differences either in the antimicrobial properties of the peptides or in their biophysical properties (Supplementary Table S2 at http://www.biochemj.org/bj/456/bj4560099add.htm). Both variants had identical activities and were able to depolarize and agglutinate bacteria with the same efficiency as their natural partners (Supplementary Table S2). Therefore we conclude that these residues are interchangeable in terms of their contribution to the antimicrobial activities of the N-terminal domain.

In summary, results obtained using model membranes correlated well with those using bacterial cultures, hence reinforcing our hypothesis that bactericidal mechanisms are consistently conserved at the N-termini of antimicrobially active hcRNases.

Vertebrate RNases share a conserved N-terminal antimicrobial domain

A next logical step was to analyse the antimicrobial activity of the N-terminus across vertebrate RNases. To this end, we analysed all RNase A homologues deposited at the Uniprot (http://www.uniprot.org) database using AMPA, an algorithm that scans proteins in search of potential antimicrobial regions (http://tcoffee.crg.cat/apps/ampa/) [15,28]. Results are summarized in Figure 3 in the form of a phylogenetic tree of vertebrate RNases together with AMPA-identified putative antimicrobial regions at the N-termini of these proteins. It can be seen that most of these N-termini are potentially antimicrobial, in congruence with the experimental data in the present study (Table 1). Results in Figure 3, particularly the evolutionary distances among different hcRNases, suggest that N-terminal regions endowed with antimicrobial properties have been central in the function of RNases since their early origins and have accordingly been conserved along evolution.

Conservation of the N-terminal antimicrobial domain among vertebrate secreted RNases

Figure 3
Conservation of the N-terminal antimicrobial domain among vertebrate secreted RNases

The evolutionary history was inferred using the UPGMA method [39]. The optimal tree is shown. The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the Poisson distribution method [40] and are in units of number of amino acid substitutions per site. All positions containing gaps and missing data were eliminated. Evolutionary analyses were conducted in MEGA5 [41]. All sequences were analysed using the AMPA algorithm [28]; detected antimicrobial regions in the N-terminal domain are depicted next to the corresponding tree branch. Each type of RNase family is depicted in a different colour.

Figure 3
Conservation of the N-terminal antimicrobial domain among vertebrate secreted RNases

The evolutionary history was inferred using the UPGMA method [39]. The optimal tree is shown. The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the Poisson distribution method [40] and are in units of number of amino acid substitutions per site. All positions containing gaps and missing data were eliminated. Evolutionary analyses were conducted in MEGA5 [41]. All sequences were analysed using the AMPA algorithm [28]; detected antimicrobial regions in the N-terminal domain are depicted next to the corresponding tree branch. Each type of RNase family is depicted in a different colour.

DISCUSSION

RNases are widely distributed in most vertebrate organs and tissues and can exert many different functions. Interestingly, many members of the RNase A superfamily, not only in mammals but also in other vertebrates, like fish and amphibians [1,14,16], show antimicrobial properties apparently unrelated to their nuclease activity. It has indeed been suggested that RNases might have emerged as a host-defence family in vertebrate evolution [16].

Although the three-dimensional structure of RNase A-like proteins is highly conserved, their amino acid sequence is quite diverse and a defined antimicrobial signature cannot be easily delineated. This is indeed the case for most antimicrobial peptides and proteins, sharing microbicidal properties despite low sequence similarity. Poor sequence homology has in practice prevented the development of robust methods for predicting antimicrobial motifs in proteins and peptides [15,28,29]. However, despite the lack of sequential or structural patterns, similar physicochemical characteristics are shared among antimicrobial peptides, like a net positive charge or the high hydrophobic amino acid content [30].

Previously, we described that the antimicrobial properties of hRNase 3 are highly retained in its N-terminal domain [20,24]. This domain is equipotent compared with the whole protein, killing both Gram-negative and Gram-positive bacteria and retains its biophysical properties, including membrane agglutination and leakage [20,24]. More interestingly, the chemically synthesized antimicrobial domain has a similar structure compared with its counterpart in the native protein in membrane-like environments [26]. However, as pointed out below, there is no significant sequence similarity between the antimicrobial N-terminus of hRNase 3 and those of other antimicrobial RNases (see the Supplementary Online Data and Supplementary Figures S1 and S2 at http://www.biochemj.org/bj/456/bj4560099add.htm). For instance, the residues reported to be critical for antimicrobial activity in hRNases 3 and 7 are not significantly conserved (Supplementary Figure S1) [17]. Nonetheless, all hcRNase N-termini have a high isolectric point and similar hydrophobicity (Table 1).

In the present study, we have shown that all N-terminal antimicrobial domains of hcRNases with described antimicrobial activity are active against both Gram-positive and Gram-negative bacteria (Table 2). Even more interesting, in the cases where enough experimental information on the antimicrobial activity of the proteins is available, the peptides tested have similar activity (Table 2). We have also analysed the biophysical characteristics of the peptides and evaluated the correlation between their ability to disrupt model membranes and their inherent antimicrobial activity (Tables 2 and 3). Liposome leakage shows a good correlation with both antimicrobial activity and membrane depolarization (Supplementary Figure S7 at http://www.biochemj.org/bj/456/bj4560099add.htm). The correspondence between leakage and depolarization is particularly interesting as it suggests that the antimicrobial peptides will eventually promote a disruptive action at the membrane level, rather than interacting with internal targets in the microbes. Interestingly, there is a significant correlation between the ability to acquire a defined structure in membrane-like environments (e.g. SDS micelles) and the antimicrobial activity (Supplementary Figures S6 and S8 at http://www.biochemj.org/bj/456/bj4560099add.htm), suggesting that peptides would adopt an α-helical structure to perturb the bacterial membrane integrity. In fact, all RNases conserve the helical structure at their N-terminus (Supplementary Figure S9 at http://www.biochemj.org/bj/456/bj4560099add.htm).

Indeed, all hcRNase N-terminal peptides with antimicrobial activity show significant agglutination activity, suggesting that this trait might be inherent to antimicrobial RNases. The agglutinating activity of hcRNase 3 has been previously attributed to a high aggregation-prone region at the N-terminus (residues 8–16) [19,25]. Removal of this corresponding hydrophobic patch inside the antimicrobial domain drastically reduces hcRNase 3 antimicrobial activity, particularly in Gram-negative species. A comparison of the computed aggregation propensity profiles of all RNases shows a significant correlation between the aggregation propensity in this particular region and the observed agglutinating activity (Supplementary Figures S3 and S4 at http://www.biochemj.org/bj/456/bj4560099add.htm), strengthening the conclusion that bacterial agglutination may be a characteristic trait of the antimicrobial mechanism of action. Bacterial agglutination activity is also associated with the ability to bind LPS (Table 4). In fact, agglutination has been shown to take place at the external layers of the Gram-negative bacterial cell wall [23,25]. Accordingly, we suggested that following RNase binding to the external layers of Gram-negative bacteria, a slight conformational change in the protein would expose the N-terminal hydrophobic patch to the surface, thereby triggering agglutination by protein–protein interaction. Bacterial agglutination would therefore correlate with the hydrophobic nature of the aggregation-prone region (Supplementary Figure S5). A side-by-side comparison of the two best-studied antimicrobial hcRNases, 3 and 7, corroborate this behaviour, showing that the agglutination potency is far more intense for RNase 3 than for RNase 7 [25]. In this light, it is reasonable to conclude that both hcRNases and their derived peptides will act by a membrane perturbation mechanism, which supports the hypothesis that RNase activity is not required for the antimicrobial action. Notwithstanding, we should also consider that within the context of the whole RNase A family the distinct members would combine their diverse abilities, and among them their catalytic activity, providing a fast and effective multifaceted mechanism to combat pathogens simultaneously at several cellular targets. Indeed, the most recent reviews on antimicrobial peptides point at such a multifunctional working mode [31,32]

Other literature reports also suggest that the N-terminus of RNases might be the main region responsible for their cytotoxicity. For example, Picone and co-workers [33] have found that increasing the positive charge at the N-terminus of bovine seminal RNase enhances membrane interaction and antitumour activity. Interestingly, we have also found comparable behaviours in other cytotoxic RNase families, such as the contribution of the N-terminus of ribotoxin in the interaction with lipid bilayers [34].

The present experimental findings suggest that RNase N-termini might have been selected in evolution to execute host-defence-related functions, particularly those dependent on a cytotoxic action. To strengthen this conclusion, we have analysed the antimicrobial profile of all secreted vertebrate RNases using the AMPA algorithm. The analysis output (Figure 3) shows that most RNases have putative antimicrobial domains at the N-terminal region whose sequences overlap with the antimicrobial-prone region described for hcRNase 3. It is thus reasonable to suggest that both antimicrobial and catalytic activities in the RNase A-like superfamily have evolved independently and that evolution has used the N-terminal domain of RNases to develop anti-pathogen properties. Hence, the results of the present study support the hypothesis that RNases might have emerged as a host-defence family.

Unavoidably, the question about the role of the C-terminal part of the protein remains and has only been partially addressed until now. Although previous studies discard a primary contribution of the C-terminus in the protein bactericidal activity [18,35], we cannot disregard the fact that the active site architecture, shared by all family members, consists of both N- and C-termini. Moreover, residues all along the protein surface are involved in the recognition of the RNA polymer, as identified in some family members studied [3638]. In this context, one may speculate on an evolutionary constraint to retain those residues implied in the RNA recognition, which could partially account for the differences observed in RNases in terms of catalytic activity [36]. Therefore further investigations are required to provide a complete scenario of the role of the distinct protein domains and their potential contribution in complementary tasks that may prove physiologically relevant.

In conclusion, we have shown that the antimicrobial properties of human RNases are retained in their N-terminal domain. Despite the low sequence conservation, the latent antimicrobial properties of this N-terminal domain in many RNase homologues suggest that this region might have evolved to provide host-defence functions for the protein. In this regard, antimicrobial RNases, and more specifically their N-terminal domains, would therefore constitute suitable scaffolds useful as lead templates for the development of wide-spectrum antibiotics.

Abbreviations

     
  • ACN

    acetonitrile

  •  
  • ANTS

    8-aminonaphthalene-1,3,6-trisulfonic acid

  •  
  • BC

    BODIPY® cadaverine

  •  
  • DIEA

    N,N-di-isopropylethylamine

  •  
  • DLS

    dynamic light scattering

  •  
  • DOPC

    1,2-dioleoyl-sn-glycero-3-phosphocholine

  •  
  • DOPG

    1,2-dioleoyl-sn-glycero-3-phosphoglycerol

  •  
  • DPX

    p-xylenebispyridinium bromide

  •  
  • HBTU

    2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate

  •  
  • hcRNase

    human canonical ribonuclease

  •  
  • hRNase

    human ribonuclease

  •  
  • LPS

    lipopolysaccharide

  •  
  • LUV

    large unilamellar vesicle

  •  
  • MAC

    minimal agglutination concentration

  •  
  • MIC

    minimal inhibitory concentration

  •  
  • RBC

    red blood cell

  •  
  • RNase

    ribonuclease

  •  
  • TFA

    trifluoroacetic acid

AUTHOR CONTRIBUTION

Marc Torrent, Ester Boix and David Andreu conceived and designed the experiments. Javier Valle synthesized the peptides. Marc Torrent and David Pulido performed the experiments. Marc Torrent, David Pulido, M. Victòria Nogués, David Andreu and Ester Boix analysed the data. M. Victòria Nogués, David Andreu and Ester Boix contributed reagents, materials and analysis tools. Marc Torrent, David Andreu and Ester Boix wrote the paper.

FUNDING

This work was supported by the European Union [grant number HEALTH-F3-2008-223414], the Ministerio de Ciencia e Innovación [grant numbers BIO2008-04487-CO3-02, BFU2009-09371, FEDER funds], the Ministerio de Economía y Competitividad [grant number BFU2012-38965] and the Generalitat de Catalunya [grant numbers SGR2009-494, SGR2009-795]. M.T. is a recipient of a Beatriu de Pinós fellowship. D.P. is a recipient of an FPU fellowship (MICINN; Ministerio de Ciencia e Innovación). Microscopy experiments were carried out at the Servei de Microscòpia (Universitat Autònoma de Barcelona) and fluorescence measurements at the Servei d’Anàlisi i Fotodocumentació (Universitat Autònoma de Barcelona). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the paper.

References

References
1
Dyer
K. D.
Rosenberg
H. F.
The RNase a superfamily: generation of diversity and innate host defense
Mol. Diversity
2006
, vol. 
10
 (pg. 
585
-
597
)
2
Sorrentino
S.
The eight human “canonical” ribonucleases: molecular diversity, catalytic properties, and special biological actions of the enzyme proteins
FEBS Lett.
2010
, vol. 
584
 (pg. 
2194
-
2200
)
3
Cho
S.
Beintema
J. J.
Zhang
J.
The ribonuclease A superfamily of mammals and birds: identifying new members and tracing evolutionary histories
Genomics
2005
, vol. 
85
 (pg. 
208
-
220
)
4
Kover
K. E.
Bruix
M.
Santoro
J.
Batta
G.
Laurents
D. V.
Rico
M.
The solution structure and dynamics of human pancreatic ribonuclease determined by NMR spectroscopy provide insight into its remarkable biological activities and inhibition
J. Mol. Biol.
2008
, vol. 
379
 (pg. 
953
-
965
)
5
Mosimann
S. C.
Newton
D. L.
Youle
R. J.
James
M. N.
X-ray crystallographic structure of recombinant eosinophil-derived neurotoxin at 1.83 Å resolution
J. Mol. Biol.
1996
, vol. 
260
 (pg. 
540
-
552
)
6
Laurents
D. V.
Bruix
M.
Jimenez
M. A.
Santoro
J.
Boix
E.
Moussaoui
M.
Nogués
M. V.
Rico
M.
The 1H, 13C, 15N resonance assignment, solution structure, and residue level stability of eosinophil cationic protein/RNase 3 determined by NMR spectroscopy
Biopolymers
2009
, vol. 
91
 (pg. 
1018
-
1028
)
7
Terzyan
S. S.
Peracaula
R.
de Llorens
R.
Tsushima
Y.
Yamada
H.
Seno
M.
Gomis-Rüth
F. X.
Coll
M.
The three-dimensional structure of human RNase 4, unliganded and complexed with d(Up), reveals the basis for its uridine selectivity
J. Mol. Biol.
1999
, vol. 
285
 (pg. 
205
-
214
)
8
Acharya
K. R.
Shapiro
R.
Allen
S. C.
Riordan
J. F.
Vallee
B. L.
Crystal structure of human angiogenin reveals the structural basis for its functional divergence from ribonuclease
Proc. Natl. Acad. Sci. U.S.A.
1994
, vol. 
91
 (pg. 
2915
-
2919
)
9
Huang
Y. C.
Lin
Y. M.
Chang
T. W.
Wu
S. J.
Lee
Y. S.
Chang
M. D.
Chen
C.
Wu
S. H.
Liao
Y. D.
The flexible and clustered lysine residues of human ribonuclease 7 are critical for membrane permeability and antimicrobial activity
J. Biol. Chem.
2007
, vol. 
282
 (pg. 
4626
-
4633
)
10
Sorrentino
S.
Libonati
M.
Structure–function relationships in human ribonucleases: main distinctive features of the major RNase types
FEBS Lett.
1997
, vol. 
404
 (pg. 
1
-
5
)
11
Boix
E.
Nogues
M. V.
Mammalian antimicrobial proteins and peptides: overview on the RNase A superfamily members involved in innate host defence
Mol. BioSyst.
2007
, vol. 
3
 (pg. 
317
-
335
)
12
Rosenberg
H. F.
Recombinant human eosinophil cationic protein. Ribonuclease activity is not essential for cytotoxicity
J. Biol. Chem.
1995
, vol. 
270
 (pg. 
7876
-
7881
)
13
Pizzo
E.
Varcamonti
M.
Di Maro
A.
Zanfardino
A.
Giancola
C.
D’Alessio
G.
Ribonucleases with angiogenic and bactericidal activities from the Atlantic salmon
FEBS J.
2008
, vol. 
275
 (pg. 
1283
-
1295
)
14
Nitto
T.
Dyer
K. D.
Czapiga
M.
Rosenberg
H. F.
Evolution and function of leukocyte RNase A ribonucleases of the avian species, Gallus gallus
J. Biol. Chem.
2006
, vol. 
281
 (pg. 
25622
-
25634
)
15
Torrent
M.
Nogues
V. M.
Boix
E.
A theoretical approach to spot active regions in antimicrobial proteins
BMC Bioinf.
2009
, vol. 
10
 pg. 
373
 
16
Cho
S.
Zhang
J.
Zebrafish ribonucleases are bactericidal: implications for the origin of the vertebrate RNase A superfamily
Mol. Biol. Evol.
2007
, vol. 
24
 (pg. 
1259
-
1268
)
17
Boix
E.
Salazar
V. A.
Torrent
M.
Pulido
D.
Nogues
M. V.
Moussaoui
M.
Structural determinants of the eosinophil cationic protein antimicrobial activity
Biol. Chem.
2012
, vol. 
393
 (pg. 
801
-
815
)
18
Sanchez
D.
Moussaoui
M.
Carreras
E.
Torrent
M.
Nogues
V.
Boix
E.
Mapping the eosinophil cationic protein antimicrobial activity by chemical and enzymatic cleavage
Biochimie
2011
, vol. 
93
 (pg. 
331
-
338
)
19
Torrent
M.
Odorizzi
F.
Nogues
M. V.
Boix
E.
Eosinophil cationic protein aggregation: identification of an N-terminus amyloid prone region
Biomacromolecules
2010
, vol. 
11
 (pg. 
1983
-
1990
)
20
Torrent
M.
Pulido
D.
de la Torre
B. G.
Garcia-Mayoral
M. F.
Nogues
M. V.
Bruix
M.
Andreu
D.
Boix
E.
Refining the eosinophil cationic protein antibacterial pharmacophore by rational structure minimization
J. Med. Chem.
2011
, vol. 
54
 (pg. 
5237
-
5244
)
21
Torrent
M.
Cuyas
E.
Carreras
E.
Navarro
S.
Lopez
O.
de la Maza
A.
Nogués
M. V.
Reshetnyak
Y. K.
Boix
E.
Topography studies on the membrane interaction mechanism of the eosinophil cationic protein
Biochemistry
2007
, vol. 
46
 (pg. 
720
-
733
)
22
Torrent
M.
Sanchez
D.
Buzon
V.
Nogues
M. V.
Cladera
J.
Boix
E.
Comparison of the membrane interaction mechanism of two antimicrobial RNases: RNase 3/ECP and RNase 7
Biochim. Biophys. Acta
2009
, vol. 
1788
 (pg. 
1116
-
1125
)
23
Torrent
M.
Navarro
S.
Moussaoui
M.
Nogues
M. V.
Boix
E.
Eosinophil cationic protein high-affinity binding to bacteria-wall lipopolysaccharides and peptidoglycans
Biochemistry
2008
, vol. 
47
 (pg. 
3544
-
3555
)
24
Torrent
M.
de la Torre
B. G.
Nogues
V. M.
Andreu
D.
Boix
E.
Bactericidal and membrane disruption activities of the eosinophil cationic protein are largely retained in an N-terminal fragment
Biochem. J.
2009
, vol. 
421
 (pg. 
425
-
434
)
25
Torrent
M.
Badia
M.
Moussaoui
M.
Sanchez
D.
Nogues
M. V.
Boix
E.
Comparison of human RNase 3 and RNase 7 bactericidal action at the Gram-negative and Gram-positive bacterial cell wall
FEBS J.
2010
, vol. 
277
 (pg. 
1713
-
1725
)
26
Garcia-Mayoral
M. F.
Moussaoui
M.
de la Torre
B. G.
Andreu
D.
Boix
E.
Nogues
M. V.
Rico
M.
Laurents
D. V.
Bruix
M.
NMR structural determinants of eosinophil cationic protein binding to membrane and heparin mimetics
Biophys. J.
2010
, vol. 
98
 (pg. 
2702
-
2711
)
27
Rudolph
B.
Podschun
R.
Sahly
H.
Schubert
S.
Schroder
J. M.
Harder
J.
Identification of RNase 8 as a novel human antimicrobial protein
Antimicrob. Agents Chemother.
2006
, vol. 
50
 (pg. 
3194
-
3196
)
27a
Torrent
M.
Pulido
D.
Nogués
M. V.
Boix
E.
Exploring new biological functions of amyloids: bacteria cell agglutination mediated by host protein aggregation
PLoS Pathog.
2012
, vol. 
8
 pg. 
e1003005
 
28
Torrent
M.
Di Tommaso
P.
Pulido
D.
Nogues
M. V.
Notredame
C.
Boix
E.
Andreu
D.
AMPA: an automated web server for prediction of protein antimicrobial regions
Bioinformatics
2012
, vol. 
28
 (pg. 
130
-
131
)
29
Torrent
M.
Nogues
M. V.
Boix
E.
Discovering new in silico tools for antimicrobial peptide prediction
Curr. Drug Targets
2012
, vol. 
13
 (pg. 
1148
-
1157
)
30
Torrent
M.
Andreu
D.
Nogues
V. M.
Boix
E.
Connecting peptide physicochemical and antimicrobial properties by a rational prediction model
PLoS ONE
2011
, vol. 
6
 pg. 
e16968
 
31
Nguyen
L. T.
Haney
E. F.
Vogel
H. J.
The expanding scope of antimicrobial peptide structures and their modes of action
Trends Biotechnol.
2011
, vol. 
29
 (pg. 
464
-
472
)
32
Yeung
A. T.
Gellatly
S. L.
Hancock
R. E.
Multifunctional cationic host defence peptides and their clinical applications
Cell. Mol. Life Sci.
2011
, vol. 
68
 (pg. 
2161
-
2176
)
33
D’Errico
G.
Ercole
C.
Lista
M.
Pizzo
E.
Falanga
A.
Galdiero
S.
Spadaccini
R.
Picone
D.
Enforcing the positive charge of N-termini enhances membrane interaction and antitumor activity of bovine seminal ribonuclease
Biochim. Biophys. Acta
2011
, vol. 
1808
 (pg. 
3007
-
3015
)
34
Alvarez-Garcia
E.
Martinez-del-Pozo
A.
Gavilanes
J. G.
Role of the basic character of α-sarcin's NH2-terminal β-hairpin in ribosome recognition and phospholipid interaction
Arch. Biochem. Biophys.
2009
, vol. 
481
 (pg. 
37
-
44
)
35
Wang
H.
Schwaderer
A. L.
Kline
J.
Spencer
J. D.
Kline
D.
Hains
D. S.
Contribution of structural domains to the activity of ribonuclease 7 against uropathogenic bacteria
Antimicrob. Agents Chemother.
2013
, vol. 
57
 (pg. 
766
-
774
)
36
Boix
E.
Blanco
J. A.
Nogues
M. V.
Moussaoui
M.
Nucleotide binding architecture for secreted cytotoxic endoribonucleases
Biochimie
2013
, vol. 
95
 (pg. 
1087
-
1097
)
37
Sikriwal
D.
Seth
D.
Batra
J. K.
Role of catalytic and non-catalytic subsite residues in ribonuclease activity of human eosinophil-derived neurotoxin
Biol. Chem.
2009
, vol. 
390
 (pg. 
225
-
234
)
38
Cuchillo
C. M.
Nogues
M. V.
Raines
R. T.
Bovine pancreatic ribonuclease: fifty years of the first enzymatic reaction mechanism
Biochemistry
2011
, vol. 
50
 (pg. 
7835
-
7841
)
39
Sneath
P. H. A.
Sokal
R. R.
Numerical Taxonomy
Freeman, San Francisco
1973
40
Zuckerkandl
E.
Pauling
L.
Brysol
V.
Vogel
H. J.
Evolutionary divergence and convergence in proteins
Evolving Genes and Proteins
1965
New York
Academic Press
(pg. 
97
-
166
)
41
Tamura
K
Peterson
D.
Peterson
N.
Stecher
G.
Nei
M.
Kumar
S.
MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods
Mol. Biol. Evol.
2011
, vol. 
28
 (pg. 
2731
-
2739
)

Author notes

1

These authors contributed equally to this work.

Supplementary data