Structure, antimicrobial activities and mode of interaction with membranes of novel [corrected] phylloseptins from the painted-belly leaf frog, Phyllomedusa sauvagii

Abstract

Transcriptomic and peptidomic analysis of skin secretions from the Painted-belly leaf frog Phyllomedusa sauvagii led to the identification of 5 novel phylloseptins (PLS-S2 to -S6) and also of phylloseptin-1 (PSN-1, here renamed PLS-S1), the only member of this family previously isolated in this frog. Synthesis and characterization of these phylloseptins revealed differences in their antimicrobial activities. PLS-S1, -S2, and -S4 (79-95% amino acid sequence identity; net charge = +2) were highly potent and cidal against Gram-positive bacteria, including multidrug resistant S. aureus strains, and killed the promastigote stage of Leishmania infantum, L. braziliensis and L. major. By contrast, PLS-S3 (95% amino acid identity with PLS-S2; net charge = +1) and -S5 (net charge = +2) were found to be almost inactive against bacteria and protozoa. PLS-S6 was not studied as this peptide was closely related to PLS-S1. Differential scanning calorimetry on anionic and zwitterionic multilamellar vesicles combined with circular dichroism spectroscopy and membrane permeabilization assays on bacterial cells indicated that PLS-S1, -S2, and -S4 are structured in an amphipathic α-helix that disrupts the acyl chain packing of anionic lipid bilayers. As a result, regions of two coexisting phases could be formed, one phase rich in peptide and the other lipid-rich. After reaching a threshold peptide concentration, the disruption of lipid packing within the bilayer may lead to local cracks and disintegration of the microbial membrane. Differences in the net charge, α-helical folding propensity, and/or degree of amphipathicity between PLS-S1, -S2 and -S4, and between PLS-S3 and -S5 appear to be responsible for their marked differences in their antimicrobial activities. In addition to the detailed characterization of novel phylloseptins from P. sauvagii, our study provides additional data on the previously isolated PLS-S1 and on the mechanism of action of phylloseptins.

Figure 1. Complete nucleotide and deduced amino acid sequences of the cDNAs encoding P. sauvagii preprophylloseptin-S1 (ppPLS-S1, EMBL accession number: AM903077), -S2 (ppPLS-S2, AM903078), -S3 (ppPLS-S3, AM903079), -S4 (ppPLS-S4, AM903080), -S5 (ppPLS-S5, AM903081), and -S6 (ppPLS-S6, HE974361).

The open reading frames (capital letters) contain the signal peptide (gray), followed by an acidic sequence ending with a pair of basic residues (in bold), and the phylloseptin progenitor sequence (black). The G residue (in italic) at the C-terminal of the progenitor sequence serves as an amide donor. The stop codon is indicated in bold. The partial 5′- and 3′- UTR are represented in italics and lowercase.

Figure 2. Reversed-phase HPLC chromatogram of P. sauvagii skin extract prepurified on a sep-pak C18 cartridge (A, full scale; B, zoom).

The sample was injected on a semi-preparative Nucleosil C18 column eluted at 4 mL/min with a 0–70% linear gradient of acetonitrile in 0.1% TFA/water (1% ACN/min). Fractions of 4 mL were collected, lyophilized, and analyzed. The position of the mature PLSs is indicated by an arrow (PLS-S3 and PLS-S6 were not detected). The identification of PLSs was achieved by MALDI-TOF-MS and MS/MS (Supporting information, Fig. S1 A–C and S2 A–C).

Figure 3. Activity of phylloseptins-S against Leishmania infantum promastigotes.

IC₅₀ values were determined from a dose-response inhibition fit using GraphPad Prism 5.0 (Table 2). Results represent the mean ± S.E.M. and are representative of three experiments carried out in triplicates.

Figure 4. Time-kill curves of PLS-S1 (S1), -S2 (S2), and -S4 (S4).

Bacteria (E. coli ML-35p and S. aureus ST1065, ∼2–3 10⁶ cfu/mL) were diluted in PBS and incubated with synthetic PLSs at concentrations corresponding to the MIC (E. coli ML-35p: 25 µM, S. aureus ST1065: 6.25 µM) and two-fold above the MIC. Controls correspond to bacteria incubated in PBS without peptide. The data are the means ± S.E.M. of two experiments carried out in triplicates.

Figure 5. Kinetics of the cytoplasmic membrane leakage of E. coli ML-35p and S. aureus ST1065 after incubation with different concentrations of active phylloseptins-S.

The membrane leakage was determined by measuring the production of o-nitrophenol at 405 nm following hydrolysis of ONPG by the cytoplasmic bacterial β-galactosidase. Data are expressed as the mean ± S.E.M of two experiments carried out in triplicates after subtraction of the negative control values (w/o peptide) from the test values. Control panels display the kinetics obtained without peptide (negative control) and with 10 µM dermaseptin B2 (Drs B2, positive control). For E. coli ML-35p, the extracellular release of β-galactosidase was also measured as the production of o-nitrophenol at 405 nm after incubation (60 min) of E. coli cells with PLS-S2 (30 and 100 µM) and PLS-S4 (25 and 100 µM), removing of bacteria by centrifugation and adding ONPG to the supernatant. The negative (w/o peptide) and positive (10 µM Drs B2) controls are indicated for comparison. Results are expressed as the mean ± S.E.M of one representative experiment performed in triplicates.

Figure 6. Circular dichroism spectra of synthetic phylloseptins-S (30 µM).

(A) DMPC/DMPG (3∶1) large unilamellar vesicles in PBS (1 mg/mL). (B) 80 mM SDS. CD measurements are reported as the dichroic increment (Δε) per residue.

Figure 7. Helical wheel diagram of the different phylloseptins-S.

Residues are circled proportionally to amino acid volume. Apolar residues are colored in grey and polar residues are in white. Basic residues are indicated in bold.

Figure 8. PLS interactions with bacterial membrane studied by molecular dynamic simulations.

PLS-S2 was simulated with an E. coli membrane bilayer model using the software Hex 6.3. PDB files generated were visualized using Rasmol 2.7.5.2. No docking is observed for the non-structured PLS-S2 (H₂O) (A) whereas an insertion into the lipid bilayer is predicted for the peptide in α-helical structure (B).

Figure 9. DSC heating thermograms for DMPG and DMPC multilamellar vesicles with or without PLS-S1, -S2 and -S4.

Scans were acquired at different peptide/lipid molar ratios (red, lipid control w/o peptide; blue, 1∶100; green, 1∶50;).