Introduction

Plant defensins are basic peptides that have molecular weight between 5 and 6 kDa and have a primary structure composed of 45–54 amino acid residues, arranged in a three-dimensional structure composed of one α-helix and three antiparallel β-sheets stabilized by four disulfide bonds (Fant et al. 1999; Song et al. 2011). These plant peptides have several biological activities (van der Weerden and Marilyn 2013) and among them the antimicrobial activity is striking. Due to several reports showing the increased plant defensins genes expression in response to pathogen infection and reports of transgenic plants expressing them that have become resistant to fungal and insect attacks (Jha and Chattoo 2010; Sarkar et al. 2017), this family of plant peptides is supposed to be involved in plant innate immunity (Carvalho and Gomes 2011; Cools et al. 2017). Based on the aforementioned data, plant defensins constitute the PR-12 group of pathogenesis-related proteins (Sels et al. 2008).

Few plant peptides have been reported to exhibit activity on parasites (McGwire and Kulkarni 2010; Torrent et al. 2012) and amidst them is Vu-Def, a defensin from Vigna unguiculata (cowpea) seed, which has activity against Leishmania amazonensis promastigotes (Souza et al. 2013). It has been described that the biological activity of some representative of the plant defensin peptides is related to an amino acid region located between the β2 and β3 strands (Schaaper et al. 2001; Sagaram et al. 2011). This region has also been identified in other defensin family peptides, for example in the Mediterranean mussel MGD1 (Mytilusgalloprovincialisdefensin one) (Romestand et al. 2003). Later, a work identified a motif encompassing few amino acid residues shared with several peptides with antimicrobial activity belonging to different families, which the authors named γ-core (Yount and Yeaman 2004). The γ-core in plant defensins is located between the β2 and β3 strands (Carvalho and Gomes 2011).

In this work, following the idea that the major determinants for the biological activity of some plant defensins is imparted by the amino acid sequence of the γ-core region (de Samblanx et al. 1997; Schaaper et al. 2001) and also by the fact that Vu-Def has activity against L. amazonensis (Souza et al. 2013), we perform the chemical synthesis of the amino acid region encompassing the β2 and β3 strands of Vu-Def that comprehend the Vu-Def γ-core. Our intention was to understand if this region is responsible for the Vu-Def’s biological activity against L. amazonensis and to unveil its mechanism of action. Plant defensins have demonstrated low toxicity to different mammal cells at up to 500 µg/mL including human umbilical vein endothelial cells, human skin-muscle fibroblasts, and red blood cells (Terras et al. 1992). In addition, the in vivo prophylactic treatment of murine model of candidiasis (Tavares et al. 2008) along with their wide antimicrobial activity against different microorganisms including bacteria (Pelegrini et al. 2011), fungi (Cools et al. 2017), protozoa (Souza et al. 2013), and even cancer cells (Figueira et al. 2017), altogether suggest that this family of peptides is a promising scaffold for the development of new therapeutic substances (Thevissen et al. 2007). As microorganism, we had chosen the protozoan belonging to the genus Leishmania because of the epidemiological importance of this genus (Pace 2014). Leishmania is a genus of intracellular flagellated protozoan belonging to the order Kinetoplastida encircling more than 20 different species. They are transmitted to humans by infected female sand flies bites that belonging to the Phlebotomus genera in Europe, Africa and Asia and Lutzomyia in the Americas. Leishmania-infected humans develop leishamaniasis, a disease continuum ranging in severity from self-healing disfiguring skin wounds to lethal visceral manifestation depending on the interaction between the Leishmania species and human immune system. The disease is spread in 98 countries of the subtropical and tropical areas of the world, with an estimated population, residents or travelers, at risk of infection of about 350 million. The disease causes considerable burden on the health system of countries with estimated morbidity of more than 3.7 disability-adjusted life years (Pace, 2014). There are drugs available for the treatment of leishmaniasis such as pentavalent antimonials (sodium stibogluconate and meglumine antimoniate), polyene macrolide (amphotericin B), alkyl phosphocholine compound (miltefosine), and lipid formulation of amphotericin B. However, despite availability, low efficacy, serious side effects including toxicity, long treatment basis, high price and additionally reports of parasites resistance impose treatment limitations. Therefore, the development of new drugs and delivery modes with improved qualities are urgent (Akbari et al. 2017). In addition, the mechanism of inhibition of this protozoan by this peptide is unknown.

In this work, we describe a small synthetic peptide, called A36,42,44γ32-46Vu-Def, which has activity on L. amazonensis and that the interaction between A36,42,44γ32-46Vu-Def-L. amazonensis results in parasite inhibition by the activation of an apoptotic-like cell death pathway.

Materials and methods

Cell cultures

Promastigote cells of the Leishmania amazonensis strain LV 79 were maintained in 5-mL Warren’s medium (90% Brain heart broth, Fluka, enriched with 10% heat-inactivated fetal bovine serum (FBS, Gibco), 0.4% hemin (Sigma-Aldrich) and 0.01% folic acid (Sigma-Aldrich)) or RPMI-1640 medium (Sigma-Aldrich) supplemented with 10% FBS at 28 °C for 3 days and a new culture was established every 3 days. The L. amazonensis LV 79 cells were provided by Laboratório de Biologia Celular e Tecidual, Centro de Biociências e Biotecnologia, Universidade Estadual do Norte Fluminense Darcy Ribeiro, Campos dos Goytacazes, Rio de Janeiro, Brazil.

Vu-Defr obtainment

The recombinant Vigna unguiculata (L. Walp.) defensin (Vu-Defr) was obtained as previously described by Souza et al. (2013). In brief, the Vu-Defr coding sequence was linked into the pET-32 EK/LIC expression vector and the construction cloned into E. coli expression strain Rosetta-gami2 (DE3) pLysS according to the instruction manual (Novagen TB055). For Vu-Defr induction, a single colony of transformed Rosetta-gami2 (DE3) pLysS was inoculated in TB medium supplemented with ampicillin (50 µg/mL) (Sigma-Aldrich) and chloramphenicol (34 µg/mL) (Sigma-Aldrich) and incubated at 37 °C for 16 h at 250 rpm. After this period, 1 mL of that culture was inoculated at 50 mL of fresh supplemented TB medium and incubated as aforementioned until the optical density at 600 nm reaches 0.5–1.0. Then, 1 mM of IPTG was added and the incubation conditions were changed to 30 °C for 3 h at 250 rpm. Cells were recovered by centrifugation at 15,400xg for 10 min at 4 °C, resuspended in resuspension buffer (50 mM sodium phosphate, pH 8.0, 300 mM NaCl) containing protease inhibitor cocktail (protease inhibitor cocktail for general use, Sigma-Aldrich) and lysed by the CelLytic B Cell Lysis Reagent according to the instruction manual (Sigma-Aldrich). An aliquot of the supernatant consisting of 30 mg of protein was applied to a Ni+-NTA agarose column (Qiagen) which was previously equilibrated in resuspension buffer and the run was developed as described in the column instruction manual (QIAexpressionist, Qiagen). The purified Vu-Defr fused with the thioredoxin and six consecutive His was cleavage with enterokinase according to the protease instruction manual (Sigma-Aldrich). After cleavage, the Vu-Defr final purification was accomplished in a C18 reversed-phase column (Shim-pack VP-ODS 250L × 4.6, Shimadzu) attached to a C8 pre-column (20 × 4.6 mm, Pelliguard, Sigma-Aldrich) in an HPLC (Prominence, Shimadzu) according to Souza et al. (2013).

Vu-Defr modeling

The structural model of Vu-Defr was modeled using the Modeller program (Šali and Blundell 1993). For this, it was necessary to have a template (a protein with known three-dimensional structure) to predict the target protein structural model and the sequence obtained in the protein sequence database (Blastp) (Altschul et al. 1990) with the highest similarity with Vu-Defr was used. After obtaining the structural model of Vu-Defr, its stereochemical quality was checked through the Procheck program and Profile 3D, available on the PARMODEL web server (Laskowski 1993; Uchôa et al. 2004).

Synthetic peptide

Based on the literature information on the γ-core of defensins (Schaaper et al. 2001; Yount and Yeaman, 2004; Sagaram et al. 2011) and on the Vu-Defr structural model (as determined in “Vu-Defr Modeling”), an amino acid stretch from the loop region between two β-strands (β2 and β3) and also part of the intervenient β2 and β3 strands of the Vu-Defr was selected to be synthesized. The synthesized peptide comprised a stretch of 15 amino acid residues and the three cysteine residues were replaced by l-alanine residues. The synthetic peptide sequence was L32SGRARDDVRAWATR46. The peptide was acquired commercially by AminoTech (Rio de Janeiro, Brazil). The synthetic peptide was dissolved in pure water to the concentration of 1.1 M. The molecular weight and isoelectric point of the synthetic peptide were determined by Expasy Compute pI/Mw tool (Bjellqvist et al. 1993, 1994; Gasteiger et al. 2005).

Analysis of the biological activity of the synthetic peptide on L. amazonensis promastigotes by culture growth inhibition assay

In our previous work (Souza et al. 2013), we established that Vu-Defr at concentration of 100 µg/mL, correspondent to 18.5 µM, and incubation time of 24 h at 28 °C caused the inhibition of 54.3% of L. amazonensis from culture. Therefore, we chose this initial concentration and time as a comparison parameter for the toxicity assays for the synthetic peptide. For culture growth inhibition assay, initially the synthetic peptide and Vu-Defr, that were previously resuspended in water, were diluted to 18.5 μM in supplemented RPMI-1640 medium containing 1.5% dimethyl sulfoxide (DMSO, Sigma-Aldrich) and sterilized by filtration (Millex-GV 0.22 μm, Millipore). For parasites, an aliquot of the L. amazonensis promastigotes cell culture was counted in a Neubauer chamber (LaborOptik). For the assay, control consisted of 4.5 µL of DMSO (1.5% at final volume of the assay) with 295.5 µL of supplemented RPMI-1640 medium with parasites. Treated samples consisted of 4.5 µL of DMSO, peptides at the concentration of 18.5 μM with supplemented RPMI-1640 medium at maximum volume of 150 µL and 150 µL of supplemented RPMI-1640 medium with parasites. The final assay volume was 300 µL, the initial cell number was 1.5 × 106 promastigotes/mL and it was incubated at 28 °C for 24 h. The number of parasite was counted in a Neubauer chamber after 24 h of incubation.

The growth inhibition assay with the synthetic peptide at concentrations of 18.5 and 74 µM was done as described above, and the parasite cell number was counted after 24 and 48 h of incubation. This assay, and all subsequent ones, was done with the supplemented Warren’s medium.

The growth inhibition assay with the synthetic peptide at concentration of 74 µM was done as described above, and the parasite cell number was counted at 0, 4, 8, 16, 24, 32 and 48 h of incubation.

Growth recovery assay

After 48 h of incubation with the synthetic peptide at 74 µM as described in “Analysis of the biological activity of the synthetic peptide on L. amazonensis promastigotes by culture growth inhibition assay”, with the exception this repetition was done with supplemented Warren’s medium, the cells of L. amazonensis were harvested by centrifugation (10 min, 500×g, 21 °C). Next, these cells were washed in Warren’s medium and incubated with new Warren’s medium at 28 °C for 24 h in the absence of the synthetic peptide. The number of parasite was determined as described in “Analysis of the biological activity of the synthetic peptide on L. amazonensis promastigotes by culture growth inhibition assay”.

Synchronization of L. amazonensis cells by 1 mM hydroxyurea

Hydroxyurea (HU) was selected to synchronize the culture because it is a known substance with this effect on cells. HU halt the cell cycle at the G1/S phase. The cell cycle progresses to S phase after cell washing with fresh medium to remove UH (Ashibara and Baserga 1979). For this assay, the HU concentration necessary for cell cycle synchronization was initially determined. For this, an aliquot of the cell culture had its cell number adjusted to 1.5 × 106 promastigotes/mL. Then, the cell culture was incubated with HU at 0.5, 1.0 and 2.0 mM for 12 h. The HU was diluted in supplemented Warren’s medium and sterilized by filtration (Millex-GV 0.22 μm, Millipore). Next, the HU-treated cells were washed by centrifugation (10 min, 500×g, 21 °C) and incubated with fresh supplemented Warren’s medium at 28 °C for 48 h. The number of parasite was counted in a Neubauer chamber at 24 and 48 h. Control consisted of supplemented Warren’s medium with parasites. The final assay volume was 300 µL. The culture was considered synchronized at the HU concentration that halts the cell number increasing determined by the cell counting.

After determining the HU concentration required for cell synchronization as described above, the assay was repeated at the chosen HU concentration (1 mM) with the following modifications: the HU-treated culture after washing was incubated with 74 µM of synthetic peptide for 48 h. The number of parasite cells was counted in a Neubauer chamber at 0, 8, 24 and 48 h.

Ultrastructural analysis of Leishmania amazonensis promastigote in the presence of synthetic peptide

The cell culture of L. amazonensis promastigotes was treated as described in “Analysis of the biological activity of the synthetic peptide on L. amazonensis promastigotes by culture growth inhibition assay” with the following modification: this repetition was done with supplemented Warren’s medium, incubation time of 8 h at 28 °C in the absence (control) and presence (treated) of 74 µM of the synthetic peptide. After incubation, cells were prepared as described by Carvalho et al. (2010) and Carvalho and Melo (2018). In brief, after the treatments, the samples were washed in phosphate-buffered saline pH 7.2 at 37 °C and fixed at room temperature with Karnovsky’s solution (1% (V/V) glutaraldehyde, 4% (V/V) paraformaldehyde, 5 mM CaCl2, 5% (W/V) sucrose, and 100 mM cacodylate buffer, pH 7.2), and then post-fixed for 1 h at room temperature in a solution composed of 2% (V/V) OsO4, 0.8% (V/V) potassium ferrocyanide, and 5 mM CaCl2. After fixations, promastigotes were rinsed with 100 mM cacodylate buffer pH 7.2, dehydrated in graded acetone, and embedded in PolyBed812 (Fluka). After, the resin was polymerized for 2 days in 60 °C. Ultra-thin sections obtained with an ultramicrotome (LEICA) were stained with uranyl acetate and lead citrate, and observed with a JEOL 1400Plus Transmission Electron Microscope (TEM) at 60 kV acceleration.

Analysis of the nuclear integrity of L. amazonensis cells treated with synthetic peptide by ethidium bromide dye

The cell culture of L. amazonensis promastigotes was treated as described in “Analysis of the biological activity of the synthetic peptide on L. amazonensis promastigotes by culture growth inhibition assay” with the following differences: this repetition was done with supplemented Warren’s medium, incubated for 8 h at 28 °C in the absence (control) and presence (treated) of 74 µM of the synthetic peptide. After incubation, the cells were treated with 100 µM of ethidium bromide dye for 20 min. This dye intercalates double-stranded DNA and RNA and became intensely fluorescent. The cells were observed on the Axioplan optical microscope (Zeiss) coupled with a DP72 digital camera (Olympus) using detection filters (excitation: 546 nm; emission: 600 nm) (Baskić et al. 2006).

Analysis of the mitochondria membrane potential in L. amazonensis cells treated with synthetic peptide using rhodamine 123 dye

The cell culture of L. amazonensis promastigotes was treated as described in “Analysis of the biological activity of the synthetic peptide on L. amazonensis promastigotes by culture growth inhibition assay” with the following differences: this repetition was done with supplemented Warren’s medium, and L. amazonensis promastigotes were incubated for 4 h at 28 °C in the absence (control) and presence (treated) of 74 µM of the synthetic peptide. To observe if the synthetic peptide caused some effect on the mitochondrial membrane potential after the incubation time, rhodamine 123 dye was added to the assay at 20 µM as described in rhodamine 123 instruction manual (Thermo Fisher). For this assay, the cells were observed on the Axioplan optical microscope (Zeiss) coupled with a DP72 digital camera (Olympus) using detection filters (excitation: 503 nm; emission: 527 nm).

Analysis of vesicular acidification of L. amazonensis cells treated with synthetic peptide by acridine orange dye

The cell culture of L. amazonensis promastigotes was treated as described in “Analysis of the biological activity of the synthetic peptide on L. amazonensis promastigotes by culture growth inhibition assay” with the following differences: this repetition was done with supplemented Warren’s medium, and L. amazonensis promastigotes were incubated for 4 h at 28 °C in the absence (control) and presence (treated) of 74 µM of synthetic peptide. After the treatment, the cells were incubated with a solution of acridine orange at 5 µg/mL (Sigma-Aldrich) for 20 min in the absence of light. Acridine orange is a cell-permeant nucleic acid binding dye that emits green fluorescence when bound to double-stranded DNA (excitation: 502 nm; emission: 525 nm) and red fluorescence when bound to single-stranded DNA or RNA (excitation: 460 nm; emission: 650 nm). This dye also enters acidic compartments such as lysosomes, where it becomes protonated and sequestered, and in low pH, acridine orange emits orange fluorescence (emission: 590 nm) when excited by blue light (475 nm) (Cayman chemical manual). For this assay, the cells were observed on the Axioplan optical microscope (Zeiss) coupled with a DP72 digital camera (Olympus) using appropriate detection filters (Carvalho et al. 2010).

Effects of temperature on the bioactivity of the synthetic peptide on L. amazonensis promastigotes

The assay was done as described in “Analysis of the biological activity of the synthetic peptide on L. amazonensis promastigotes by culture growth inhibition assay” with the following differences: this repetition was done with supplemented Warren’s medium, the cell culture was incubated the synthetic peptide at concentration of 74 µM for 4 h at 4 °C and after the first 4 h at 4 °C, the cell culture was incubated for another 4 h at 28 °C.

Statistical analysis

Each assay was done three times and in triplicate. Statistical analyses were performed by Tukey’s or Sidak tests using Prism software (version 5.0).

Results and discussion

To synthesize a peptide containing the Vu-Def γ-core, first we obtained the Vu-Defr structural model based on the VrD2 (Vigna radiata defensin two) (PDB ID: 2GL1) three-dimensional structure as protein model. VrD2 is 91.6% similar to Vu-Defr, differing in only four positions (green amino acids in Fig. 1a), and by the first amino acid residue, a methionine (red amino acid in Fig. 1a). This methionine was added to the Vu-Defr sequence as a requirement for cloning it into the pET-32 EK/LIC vector (Santos et al. 2010). The Ramachandran plot values of the Vu-Defr structural model are shown in Supplementary Fig. 1. The structural model of Vu-Defr consists of one α-helix and three antiparallel β-sheets (Fig. 1b) as reported for other plant defensins (Fant et al. 1999) and, in addition, as previously demonstrated by Santos et al. (2010), the addition of the Met1 did not interfere in the structural model of Vu-Defr. Based on the structural model of Vu-Defr, we were able to select the amino acid residues that encompass the γ-core (highlighted in blue in Fig. 1a), part of the β2 strand (Lys32 and Ser33), and part of the β3 strand (Trp43, Cys44, Thr45 and Arg46) (Fig. 1a). The three cysteine residues were replaced by alanine in the selected sequence, yielding the peptide L32SGRARDDVRAWATR46 (Fig. 1c) which was commercially acquired. This synthetic peptide was called A36,42,44γ32-46Vu-Def and its chemical characteristics are shown in Fig. 1d. The position of the A36,42,44γ32-46Vu-Def in consideration of to the Vu-Def structural model is shown in Fig. 1e which shows only the position of A36,42,44γ32-46Vu-Def in the Vu-Def structural model and not its conformation. A36,42,44γ32-46Vu-Def had their cysteine residues exchanged by alanine residues to prevent the formation of non-existent disulfide bonds in the natural peptide and also to avoid the peptide from having free cysteine residues. The other authors, to solve this same problem, substituted the cysteine residues for the non-protein amino acid α-aminobutyric acid (Schaaper et al. 2001). We exchange cysteine for alanine because alanine is a nonpolar amino acid (has hydrophobic characteristic and is neutral in charge) and has a small volume and non-reactive side chain. Therefore, this substitution cancels side-chain interactions without altering main-chain conformation or including steric or electrostatic effects. In our synthetic peptide, the Vu-Def γ-core, as defined by Yount and Yeaman (2004), comprises the amino acid residues G34RCRDDVRC42, but the entire sequence of our synthetic peptide is longer, i.e., L32SGRARDDVRAWATR46 (shown with the Cys substituted by Ala). We chose a longer sequence because the other authors have shown the importance of synthesizing peptides larger than the γ-core itself to improve their antifungal activity (Schaaper et al. 2001; Sagaram et al. 2011). Schaaper et al. (2001) has demonstrated by synthesizing various peptides of different sizes, ranging from 13 to 20 amino acid residues, based on the primary structure of Rs-AFP2 (Raphanus sativus antifungal peptide two) that the size of the peptide influences its antifungal activity; decreasing the size of the peptide also decreased its antifungal activity and the highest activity was found in peptides encompassing the whole β2 and β3 strands and its connecting loop, where the γ-core is inserted. After designing the experiment, microbiological assays were done to confirm the biological activity of A36,42,44γ32-46Vu-Def against Leishmania amazonensis promastigotes.

Fig. 1
figure 1

a Alignment of the Vu-Defr and Vigna radiata defensin 2 (VrD2) (PDB ID: 2GL1) primary structures. The lines below the Cys residues (highlighted in bold) in Vu-Defr indicate the arrangement of disulfide bonds. The four amino acid residues that differ between the two primary structures are marked in green in the VrD2 sequence. One additional methionine residue was included in the primary structure of Vu-Defr as a cloning requirement is marked in red. The amino acid residues marked in dark blue in Vu-Defr correspond to the γ-core region as determined by Yount and Yeaman (2004). The amino acid residues marked in a box correspond to a loop region of Vu-Defr. The numbers above the sequence indicate the peptide sizes in amino acids according to Vu-Defr. I, indicates the percentage of identical amino acids residues between the two peptides. b Structural model of Vu-Defr modeled with Modeller, based on the three-dimensional structure of VrD2. Gray represents unstructured elements, red represents α-helices, light blue represents β sheets, and dark blue represents the γ-core region. c Steps of the synthetic peptide design. d Chemical properties of the synthetic peptide A36,42,44γ32-46Vu-Def. e The same as in b with the synthetic peptide A36,42,44γ32-46Vu-Def indicated in purple

Full size image

The A36,42,44γ32-46Vu-Def activity was initially checked by comparing it with the Vu-Defr activity in a culture growth inhibition assay against L. amazonensis with both peptides at 18.5 µM and 24 h of incubation (Fig. 2a). The A36,42,44γ32-46Vu-Def was capable of inhibiting 41% of the culture growth compared to the control and Vu-Defr was able to inhibit 51%. Despite this difference, it was not statistically significant. Therefore, A36,42,44γ32-46Vu-Def retains biological activity of Vu-Defr (Souza et al. 2018). In our previous work (Souza et al. 2013), we established that Vu-Defr at a concentration of 100 µg/mL, corresponding to 18.5 µM, and an incubation time of 24 h caused 54.3% inhibition of L. amazonensis culture. For this reason, we have chosen this initial concentration and time as parameters for comparison for synthetic peptide toxicity assays. The strategy of identification and subsequent use of smaller peptides which retain the biological activity of the entire peptide has some advantages to the medical and pharmaceutical areas. First, the identification of the peptide itself unravels a small structure that retains full biological activity. Second, shorter biologically active sequence enable pharmaceutical and medical uses because of lower cost on chemical synthesis and the lower likelihood of immune reaction compared to the whole peptides (Seo et al. 2012). The in vitro biological activity of the A36,42,44γ32-46Vu-Def also point out another good evidence that the A36,42,44γ32-46Vu-Def could maintain its antimicrobial efficacy in in vivo disease models because its activity is retained in complex media (Warren’s and RPMI-1640 media) and also in the presence of serum. This last characteristic reinforces this possibility since other AMPs completely lose their antimicrobial activities in the presence of serum, such as LL-37 (AMP from Homo sapiens) (Johansson et al. 1998) and gomesin (AMP from the spider Acanthoscurria gomesiana) (Fázio et al. 2006), turning their pharmaceutical uses unfeasible. After confirming the biological activity of the A36,42,44γ32-46Vu-Def, all other experiments were done only with it.

Fig. 2
figure 2

a Growth of Leishmania amazonensis in the absence (control) and in the presence of Vu-Defr and of A36,42,44γ32-46Vu-Def. The percentage of protozoan inhibition is shown above the bars. This result indicates that the synthetic peptide A36,42,44γ32-46Vu-Def is biologically active. (*) Indicates significance (p = 0.0058 for Vu-Def; p = 0.0106 for A36,42,44γ32-46Vu-Def) by the Tukey test (p < 0.05) among experiments and control. (**) Indicate no significance (p = 0.3394) by the Tukey test (p < 0.05) among the Vu-Defr and A36,42,44γ32-46Vu-Def. The graphic was done with the mean and standard deviation of one independent assay. b Growth of L. amazonensis in the absence (control) and presence of different A36,42,44γ32-46Vu-Def concentrations. The percentage of protozoan inhibition is shown above the bars. (*) Indicates significance (p = 0.0344 for 18.5 μM and 24 h; p = 0.0001 for 74 μM and 24 h; p = 0.0013 for 18.5 and 48 h; p < 0.0001 for 74 μM and 48 h) by the Tukey test (p < 0.05) among the experiments and their respective controls. The graphic was done with the mean and standard deviation of one independent assay. cL. amazonensis growth recovery assay after culture growth inhibition assay in the presence of A36,42,44γ32-46Vu-Def. Under the conditions tested, 74 μM is a lethal concentration for the parasite. The percentage of protozoan inhibition is shown above the bars. (*) Indicates significance (p < 0.0001) by the Tukey test (p < 0.05) among the experiments and their respective controls. The graphic was done with the mean and standard deviation of one independent assay

Full size image

To discover the lethal concentration of A36,42,44γ32-46Vu-Def on L. amazonensis promastigotes, a new culture growth inhibition assay was done at concentrations of 18.5 and 74 µM (Fig. 2b). A36,42,44γ32-46Vu-Def, at concentrations of 18.5 and 74 µM, inhibited 40 and 80% of L. amazonensis culture in 24 h of incubation, respectively, comparing to the control. In addition, after 48 h of incubation A36,42,44γ32-46Vu-Def was able to inhibit 65 and 93%, respectively. This assay was done with supplemented Warren’s medium, and despite this medium change, it did not affect the A36,42,44γ32-46Vu-Def activity because the inhibition percentage of A36,42,44γ32-46Vu-Def at 18.5 µM after 24 h in this medium, i.e., 40%, was practically the same in the RPMI-1640 medium, i.e., 41% (Fig. 2a). The effect of the peptide was tested in RPMI-1640 medium because in future studies, we will test the activity of A36,42,44γ32-46Vu-Def against intracellular amastigotes of L. amazonensis that are grown in infected host cells that are cultivated in this medium. All subsequent assays were done with supplemented Warren’s medium.

Next, we performed a growth recovery assay to determine whether the remaining parasites in the treated culture were able to recover their viability after A36,42,44γ32-46Vu-Def treatment. We found that A36,42,44γ32-46Vu-Def at the concentration of 18.5 μM caused 65% of the culture growth inhibition at 48 h (Fig. 2b), the remaining 35% of L. amazonensis cells were able to regrow when transferred to a new medium without A36,42,44γ32-46Vu-Def (Fig. 2c). A36,42,44γ32-46Vu-Def at the concentration of 74 μM caused 93% culture growth inhibition at 48 h (Fig. 2b), the remaining 7% of L. amazonensis cells were unable to regrow when transferred to a new medium without A36,42,44γ32-46Vu-Def (Fig. 2c). Therefore, under the conditions tested, 74 μM is a lethal dose to the parasite. Based on these results, we started working with the most toxic concentrations as determined in our conditions to understand the mechanisms involved in the parasite cell death. The importance of using a lethal concentration of an antimicrobial peptide (AMP) is that there is no misunderstanding on the response signals which lead to the microorganism death with response signals that lead to microorganism recovering as described by Soares et al. (2017). Those authors worked with the budding yeast Saccharomyces cerevisiae and the defensin ApDef1 (Adenanthera pavonina defensin one) and suggested that, at lower ApDef1 concentrations, the yeast cells were able to mount a defense response that overcome the pathway leading to induced cell death by ApDef1. Moreover, the use of a lethal concentration of an antimicrobial agent is desirable because it reduces the risk of resistance development by target microorganism (Vieira et al. 2015).

To better understand the growth inhibition process of L. amazonensis culture in the presence of 74 μM of the A36,42,44γ32-46Vu-Def, an assay was carried out in which the culture growth was evaluated at different times (Fig. 3a). The assay indicated that at zero time, no inhibition occurred, but at 4, 8, 16, 32, 40 and 48 h of A36,42,44γ32-46Vu-Def treatment, levels of 30, 45, 70, 73, 79.7, 80 and 97% of parasite inhibition were observed, respectively, compared to the control. By the analysis of the growth culture inhibition at different times (Fig. 3a), we observe that the inhibition of the parasite occurred mostly gradually; at 8–16 h, 1.5–2.3 times greater inhibition of parasites was seen when compared to that at 0–4 h. This higher inhibition index at 8–16 h is coincident to the L. amazonensis cell division time, i.e., 8 h, in the condition tested (Result not shown). Taking these results together, we suggest that inhibition of the parasite may be related to some cellular target which is exposed only at a certain stage of the cell cycle.

Fig. 3
figure 3

a Time course of Leishmania amazonensis promastigote growth in the presence of 74 μM of A36,42,44γ32-46Vu-Def synthetic peptide. (Dotted circle) corresponds to the protozoan growth and (dotted triangle) corresponds the protozoan inhibition. The percentage of inhibition is shown above the (dotted triangle). (*) Indicates significance (p = 0.0149 for 8 h; p = 0.0013 for 16 h; p = 0.0003 for 24 h; p < 0.0001 for 32 h; p = 0.003 for 40 h, and p < 0.001 for 48 h) by the Sidak’s test (p < 0.05) among the tests and their respective controls. b Growth of L. amazonensis synchronized cell culture in the absence (dotted circle, control) and presence of 74 µM of the A36,42,44γ32-46Vu-Def (dotted triangle). The percentage of protozoan inhibition is shown above the (dotted triangle). (*) Indicates significance (p = 0.0198 for 24 h; p = 0.0006 for 48 h) by the Tukey’s test (p < 0.05) among the tests and their respective controls. The graphics were done with the mean and standard deviation of one independent assay

Full size image

To verify whether the cell cycle was related or not to the parasite inhibition caused by A36,42,44γ32-46Vu-Def, L. amazonensis culture cells were synchronized with 1 mM HU (Fig. 3b). We observe that after 8 h of incubation, the synthetic peptide was able to inhibit 19% of parasites present in the culture as compared to the control. At 24 and 48 h, the peptide inhibited 58 and 97%, respectively. For the unsynchronized culture, the data were: for 8 h, 45.71%, for 24 h, 73% and for 48 h, 97% (see Fig. 3a). Some authors reported that some plant defensins like PsD1 (Pisum sativum defensin one) (Lobo et al. 2007) get accesses to the fungal cytoplasm, and, therefore, these defensins are able to interact with intracellular targets. Amidst the intracellular targets that are affected are components of the cell cycle. PsD1 enters in the cytoplasm of the fungus Neurospora crassa and binds to cyclin F impairing cell cycle progression (Lobo et al. 2007). Another plant defensin whose mechanism of action was related to the S. cerevisiae cell cycle is ApDef1 (Soares et al. 2017). With regard to the involvement of cell cycle in L. amazonensis culture growth inhibition, the result of synchronized culture (Fig. 3b) indicates that the peptide action in the first 24 h is 2.3 times less expressive than in the synchronized culture, however, at the time of 48 h, the synchronized protozoan cells are inhibited at the same proportion of the protozoan cells of the unsynchronized culture. This result suggests that L. amazonensis inhibition maybe cell cycle dependent, although more studies are needed to unravel the cell cycle involvement.

To understand the toxic effect of A36,42,44γ32-46Vu-Def on the cell, the analysis of the parasite cellular structures was performed by TEM after 8 h of incubation at 28 °C (Fig. 4). The control cells were slightly slender and present an organized cytoplasm with well-characterized organelles such as nuclei, mitochondria, kinetoplast, and flagellar pocket (Fig. 4). Particularly, the nuclei had a large centrally located nucleolus and chromatin evenly distributed along the nuclear membrane. Treated cells were more rounded, the cytoplasm was granulated, less electron-dense and vacuolated. The plasma membrane appears to be intact. The nuclei present a disorganized chromatin, reduced and distorted nucleolus, and chromatin condensation in the nuclear periphery. Other cell organelles also presented ultrastructural alterations such as swollen mitochondria and split kinetoplasts which were also more electron-dense (Fig. 4). This ultrastructural analysis of L. amazonensis cell by TEM revealed a number of cellular alterations after 8 h of incubation with A36,42,44γ32-46Vu-Def (Fig. 4). We chose this initial time because we were interested in determining of the first alterations caused in the L. amazonensis cells, because the antimicrobial assay had demonstrated that since 4 h A36,42,44γ32-46Vu-Def causes 30% of promastigote inhibition and at 8 h, it causes 45% of promastigotes inhibition (Fig. 3a). The other authors found similar results after incubation of Leishmania species with different compounds (Mangoni et al. 2005; Rodrigues et al. 2014).

Fig. 4
figure 4

Transmission electron microscopy images of Leishmania amazonensis promastigote treated with 74 μM of A36,42,44γ32-46Vu-Def for 8 h. The treated protozoan cells showed an altered ultrastructural morphology with change in shape which the cells are rounded, the cytoplasm is granulated, less electron-dense and vacuolated. The nuclei presented disorganized chromatin, reduced and distorted nucleolus and chromatin condensation in the nuclear periphery. The mitochondria were swollen and kinetoplast more electron-dense and in some cases, split. The plasma membrane appears to be intact. N nucleus, n nucleolus, M mitochondrion, K kinetoplast, F flagellum, FP flagellar pocket, CMC condensed and marginated chromatin, star cytoplasm not electron-dense, → split kinetoplast, VC vacuolated cytoplasm. The images are representative of one independent assay

Full size image

After observing the effects caused on the nucleus and kinetoplast of L. amazonensis cells by A36,42,44γ32-46Vu-Def (Fig. 4), after 8 h of incubation, we performed a fluorescence microscopy assay in the presence of ethidium bromide (Fig. 5). The control samples showed a punctual labeling in relation to the nucleus and kinetoplast, which is different from that observed in the treated ones, which presented a diffuse labeling (Fig. 5). Taking together the results of loss of viability (Fig. 2c) and structural alterations in the nuclei (Fig. 4) of the treated cells, it is possible to suggest that an apoptotic cell death pathway might be activated after A36,42,44γ32-46Vu-Def-L. amazonensis interaction. The ethidium bromide-labeling pattern (Fig. 5) suggests DNA fragmentation and nucleus disorganization. The low cell density observed in Fig. 5 is due to the 45% inhibition of L. amazonensis cell culture caused by its incubation with 74 µM of the A36,42,44γ32-46Vu-Def for 8 h (see Fig. 3a). The same phenomenon is observed in Figs. 6 and 7, which also has inhibition as a cause for the low number of cells observed. This result (Fig. 5) corroborates with the results of electron microscopy (Fig. 4) as well as it is one well-characterized feature of apoptosis-induced death (Reece et al. 2011).

Fig. 5
figure 5

Images of Leishmania amazonensis promastigotes treated with 74 µM of A36,42,44γ32-46Vu-Def for 8 h and incubated with ethidium bromide fluorescent dye. Note the diffuse labeling in the presence of A36,42,44γ32-46Vu-Def. This labeling pattern suggests DNA fragmentation and nuclear disorganization. Bars = 20 µm. The images are representative of one independent assay

Full size image
Fig. 6
figure 6

Images of Leishmania amazonensis promastigotes treated with 74 µM of A36,42,44γ32-46Vu-Def for 4 h and incubated with the rhodamine 123 fluorescent dye. Note the diffuse labeling in the presence of A36,42,44γ32-46Vu-Def. Bars = 20 µm. The images are representative of one independent assay

Full size image
Fig. 7
figure 7

Images of Leishmania amazonensis promastigotes treated with 74 µM of A36,42,44γ32-46Vu-Def for 4 h and incubated with the acridine orange fluorescent dye. Our result indicates that cytoplasm and some other organelles lumen became acidified. K kinetoplast, N nucleus. Bars = 20 µm. The images are representative of one independent assay

Full size image

Based on the effect that A36,42,44γ32-46Vu-Def has on mitochondria (Fig. 4), we performed a fluorescence microscopy assay in the presence of rhodamine 123 dye. We could observe that control cells had punctual dye label along the cell cytoplasm (Fig. 6). In the presence of A36,42,44γ32-46Vu-Def within 4 h of incubation, these punctual labelings were not observed, and instead a diffuse labeling was observed throughout the cell cytoplasm. This diffuse labeling may be related to mitochondrial fragmentation and swelling, and to the vacuolization observed in the TEM images (Fig. 4). The effect of A36,42,44γ32-46Vu-Def caused in the mitochondria (Figs. 4, 6) indicates that the mitochondria were depolarized and that the membrane potential of other organelles or vesicles may be stimulated due to the higher fluorescence signal observed in the cytoplasm of the treated cells. This interpretation is due to the features of the rhodamine 123 dye that accumulates in places where the electrochemical potential is high. Mitochondrial membrane depolarization has been previously associated with apoptosis cell death in Leishmania (Mukherjee et al. 2002), which reinforces our observation of cell death (Fig. 2c) and the morphological features of apoptosis seen by TEM (Fig. 4). Although within 4 h of incubation time, protozoan inhibition was not statistically significant from the control (Fig. 3a), we chose this time point because TEM analysis at 8 h of incubation time indicated great morphological alteration in protozoan cells (Fig. 4) and also because we wanted to discover the first organelles to be affected. This result indicates that in L. amazonensis promastigotes treated with A36,42,44γ32-46Vu-Def, mitochondria are among the first organelles to be affected.

Based on the rhodamine 123 labeling in the cytoplasm which indicated a higher electrochemical potential of organelles or vesicular membranes, we investigated whether this membrane energization could be due to H+ fluxes and thus we analyzed their acidification by acridine orange dye. This dye labels acid compartments when excited with blue light (475 nm) and in control samples, the labeling was stronger on nuclei and kinetoplast and in the cytoplasm, the dye labeling was sparse and some regions were not labeled. Instead, in the treated samples, we still observe the nuclei and kinetoplast labelings, but with morphological alterations, and the main alteration occurs in the cytoplasm where the labeling has become diffuse and more intense (Fig. 7, 475 nm). As expected and corroborating the rhodamine 123 labeling analysis (Fig. 6), our result with the acridine orange dye labeling indicates that the cytoplasm and the lumen of some other organelles became acidified already after 4 h incubation with A36,42,44γ32-46Vu-Def. Similar results were obtained for L. amazonensis treated with a fraction of Bertholletia excelsa (Brazil nut) where the cytoplasm of treated cells presents a stronger labeling for rhodamine 123 in comparison to the control cells, in which case the labeling indicated that the membrane potential of organelles were altered by the treatment (Fardin et al. 2016). Taking advantage of the capability of acridine orange dye to also emit green fluorescence when bound to double-stranded DNA and excited by green light (502 nm), and since A36,42,44γ32-46Vu-Def was able to cause morphological alterations in the nuclei (Fig. 4), a fluorescence microscopy assay, after 4 h incubation with A36,42,44γ32-46Vu-Def, was performed. Figure 7 (502 nm) shows that some treated cells had a fragmented labeling than that observed in the control where it is possible to see the labeling of the entire nucleus and kinetoplast. This observation is an indication of chromatin fragmentation, which corroborates the results observed by TEM (Fig. 4) and ethidium bromide dye (Fig. 5). These results show that the cellular morphological alterations had already begun within 4 h of contact with A36,42,44γ32-46Vu-Def and are indicative that the nucleus is one of the first organelles to be affected along with mitochondria (Fig. 6). Neto et al. (2011) suggests that the alteration in the green fluorescence of acridine orange observed in L. chagasi and L. amazonensis treated with Ocotea duckei Vattimo lignan occurs due to damage or to conformational changes in nucleic acids.

An assay incubating the parasite for 4 h at 4 °C was performed to understand whether the L. amazonensis-A36,42,44γ32-46Vu-Def interaction is energy dependent. As can be seen from Fig. 8 that when promastigotes were incubated at 4 °C, A36,42,44γ32-46Vu-Def was still able to inhibit 48% of culture growth at 4 h and when the promastigotes were transferred to 28 °C for further 4 h (total incubation time of 8 h), A36,42,44γ32-46Vu-Def was capable to inhibit 93% of the parasites culture growth. When the cells are maintained at 28 °C all the time, in the first 4 h of incubation, the peptide was capable of inhibiting 30% growth of the parasites culture and 45% inhibition at 8 h (Fig. 3a). It has already been described that the interaction of some microorganism-AMPs occurs through the electrical potential of microorganism plasma membrane. This was observed for peptides ApDef1 and histatin 5 (an AMP from human saliva) which, when the microorganism was previously treated with carbonyl cyanide 3-chlorophenylhydrazone (CCCP), a membrane depolarizing agent, or the assay was incubated at 4 °C, had their antimicrobial activity inhibited (Soares et al. 2017). Also to NaD1, its entry into Candida albicans cells is blocked by CCCP and 4 °C, and the authors concluded that the process by which NaD1 gets access to the yeast cytoplasm is endocytosis (Hayes et al. 2018). Both the depolarizing agent and low temperature decrease cell metabolism and thus inhibit energy-dependent cellular process such as endocytosis. In addition, it was observed that, at 4 °C, the parasite Trypanosoma cruzi, belonging to Kinetoplastida order, as well as Leishmania spp., had affected the receptor-mediated endocytosis and fluid-phase pinocytosis pathways (de Figueiredo and Soares 2000), preventing the entry of nutrients and even peptides. In view of this, it was expected that at 4 °C, the L. amazonensis cells should be protected from the A36,42,44γ32-46Vu-Def action. However, our result showed the contrary to expected (Fig. 8) because when the parasites were incubated at 4 °C, the synthetic peptide became more effective in its action, being able to inhibit 93% of the parasites in 8 h. This inhibition rate for the parasite incubated all the time at 28 °C with the peptide is only observed for 48 h (Figs. 2b, 3a, b). These results are an indicative that the peptide does not enter the parasite through transporters, which depend on the membrane potential, neither through receptor-mediated endocytosis nor fluid-phase pinocytosis at least to the studied period. Another possibility is that the low temperature is interfering with some unknown L. amazonensis cellular component that affects the interaction of A36,42,44γ32-46Vu-Def. However, further studies should be performed to better understand the mechanism behind L. amazonensis susceptibility to A36,42,44γ32-46Vu-Def at 4 °C.

Fig. 8
figure 8

Growth of Leishmania amazonensis promastigotes at 4 °C in the absence (control) and presence of 74 μM of the A36,42,44γ32-46Vu-Def synthetic peptide. Growth was observed after 4 h incubation at 4 °C and after further 4 h at 28 °C. The percentage of protozoan inhibition is shown above the test bars. (*) Indicates significance (p = 0.0023 for 4 °C; p < 0.0001 for 28 °C) by the Tukey test (p < 0.05) among the experiments and their controls. The graphic was done with the mean and standard deviation of one independent assay

Full size image

Conclusion

The results presented in the work indicate that the design and chemical synthesis of a 15 amino acid residues long peptide with amino acid substitutions based on the Vu-Defr plant defensin sequence, called A36,42,44γ32-46Vu-Def, is biologically active and causes L. amazonensis culture inhibition by activation of an apoptotic-like cell death pathway. The results presented in this work reinforce the importance of the general strategy of amino acid substitution through mutagenesis or chemical synthesis which renders peptides with modified or improved biological activity to meet medical demand.