Metacaspase-binding peptide inhibits heat shock-induced death in Leishmania (L.) amazonensis

Leishmania (Leishmania) amazonensis is an important agent of cutaneous leishmaniasis in Brazil. This parasite faces cell death in some situations during transmission to the vertebrate host, and this process seems to be dependent on the activity of metacaspase (MCA), an enzyme bearing trypsin-like activity present in protozoans, plants and fungi. In fact, the association between MCA expression and cell death induced by different stimuli has been demonstrated for several Leishmania species. Regulators and natural substrates of MCA are poorly known. To fulfill this gap, we have employed phage display over recombinant L. (L.) amazonensis MCA to identify peptides that could interact with the enzyme and modulate its activity. Four peptides were selected for their capacity to specifically bind to MCA and interfere with its activity. One of these peptides, similar to ecotin-like ISP3 of L. (L.) major, decreases trypsin-like activity of promastigotes under heat shock, and significantly decreases parasite heat shock-induced death. These findings indicate that peptide ligands identified by phage display affect trypsin-like activity and parasite death, and that an endogenous peptidase inhibitor is a possible natural regulator of the enzyme.

Leishmania (Leishmania) amazonensis is the second most frequent agent of cutaneous leishmaniasis in Brazil, 1 a country with high incidence of this disease (WHO 2016).
The promastigote forms of Leishmania are transmitted to man and other mammals by the bite of an infected female sand fly. Once inside vertebrate phagocytes, promastigotes convert into the amastigote forms, responsible for disease progression. 2,3 Macrophages are the parasite's main host cells that nevertheless possess several mechanisms to restrain the infection such as by nitric oxide (NO) and reactive oxygen species (ROS) production. 4,5 After transmission to the vertebrate host, promastigotes face a heat shock and an oxidative attack of the innate immune system. The exposure of promastigotes to NO, ROS, hydrogen peroxide, 6 heat shock 7,8 and drugs [9][10][11] induces phenotypical changes characteristic of programmed cell death (PCD) such as cell shrinkage, DNA fragmentation, activation of peptidases 12 and exposition of a 'PS-like' (annexin V binding) phospholipid. 13,14 PCD by apoptosis is considered to be dependent on the activation of caspases, cysteine-dependent peptidases. 15 Plants, fungi and protozoa do not code for caspases but express metacaspases (MCAs). 16 MCAs are cysteine peptidases from clan CD, family C14, that share the histidinecysteine catalytic dyad with caspases, but differently from caspases, which are specific to aspartic acid at P1, are specific to arginine/lysine, that is, have trypsin-like activity. [16][17][18][19] MCAs have been shown to have caspase-like auto-processing, 20 and for some Leishmania species, 20,21 as well as yeast 22 and Arabidopsis, 23,24 processing of the enzyme is essential for activity.
Leishmania MCA was first described in Leishmania (L.) major 20 and in Leishmania (L.) donovani, 25 and was later identified in Leishmania (L.) mexicana 26 and Leishmania (L.) infantum. 9 Most Leishmania species have a single MCA gene, while L. (L.) infantum and L. (L.) donovani have two genes that code for proteins with 96% identity. 27 MCA has important roles not only in cell death but also in the control of amastigote intracellular proliferation 26 and autophagy. 26,28 The association between MCA and cell death has first demonstrated in yeast 22 and Trypanosoma brucei, 29 and later in Arabidopsis 30 and different Leishmania species. Oxidative stress caused MCA-dependent cell death in Saccharomyces cerevisiae, and the yeast knock out strain was efficiently complemented by Leishmania (L.) major MCA. 20 Accordingly, oxidative stress induced higher 'PS-like' exposure in a Leishmania (L.) major lineage that over expressed MCA catalytic region. 21 Drugs such as miltefosine induced death in Leishmania (L.) infantum associated with MCA overexpression, 9 and a Leishmania (L.) major MCA-deficient lineage was resistant to miltefosine and curcumin. 28 Regulators and natural substrates of MCA were not frequently studied. It was recently shown by two-hybrid system that Leishmania (L.) major mitogen-activated protein kinase MPK7 and calpain interact with the C-terminal domain of MCA, probably participating on the induction of parasite death. 28 No similar study was ever performed on Leishmania (L.) amazonensis MCA or other Leishmania species.
Phage display is an effective tool for searching for protein ligands, allowing the identification of natural protein interactions and of potential synthetic modulators. 31,32 In this work we have employed a commercial phage display library over recombinant L. (L.) amazonensis MCA to identify peptides that could interact with the enzyme, modulating its activity and eventually affecting parasite survival. We identified peptides that specifically bind to MCA. One of them decreases trypsinlike activity in promastigotes under heat shock and reduces parasite heat shock-induced death. This peptide is similar to the L. (L.) major ecotin-like ISP3, 33 a potent inhibitor of several serine proteases. 34 Our study is the first to show that MCA can be inhibited by a peptide similar to a parasite-coded serine protease inhibitor. Besides, we show that phage display can be effectively used to identify modulators of specific Leishmania targets, suggesting that this technique can be employed to decipher poorly known processes and to search for potential parasitespecific drugs.

Results
Heat shock induces L. (L.) amazonensis promastigote death and increases trypsin-like activity. Leishmania faces heat shock during transmission to the vertebrate host, and this event induces death of a part of the parasite's population. 8 We analyzed the effect of heat shock at 37°C (compared with 22°C, the sand fly temperature) for different periods in promastigote death employing MTT assay and annexin V and PI labeling. The results shown in Figures 1 (a-c) indicate that viability is significantly diminished after 2 (based on MTT, c) and 3 h (based on annexin V and PI, a, b) of heat shock.
MCA has already been suggested to be involved in Leishmania heat shock-induced cell death. 21 We thus quantified trypsin-like activity in parasite extracts after incubation at 37°C. Figure 1c demonstrates that heat shock for 1, 2, 3 and 4 h significantly increased activity, reinforcing the possible role for the enzyme in this process.
Production of active recombinant L. (L.) amazonensis MCA. Little is known about MCA substrates and regulators. Aiming to find natural or artificial ligands that could modulate enzyme activity and eventually parasite death, we employed the phage display technique on the recombinant enzyme. Bacteria were transformed with pET-28a vector containing L. (L.) amazonensis MCA gene and trypsin-like activity was assayed in extracts ( Figure 2a). Activity was higher in bacteria transformed with MCA containing plasmid, as already described for plant MCA in the same system. 24 This result indicated that bacteria could produce a functional (active) recombinant protein. His-tagged protein with approximately 50 kDa was identified in bacterial extracts after induction with IPTG (Figure 2b), and was purified along with a processed form of~25 kDa (Figure 2c), already described. 23 Identification of MCA-binding peptides able to modulate enzyme activity. Panning using the commercial Ph.D.-7 Phage Library (phages containing seven random amino acids in pIII protein) was performed over the immobilized MCA protein. Three cycles were performed, and a small enrichment in the number of bound phages was observed after each cycle (Supplementary Table I). After three cycles, 50 bound phages were randomly picked and sequenced. The sequences of the corresponding peptides and their frequencies are shown in Table 1.
Phage display selection usually leads to a large number of different bound phages, and selection of the ones to be validated is usually based on their frequencies. [35][36][37] Thirtythree different peptides were encoded by the 50 phages sequenced. Most of them were observed in only one phage, seven were present in two phages, one in four phages and two in five phages (see Table 1). We thus selected 13 peptides (named 1-13, shown in Table 1) using as criteria the identification of the peptide in more than one phage or the presence of a three amino acid repeat shared with another peptide. These peptides were synthesized and the effect of each of them on trypsin-like activity in promastigote extracts and in extracts of bacteria transformed with control vector or MCA plasmid is shown in Figure 3. Figure 3a shows that peptides 1, 2, 10, 11 and 12 significantly increased and peptide 3 significantly decreased trypsin-like activity in promastigotes compared with control. To prove that the peptides' effects on the promastigote trypsinlike activity were due to their interaction with MCA and not with other enzymes, we analyzed the effects of peptides 1, 3, 10 and 11 on trypsin-like activity in bacteria extracts. Trypsin-like activity of bacteria expressing or not MCA in the presence of the peptides is shown in Figure 3b. Activity was again higher in bacteria transformed with MCA plasmid (control in grey) than with control vector (control in white), as previously shown in Figure 2a. As observed in promastigote extracts (Figure 3a), peptide 3 decreased trypsin-like activity of MCA expressing bacteria ( Figure 3b). Peptides 10 and 11 showed no significant increase in activity, and peptide 1 decreased activity in bacteria, differently from promastigote data. Peptides had no effect on trypsin-like activity of bacteria transformed with control vector.
Peptides bind to recombinant MCA and to fixed and live parasites. To confirm that peptides modulated activity due to direct interaction with the enzyme we evaluated the binding of peptides 1, 3, 10 and 11 to MCA. Peptides were conjugated with Alexa 488 and incubated with either immobilized recombinant MCA or streptavidin. The results shown in Figure 4 indicate that all peptides have significant higher binding to MCA than to streptavidin, indicating that the phage selection was effective and specific.
We then tested if the peptides were able to recognize and bind to MCA in parasites. To answer this question, we incubated fixed non-permeabilized promastigotes or fixed promastigotes after heat shock (1 h at 37°C) with peptides conjugated with Alexa 546 and analyzed labeling under the microscope. Representative images shown in Figure 5a and the corresponding quantifications shown in Figure 5b indicate that peptide 3 binds to fixed parasites previously submitted or not to heat shock, while peptides 1, 10 and 11 bind weakly in the absence of heat shock and more intensely to promastigotes submitted to incubation at 37°C for 1 h. The lower binding of the other peptides compared with peptide 3 suggests that its stronger binding is not artefactual.    Table 1 Peptide sequences and frequencies of the 50 phages Peptide numbers (column 1), peptide sequences showing polar amino acids in green, hydrophobic in red, acid in blue and basic in pink (column 2), frequencies of the sequences (column 3) parasite extracts, but only two significantly affected trypsinlike activity in bacteria expressing MCA, both decreasing enzyme activity. Since none of the peptides affected the activity in bacteria not expressing MCA, we concluded that the effects observed were specific for MCA. Promastigotes have enzymes other than MCA with trypsin-like activity, such as cysteine proteinase C, 38 and have higher activity than MCA expressing bacteria (data not shown). These facts may explain the differences between the results.
The ability of the peptides to bind to MCA was confirmed using recombinant enzyme and fixed parasites submitted or not to heat shock. Heat shock is known to promote enzyme processing, and could also alter enzyme structure and affinity to the peptides, and even open pores in the parasite membrane, facilitating peptide binding. In fact, all peptides showed visible labeling to parasites after heat shock, while peptide 3 was the only peptide with visible binding to parasites at 22°C, suggesting a more effective interaction with the 'native' MCA inside the parasite.
Peptide 3 inhibits MCA and reduces parasite death. The four peptides could represent natural regulators of MCA or could be artificial synthetic mimotopes. To perform functional assays on the effect of the peptides in parasite death, we first checked their ability to bind/enter live cells. Although none of them had the extremely basic amino acid composition or amphipatic membranotropic composition usually found in cell penetrating peptides (CPPs), 39,40 peptides 1 and 3 are predominantly basic (three out of seven amino acids) and peptide 11 predominantly nonpolar/hydrophobic (three out of seven amino acids). Peptide labeling was observed in a small proportion of the parasites, suggesting that they bound/entered cells and that functional tests in live parasites could be done.
Functional assays were performed with peptide 3, which significantly reduced trypsin-like activity and parasite death after heat shock. This data indicates that even a small inhibition of MCA by the peptide has an impact on parasite death induced by stimuli such as heat shock. We believe similar effects would be observed if parasite death was triggered by stimuli such as oxidative stress and drugs.
Although peptide 3 decreases cell death and thus could not be used as drug for parasite killing, it may shed light on natural regulators of Leishmania MCA. In fact, its similarity to parasite ecotin, more specifically ISP3, suggests that this protein regulates endogenous MCA activity and thus parasite death. First described in E. coli, ecotin is a potent inhibitor of several serine proteases. 34 L. (L.) major was shown to have three ecotin-like genes, named ISP 1, 2, 3 for inhibitor of serine peptidase, and ISPs 2 and 3 seem to have a role on the interaction with the host cell. 33 ISP3 seems to have a very low expression in L. (L.) major, 33 but we were able to identify its transcript in L. (L.) amazonensis promastigotes. An ISP was recently identified in L. (L.) donovani, and shown to inhibit trypsin activity but not the activity of a L. (L.) donovani serine protease. 41 Ecotin and ISPs are considered inhibitors of serine peptidase from S1A family. Genomic studies have proven that L. (L.) major has several serine peptidases belonging to six families but not to S1A family. 42 Our findings suggest that ecotin may also inhibit cysteine proteases such as MCA. Besides, we show that phage display is an efficient tool to search for ligands and regulators of proteins and enzymes with poorly known pathways.

Conclusion
Our data demonstrate that L. (L.) amazonensis trypsin-like activity and promastigote death are induced by heat shock. We showed that peptides that bind to recombinant MCA, selected by phage display, may affect enzyme activity. One of the peptides, similar to Leishmania ISP3, is able to reduce parasite trypsin-like activity induced by heat shock and promastigote death after the shock. We suggest for the first time that ISP3, considered a serine peptidase inhibitor, may also inhibit cysteine proteases such as MCAs.   Annexin V and PI labeling by flow cytometry. Promastigotes at log phase (day 3) were centrifuged at 4000 × g for 5 min and washed three times in Hepes buffer (100 mM Hepes, 150 mM NaCl, 5 mM KCl, 3 mM CaCl 2 , 1 mM MgCl 2 , pH 7,2). Parasites were resuspended at 5 × 10 6 cells/ml in the same buffer and 200 μl were incubated with annexin V Alexa Fluor 488 1 : 200 for 20 min on ice, then washed three times and incubated with 10 μg/ml propidium iodide (IP) for 20 min. As positive control we treated parasites with 100 μM digitonin in annexin V reaction. 30.000 events were captured for each sample in Guava easycyte (Millipore, Bedford, MA, USA).
10 ml of E. coli BL21 (DE3) expressing MCA or containing pET28a plasmid were centrifuged at 16 000 × g at 4°C for 20 min, resuspended in 250 μl lysis buffer containing 0,4 mg/ml lysozyme, and kept on ice for 1 h with vortexing every 10 min. The lysate was then centrifuged as described above.
For both promastigote and bacteria lysates, activity was assayed in 96-well Costar 3603 plates (Costar-Sigma, SP, Brazil) in 100 μl of buffer containing 50 mM Tris-HCl, 15 mM NaCl, 5 mM DTT, 10 mM de CaCl, pH 8.0. Extracts were added in the presence or not of peptides or DMSO and incubated for 2 h at 22°C. Z-Arg-Arg-AMC substrate was added for 10 μM and capture was performed at 30.5°C in a POLARstarOmega (BMG, Ortenberg, Germany) fluorimeter excitation at 380 nm and emission at 460 nm.
SDS-PAGE and western blot. Gels and membranes were prepared as described before, 43 using 10 μg of proteins and 1 μg of recombinant MCA.
Binding of peptides to metacaspase. 15 μg of MCA or 15 ug streptavidin in 150 μl of 0.1 M NaHCO 3 pH 8,6 were incubated in 96-well plates (Costar3603) at 4°C o/n. Wells were blocked with 150 μl of 1% BSA in 0.1 M NaHCO 3 pH 8,6 for 1 h at 4°C, washed six times with PBS and incubated with 100 μM of Alexa Fluor 488 conjugated peptide (labeled with Protein Labeling Kit-Thermofisher, SP, Brazil) for 1 h at 4°C. Binding was estimated after analysis in POLARstar Omega at 488 nm for excitation and 520 nm for emission.
Labeling of parasites with fluorescent peptides. For the binding analysis, promastigotes were washed in PBS, fixed in 4% paraformaldehyde for 30 min, resuspended in PBS and applied in glass slides. After drying, slides were blocked with 1% BSA in PBS for 1 h, washed in PBS and incubated with 100 μM Alexa Fluor 546 peptide in PBS 1% BSA o/n. Slides were then washed in PBS, incubated with 1 μM DAPI in PBS 1% BSA for 1 h, dried and mounted in ProLong (Molecular Probes,Thermo Fisher, SP, Brazil). Images were captured in ZEISS Axio Imager M2 Imaging System (Oberkochen, Germany) and quantification was performed using ImageJ (Bethesda, MD, USA), and expressed as of fluorescence intensity/μm2 (mean values of three parasites normalized by parasite area).
For the evaluation of peptide entry in intact promastigotes we used flow cytometry and the same Alexa Fluor 546 peptides. Promastigotes were incubated for 2 h at 22°C with 100 μM Alexa Fluor 546 peptide in HEPES buffer 10 mM, washed in PBS and analyzed in Guava easycyte cytometer (Millipore).
RT-PCR. RNA was isolated from 5 × 10 7 promastigotes using Trizol reagent (Life Technologies). cDNA was prepared from 2 μg of RNA using random primers, oligodT and Superscript II Reverse Transcriptase (Life Technologies) in 20 μl. PCR was performed with 2 μl cDNA reaction 1 : 10, the corresponding RNA mass or 70 ng of genomic DNA, using 0.7 U Taq polymerase (Life Technologies) and the following cycle conditions: 94°C for 4 min, 35 cycles of 94°C 30 s, 60°C 45 s, 72°C 2 min and 72°C for 10 min.
Statistical analysis. GraphPad software (San Diego, CA, USA) was used to perform all analysis. We employed one way ANOVA followed by Tukey's multiple comparison test (for three or more samples), or t-test (for comparison of two conditions).