**3. Structural Aspects of Jaburetox and Soyuretox and Its Interaction with Membranes**

Since the discovery of the insecticidal effect of Jaburetox, our group carried out different approaches aimed to identify the active region of the molecule and characterize the structure involved in this toxicity. In this direction, a first modeling for Jaburetox was performed employing an ab initio approach. Using this strategy, 10 different models of possible conformations of the peptide were proposed. In all of them, a β-hairpin motif appeared in the C-terminal region of the molecule, similar to those found in pore-forming peptides. Based on this, it was hypothesized that the β-hairpin motif could be involved in the entomotoxic activity of Jaburetox [57].

Two years later, molecular aspects of Jaburetox and its capability to interact with model membranes were demonstrated for the first time. In this study, Jaburetox displayed the ability to interact with and disrupt large unilamellar vesicles (LUVs) composed mainly by acidic lipids. Moreover, the interaction with LUVs were found to depend on the aggregation state of the polypeptide [66]. Dynamic light scattering and small angle X-ray scattering techniques were used to demonstrate the Jaburetox-lipid interaction with platelet-like multilamellar liposomes (PML) [67]. A tridimensional model indicated that Jaburetox could anchor at a polar/non polar lipid interface, and it was suggested that the β-hairpin motif could be the responsible for this membrane-disturbing effect [66]. The presence of the Jaburetox's β-hairpin motif in the structure of JBU was confirmed by crystallography by the Ponnuraj group in India, who elucidated the 3D structure of JBU [68]. β β

**Figure 1.** Structural representation of entomotoxic peptides and Jack Bean Urease. (**A**) Graphical representation of the location of the peptides (red) in the protein structure of Jack Bean Urease (pdb: 3LA4). Each monomer of the JBU hexamer is represented with a different shade of grey. (**B**) Jaburetox, (**C**) Soyuretox and (**D**) comparison of the primary sequences.

β Δβ Δβ β β In order to identify the regions of the molecule involved in the biological activities of Jaburetox, truncated versions of the peptide were obtained by site-directed mutagenesis [58]. Three mutated versions were constructed and overexpressed in *E. coli*: (1) a peptide lacking the β-hairpin motif (residues 61–74 was deleted, the resulting peptide was called jbtx ∆β-hairpin); (2) a peptide corresponding to the N-terminal portion (residues 1–44, called jbtx N-ter); and (3) a peptide corresponding to the C-terminal region (residue 45–93, called jbtx C-ter). All these peptides were tested for their insecticidal activity and the ability to interact with LUVs. The peptide jbtx ∆β-hairpin displayed the same insecticidal effect of the whole peptide, demonstrating that β-hairpin was not involved in this toxicity. The peptide jbtx N-ter, corresponding to the N-terminal half of Jaburetox, showed insecticidal effect comparable to that of the native peptide, while jbtx C-ter, representing the C-terminal half, was largely inactive on two different insect models [43]. On the other hand, both jbtx N-ter and jbtx C-ter were able to interact with LUVs, the C-terminal half being more effective. As the amphiphilic β-hairpin is present in the C-terminal domain, this motif could be in part involved in membrane interaction. In contrast, for the entomotoxic activity, the half peptide jbtx N-ter contains the most important domain, as observed in the experiments. Three dimensional models for Jaburetox were proposed using bioinformatics and the polypeptide and its mutants were submitted to a molecular dynamic simulation in aqueous system for 500 ns. The results suggested that the whole peptide becomes more unstructured, particularly at its N-terminal portion, and contained a few secondary structure elements with a major part of molecule in random coil conformation at the end of the 500 ns

β

simulation. Interestingly, the β-hairpin structure was conserved in the C-terminal half. When the mutants were submitted to simulation, the jbtx N-ter peptide became totally unfolded while the jbtx C-ter showed a stabilization with β-sheet structures after 500 ns molecular dynamics [58].

Jaburetox and its truncated peptides were also studied using an electrophysiological approach to test their ability to form channels in planar lipids bilayers (PLBs). Two different membrane compositions, to produce neutral net charged and negative net charged lipid interfaces, were tested. All peptides, Jaburetox, jbtx N-ter, jbtx C-ter and jbtx ∆β-hairpin, were able to form channels in both types of bilayers, observed within 30 min after addition of 5–15 µg/mL of each peptide. All channels showed similar biophysical properties, being highly selective to cations, and displayed two conducting states: 7 pS–18 pS and 32 pS–79 pS (smaller channel and main channels, respectively). Similar to Jaburetox, jbtx N-ter was more active at negative voltages while the others did not show voltage dependence. Multiple levels of currents were observed during the experiments using high doses of peptides, suggesting the presence of several identical channels or simultaneous activity of oligomers, corroborating previously reported data on the tendency of Jaburetox and its truncated peptides to form aggregates and the fact that the peptides' oligomerization state influences their biological effects [58,66].

A structural characterization of Jaburetox in solution was carried out using light scattering, circular dichroism (CD) spectroscopy and NMR, and demonstrated the intrinsically disordered nature of the polypeptide [69]. Light scattering studies of the hydrodynamics properties of Jaburetox showed that the peptide in a neutral solution is found in a single oligomeric form with a molar mass of 11.03 kDa. It exhibited a large hydrodynamic radius for a peptide of this molecular mass, a feature suggestive of a disordered polypeptide. CD spectroscopy revealed a typical random coil conformation of Jaburetox and no strong negative signals above 205 nm, thus indicating that Jaburetox presents small amounts of secondary structure. Computational analysis predicted a propensity to disorder mainly at the Jaburetox's N-terminal domain (Figure 1). As Jaburetox tends to aggregate, a thermal scanning fluorimetric assay was performed to identify the best conditions to stabilize the peptide. In the presence of the reducing agent tris(2-carboxyethyl)phosphine (TCEP), the tendency of Jaburetox to aggregate was reduced allowing the NMR assays. The heteronuclear single quantum coherence (HSQC) spectrum unveiled low signal dispersion in the proton dimension which is an indicative of a disordered state of the polypeptide. The chemical shifts were used to predict the residue-specific propensity to form a secondary structure (SSP). SSP analysis predicted that Jaburetox is widely disordered with a small tendency to form α- structures, in addition, with slightly smaller SSP values in the N-terminal (larger predicted disorder) compared to the C-terminal portion (smaller predicted disorder), corroborating the disorder predictors (Figure 2). Data acquired from the nuclear Overhauser enhancement (NOE), which reflect the global fold state of the protein, did not demonstrate the presence of a stable tertiary structure. Still, some elements of secondary structure were detected in parts of the molecule, as a small α-helical motif in the N-terminal region (residues A12-V16), and two turn-like structures, one located in the middle of the polypeptide (residues R48-G56) and the other in its C-terminal region (residues I63-E74). However, the β-hairpin, evidenced in the JBU crystal and in the bioinformatics studies, was not observed in the 3D structure. An in-cell NMR spectroscopy was employed to investigate the molecular folding of the peptide in a physiological condition. A HSQC spectra recorded on *E. coli* cells overexpressing Jaburetox confirmed its disordered folding inside the cells [69].

In order to elucidate if Jaburetox could acquire some conformation when in contact with membranes, a study using artificial and biological membranes was performed [62]. When incubated with sodium dodecyl sulfate (SDS) micelles, a change in the secondary structure of Jaburetox was detected in its CD spectra. Moreover, NMR HSQC confirmed changes in its tertiary structure. Some conformational changes were observed in the molecule, mainly in its N-terminal region, when in contact with LUVs and micelles, prepared with different net charges and molar ratios of phospholipids. This effect was more visible with negatively charged LUVs and micelles, despite without major acquisition of tertiary structure, as determined by NMR. Fluorescence microscopy was employed

to show the interaction between Jaburetox labelled with fluorescein isothiocyanate (Jaburetox-FITC) and cockroach's nervous cord (NC) as a biological membrane, revealing a great intensity of Jaburetox-FITC attached to the ganglia. When Jaburetox-FITC was pre-incubated with LUVs and bicelles, before addition to insect NC membrane, the fluorescence decreased about 50% and 70%, respectively, compared to the initial values, suggesting that the vesicles competed with the insect membranes as a target for the binding of Jaburetox. These data provided evidence that the interaction between Jaburetox and phospholipids did not induce a complete transition from unfolded to folded state but it could have facilitated the peptide's anchorage in cell membranes for posterior acquisition of a folded state [62].

**Figure 2.** Disorder profile of Jaburetox and Soyuretox. Algorithm of disorder prediction VL-XT PONDR® was applied to compare Jaburetox and Soyuretox amino acid sequences. The amino acid sequences (including the 6-His tags) were submitted to http://www.pondr.com/ to generate the graphics. PONDR score above 0.5 indicates disorder. Jaburetox is slightly more disordered than Soyuretox, especially in its N-terminal region. Jaburetox structure is based on conformer 1 (PDB 2MM8). The modeled structure of Soyuretox was obtained with Modeller 9.19 using Jaburetox as a model. The primary sequences of the peptides appear below each 3D-structure, negatively charged amino acids are highlighted in blue and positively charged in red.

The secondary and tertiary structures of Soyuretox were determined using bioinformatics tools, CD spectra and NMR experiments. Jaburetox was used in the same study for comparison. Both peptides share 68% of identity in their amino acid sequences (Figure 1). The secondary structure of Soyuretox and Jaburetox was analyzed by CD spectroscopy at pH 6.5 and 8.0. At pH 6.5 solution, both peptides maintained a disordered behavior. Nonetheless, while Jaburetox kept its disordered state at pH 8.0, Soyuretox acquired some secondary structure in the alkaline medium. A 3D model of Soyuretox was obtained using the structure of JBU as template and a molecular dynamics (MD) was carried out. After 500 ns of MD simulation, Soyuretox became more globular in solution and showed changes in its secondary structure, with loss of helices and beta strands [64]. In spite of the fact that Jaburetox

and Soyuretox behave similarly, Soyuretox demonstrated a tendency to aggregate at pH 6.5 and 8.0 (at the protein concentrations necessary for NMR the experiment), preventing the assignment of the NMR signals [64,69]. The HSQC NMR spectrum obtained for Soyuretox showed a low signal dispersion in the proton dimension, typical of intrinsically disordered states. The ability of Soyuretox to interact with SDS micelles was studied by CD, revealing that in presence of 10 mM SDS (above critical micellar concentration) there was an increase in the peptide's content of secondary structure. A more ordered structure of Soyuretox in the presence of SDS micelles (10 mM) was also confirmed in the peptide's HSQC NMR spectrum, exhibiting a widening of signal dispersion. Under these conditions, Soyuretox kept its intrinsically disordered state [64]. A compilation of the main data obtained in the structural studies of Jaburetox and Soyuretox is present in Table 1.


**Table 1.** Compiled data from structural studies of Jaburetox and Soyuretox.
