Structural Similarity between Defense Peptide from Wheat and Scorpion Neurotoxin Permits Rational Functional Design

Background: Protein folds differ in size and complexity and hence in their utility for engineering purposes.

Results: The three-dimensional structure of wheat antifungal peptide Tk-AMP-X2 was investigated, and a new functionality was engineered based on its α-hairpin scaffold.

Conclusion: α-Hairpinins are an attractive simple structural template for functional engineering and drug design.

Significance: The repertoire of available scaffolds for protein engineering is broadened.

Abstract

In this study, we present the spatial structure of the wheat antimicrobial peptide (AMP) Tk-AMP-X2 studied using NMR spectroscopy. This peptide was found to adopt a disulfide-stabilized α-helical hairpin fold and therefore belongs to the α-hairpinin family of plant defense peptides. Based on Tk-AMP-X2 structural similarity to cone snail and scorpion potassium channel blockers, a mutant molecule, Tk-hefu, was engineered by incorporating the functionally important residues from κ-hefutoxin 1 onto the Tk-AMP-X2 scaffold. The designed peptide contained the so-called essential dyad of amino acid residues significant for channel-blocking activity. Electrophysiological studies showed that although the parent peptide Tk-AMP-X2 did not present any activity against potassium channels, Tk-hefu blocked Kv1.3 channels with similar potency (IC₅₀ ∼ 35 μm) to κ-hefutoxin 1 (IC₅₀ ∼ 40 μm). We conclude that α-hairpinins are attractive in their simplicity as structural templates, which may be used for functional engineering and drug design.

Introduction

During folding and maturation, polypeptide chains are able to realize a large variety of spatial conformations, depending on their primary structure and environmental conditions. Elements of secondary structure form regular motifs that are further organized into functional domains and large sophisticated polydomain conglomerates. Domains are universal structural units that may exist and evolve separately as an independent protein molecule or part of a larger polypeptide. Their length may vary between as little as 25 and as many as 500 residues. To systemize available structural information, some databases like SCOP (Structural Characterization of Proteins) (1) and CATH (Class, Architecture, Topology, Homologous superfamily) (2) suggest different approaches to its analysis and classification. Although it would appear that the number of possible spatial folds is unlimited, only a few are actually realized. Moreover, protein folds are not equally populated: some of them are either more structurally or sequentially diverse, and these effects are not necessarily correlated. It is still unclear whether highly populated folds arise because of convergent evolution from multiple origins, which perhaps might suggest that these folds are more stable in the evolution process, or divergent evolution from a single source. Often families of sequences without any detectable similarity can adopt the same structural fold. For instance, the TIM barrel is the most common enzyme fold in the Protein Data Bank of known protein structures. It has been detected in many different enzyme families, catalyzing completely unrelated reactions in metabolic processes (3). The TIM barrel is the only barrel fold with a completely parallel sheet that has the topology (βα)₈ and typically some 250 residues. TIM barrel enzymes can be small (only one domain) or very large (consisting of as many as five domains) (4). Based on recent directed evolution selection experiments, it has become clear that small sequence modifications can change protein function: for instance, Sterner and co-workers (5) found that one point mutation is sufficient to convert the HisA protein into a protein with TrpF activity. Both HisA and TrpF catalyze the irreversible isomerization of an aminoaldose into an aminoketose but in metabolism of histidine and tryptophan, correspondingly, and have no detectable sequence similarity.

Another example is the immunoglobulin (Ig) fold. This natural scaffold is utilized by the immune system of higher organisms to withstand pathogenic invaders. Antibodies, molecules of the humoral response, consist of two types of domains, constant and variable, each based on the Ig fold. Six hypervariable loops, three within each variable domain, form the active part of the antibody surface that contacts the antigen upon complex formation. These loops protrude from a sandwich of two antiparallel β-sheets that forms the framework region of the variable domain (6). Dozens of antibody-based products have already been approved as drugs, and many more are under clinical trials (7). Unfortunately, antibodies are rather large molecules consisting of separate polypeptide chains, which are difficult and expensive to manufacture, an obvious limit to their clinical use.

Engineering binders from antibodies has been a standard practice during the last two decades, but the use of alternative scaffolds has lately become an emerging field in protein engineering (8). Alternative scaffolds such as fibronectin domains are small proteins that have certain surface regions that can be highly diversified to bind to a variety of targets (9). The 10th fibronectin type III domain (¹⁰FN3)³ was utilized as the starting point for the design of a family of target-binding proteins because of its structural similarity to antibody variable domains. Despite the lack of significant sequence homology, antibody variable domains and ¹⁰FN3 have similar structures (10). FN3 is a monomeric natural β-sandwich protein with resemblance to a trimmed chain of a variable Ig domain. It consists of ∼90 residues and possesses seven β-strands with three loops connecting the strands in a pairwise fashion at one end of the β-sheet. ¹⁰FN3 is one of the rare members of the Ig superfamily devoid of disulfide bonds. FN3 domains are ubiquitous and occur in cell adhesion molecules, cell surface hormone and cytokine receptors, chaperonins, and carbohydrate-binding proteins, all of which are involved in molecular recognition. Adnectins are a new family of therapeutic proteins based on ¹⁰FN3 and designed to bind to therapeutically relevant targets with high affinity and specificity (11).

Albebetin is an artificial small protein (∼70 residues) that was constructed de novo. The albebetin fold consists of two halves, each of which includes an α-helix of three turns (11 residues) and a β-hairpin from a four-stranded antiparallel β-sheet (12, 13). It is interesting that after the first draft of the primary design of albebetin was proposed a similar fold was found in some natural proteins (14, 15). It was further shown that this artificial protein can be used as a scaffold for engineering desired biological functions (16).

It is apparent that not only huge protein domains but also rather small peptides may adopt very stable common folds carrying out a variety of functions. In this study, we explore a small α-hairpin peptide Tk-AMP-X2 (28 residues) from wheat (Triticum kiharae) seeds (17). This peptide belongs to the α-hairpinin family of plant defense peptides (18), which includes both antimicrobial peptides (AMPs) and protease inhibitors (18–20). Its main structural features are two α-helices stabilized by two disulfide bonds and two characteristic cysteine motifs CXXXC where X is any residue; the disulfide pairing is C¹–C⁴ and C²–C³. A number of peptide toxins from venomous animals such as scorpions and sea snails (21–23) present the same fold. Thus, this kind of peptide structure seems to be a rather universal scaffold for a variety of biological functions. Therefore, we decided to create a novel molecule with predicted structure and function using Tk-AMP-X2 as a template.

EXPERIMENTAL PROCEDURES

Peptide gene expression was performed essentially by following a procedure elaborated previously (24).

Expression Vector Construction

DNA sequences encoding Tk-AMP-X2 and Tk-hefu were constructed from a number of synthetic oligonucleotides (Table 1) using a combination of PCR and ligation techniques. The target PCR fragments were amplified using a forward primer containing a KpnI restriction site and an enteropeptidase cleavage site for fusion protein hydrolysis and a reverse primer containing a BamHI restriction site and a stop codon. The PCR fragments were cloned into the expression vector pET-32b (Novagen) to produce pET-32b-Tk-AMP-X2 and pET-32b-Tk-hefu.

TABLE 1.

Oligonucleotides used in Tk-AMP-X2 and Tk-hefu gene synthesis

Name	Sequence	Length
		nucleotides
f1	ACGAGATCTGGTACCGACGACGACGACAAAGCTGACGACCGTTGCGAAC	49
f1-hefu	ACGAGATCTGGTACCGACGACGACGACAAAGCTGACGACCGTTGCTACC	49
f2	GTATGTGCCAGCGTTACCACGACCGTCGTGAGAAGAAACAGTGCATGAA	49
f2-hefu	GTATGTGCCAGCGTTACCACGACCGTCGTGAGAAGAAACAGTGCAAAGA	49
rev1/2	GGTAACGCTGGCACATACGTTCGCAACGGTCGTCAG	36
rev1/2-hefu	GGTAACGCTGGCACATACGGTAGCAACGGTCGTCAG	36
rev	ATCGAATTCGGATCCCTAACCGTAACGGCAACCTTTCATGCACTGTTTC	49
rev-hefu	ATCGAATTCGGATCCCTAACCGTAACGGCAACCTTCTTTGCACTGTTTC	49

Fusion Protein Expression and Purification

Escherichia coli BL21(DE3) cells transformed with the expression vector pET-32b-Tk-AMP-X2/pET-32b-Tk-hefu were cultured at 37 °C in Luria-Bertani medium to the midlog phase. Expression was then induced by 0.2 mm isopropyl β-d-thiogalactopyranoside. Cells were cultured at room temperature (24 °C) overnight (16 h) and harvested by centrifugation. The cell pellet was resuspended in 300 mm NaCl, 50 mm Tris-HCl buffer, pH 6.8 and ultrasonicated. The lysate was applied to TALON Superflow resin (Clontech), and the fusion protein (Trx-Tk-AMP-X2/Trx-Tk-hefu) was purified according to the protocol supplied by the manufacturer. To produce ¹⁵N-labeled Tk-AMP-X2, a procedure similar to the one above was followed using M9 minimal medium with ISOGRO (Sigma-Aldrich) instead of Luria-Bertani medium.

Fusion Protein Cleavage and Purification of the Target Peptide

Chimeric proteins were dissolved in 50 mm Tris-HCl, pH 8.0 to a concentration of 1 mg/ml. Protein cleavage with human enteropeptidase light chain (25) (1 unit of enzyme/1 mg of substrate) was performed overnight (16 h) at 37 °C. Recombinant Tk-AMP-X2/Tk-hefu was purified by reversed-phase HPLC on a Jupiter C₅ column (250 × 10 mm; Phenomenex) using a linear gradient of acetonitrile concentration (0–60% in 60 min) in the presence of 0.1% trifluoroacetic acid. The purity of the target peptide was checked by MS, N-terminal sequencing, and analytical chromatography on a Vydac 218TP54 C₁₈ column (4.6 × 250 mm; Separations Group) in a shallow acetonitrile gradient.

Selective Proteolysis

Our protocols have been published previously (26). Tk-hefu was first digested by endoproteinase Glu-C (Sigma-Aldrich) and then by CNBr to ensure cleavage between all cysteine residues. Digestion of the peptide with Glu-C was performed in 20 μl of 50 mm ammonium bicarbonate buffer, pH 8.0. The probe was incubated at 37 °C for 4 h. The resulting cleavage products were then separated using reversed-phase HPLC, and their masses were determined by MS using an Ultraflex TOF-TOF mass spectrometer (Bruker Daltonics, Germany). The major peptide product was dissolved in 20 μl of 80% aqueous TFA, and 1 μl of 5 m CNBr in acetonitrile (Sigma-Aldrich) was added. The sample was incubated for 24 h at room temperature in the dark. Products of hydrolysis were separated by HPLC and analyzed by MS.

NMR Spectroscopy

Two samples were prepared for Tk-AMP-X2 structure determination: ¹⁵N-labeled peptide in H₂O/D₂O (19:1) and unlabeled Tk-AMP-X2 in 100% D₂O. In both cases, the concentration of peptide was 5 mm, and pH was adjusted to 5.6. All NMR experiments were performed on an Avance 700-MHz spectrometer (Bruker Biospin) equipped with a cryoprobe at 30 °C. Unless otherwise stated, a relaxation delay of 1.4 s was used. The Watergate (27) technique was applied to suppress strong solvent resonance in some spectra acquired in H₂O solution. ¹H chemical shifts were measured relative to the protons of H₂O; the chemical shift of their signal was arbitrary chosen as 4.75. Chemical shifts of ¹³C and ¹⁵N were calculated from the respective gyromagnetic ratios. Proton and ¹⁵N resonance assignments for Tk-AMP-X2 were obtained by a standard procedure (28, 29) using ¹⁵N HSQC, three-dimensional ¹⁵N TOCSY-HSQC, three-dimensional ¹⁵N NOESY-HSQC, and ¹³C HSQC in the CARA software (30). The ³J_HNHα coupling constants were determined from the intensity ratio of cross- and diagonal peaks in the three-dimensional HNHA spectrum (31). The ³J_HαHβ coupling constants were measured by line shape analysis of H_α-H_β cross-peaks in the double quantum filter COSY spectrum of Tk-AMP-X2 in D₂O (relaxation delay of 3 s). ³J_NHβ constants were calculated from the intensities of cross- and diagonal peaks in the three-dimensional HNHB experiment (32).

Three-dimensional structure calculation was performed using the simulated annealing/molecular dynamics protocol as implemented in the CYANA software package version 3.0 (33). Upper interproton distance constraints were derived from NOESY (τ_m = 80 ms) cross-peaks via a 1/r⁶ calibration. Torsion angle restraints and stereospecific assignments were obtained from J coupling constants and NOE intensities. Hydrogen bonds were introduced based on temperature coefficients and water exchange rates of H_N protons (protons with gradients less than 4.5 ppb/K were supposed to participate in hydrogen bonding). The disulfide linkages were ascribed according to the experimental data on chemical cleavage published for Tk-AMP-X2 (17) and then were confirmed by CYANA simulation.

The dissociation constant of the H_N proton of the His-13 imidazole ring was determined from the dependence of chemical shift of Hϵ1 on the ambient pH, which was approximated by the equation

where δ is the measured chemical shift and δ_a and δ_b are the chemical shifts of fully protonated and deprotonated states, respectively. pK_a, δ_a, and δ_b were the parameters of approximation (34).

Visual analysis of the calculated structures and figure drawings were performed using PyMOL (35) and MOLMOL (36) software. The molecular hydrophobicity potential approach (37) was used to analyze probable amphiphilic properties of the peptide.

Expression in Xenopus Oocytes

For the expression of voltage-gated potassium channels (Kvs) (rKv1.1, rKv1.2, hKv1.3, rKv1.4, rKv1.5, rKv1.6, Shaker IR, rKv2.1, hKv3.1, and hERG) in Xenopus oocytes, the linearized plasmids were transcribed using the T7 or SP6 mMESSAGE-mMACHINE transcription kit (Ambion). The harvesting of stage V-VI oocytes from anesthetized female Xenopus laevis frog was described previously (38). Oocytes were injected with 50 nl of cRNA at a concentration of 1 ng/nl using a microinjector (Drummond Scientific). The oocytes were incubated in ND96 solution containing 96 mm NaCl, 2 mm KCl, 1.8 mm CaCl₂, 2 mm MgCl₂, and 5 mm HEPES, pH 7.4 supplemented with 50 mg/liter gentamycin sulfate.

Electrophysiological Recordings

Two-electrode voltage clamp recordings were performed at room temperature (18–22 °C) using a Geneclamp 500 amplifier (Molecular Devices) controlled by a pClamp data acquisition system (Axon Instruments). Whole-cell currents from oocytes were recorded in 1–4 days after injection. The bath solution composition was ND96 or HK containing 2 mm NaCl, 96 mm KCl, 1.8 mm CaCl₂, 2 mm MgCl₂, and 5 mm HEPES, pH 7.4. Voltage and current electrodes were filled with 3 m KCl. The resistance of both electrodes was kept between 0.5 and 1.5 megaohms. The elicited currents were filtered at 500 Hz and sampled at 2 kHz using a four-pole low pass Bessel filter. Leak subtraction was performed using a −P/4 protocol. Kv1.1–Kv1.6 and Shaker currents were evoked by 500-ms depolarizations to 0 mV followed by a 500-ms pulse to −50 mV from a holding potential of −90 mV. Current traces of hERG channels were elicited by applying a +40-mV prepulse for 2 s followed by a step to −120 mV for 2 s. Kv2.1 and Kv3.1 currents were elicited by 500-ms pulses to +20 mV from a holding potential of −90 mV. To investigate the current-voltage relationship, current traces were evoked by 10-mV depolarization steps from a holding potential of −90 mV. To assess the concentration dependence of toxin-induced inhibitory effects, a concentration-response curve was constructed in which the percentage of current inhibition was plotted as a function of toxin concentration. Data were fitted with the Hill equation: y = 100/[1 + (IC₅₀/[toxin])^h] where y is the amplitude of the toxin-induced effect, IC₅₀ is the toxin concentration at half-maximal efficacy, [toxin] is the toxin concentration, and h is the Hill coefficient. All data represent at least three independent experiments (n ≥ 3) and are presented as means ± S.E.

RESULTS

Recombinant Peptide Tk-AMP-X2 Production

To provide enough material for Tk-AMP-X2 structure investigation, recombinant peptide and its ¹⁵N-labeled analog were produced in a bacterial expression system. Thioredoxin (Trx) was used as the fusion partner to ensure high yield of the peptide in the native conformation (39). A synthetic gene coding for Tk-AMP-X2 was produced from oligonucleotides and cloned into the pET-32b expression vector. To enable chimeric protein (Trx-Tk-AMP-X2) hydrolysis, sequence encoding an enteropeptidase cleavage site was introduced to the 3′ terminus of the synthetic gene. The resulting plasmid, pET-32b-Tk-AMP-X2, was used to transform E. coli BL21(D3E) cells. Fusion protein isolation by affinity chromatography was followed by SDS-PAGE (Fig. 1A). Pure peptide was achieved as the result of Trx-Tk-AMP-X2 cleavage followed by reversed-phase HPLC (Fig. 1B). The recombinant peptide was shown to have the same retention time and to co-elute with the native Tk-AMP-X2 in analytical HPLC experiments. Moreover, its N-terminal amino acid sequence and molecular mass corresponded to those of the native peptide. For structural studies, the ¹⁵N-labeled peptide was produced using M9 minimal medium with ISOGRO. The final yields of both Tk-AMP-X2 and its ¹⁵N-labeled analog were ∼5 mg/1 liter of bacterial culture.

FIGURE 1.

Production of recombinant Tk-AMP-X2. A, expression and purification of Trx-Tk-AMP-X2 fusion protein as followed by SDS-PAGE (10%). Lane 1, whole-cell lysate of E. coli BL21(DE3) cells carrying the plasmid pET-32b-Tk-AMP-X2 before isopropyl β-d-thiogalactopyranoside treatment; lane 2, induced with 0.2 mm isopropyl β-d-thiogalactopyranoside; lane 3, affinity column flow-through; lane 4, purified fusion protein; lane 5, molecular mass markers (LMW-SDS Marker kit from GE Healthcare); the corresponding molecular mass values are labeled in kDa. The desired product is labeled by an arrow. B, reversed-phase HPLC of the fusion protein Trx-Tk-AMP-X2 cleaved with CNBr on a Jupiter C₅ column. The faction corresponding to Tk-AMP-X2 is indicated with an arrow.

Three-dimensional Structure of Tk-AMP-X2 in Solution

The three-dimensional structure of Tk-AMP-X2 was investigated by NMR spectroscopy. A summary of NMR data obtained for Tk-AMP-X2 is shown in Fig. 2. Analysis of these data indicates that the peptide contains a pair of α-helices connected with a loop. A set of 10 Tk-AMP-X2 structures was calculated in CYANA from 100 random start points using the following experimental data: upper and lower NOE-based distance restraints, J coupling-based torsion angle restraints, and hydrogen bond restraints. The statistics for the obtained set of Tk-AMP-X2 structures are shown in Table 2. Structures from the set are characterized by low CYANA target function values and residual restraint violations and quite a low root mean square deviation value for backbone atoms, thus indicating that the structure of Tk-AMP-X2 is defined accurately and precisely by experimental data in the region corresponding to residues 3–26.

FIGURE 2.

Overview of NMR data. Data determining the secondary structure of Tk-AMP-X2 are shown. ³J_HαHβ2 and ³J_HαHβ3, ³J_NHβ2 and ³J_NHβ3, and ³J_HNHα coupling constants; temperature coefficients of H_N chemical shifts (TempGrad); and NOE connectivities are shown versus the Tk-AMP-X2 sequence. ³J_HαHβ are marked by closed squares if <5 Hz and by open squares if >10 Hz, ³J_NHβ are marked by closed squares if <2 Hz and by open squares if >3 Hz, and ³J_HNHα are marked by closed squares if <5 Hz. Closed triangles correspond to residues, H_N protons of which demonstrate temperature gradients higher than −4.5 ppb/K. Elements of secondary structure are shown on a separate line; helices are shown by rectangles: α-helices, closed, 3₁₀ helix, open.

TABLE 2.

Statistics for the 10 best Tk-AMP-X2 NMR structures

r.m.s.d., root mean square deviation.

Parameter	Value
Distance and angle restraints
Total NOEs	312
Intraresidual	192
Interresidual	120
Sequential (|I − J| + 1)	75
Medium range (1 < |I − J| ≤ 4)	41
Long range (|I − J| > 4)	4
Hydrogen bond restraints (upper/lower)	30/30
S–S bond restraints (upper/lower)	6/6
Angles
ϕ	26
χ1	12
Total restraints per residue	15
Statistics for the calculated set of structures
CYANA target function (Å2)	0.15 ± 0.03
Restraint violations
Distance (≥0.2 Å)	5
Angle (≥5°)	0
r.m.s.d. (Å)
Elements of secondary structure
Backbone	1.05 ± 0.37
All heavy atoms	1.56 ± 0.31
Ramachandran analysis
Residues in most favored regionsa (%)	86
Residues in additional allowed regions (%)	13.6
Residues in generally allowed regions (%)	0
Residues in disallowed regions (%)	0.4

^a Ramachandran analysis was performed with PROCHECK tool at RCSB validation server.

A ribbon representation of Tk-AMP-X2 three-dimensional structure is shown in Fig. 3A. The peptide contains two α-helices (Arg-4 through Met-8 and Arg-15 through Gly-24) and a short segment in a 3₁₀ helix conformation (Cys-9 through Arg-11). The angle between the axes of the helices (θ) equals 133°. Two disulfide bridges stabilize the structure: Cys-5 to Cys-25 and Cys-9 to Cys-21. Alternative disulfide connections were tested in the process of structure calculation, but they did not allow formation of reasonable structures with low target functions. Despite the presence of a number of charged amino acids in the Tk-AMP-X2 amino acid sequence, analysis of the NOE signals and calculated spatial structure did not reveal the formation of salt bridges. The surface of Tk-AMP-X2 is entirely hydrophilic, suggesting that the peptide does not possess membrane-penetrating or pore-forming activity (Fig. 3, B and C).

FIGURE 3.

Three-dimensional structure of Tk-AMP-X2. A, ribbon representation of spatial structure of Tk-AMP-X2. Disulfide bonds (yellow) and positively charged (blue) and negatively charged (red) side chains are shown as sticks. B and C, surface of Tk-AMP-X2 colored with respect to molecular hydrophobicity potential (37) (B) and electrostatic surface potential (C).

Design of a Potassium Channel Blocker Based on the Scaffold of Antimicrobial Peptide

The determined spatial structure confirms that Tk-AMP-X2 belongs to the α-hairpinin plant defense peptide family (18). It should be highlighted that this scaffold is shared also by potassium channel blockers κ-hefutoxin 1 from Heterometrus fulvipes (23) and OmTx1–3 from Opisthacanthus madagascariensis (21).

Although sequence similarity between Tk-AMP-X2 and κ-hefutoxin 1 is very low (∼27% of identical residues), based on their spatial resemblance (Fig. 4), a mutant molecule named Tk-hefu was engineered. The dyad (one lysine and one tyrosine residue, Tyr-5 and Lys-19) crucial for κ-hefutoxin 1 function as a potassium channel blocker was introduced onto the Tk-AMP-X2 scaffold (E6Y and M22K replacements; Table 3). In addition, the adjacent lysine residue 23 in Tk-AMP-X2 was changed to glutamate (K23E replacement) because a positive charge at this position is known to significantly diminish the activity (40).

FIGURE 4.

Comparison of Tk-AMP-X2 and κ-hefutoxin 1 structures. Tk-AMP-X2 is shown in light brown (left), and κ-hefutoxin 1 is shown in light cyan (middle). Cysteine residues are colored yellow. The functional dyad of κ-hefutoxin 1 (Tyr-5 and Lys-19) is colored violet, and corresponding residues of Tk-AMP-X2 (Glu-6 and Met-22) are colored red. The three-dimensional structure alignment is shown at right.

TABLE 3.

Tk-AMP-X2, κ-hefutoxin 1, and Tk-hefu sequences

aa, amino acids.

Peptide name	Amino acid sequencea	Length
		aa
κ-Hefutoxin 1	-GHACYRNCWR--EGNDEETCKERC---	22
Tk-AMP-X2	ADDRCERMCQRYHDRREKKQCMKGCRYG	28
Tk-hefu	ADDRCYRMCQRYHDRREKKQCKEGCRYG	28

^a Residues constituting the functional dyad are underlined. Glu-23 in Tk-hefu corresponding to Glu-20 in κ-hefutoxin 1 is in italics.

The mutant molecule Tk-hefu was produced in E. coli with a yield of ∼4 mg/1 liter of bacterial culture similarly to the parent peptide Tk-AMP-X2. Its measured monoisotopic molecular mass (3548.3 Da) matched the calculated value (3548.6 Da). Correct disulfide linkages were assigned by chemical cleavage: recombinant Tk-AMP-X2 was analyzed as reported previously for the native peptide (17), and Tk-hefu was successively digested by endoproteinase Glu-C and CNBr. The masses of the resulting products unequivocally corresponded to fragments connected by the C¹–C⁴ and C²–C³ bridges. Recombinant Tk-AMP-X2 co-eluted with the native peptide in analytical HPLC. Both peptides were then used in electrophysiological studies to determine potassium channel-blocking activity.

Electrophysiological Recordings

At a concentration of 40 μm, Tk-AMP-X2 and the chimera Tk-hefu were submitted to a wide screening on a panel of Kvs belonging to different families. Their activity was investigated against members of the Shaker (Kv1.1–Kv1.6 and Shaker IR), Shab (Kv2.1), Shaw (Kv3.1), and erg (hERG) families (Fig. 5). Although the parent molecule Tk-AMP-X2 did not exert any activity against the tested Kvs even at concentrations up to 250 μm, the chimera Tk-hefu did selectively target members of the Shaker family. At 40 μm, it inhibited the potassium currents through Kv1.2, Kv1.3, and Kv1.6 channels by 8.3 ± 1.3, 58.4 ± 1.6, and 7.3 ± 2.3%, respectively. No activity on other channels was noticed. To investigate the affinity of Tk-hefu to Kv1.3 channels, a concentration-response curve was constructed (Fig. 6A). The IC₅₀ value was 34.0 ± 2.8 μm.

FIGURE 5.

Activity profile of Tk-hefu on several Kv channel isoforms. Representative whole-cell current traces in control and toxin conditions are shown. The dotted line indicates the zero current level. The asterisk (*) marks the steady-state current traces after application of 40 μm toxin. Traces shown are representatives of at least three independent experiments (n ≥ 3).

FIGURE 6.

Concentration and voltage dependence of the Tk-hefu-induced decrease of potassium currents (I). A, concentration-response curves for Tk-hefu (squares) and Tk-AMP-X2 (circles) on Kv1.3 channels. B, steady-state activation curves in control (squares) and after toxin application (40 μm Tk-hefu; circles). Error bars represent S.E.

Further experiments demonstrated that no significant alteration of the midpoint of activation was observed for Kv1.3 channels because the V_1/2 was determined at −8.4 ± 1.8 mV in control conditions and −5.5 ± 2.4 mV after application of 40 μm Tk-hefu (Fig. 6B). In concordance to what has been reported for native κ-hefutoxin 1, the mutant peptide inhibits the current through Kv1.3 channels in a dose-dependent, but voltage-independent, manner. No difference in the degree of block was observed in the range of depolarizing potentials from −30 to +30 mV (data not shown). Moreover, no modification of gating kinetics of Kv1.3 channels was observed.

DISCUSSION

Tk-AMP-X2 Is a Member of the α-Hairpinin Structural Family

At present, the classification of plant defense peptides is based on certain structural features, cysteine motif and spatial organization in particular, and includes several commonly known families such as defensins, thionins, and knottins (41). The deduced Tk-AMP-X2 spatial structure (Fig. 3A) confirmed that this peptide belongs to the recently described family of plant defense peptides called α-hairpinins (18). Representatives of this family possess a unique cysteine pattern, CX₃CX_nCX₃C (where X stands for any residue), and specific three-dimensional structure motif, which is formed by two antiparallel α-helices stabilized by two intramolecular disulfide bridges. Of note, the cysteine pattern itself is not sufficient for reliable spatial structure prediction and, therefore, peptide classification. For instance, the CX₃CX_nCX₃C motif was found in other AMPs that, in contrast to α-hairpinins, possess β-hairpin structure (42, 43).

According to available data, the three-dimensional structures of four other α-hairpinins except for Tk-AMP-X2 are known; they are the antifungal peptide EcAMP1 from Echinochloa crus-galli (20); two inhibitors of serine proteases, namely BWI-2c from Fagopyrum esculentum (18) and VhTI from Veronica hederifolia (19); and the ribosome-inactivating peptide luffin P1 from Luffa aegyptiaca (44). Several other molecules are likely to belong to α-hairpinins based on their cysteine motif and, for some, circular dichroism data (45–49).

The Cysteine-stabilized Helical Hairpin Fold Is Characteristic of Some Animal Toxins

The search for plant α-hairpinins structural analogs revealed a number of peptides adopting the α-helical hairpin fold. Some of them possess the same cysteine pattern as the plant α-hairpinins (21–23, 50–52). Most of them belong or are presumed to belong to the group of potassium channel blockers, including κ-hefutoxins isolated from the venom of scorpion H. fulvipes, OmTx1–3 from the scorpion O. madagascariensis, flf14a–c from the cone snail Conus floridanus floridensis, and vil14a from the cone snail Conus villepinii.

Structural similarity between α-hairpinins and potassium channel blockers (Fig. 4) may be due to common descent. Similar to the peptide MsDef1, which was shown to block calcium channels (53), α-hairpinins could be blockers of potassium channels in fungi. During further evolution, a more specific and efficient mechanism was probably developed because only a couple of mutations in the DNA code were enough to turn a potassium channel blocker into an antifungal peptide (or vice versa) as exemplified by Tk-AMP-X2 and κ-hefutoxin 1 in this work. The presented data support a divergent evolution scenario.

There are also several examples of peptides sharing the helical hairpin fold with a different cysteine motif (54–56). The most interesting of them in terms of comparison with α-hairpinins is the potassium channel blocker κ-BUTX-Tt2b from the venom of the scorpion Tityus trivittatus (56). It possesses the cysteine motif C¹X₄C²X₃C³X₈C⁴X₃C⁵XC⁶, which in the underlined region resembles plant α-hairpinins. However, the disulfide pairing is different: C¹–C⁵, C²–C⁴, and C³–C⁶.

Remarkably, there are some other examples of structural resemblance between peptides from different organisms. For instance, plant defensins together with fungal and insect defensins assume the CSαβ fold just like scorpion toxins (57). Plant knottins and cyclotides adopt the inhibitor cystine knot fold characteristic of spider neurotoxins (58). The instances of structural similarity between plant defense peptides and toxins from venomous animals, which are separated by millions of years of evolution, are intriguing and may serve as an indication of a very early emergence of the listed structural motifs and further divergent evolution. The alternative scenario is independent recruitment of the same fold in distant organisms.

The Helical Hairpin Fold Can Be Used as a Template for Engineering Desired Functionalities

The structure-function relations in blockers of the Shaker family of Kvs are more or less understood or at least well probed. It was shown that peptides containing the so-called “functional dyad” usually display potassium channel-blocking activity, although for several molecules, the dyad was shown to be insufficient (59–62). Moreover, the importance of some other residues, which often serve for higher binding affinity and better blocking ability, was also shown (63–65).

We introduced the lysine-tyrosine dyad into Tk-AMP-X2 to obtain potassium channel-blocking function on its helical hairpin scaffold (Table 3). The electrophysiological studies demonstrated that the mutant molecule, Tk-hefu, possessed the predicted properties (Figs. 5 and 6). κ-Hefutoxin 1 is known to inhibit Kv1.2, Kv1.3, and Kv1.6 channels (23, 66). A similar selectivity profile was observed for Tk-hefu (Fig. 5). Moreover, its activity was even slightly greater than that of the dyad donor κ-hefutoxin 1: at 40 μm, the extent of Kv1.3 current inhibition was 58% for Tk-hefu versus 50% for κ-hefutoxin 1 (23).

A more detailed comparison of κ-hefutoxin 1 and Tk-hefu was performed. Although κ-hefutoxin 1 is capable of modifying the kinetics of Kv1.3 channels in a voltage-independent manner (66), Tk-hefu failed to exert a similar activity. However, several other κ-hefutoxin-related toxins, members of the so-called κ-KTX family, do not modify the activation kinetics of Kv1.3 channels either (67). Therefore, it seems that specific residues differing between κ-hefutoxin 1 and other members of the family are responsible for the modulation of Kv1.3 channels. Curiously, the parent molecule Tk-AMP-X2 contains lysine and tyrosine residues separated by four amino acid residues like for instance in the ShK toxin (68). However, no activity was found in the electrophysiological tests. It may be that location in the same helix gives insufficient disposition of these two residues, thus precluding block of potassium channels.

In this work, the potassium channel-blocking function was engineered based on structural resemblance. These results should not be underestimated. In fact, it was shown that similarly to the well known scaffolds such as Ig or FN a quite minimalistic α-hairpin scaffold could be successfully used for functional engineering, which permits its future usage for drug design.

Briefly summarized, in this study, the spatial structure of Tk-AMP-X2, an AMP from wheat seeds, was studied. It was shown that Tk-AMP-X2 adopts the α-hairpinin fold, which is also shared by a number of plant defense peptides and invertebrate toxins. Based on this observation, we assumed that the α-hairpinin fold is a universal type of peptide fold. Successful construction of a novel molecule with predicted function indicates an opportunity for further utilization of this scaffold as a framework for rational design and protein engineering.

It is important to point out that the development of new antimicrobial agents has become one of the most necessary needs of current biotechnology and clinical drug design in the last decades. The number of multiresistant pathogens and plant pests keeps increasing. Generation of novel, highly stable peptide antibiotics and transgenic resistant crops is often considered as a possible solution to the problem. A scaffold engineering approach appears to offer a useful technique for drug development and has already shown some success. Small size, high stability, and a wide spectrum of biological functions make the α-helical hairpin peptides promising subjects for this approach.