We first evaluated Tsv activity with peptidase inhibitors, using two fluorescent substrates: Abz-FRK(Dnp)P-OH, which was designed for ACE studies [31], and Abz-GGFLRRV-EDDnp, homologous to the dynorphin 1–13 sequence and previously determined as substrate for Tsv endopeptidases [18]. Results are presented in Figure 1.

Although only metallopeptidases were detected in Tsv (due to inhibition of EDTA and 1,10-phenantroline), we observed that different proteases act on each substrate. For Abz-FRK(Dnp)P-OH, the two classical ACE inhibitors, captopril (100 nM) and BPP 10c (16 µM), significantly inhibited Tsv activity (98% and 90%, respectively), while Abz-GGFLRRV-EDDnp hydrolysis was not affected by these inhibitors at the same concentrations. However, PMSF (2 mM) had no effect on the activity of the venom on any fluorescent substrates tested. Thus, it seems that Abz-GGFLRRV-EDDnp is cleaved mainly by metalloendopeptidases and Abz-FRK(Dnp)P-OH by metallo dipeptidyl carboxypeptidases.

We tested nine different concentrations of arachnid (AAV) and scorpion (SAV) antivenoms to determine their efficacy in neutralizing Tsv hydrolysis of Abz-FRK(Dnp)P-OH (Figure 2).

SAV partially inhibited the activity even at low venom:antivenom ratios, whereas AAV started inhibiting Tsv at a higher dose, after 25 µg of antivenom (1:25), but also reached higher levels of inhibition than SAV at concentrations of 1:500 µg. Neither antivenom fully neutralized Tsv activity at the highest dose tested (1:1000), with inhibition reaching 94% for SAV and 98% for AAV.

As chloride ions are known to affect ACE activity [32], we compared Tsv and ACE hydrolysis of the Abz-FRK(Dnp)P-OH substrate in the presence of different Cl⁻ concentrations (Figure 3).

As showed in Figure 3, both enzymes were already active in the absence of NaCl, with higher proteolytic activities observed as NaCl concentrations increased. At 10 mM NaCl, Tsv and ACE activities increased by 15% and 25%, respectively, and, at the highest NaCl concentration tested (50 mM) hydrolysis of Abz-FRK(Dnp)P-OH by Tsv increased by more than 35%, and around 46% by ACE.

Angiotensin I and bradykinin were incubated with Tsv in the classical ACE working buffer, 100 mM Tris, 50 mM NaCl, 10 µM ZnCl₂, pH 7.0. After verifying, with reserve phase chromatography, that both peptides were substrates for Tsv, we manually collected the products of hydrolysis and analyzed them by mass spectrometry. Additionally, we determined the specific activity of the venom on each peptide, including hemopressin, and determined the cleavage sites (Table 1).

Hemopressin was the best substrate for the whole venom (0.40 µM/µg/min), followed by angiotensin I (0.05 µM/µg/min) and bradykinin (0.045 µM/µg/min). One of the fragments collected from angiotensin I corresponded to angiotensin II, due to the removal of His⁹-Leu¹⁰. Other fragments were also formed from angiotensin I hydrolysis, such as Ang_(1–4), Ang_(5–10), Ang_(1–7), and Ang_(8–10), probably due to endopeptidase activity; and Ang III and Ang_(6–10), probably due to aminopeptidase activity. Bradykinin hydrolysis by the venom formed the fragments BK_(1–7), BK_(1–5), the expected products of ACE-like activity, and BK_(3–7). Hemopressin cleavage sites were determined previously [18], but the hydrolysis rate increased 5.6 times when the buffer containing 10 µM ZnCl₂ was used.

As new substrates and hydrolysis rates were observed for Tsv when using the ACE buffer, we evaluated inhibition by captopril, a potent and specific human ACE inhibitor, and by EDTA (Table 2).

EDTA fully inhibited the activity of Tsv on all substrates. Captopril, in either of the two concentrations tested, failed to inhibit the cleavage of dynorphin 1–13 in any concentration, which was expected, as a similar result was observed with the analog fluorescent substrate, Abz-GGFLRRV-EDDnp (Figure 1). On the other hand, captopril inhibited bradykinin and angiotensin I hydrolyses in a dose-dependent manner (Table 2). Hydrolysis of hemopressin was also strongly inhibited (37.8%) by 100 nM captopril, but less so with a higher dose (1 µM) of this inhibitor. The RP-HPLC profile of hemopressin hydrolysis by Tsv (Figure 4), shows that, when captopril was present (in both doses), one of the products (PVNFKFL) was not formed, indicating that the dipeptidyl carboxypeptidase activity of the whole venom was abolished. Nevertheless, the other products (PVNFK and FLSH) were not affected by this inhibitor.

In order to purify the ACE-like peptidase present in Tsv, we performed three chromatographic steps to obtain a homogeneous and active enzyme. First, the whole Tsv was applied to an HPLC-DEAE column, and the ACE-like activity was identified (77 UF/µg/min) in fraction 1, which did not interact with the column (data not show). Then, fraction 1 was loaded onto a Diol-300 column, from which fractions were collected (Figure 5, panel A) and screened by Abz-FRK(Dnp)P-OH hydrolysis. Since fraction F1-2 showed the best hydrolysis rates (88 UF/µg/min), it was further fractioned with PA-CM ion exchange chromatography. After this last step, we observed one active fraction, presenting a single band of 70 kDa (Figure 5, panel B), with a specific activity of 220 UF/µg/min on the FRET substrate. In the final chromatography step, the purification factor obtained was 22.9 with a yield of 1% (Table S1).

The protein content from the SDS-PAGE band was extracted and subjected to MS/MS for peptide fingerprint analysis. Using a database that combined sequences restricted to Tityus genus from UNIPROT and the transcript sequences of ACE-like peptidase from Tityus scorpion venom glands, PeaksDB was able to identify, with high confidence (FDR ≤ 1%), two unique ACE-like peptides from Tityus serrulatus (TserSP00939) (Figure 6). Additionally, the pure enzyme was able to convert angiotensin I into angiotensin II, in addition to Abz-FRK(Dnp)P-OH, with specific activity of 0.01 µM/µg/min (Figure 5, panel C).

We aligned the amino acid sequences of ACE-like peptidases obtained from transcriptomics data analysis from Tityus serrulatus, T. obscurus and T. bahiensis with each other and with human testicular ACE (Table 3).

The identity with human tACE varied from 23% to 39%; however, the metallopeptidase motif was highly conserved throughout the species (Figure S1). The ACE-like sequences of Tityus serrulatus (TserSP00939) and Tityus obscurus (Tobs01141) had 92.51% of similarity, while the tbahACE-like sequence shared 62.31% and 67.34% similarity with T. obscurus and T. serrulatus, respectively.

Phenylmethanesulfonylfluoride (PMSF), 1,10-phenanthroline, dynorphin 1–13 (Dyn 1–13), hemopressin (hemo), angiotensin I (Ang I), bradykinin (BK), captopril, and rabbit lung somatic angiotensin I-converting enzyme (ACE) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Acetonitrile and trifluoroacetic acid (TFA) were acquired from J.T. Baker. The fluorescent resonance energy transfer (FRET) substrates Abz-GGFLRRV-EDDnp and Abz-FRK(Dnp)P-OH, synthesized with automated solid-phase synthesis [52], were kindly provided by Dr. Adriana Carmona, from the Department of Biophysics of UNIFESP-EPM, São Paulo, Brazil.

The lyophilized venom of Tityus serrulatus (Batch No. 2146 and batch No. 744) were provided by the Venom Section of Instituto Butantan, SP, Brazil. The venoms were submitted to a 10 kDa molecular weight cutoff (Amicon Ultra-15 Centrifugal Filter Devices) and stock solution was prepared in 50 mM sodium phosphate and 50 mM NaCl, pH 6.0. The scorpion and arachnid antivenoms (SAV and AAV, respectively) were from the Hyperimmune Plasmas Processing Section, Instituto Butantan, SP, Brazil. The SAV (batch No. 0905104/A) and the AAV (batch No. 0706121) protein concentrations were 8.43 g/dL and 15.4 g/dL, respectively. The antivenoms from Instituto Butantan are produced through hyperimmunization of horses, with a pool of T. serrulatus (50%) and T. bahienses (50%) venoms for SAV, and T. serrulatus (57%), Phoneutria nigriventer (21.5%) and Loxosceles gaucho (21.5%) venoms for AAV.

The substrate-specific assay was carried out using 1 µg Tityus serrulatus venom (Tsv) and the fluorescent substrates Abz-GGFLRRV-EDDnp (5 µM) [18] and Abz-FRK(Dnp)P-OH (4 µM), in 100 mM Tris, 50 mM NaCl, 10 µM ZnCl₂ buffer, pH 7.0. All reactions were monitored by measuring hydrolyses using a fluorimeter (Victor 3™, Perkin-Elmer, Waltham, MA, USA; λem 420 nm and λex 320 nm), at a stable temperature of 37 °C. The measurements of peptidase activity were made for 15 min continuously (one read per minute). The fluorometric assays were analyzed using Grafit 5.0 from Erithacus Software (version 5.0.6, 1989-2003, Erithacus Software, West Sussex, UK), and the hydrolysis rates (UF/min) were determined. All fluorometric measurements were made in duplicate, and the results are shown as the mean with SD.

For the inhibition of Tsv, we used the serine peptidase inhibitor PMSF (2 mM), the metallopeptidase inhibitors 1,10-phenantroline (2 mM) and EDTA (50 mM), and the peptidase-specific inhibitors captopril (100 nM) and BPP 10c (Bradykinin Potentiating Peptide, <ENWPHPQIPP, 16 µM). Control samples included a volume of ethanol equal to that used in the PMSF and 1,10-phenantroline stock solutions. PMSF and 1,10-phenantroline were pre-incubated for 30 min at room temperature before the test.

The effects of ion chloride on Tsv were evaluated in parallel with angiotensin I-converting enzyme (ACE) on fluorometric experiments. For this, Tsv and ACE were added to 96-well plates at concentrations of 1 µg and 50 ng, respectively, in 100 mM Tris, 10 µM ZnCl₂, pH 7.0 buffer, at a final volume of 100 µL, with 4 µM Abz-FRK(Dnp)P-OH. Four concentrations of NaCl were tested on Tsv and ACE: 0 (control), 10 mM, 20 mM, and 50 mM. All the assays were performed in triplicate, and the specific venom peptidase activities were expressed as units of free fluorescence of the cleaved substrate per µg of venom per min (UF/µg/min).

The serum neutralization assay of Tsv activity upon the Abz-FRK(Dnp)P-OH substrate was performed using different doses of the antivenoms. The Tsv (1 µg) was incubated at room temperature with the antivenoms for 30 min in the following concentrations (weight ratio of venom and antivenom, respectively): 0 (control), 1:1; 1:2; 1:10; 1:25; 1:50; 1:100; 1:250; 1:500, and 1:1000. After incubation, the FRET substrate was added, and the residual peptidase activity of the venom measured, in duplicate, as described (Section 4.3).

Tsv (0.5 µg) was incubated in 100 mM Tris, 50 mM NaCl, 10 µM ZnCl₂ buffer, pH 7.0, at 37 °C with dynorphin 1–13 (30 µM), for 70 min; hemopressin (30 µM) for 2 h; angiotensin I (30 µM) for 4 h; and bradykinin (30 µM) for 6 h. Captopril was used at 100 nM and 1 µM final concentrations. EDTA was also used at 100 mM final concentration. All experiments were performed in duplicate, and the results are shown as the mean with SD. Hydrolyses were analyzed by reverse-phase HPLC (Prominence, Shimadzu, Japan), with 0.1% trifluoroacetic acid (TFA) in water, as solvent A, and acetonitrile and solvent A (9:1), as solvent B. Separations were performed at a flow rate of 1 mL/min, using a Restek Ultra C-18 column (4.6 mm × 250 mm) and a 20%–60% gradient of solvent B over 30 min. In all cases, elution was followed by measurement of ultraviolet absorption (214 nm). The specific activities were expressed in µM of hydrolyzed substrate per µg of venom per minute (µM/µg/min) and the inhibition of venom peptidase activity was calculated by comparing the peptide areas.

The products of the hydrolysis of angiotensin-I and bradykinin, catalyzed by Tsv, were manually collected and subjected to mass spectrometry analysis. The reverse phase chromatography fractions were resuspended in 0.1% formic acid and analyzed by online liquid chromatography in an Easy-nLC Proxeon nanoHPLC system coupled to an LTQ-Orbitrap Velos (Thermo Fisher Scientific, Bremen, Germany) through a nanoelectrospray ion source. Separation was carried out in a 10-cm column (75 µm i.d. × 350 µm e.d.) packed in-house with 5-µm Jupiter^® C-18 beads (Phenomenex, Torrance, CA, USA). Peptides were eluted with a linear gradient of 5%–30% acetonitrile, in 0.1% formic acid, in 45 min at 300 nL/min. The spectrometer was operated in data-dependent mode and the 10 most intense peaks were selected for CID fragmentation after acquiring each full scan. The settings for the spectrometer were defined as: high-resolution full MS parameters (1 μscan; full MS mass range m/z of 200–2000 with an R = 30,000 and a target value of 1× 104 ions; max injection time = 10 ms). For fragment scans the settings were: an isolation window of 2 Da, a max list size of 500, a time window of 30 s, a minimum signal of 5000, activation time = 10 ms and normalized collision energy = 35%.

The raw data files were submitted to searches against the angiotensin I and bradykinin sequences using PEAKS Studio (version 8, Bioinformatics Solution, Waterloo, ON, Canada) [53,54]. A decoy database was also searched to calculate False Discovery Rate (FDR) using the decoy-fusion method [53,54]. The search parameters were: no enzyme specificity; precursor mass tolerance set to ±10 ppm and a fragment ion mass tolerance of ±0.5 Da; oxidized methionine (M + 15.994915 Da) was set as variable modification. The identified peptides were then sorted by their Average of Local Confidence (ALC) to select the best spectra to annotate, and were filtered by FDR ≤ 1%.

All purification steps were followed by buffer exchange to 100 mM Tris, 50 mM NaCl, 10 µM ZnCl₂ pH 7.0, using a 10 kDa molecular weight cut off (Amicon Ultra-15 Centrifugal Filter Devices), taking into account that ACE activity is sensitive to chloride concentrations [32]. Moreover, the protein content of all samples was quantified with the commercial kit Quick Start™ Bradford Protein Assay (Bio Rad, Hercules, CA, USA). In order to screen for ACE-like activity, each collected fraction was tested with a fluorometric assay (5% of final volume) using the fluorescent substrate Abz-FRK(Dnp)P-OH (4 µM), as described in Section 4.3.

The lyophilized Tityus serrulatus venom was dissolved in 5 mL of 20 mM Tris, 20 mM NaCl, pH 8.2 buffer (final concentration 10 mg/mL). Tsv was first submitted to anion exchange chromatography in an HPLC system (Shimadzu Co., Kyoto, Japan) using a Shim-Pack PA-DEAE column (20 mm × 100 mm) at 5 mL/min flow. The gradient used was 0%–40% B for 40 min (buffer A containing 20 mM Tris, 20 mM NaCl, pH 8.2 and buffer B composed of buffer A containing 500 mM NaCl, pH 8.2). Next, fraction 1 was applied to a Shim-pack Diol-300 (7.9 mm × 50 cm) gel filtration column coupled to an HPLC system (Shimadzu Co., Kyoto, Japan) and was eluted with 200 mM sodium sulfate, 10 mM sodium phosphate, pH 7.0 buffer at 0.5 mL/min flow rate. Then, F1-2, which is the active fraction upon the FRET substrate, was injected into a cation exchange Shim-Pack PA-CM column (20 mm × 100 mm) in an HPLC system (Shimadzu Co., Kyoto, Japan) at a flow rate of 5 mL/min. The gradient used was 0%–40% B in 40 min, with buffer A containing 20 mM Tris, 20 mM NaCl, pH 7.0 and buffer B the same as buffer A but with 500 mM NaCl, pH 7.0. For all chromatographic steps, UV detection was at 280 nm. Lastly, the homogeneous ACE-like peptidase (100 ng) was incubated with angiotensin I (30 µM) overnight, in 100 mM Tris, 50 mM NaCl, 10 µM ZnCl₂ buffer, pH 7.0 and the hydrolysis was verified by HPLC, as described in Section 4.6.