Enzyme sequences were obtained by retrieving the corresponding domain sequence from UniProt [serine peptidase Phoneutria nigriventer P84033 (Phoni) and astacin-like metalloprotease L. intermedia C9D7R2 (Loxin)]. Translation and codon optimization for E. coli were carried out using EMBOSS Backtranseq by EMBL-EBI (Madeira et al., 2024). To enable subsequent Gibson cloning, 20 base pairs matching the vector sequence were added at the 3′ and 5′ ends during synthesis. Codon-optimized genes were synthesized using Thermo Fisher GeneArt Gene and Protein Synthesis Services (Germany), resulting in the delivery of Loxin as a gene fragment, whereas the gene fragment for Phoni was delivered within a plasmid. The E. coli expression vector pET-32a (+) (Novagen, Germany) and the plasmid containing the gene fragment of Phoni were prepared through digestion with EcoRI-HF and KpnI-HF restriction enzymes (New England Biolabs, Germany). The resulting opened vector and desired Phoni gene fragment were purified using the DNA recovery kit (Zymoclean Gel DNA Recovery Kit; Cat. No.: D4007, Zymo Research, Germany), according to the manufacturer’s instructions. For Loxin, the obtained enzyme sequence was transferred to the opened expression vector pET-32a (+) using the Gibson Assembly Cloning Kit (Cat. No. E5510S, New England Biolabs, Germany), following the manufacturer’s instructions. The gene fragment of Phoni was transferred to the open vector by ligation using a T4 DNA Ligase (Promega, M179A) in a vector: insert ratio of 1:3, according to the manufacturer’s instructions. The resulting plasmid (Figure 1) was used to transform chemically competent E. coli NEB® 5-alpha cells (Novagen, Germany) by mixing 2 µL plasmid with 25 µL competent E. coli NEB® 5-alpha cells. After a 15-min incubation on ice, a 45-s heat shock at 42 °C, followed by a 2-min incubation on ice, was carried out. A volume of 250 μL of the included SOC-medium was added to the sample and incubated for 1 h at 37 °C and 300 rpm agitation (Thermoblock, Eppendorf, Germany). Afterward, the cells were plated out on Luria-Bertani (LB) agar plates containing 100 μg/mL ampicillin (LB_amp) and incubated overnight at 37 °C. One clone was used to inoculate 3 mL liquid LB_amp medium and incubated overnight at 37 °C and 180 rpm. The plasmid was purified using the NucleoSpin® Plasmid EasyPure Kit (Cat. No. 740.727.250, Macherey-Nagel GmbH & Co. KG, Germany), following the manufacturer’s instructions. DNA concentration and purity were determined photometrically with A₂₆₀/₂₈₀ using the BioTek® EON^TM Microplate Reader (BioTek Instruments, United States), and the plasmid sequence was confirmed using Oxford Nanopore sequencing (Eurofins Genomics, Germany).

Purified plasmids were transformed into six different chemically competent E. coli strains [Origami^TM 2, Origami^TM B, BL21 (DE3), Rosetta-gami^TM B, and Rosetta-Gami^TM 2 from Merck Millipore/Novagen (Germany) and SHuffle T7 Express lysY from New England Labs (Germany)]. For transformation, 10 ng of plasmid DNA was added to 100 µL chemically competent cells, and the procedure was carried out as described in Section 2.1. Afterward, cells were plated out on LB_amp agar plates and incubated overnight at 37 °C. One clone per successfully transformed strain was used to inoculate 3 mL of Terrific Broth (TB) medium with 100 μg/mL ampicillin (TB_amp) and incubated at 37 °C and 180 rpm overnight. The culture was used to prepare cryo-conserved stocks by combining 500 µL of culture with 500 µL of 30% glycerol and stored at −80 °C.

Fusion protein production was examined by cultivating transformed E. coli strains using three distinct production temperatures. Fusion protein biosynthesis was confirmed using one-dimensional sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). This technique enabled the comparison of the expected protein bands with the present bands (Table 1). For this, overnight cultures of 3 mL TB_amp medium with 3 µL cryo stock were incubated at 37 °C and 180 rpm and used to inoculate 50 mL TB_amp medium in 250 mL baffled flasks. The cultures were incubated at 37 °C and 180 rpm until the bacterial growth reached a optical density measured at 600 nm (ΔOD₆₀₀) of 0.8–1.0. Isopropyl-β-D-thiogalactopyranosid (IPTG) was added to a final concentration of 1 mM to induce the fusion protein biosynthesis under the control of a T7 promoter. After induction, cultures were incubated at three distinct temperatures (16 °C, 28 °C, or 37 °C). Cultures were incubated at 37 °C, 28 °C, and 16 °C were then incubated for 4 h, 16 h, and 24 h, respectively, and samples for SDS-PAGE were collected throughout the incubation period. For this, ΔOD₆₀₀ was measured, and a cell volume corresponding to 500 μL cell culture/ΔOD₆₀₀ per sample was collected. This standardized cell concentrations, enabling accurate estimation of protein yield without interference from variations in cell density. Cultivation medium was removed by centrifugation at 12,000 × g for 5 min, and the pellet was stored at −20 °C until analysis using SDS-PAGE.

For the SDS-PAGE, the sampled pellets were resuspended in 25 µL ultra-pure water and 25 µL 2× Laemmli buffer with 6% β-mercaptoethanol. Afterward, samples were incubated at 95 °C for 5 min and centrifuged for 3 min at 12,000 × g. An aliquot of 3 μL of each sample, together with Precision Plus Protein™ Dual Xtra Prestained Protein Standards (Cat. No. 1610377, BioRad, Germany), was loaded onto a 4%–20% Criterion™ Tris-HCl Protein Gel (Cat. No. 3450034, BioRad, Germany), and gel electrophoresis was carried out with 150 V over 55 min. The gels were stained using Roti® Blue quick ready-to-use 1X solution (Carl Roth, Germany) for 2.5 h and destained with 10% ethanol to visualize protein bands on the gel.

Expression samples were separated by reducing SDS-PAGE and transferred to a polyvinylidene difluoride (PVDF) membrane (Cat. No. 1704157, BioRad, Germany) using the Trans-Blot® Turbo™ Transfer System (BioRad, Germany). The membrane was blocked overnight at room temperature in 5% BSA solution, washed thrice with PBS +0.1% Tween 20, and incubated for 2 h at room temperature with Penta-His-HRP Conjugate (1:2000 dilution) and Precision Protein StrepTactin HRP (1:10000). Afterward, the membrane was washed thrice with PBS +0.1% Tween 20. Visualization was carried out using the Opti-4CN Substrate Kit (Cat. No. 1708235, BioRad, Germany).

As E. coli BL21 (DE3) was identified as the optimal production strain, fusion protein production was scaled up to 400 mL. For this, 3 mL of TB_amp medium was inoculated with 10 µL of cryo-conserved stock at 37 °C and 180 rpm. After 7 h of incubation, 100 µL of the culture was transferred into 50 mL of TB_amp medium in a 250-mL baffled flask and incubated overnight under the same conditions. To produce the fusion protein six times, 400 mL TB_amp medium in 1 L baffled flasks was inoculated with each 1 mL overnight culture. Bacterial growth was monitored using ΔOD₆₀₀, and protein production was induced with 1 mM IPTG when cultures reached a ΔOD₆₀₀ value of 0.8–1.0. After induction, the cultures were incubated at 28 °C for 5 h, and the cells were harvested by centrifugation for 15 min at 10,000 × g at 4 °C. The supernatant was discarded, and the cell pellet was stored at −20 °C until further processing. The cell pellet was resuspended with 100 mL lysis buffer (30 mM Tris and 100 mM NaCl; pH 7.5), and cell lysis was carried out using a Microfluidizer® (Microfluidics, United States) in two rounds. The lysate was centrifuged at 75,000 × g for 30 min at 10 °C; the resulting supernatant was discarded, and the pellet was stored at −20 °C. For solubilization, 10 g of wet pellet was suspended in 50 mL of solubilization buffer (8 M guanidinium chloride and 30 mM Tris; pH 7.5) and stirred overnight at room temperature. Afterward, the solution was centrifuged twice at 75,000 × g for 30 min at 10 °C. Due to the high viscosity, 9.5 mL of the solution was diluted with 240.5 mL solubilization buffer and applied to an immobilized metal affinity chromatography (IMAC) system (SE-04 FPLC) equipped with a 10 mL TALON® Superflow Resin Column (Cat. No. 635670, Clontech Laboratories, United States). The system was equilibrated with IMAC sample buffer (6 M urea, 30 mM Tris, and 100 mM NaCl; pH 7.5). After the sample application using 1 mL/min, the column was washed with IMAC sample buffer at a flow rate of 2 mL/min. Elution was performed at a flow rate of 1 mL/min using a stepwise gradient of 10%, 20%, 50%, and 100% IMAC elution buffer (6 M urea, 30 mM Tris, 100 mM NaCl, and 200 mM imidazole; pH 7.5) until baseline was reached. The process was monitored by UV absorbance at 280 nm, and fractions were collected for analysis by SDS-PAGE as described above.

Before refolding, the fusion protein (∼0.3 g/L) was supplemented with 1% β-mercaptoethanol and incubated at 37 °C for 20 min. Refolding was carried out by dialysis using a Roth dialysis tube with a 14 kDa cut off (Cat. No. 1784.1, Carl Roth). For this, several dialysis approaches were compared. Dialysis was carried out at 4 °C with continuous stirring using cold dialysis buffer at a ratio of 2 L buffer per 10 mL sample. After 18 h of incubation, the buffer was exchanged, and a second round of dialysis against fresh dialysis buffer was conducted for an additional 18 h. The solution was clarified by centrifugation (18,000 × g, 4 °C, 30 min). For the first dialysis approach, the same buffer was used for both rounds. For the second and third approaches, three rounds with different buffers were applied (Supplementary Table S1).

The first dialysis approach used Tris–NaCl buffer (20 mM Tris and 150 mM NaCl) at different pH levels (7.0, 7.5, 8.5, and 9.0) and tested initial fusion protein concentrations of 0.3 g/L, 0.2 g/L, and 0.1 g/L. Additionally, HEPES buffer (20 mM HEPES buffer, pH 6.8) and MES buffer (20 mM MES buffer, pH 6.5) were tested. The second dialysis approach used a Tris–NaCl buffer at pH 8.5, supplemented with 0.8 M L-arginine as the first buffer, 0.1 µM zinc sulfate as the second buffer, and no supplements as the third buffer. The third dialysis approach explored a Tris buffer (50 mM Tris, pH 8.5) following the same addition of L-arginine, zinc sulfate, or none as described in the second dialysis approach.

The Loxin fusion protein showed surface adsorption after refolding, leading to a loss of protein. Therefore, two storage tubes and two detergents, commonly used for membrane proteins, were analyzed for their ability to prevent the attachment processes. For this, 20 µL aliquots of dialyzed Loxin fusion protein were transferred into either a low-binding tube (Sarstedt AG & Co. KG, Germany) or an HPLC glass vial (Fisher Scientific GmbH, Germany) and incubated. Additionally, Tween 20 or Triton X-100 was added directly to the dialysis tube to final concentrations of 0.8 mM or 0.5 mM, respectively. The mixtures were thoroughly mixed, and the solutions were transferred into low-binding tubes (Sarstedt AG & Co. KG, Germany). Samples were collected before transfer into the storage tube (S), immediately after transfer into the storage tube (t0), and after 20 h at 15 °C and 300 rpm (t1). Surface adsorption was explored by removing the solution and rinsing the tubes with 4 µL PBS buffer (PBS) and 8 µL Laemmli SDS buffer (SDS). Samples for SDS-PAGE analysis were prepared by combining 4 µL of the sample with 4 µL of 2× Laemmli and 6% β-mercaptoethanol. Samples obtained through the process of rinsing with Laemmli SDS buffer were collected without additional processing. Sample incubation and SDS-PAGE analysis were carried out as described before.

To separate the fusion partners from the target enzyme Loxin, the fusion protein contains a cleavage site for activated factor X (FXa). Hence, the fusion protein was cleaved using FXa from bovine plasma (F9302, Sigma Aldrich, United States) in a 1:20 ratio (0.25 µg FXa +5 µg fusion protein) in FXa cleavage buffer (10 mM TRIS and 2 mM CaCl₂; pH 8.0) at 15 °C and 300 rpm (Thermoblock, Eppendorf). To compare cleavage success, Loxin fusion protein with and without supplemented Tween 20 (0.8 mM) was incubated for 6 and 14 days, respectively. Evaluation of the cleavage was conducted using SDS-PAGE as described before.

Following the successful cleavage of the fusion protein, the activity of the target enzyme Loxin was analyzed as part of the development process. For this, refolded Loxin fusion protein from each of the three explored dialysis approaches (see Section 2.6/Section 3.5) was analyzed. In brief, refolded Loxin fusion protein supplemented with 0.8 mM Tween 20 was cleaved with 0.05 g FXa per g fusion protein for 48 h at 15 °C and 300 rpm. The protease activity was photometrically assessed using the non-specific Pierce Fluorescent Protease Assay Kit (Thermo Scientific, Germany), which utilizes FTC-casein as substrate. Loxin samples were analyzed without further processing, whereas 1 μg/mL trypsin and 4 μg/mL FXa (the corresponding concentration used for the digestion of the Loxin fusion protein) served as positive controls, and reaction buffer (25 mM Tris and 150 mM NaCl; pH 7.2) served as the negative control. After incubation for 1 h at room temperature, fluorescein excitation/emission at 485/538 nm with a gain of 50 was measured using the Synergy H4 Hybrid Microplate Reader (BioTek Instruments, United States).

In-gel trypsin digestion followed by tandem mass spectrometry analysis was carried out for protein confirmation. For this, an SDS-PAGE was performed, and gel spots containing Loxin were cut out and fully destained through several washing steps using 50% methanol and 50 mM ammonium bicarbonate (ABC). To reduce and alkylate cysteine residues, the gel pieces were incubated in 10 mM TCEP, 40 mM chloroacetamide, and 50 mM ABC for 45 min at 500 rpm in the absence of light. After a shrinking step for 15 min in 50% methanol and 50 mM ABC, the gel pieces were air-dried under a laminar flow hood for 30 min. The protein was digested using 100 ng of sequencing-grade modified trypsin (Promega) dissolved in 70 µL of 50 mM ABC for 12 h at 37 °C, and the resulting peptides were collected in a clean PCR plate. To increase recovery, the gel pieces were incubated with 50% methanol and 1.5% formic acid (FA), and the remaining peptides were centrifuged and combined with the first flow through. The peptides were dried in a Concentrator Plus (Eppendorf, Germany) device for 3 h at 45 °C (V-AQ Mode) and finally resolubilized in 2% acetonitrile (ACN) and 0.1% FA.

Nano liquid chromatography and mass spectrometry (MS) were carried out using a Thermo Fisher Orbitrap Eclipse equipped with an ultrahigh-performance liquid chromatography unit Dionex UltiMate 3000 (Thermo Fisher, Germany), a Nanospray Flex Ion Source (Thermo Fisher, Germany), a C18 precolumn (Thermo Scientific, Germany), and a 1.9-µm C18 column (Aeros 25 cm × 150 µm ID, ProtumLink, Germany). MS was carried out using a linear gradient from 2% ACN and 0.1% FA to 35% ACN and 0.1% FA for 40 min, followed by a gradient from 35% to 95% ACN in 3 min. The analytical column was washed with 90% ACN and 0.1% FA for 5 min and then equilibrated with 2% ACN and 0.1% FA for at least 5 min. The flow rates were set as 500 nL/min (first 45 min) and 800 nL/min (last 15 min). MS data were recorded through data-dependent acquisition. The full MS scan range was 375–1500 m/z with a resolution of 75,000, an automatic gain control (AGC) value of 3E6, and a maximal ion injection duration of 148 ms. MS1 data were acquired in profile mode. The 20 most abundant precursors (charge 2–7) were selected for MS/MS and detected using the ion trap (Rapid), with an isolation window of 1.6 m/z, an automatic gain control value set to 25,000, and a maximal ion injection duration of 35 ms.

FragPipe v23.0 (Kong et al., 2017; Yu et al., 2021) was used to perform the proteome search. N-terminal acetylation (+42.01) and oxidation of methionine (+15.99) were selected as variable modifications, and carbamidomethylation (+57.02) on cysteines was selected as a fixed modification. The list of common contaminants was supplemented manually with the sequence of the protein of interest used to identify peptides and proteins with a false discovery rate (FDR) less than 1%. The workflow “LFQ-MBR” was selected for the search, with all settings kept at default.

The expression in E. coli is influenced by multiple parameters that can be adapted when optimizing an expression protocol. These factors include the selection of the expression vector, the expression cell line, and the process conditions, among others.

In this study, we selected the E. coli expression vector pET-32a (+) that utilizes the T7 lacO promoter system and contains an N-terminal 6x-His-Tag, thioredoxin A, and a cleavage site for FXa (Figure 1). The T7 promoter grants control over the expression of the fusion protein, while 6x-His-Tags are commonly used to facilitate purification via affinity chromatography. Furthermore, thioredoxin A helps increase the solubility of expressed proteins, and the FXa cleavage site allows the proteolytic removal of all fusion partners, yielding the target enzyme without leaving non-native residues at the expressed protein of interest (Francis and Page, 2010). The domain sequence for the selected spider venom enzymes was cloned into the multiple cloning sites of the pET-32a(+) vector by Gibson cloning and ligation. Sufficient plasmid amounts were achieved through transformation of chemically competent E. coli NEB® 5-alpha cells. This approach was successful for all expression plasmids, as evidenced by plasmid production and subsequent confirmation through Oxford Nanopore sequencing (Supplementary Figures S1, S2).

To ascertain the most efficacious expression system, six E. coli strains were transformed with the generated expression plasmids. As outlined in Table 2, the transformation was successful for all E. coli strains with Loxin. In contrast, only Ori 2 and ST7 were transformed using the Phoni plasmid. Following cultivation on agar plates, a single clone from each construct was used to inoculate liquid medium. This resulted in the exclusion of RG B for Loxin due to the absence of observable growth in this environment.

After successful transformation and cultivation in liquid media, the production of the fusion proteins was carried out at three different expression temperatures to identify the optimal production conditions. In particular, we compared fusion protein expression between 16 °C, 28 °C, and 37 °C. For this, each expression strain was cultivated at 37 °C until an ΔOD₆₀₀ of 0.8–1.0, then induced with IPTG, and transferred to the corresponding expression temperature, and the presence of fusion proteins was validated through SDS-PAGE and Western blot analysis. Table 3 outlines the success of fusion protein expression at the tested temperatures across available expression strains for both constructs. Loxin fusion protein was successfully produced at all temperatures using all E. coli strains except Ori 2. Although RG 2 could be cultivated for the purpose of cryo-conservation stocks, its growth was severely reduced, which impeded induction. In contrast to the high transformation efficiency of Loxin, only Ori 2, BL21, and ST7 were successfully transformed using the Phoni plasmid, and protein production was detected in BL21 and ST7 at all temperatures.

Based on the previous results concerning strain selection and suitable expression temperatures, we opted to henceforth focus on exploring avenues of production for Loxin as this protein appeared more promising in the experiments conducted to date. Therefore, following the successful production of the Loxin fusion protein, we proceeded with the development of subsequent purification steps.

SDS-PAGE analysis revealed that the fusion protein was produced as inclusion bodies under all tested E. coli strains and conditions (Supplementary Figure S3). Consequently, the optimal production conditions for Loxin were determined as expression at 28 °C over a duration of 5–6 h using BL21 due to the highest fusion protein yield (Table 3, Supplementary Figure S4). After cell lysis, the fusion protein was dissolved and purified using IMAC under denaturing conditions (Figure 2). The flow-through and collected IMAC fractions were analyzed by SDS-PAGE, which revealed the presence of the Loxin fusion protein in the flow-through and the first elution peak. Subsequent analysis indicated that the capacity of the IMAC resin had been exceeded, resulting in the presence of the fusion protein in the flow-through. The elution fractions 4–8 contained the Loxin fusion protein in an estimated concentration of approximately 0.3 g/L, resulting in a total yield of approximately 18.5 mg. As the obtained fusion protein required refolding, a range of dialysis approaches was subsequently explored.

Refolding of the Loxin fusion protein was necessary because it was expressed as inclusion bodies; however, a portfolio of available refolding approaches for spider venom enzymes is virtually unavailable. Accordingly, we compared three different dialysis approaches to identify the most suitable protocol for refolding the Loxin fusion protein.

First, considering its theoretical isoelectric point (pI) of 7.7 and the pI of the natural enzyme (9.0), a Tris–NaCl dialysis buffer at different pH levels (7.5–9.0) and an initial protein concentration of 0.1 g/L were used. This revealed high precipitation at pH 7.5, as indicated by the absence of soluble fusion protein (Supplementary Figure S5A), while less precipitation was observed at pH 9.0 (Supplementary Figure S5B). The utilization of HEPES dialysis buffer and MES dialysis buffer at pH levels 6.8 and 6.5, respectively, also resulted in elevated precipitations, leading to no soluble fusion protein (Supplementary Figure S5C). Accordingly, both buffers were excluded from subsequent trials. Consequently, Tris–NaCl dialysis buffer at pH 7.0 or 8.5, with inital protein concentrations ranging from 0.1 g/L to 0.3 g/L revealed similar results (Figure 3A). The highest solubilized protein concentration was achieved using an initial concentration of 0.3 g/L fusion protein at pH 8.5.

To minimize precipitation, two additional approaches were explored. First, dialysis was carried out using a buffer containing L-arginine, which was subsequently exchanged for a second buffer containing zinc sulfate, followed by a buffer without supplements at pH 8.5. These approaches yielded no visible precipitation after the first dialysis step and only minuscule precipitation after the second and third steps. Figure 3B illustrates refolded Loxin fusion protein after the second and third dialysis steps, revealing increased refolding yields of estimated ∼0.2 g/L protein.

Following refolding, surface adsorption of the Loxin fusion protein was detected. This was evidenced by protein loss during storage, which may be attributed to the adsorption of the protein on the surface of the storage tube. To confirm this hypothesis, the storage tube was rinsed using PBS and Laemmli SDS buffer, and the samples were analyzed using SDS-PAGE. By rinsing with PBS, remaining protein traces were removed from the tube, while the SDS in Laemmli SDS buffer causes proteins to detach from surfaces by denaturation (Winogradoff et al., 2020). As protein bands could be observed after rinsing the tube with Laemmli SDS buffer, surface adsorption was verified. Hence, low-binding tubes and glass vials were evaluated for their potential to prevent protein adsorption (Figure 4A). The utilization of low-binding tubes revealed protein bands after rinsing with Laemmli SDS buffer (Figure 4A, lane SDS), while no soluble fusion protein could be detected in the glass vial (Figure 4A, lane t1 and SDS). Consequently, two detergents typically applied for membrane proteins—0.8 mM Tween 20 and 0.5 mM Triton X-100—were employed, and the surface attachment was examined. As illustrated in Figure 4B, the two detergents result in the formation of the solubilized fusion protein. This is evidenced by the absence of a protein band following the rinsing of the low-binding tube with PBS and Laemmli SDS buffer (Figure 4B, lanes PBS and SDS). As Tween 20 is known to increase enzyme stability and activity, it was selected for further experiments (Duskey et al., 2021).

To separate the fusion partner thioredoxin A, including His-Tag and the FXa cleavage site, from the target enzyme Loxin, digestion with FXa is necessary. Figure 5 displays the cleavage of the Loxin fusion protein with and without Tween 20. The comparison reveals complete cleavage after 4 days of incubation with supplemented Tween 20, while in the absence of Tween 20, even after 14 days, no complete digestion was observed. Furthermore, SDS-PAGE analysis showed an additional protein band at ∼15 kDa, which is an unknown digestion byproduct possibly containing cleaved thioredoxin or Loxin (black arrow, Figure 5). This highlights the importance of preventing protein adsorption to surfaces as it appears to reduce the cleavage efficiency of FXa. Prior to the purification, the protease activity was photometrically assessed using the digestion products from each of the different dialysis approaches. In all cases, we detected activity ranging from 64% to 92% of the Loxin digestive approaches compared to the 1 μg/mL trypsin control, whereas the FXa control revealed 20% activity (Supplementary Figure S6). Thus, we concluded an activity of the Loxin enzyme.

The purification of the target enzyme Loxin was investigated by comparing two distinct chromatography approaches to establish the most effective protocol for future studies.

Initially, non-absorptive IMAC was conducted. As exclusively the fusion partner contains a His-Tag, the target enzyme should elute with the flow-through (lane F, Figure 6A). As demonstrated in Figure 6, the flow-through contains Loxin (blue arrow). However, subsequent SDS-PAGE analysis of elution fractions also reveals protein bands corresponding to thioredoxin A (green arrow) and an unknown digestion byproduct (black arrow). Following three washing steps (lanes W1–W3, Figure 6A), elution with imidazole further revealed that the target enzyme had adsorbed to the resin (lanes E1 and E2, Figure 6A).

As non-absorptive IMAC proved unsuccessful, CIEX was examined next. It is anticipated that this method will result in the charge-mediated binding of Loxin to the resin at different pH levels. CIEX was conducted at pH 8.0 and 6.0, which should result in a positively charged Loxin due to its pI of 9.0 while causing anionic charges on thioredoxin A (pI 4.7) and fusion protein (pI 7.7). The application of CIEX at pH 8.0 proved to be unsuccessful as no protein could be detected through SDS-PAGE analysis (Figure 6B, blue box). Conversely, the adjustment to pH 6.0 resulted in the elution of Loxin and an unknown digestion byproduct (Figure 6C). Consequently, another CIEX was conducted using a fast protein liquid chromatography system equipped with a HiScreen Capto S column at pH 6.0 and a linear elution gradient (Figure 7A). The analysis of CIEX fractions by SDS-PAGE (Figure 7B) demonstrated an elution of Loxin between 35% (=350 mM NaCl) and 60% (=600 mM NaCl) elution buffer. A total of 1.5 mg of fusion protein was prepared for this CIEX, of which approximately 0.3–0.5 mg of pure Loxin was recovered. This result is consistent with our calculations, which predict that 1.5 mg of fusion protein should yield approximately 0.9 mg of Loxin, of which a portion was observed to precipitate during the FXa cleavage. Finally, we used mass spectrometry to verify the identity of the purified protein. The identity of Loxin was confirmed by MS based in a bottom-up proteomics experiment, achieving 76% sequence coverage (Supplementary File S1).

Spider venoms constitute a vast and largely untapped resource of novel enzymes with the potential to address fundamental biological questions and provide value for various biotechnological purposes. Enzymes already find application in pharmaceutical, analytical, and nutrition industries. Due to their high selectivity, low by-product formation, and eco-friendly characteristics, their utilization has rapidly increased (Singh et al., 2016; Industrial Enzymes Market, 2025). Venom enzymes, from spiders and other organisms, are active under challenging environmental conditions and often catalyze specific reactions. They, therefore, provide desired properties for industrial development (Lorenz and Eck, 2005; Industrial Enzymes Market, 2025). However, the diversity and biological significance of spider venom enzymes have only recently been revealed. Consequently, their characterization, which features a baseline to understand their biological significance and explore their translational value, remains pending.

Due to limited venom yields, recombinant expression emerged as the primary means to access these enzymes at the laboratory scale. However, compared to other venom components, the expression of spider venom enzymes remains largely underexplored. Particularly for unstudied biomolecules, the exploration of multiple factors needs to be considered during the development process toward a methodology for expression and screening. The features of the expression vector, such as the origin of replication and promoter, and protein properties, including the size and complexity, directly influence the transcription and translation of the target protein. Moreover, the choice of fusion partners has been demonstrated to have major effects on expression success and protein solubility but will, in turn, also modify the characteristics of the expressed protein (Klint et al., 2013; Nozach et al., 2013; Turchetto et al., 2017). Hence, in this study, we focused on the exploration and optimization of a recombinant expression workflow for spider venom enzymes using bacterial systems. In our study, the expression of an astacin-like metalloprotease from L. intermedia venom and a serine peptidase from P. nigriventer venom were explored as a case study. The major goal of this study was to evaluate various technical aspects of spider venom enzyme expression prior to functional characterization, which is an important but often neglected facet of venom biotechnology. As it stands, our data may be valuable to colleagues interested in characterizing spider venom enzymes and will contribute to a more efficient research environment. By investigating factors influencing the expression success, such as strain selection and expression conditions, this study may provide valuable alternatives to facilitate production, particularly when no target protein expression is initially observed. In particular, during our work, we encountered several pitfalls and issues on various process steps, which are meticulously documented and outlined hereafter.