Aspergillus niger Ochratoxinase Is a Highly Specific, Metal-Dependent Amidohydrolase Suitable for OTA Biodetoxification in Food and Feed

Microbial enzymes can be used as processing aids or additives in food and feed industries. Enzymatic detoxification of ochratoxin A (OTA) is a promising method to reduce OTA content. Here, we characterize the full-length enzyme ochratoxinase (AnOTA), an amidohydrolase from Aspergillus niger. AnOTA hydrolyzes OTA and ochratoxin B (OTB) mycotoxins efficiently and also other substrates containing phenylalanine, alanine, or leucine residues at their C-terminal position, revealing a narrow specificity profile. AnOTA lacks endopeptidase or aminoacylase activities. The structural basis of the molecular recognition by AnOTA of OTA, OTB, and a wide array of model substrates has been investigated by molecular docking simulation. AnOTA shows maximal hydrolytic activity at neutral pH and high temperature (65 °C) and retained high activity after prolonged incubation at 45 °C. The reduction of OTA levels in food products by AnOTA has been investigated using several commercial plant-based beverages. The results showed complete degradation of OTA with no detectable modification of beverage proteins. Therefore, the addition of AnOTA seems to be a useful procedure to eliminate OTA in plant-based beverages. Moreover, computational predictions of in vivo characteristics indicated that AnOTA is neither an allergenic nor antigenic protein. All characteristics found for AnOTA supported the suitability of its use for OTA detoxification in food and feed.


■ INTRODUCTION
Mycotoxins are toxic chemical compounds produced by molds that cause huge economic losses to global agriculture.Ochratoxin A (OTA) is an important mycotoxin mostly found in dried fruits, coffee and cocoa beans, cereals, and spice. 1 To control OTA contamination, several countries have established regulations for OTA levels in a variety of food products. 2Strategies for detoxifying OTA generally include physical, chemical, and biological methods, including enzymatic treatment. 3Enzymes are included in EC regulation 386/ 2009 which establishes a new functional group of feed additives for the reduction of mycotoxin contamination, being substances that can suppress or reduce the absorption, promote the excretion of mycotoxins, or modify their mode of action. 4Enzymatic transformation of mycotoxins presents the advantages of high efficiency, strong specificity, and lack of damage to nutrients. 5icrobial enzymes are known to play a crucial role in numerous industries and applications.The demand for industrial enzymes is increasing due to the growing need for sustainable solutions.Food enzymes are mainly used to improve food taste, texture, digestibility, and nutritional value, being considered as processing aids.In general, processing aids are used during the food manufacturing process and do not have a technological function demand in the final food.Food enzymes are often not sold as pure enzymes but as enzyme preparations containing not only the desired enzyme but also various added substances.All these components are expected to be safe according to good manufacturing practice guidelines. 6n this regard, Aspergillus is a ubiquitous fungus found in nature.Species from this genus are widespread, being frequently isolated from soil, plant debris, and indoor air environments.Due to their prevalence in the natural environment, their ease of cultivation on laboratory media, and the economic importance of several of its species, this organism is among the most successful groups of molds involved in many industrial processes. 7spergillus niger is generally regarded as a safe organism and a nonpathogenic fungus widely distributed in nature. 8A. niger is one of the most important microorganisms used in biotechnology, having been used for many decades to produce food enzymes.Since the 1960s, A. niger has become a source of a variety of enzymes that are well established as technical aids in fruit processing, baking, and in the starch and food industries. 7n A. niger enzyme, defined as amidohydrolase 2 (here, AnOTA), has been described to transform OTA through the hydrolysis of the amide bond to generate OTα and L-β- phenylalanine (Figure S1), virtually nontoxic compounds.9 When the enzyme was homologously produced in A. niger, the recombinant amidase was N-terminally truncated, lacking the first 42 amino acid residues.This truncated protein was partially characterized and its three-dimensional structure was resolved.9 This structure reveals that AnOTA is a homooctamer with D4 dihedral symmetry, in which the subunits fold into a two-domain structure characteristic of metal-dependent amidohydrolases.9 Since knowledge of the substrate specificity profile is of utmost relevance when an enzyme is to be used in detoxification processes involving complex, protein-rich environments, such as food, feed, and gut, here, we aim to analyze in detail the specificity profile of full-length AnOTA, extending the analyses carried out previously demonstrating a preference for phenylalanine at the C-terminal end of the scissile amide bond.9 Therefore, in this work, a biochemical characterization including the substrate specificity of the fulllength AnOTA, heterologously produced in Escherichia coli, was determined by using synthetic substrates.We have also analyzed the structural bases of this substrate specificity by molecular docking simulations.Moreover, the use of AnOTA for OTA detoxification in plant-based beverages was assayed.Finally, the prediction of some in vivo AnOTA characteristics was also included in the study.All these results contribute to elucidating the suitability of AnOTA for its use as food and feed additives.

■ MATERIALS AND METHODS
Strains and Growth Conditions.The A. niger CBS 513.88 strain used in this study was purchased from the Westerdijk Fungal Biodiversity Institute (The Netherlands) and was routinely grown in the media and conditions recommended.
E. coli DH10B was used for DNA manipulations.E. coli BL21(DE3) was used for protein expression in the pURI3-Cter vector. 10E. coli strains were cultured in Luria−Bertani (LB) medium at 37 °C and shaken.When required, ampicillin was added to the medium at 100 μg/mL.
Production and Purification of Ochratoxinase from A. niger CBS 513.88 (AnOTA).The full-length gene encoding AnOTA (amidase 2) from A. niger CBS 513.88 (AM270317.1,contig An14c0100, 6231−7673 positions) was amplified by PCR using oligonucleotides Fw (forward) (5′-AACTTTAAGAAGGAGATATA-CATatggtccgccgaattgct tcagctac) and Rv (reverse) (5′-GCTATTAAT-GATGATGATGATGATGcagaaaaggattacgtgcat cttc) (the nucleotides pairing the sequence of the expression vector are designated in italics, and the nucleotides pairing the sequence of the gene are designated in lowercase letters).The 1.4-kb amplified PCR product was purified and inserted into the pURI3-Cter vector by a restriction enzyme-and ligation-free cloning method. 10This vector delivers recombinant proteins displaying a six-histidine affinity tag in their C-terminal end.E. coli DH10B cells were transformed and the recombinant plasmids were purified.Next, they were transformed into E. coli BL21(DE3) cells for protein expression.Since the obtained protein yield in the expression assays was low, a pMA plasmid derivative containing the A. niger gene with the codon usage adapted to the codon bias of E. coli was synthesized by Invitrogen (Thermo Fisher Scientific).The AnOTA encoding gene with the optimized E. coli codon usage was amplified by oligonucleotides Fw and Rv from the pMA plasmid.The PCR fragment was then inserted into the pURI3-Cter vector and transformed into E. coli DH10B cells for cloning and, later, into E. coli BL21 (DE3) for protein expression.
E. coli BL21(DE3) cells harboring the recombinant vector pURI3-Cter-AnOTA were grown at 22 °C in the LB medium containing 1 M D-sorbitol, 25 mM glycine betaine, and 0.25 mM isopropyl-β-Dthiogalactopyranoside to avoid inclusion body formation. 11The cells were disrupted by French press lysis; afterward, the insoluble fraction of the lysate was separated by centrifugation at 47,000g for 30 min at 4 °C.The amidase AnOTA was purified by batch affinity chromatography using TALON Superflow resin (Clontech) and eluted using 50 mM MOPS, pH 7.0, supplemented with 150 mM NaCl and 150 mM imidazole.The eluted His-tagged AnOTA was dialyzed overnight at 4 °C against 50 mM MOPS buffer, pH 7.0, containing 150 mM NaCl.Enzyme purity was verified by 12.5% SDS-PAGE in Tris-glycine buffer.
Effect of Temperature and pH on AnOTA Amidohydrolase Activity.The biochemical characterization of the full-length AnOTA was performed by using the OTA analogue N-(4-methoxyphenylazoformyl)-phenylalanine (4MF) (Bachem, Switzerland) as a substrate. 9his activity assay consists of the measurement of loss of absorbance at a wavelength of 350 nm, caused by the hydrolysis of 4MF.A 4MF stock solution (55 mM) was prepared in DMSO.To examine the influence of temperature, reactions were performed with 0.1 mM 4MF in 50 mM buffer MOPS (pH 7.0) containing 20 mM NaCl at 5, 20, 30, 37, 40, 45, 55, 65, 75, and 85 °C, using 100 ng of the enzyme.The reactions were incubated for 15 min, after which the absorbance was measured.The influence of pH was investigated by assaying amidase activity in a set of buffers with pH values from 3.0 to 9.0.The buffers (50 mM) utilized were sodium citrate buffer (pH 3.0−6.0)and Bis-Tris propane buffer (pH 6.5−9.0).Reactions containing 100 ng of the enzyme and 0.1 mM 4MF were incubated at 37 °C for 15 min, after which the absorbance was measured.For the measurement of the thermal stability, AnOTA was incubated in 50 mM buffer MOPS (pH 7.0) containing 20 mM NaCl at 20, 30, 37, 45, 55, and 65 °C for 15 and 30 min and 1, 2, 4, 6, and 20 h.After incubation, the residual activity was measured.Reaction mixtures with no added enzyme were used as negative controls.All reactions were carried out in triplicates.
Determination of the Influence of Different Metals on AnOTA Activity.The effects of the presence of different metal ions were analyzed in untreated, native (AnOTA − ), and EDTA-treated AnOTA (AnOTA + ).The chelating agent EDTA was used in an attempt to obtain the metal-free enzyme.Purified, full-length AnOTA was treated with 20 mM EDTA at room temperature for 2 h to remove metal ions.Subsequently, the chelating agent was removed by extensive, overnight dialysis against 50 mM buffer MOPS (pH 7.0) containing 20 mM NaCl at 4 °C with three changes of buffer. 12econstitution experiments were performed by adding the metal salt (1 mM) to AnOTA + and preincubating it at room temperature for 15 min, before the activity assay.Then, 4MF was added to a final concentration of 0.1 mM.The reactions were incubated for 15 min at 37 °C, and the absorbance was then measured.The residual amidase activity was measured after the incubation of AnOTA with each metal.The compounds assayed were MgCl 2 , ZnCl 2 , NiCl 2 , CoCl 2 , and MnCl 2 .In all assays, the amidase activity measured in the absence of any additive was taken as a control and was given a value of 100%.The experiments were done in triplicate.
Ochratoxin-Detoxification Assays.Enzymatic transformation of OTA and ochratoxin B (OTB) by AnOTA was confirmed by HPLC.Stock solutions (1 mg/mL) were prepared dissolving OTA and OTB in methanol and stored at −20 °C.Reactions were performed in phosphate buffer 50 mM (pH 7.0), using 5 μg of the enzyme and adding OTA or OTB to a 5 μM final concentration.Reaction mixtures were incubated at 37 °C overnight and stopped by heating for 5 min at 95 °C.Afterward, samples were centrifuged at 14,000g for 5 min.The supernatants were filtered through 0.45 μm syringe filters (Millipore, USA) and analyzed by HPLC as previously described. 13etermination of AnOTA Substrate Specificity.Endopeptidase activity was evaluated by incubating 1 μg of AnOTA with 15 μM bovine serum albumin (BSA) in 50 mM MOPS buffer containing 20 mM NaCl (pH 7.0) at 37 °C overnight.A reaction mixture with no added AnOTA was employed as a negative control.BSA degradation was examined by SDS-PAGE analysis of the reaction mixture after incubation.Coomassie brilliant blue R250 stain was used for gel staining.
OTA Detoxification in Plant-Based Beverages by AnOTA.Three commercial UHT plant-based beverages (almond, oat, and soy) were purchased from a local supermarket (Hacendado, Mercadona, Spain).The composition and nutritional values of the beverages are stated on their labels.Almond beverage contained 3% almond and 0.5% protein; oat beverage contained 8% oat and 0.7% protein, and soy beverage contained 13% soybean and 3.1% protein.
All samples contained other secondary ingredients such as stabilizers, thickeners, aromas, salt, and sugars.The protein content of the plantfood beverages was spectrophotometrically determined at 260 nm in a NanoDrop spectrophotometer.
The OTA detoxification assay of the three plant-based beverages was performed as described previously.Reactions were performed in the beverages by adding OTA at 5 μM final concentration and using 5 μg of the AnOTA enzyme.Reactions were incubated at 20 and 37 °C for 4 h.After incubation, OTA was extracted by using an OchraTest WB (VICAM, USA) immunoaffinity column, following the protocol described for cereal samples.Briefly, 2 mL of the beverage sample was diluted with 8 mL of methanol.The solution was mixed under stirring for 15 min.Next, it was passed through a filter paper.The filtered extract was then diluted with 40 mL PBS (phosphate buffered saline) and mixed well.The extract was filtered through a 1.5 μm glass microfiber filter (VICAM, USA).A 10 mL volume of diluted extract was applied to an OchraTest WB immunoaffinity column at a flow rate of about 1 drop/second.The column was washed with 10 mL of PBS followed by 10 mL of distilled water.OTA or its degradation product, OTα, was eluted with 1.5 mL of methanol and collected in a clean vial.The eluted extract was evaporated overnight at 60 °C and reconstituted with 500 μL of the HPLC mobile phase.
Protein profiles of plant-based beverages were examined by SDS-PAGE analysis of the reaction mixtures after incubation.Coomassie brilliant blue R250 stain was used for gel staining.To determine the free-amino acid composition of the plant-based beverages treated with AnOTA (15 μg of enzyme, 4 h at 37 °C) and nontreated, in the first place, after the incubation period, samples were freeze-dried using a Beta 2−8 LDplus (Christ, Germany) lyophilizer.Subsequently, freeamino acids were extracted by shaking 0.3 g of the freeze-dried sample with 2 mL of HCl 0.1 M during 1 h.The samples were then centrifuged at 12,000g for 10 min at 4 °C and the supernatant was filtered through a 0.45 μm filter (Millipore, USA).The free-amino acid composition was evaluated by the use of a Biochrom 30 amino acid analyzer.The Biochrom 30 uses the classical amino acid analysis methodology based on ion-exchange liquid chromatography and postcolumn continuous reaction with ninhydrin to provide qualitative and quantitative compositional analysis.The amino acid composition analysis was carried out in triplicate at the Protein Chemistry facility of the Center for Biological Research (CIB-CSIC).
Molecular Docking Simulations.Molecular docking was performed using AutoDock Vina 16 and UCSF Chimera 1.17 17 as the interface for the preparation of pdbqt files.The crystal structure of the truncated AnOTA determined at 2.5 Å resolution (PDB code: 4C5Z) was used as the receptor molecule.Its preparation for docking experiments was as reported before 13 and involved the removal of waters, addition of polar hydrogens, and merging of charges.The atomic coordinates were retrieved from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/) for the substrates ZF (Pub-Chem CID: 70878), ZI (PubChem CID: 2724772), ZA (PubChem CID: 736104), HF (PubChem CID: 6994977), and HR (PubChem CID: 96815).The rest of the structures, namely, ZAF, ZAL, ZAA, ZIF, and ZFI, were built as described previously. 13The structures of the substrates were energy minimized with USCF Chimera (default values) before the docking experiments.A search exhaustiveness of 32 and a maximum energy difference of 3 kcal/mol were used.The number of binding modes was 10.USCF Chimera 17 and PyMOL 18 were used for the visualization of the structures, analysis of the interactions, and figure preparation.

Biochemical Characterization of Full-Length A. niger Amidase 2 Heterologously Produced in E. coli (AnOTA).
The first microbial enzyme reported to effectively hydrolyze OTA was an amidase from A. niger referred to as amidohydrolase 2 or amidase 2 to differentiate it from all other A. niger amidases reported at that time to which it showed no sequence homology. 9This amidase 2 was identified as the OTA-hydrolyzing component of the "Amano lipase" product and subsequently annotated as the first microbial ochratoxinase (OTAse).The 480 amino acid residue amidase 2 from A. niger UVK143 (AnOTA) (100% identical to A. niger CBS 513.88 amidase 2) was homologously produced in the A. niger AP4 strain. 9When the recombinant AnOTA was purified from the fermentation broth, peptide mass fingerprinting revealed that the extracellular amidase was N-terminally Journal of Agricultural and Food Chemistry truncated, possessing only 438 amino acid residues, with the first identified residue being Ser-43.Moreover, the authors indicated that this truncation was also observed in purified intracellular A. niger amidase. 9Additionally, it was described later that the full-length 480 amino acid AnOTA (referred to as "amidase 2 sig") also hydrolyzed OTA, similar to the secreted mature 438 amino acid amidase 2 (referred to as "amidase 2 mat"). 24It was also observed that the loss of the N-terminal segment containing 42 amino acids of AnOTA resulted in sensitivity to Ca 2+ inhibition.These results suggested that although truncated AnOTA could be used to detoxify OTA, it shows certain limitations due to its sensitivity to divalent metal ions and chelators, widely present in feed additives.In contrast, full-length AnOTA was described to be much less sensitive to metal ions and chelators, revealing the functional relevance of the truncation of the N-terminal segment. 24onsidering these results, and that truncated AnOTA has been partially characterized before, we decided to heterologously produce full-length AnOTA and perform its enzymatic characterization.To characterize the enzyme, its encoding gene was cloned into the pURI3-Cter expression vector by a ligation-free cloning strategy described previously. 10The decision to include the His-tag at the Cterminal end of the protein is twofold: first, the structure of the homo-octameric AnOTA currently available corresponded to the truncated AnOTA variant and therefore no structural information from the N-terminal end is known, and second, the presence of the His-tag at the C-terminal end is a priori compatible with the integrity of the oligomer since the tag would protrude toward the solvent.
Hence, the 1.4-kb AnOTA gene, PCR amplified with oligonucleotides Fw and Rv, was expressed in E. coli, and the hyperproduced, C-terminally His-tagged recombinant protein was purified by IMAC, as described in Materials and Methods.In these first approaches for AnOTA production, we observed a very poor production yield, which agrees well with previous attempts by others to increase the yield by coexpression of the enzyme gene with those coding for molecular chaperones. 25ere, we obtained a notable increase in the production yield of the full-length AnOTA following another approach, namely, E. coli codon optimization.This procedure rendered a production yield of 3.72 mg of recombinant protein per liter of culture (Figure S2).
The dependence on pH and temperature of OTA degradation was previously assayed in recombinant A. niger amidase 2. The enzyme was active within the pH range between 3.0 and 9.0, with an optimum pH of 7.0. 24This same study revealed that recombinant AnOTA was heat stable since more than 80% of its activity remained after preincubation for 5−30 min at 80 °C. 24Additionally, the effect of pH and temperature on truncated AnOTA has been also described, using the OTA analogue 4MF as a substrate. 9Truncated AnOTA hydrolyzes 4MF optimally at pH 5.6−6.0, and at pH 9.0, approximately 50% of the maximum activity is retained.The optimal temperature for 4MF hydrolysis was determined to be approximately 66 °C.Our results with the full-length AnOTA, using 4MF as a substrate, revealed pH and temperature dependences similar to those from the truncated protein: 9 the optimal pH determined here was 6.0−6.5, exhibiting 40% of its maximal activity at pH 9.0 (Figure 1A), and maximal activity at 65 °C, retaining 60% of maximal activity at 75 °C (Figure 1B).However, AnOTA is fully inactivated after 15 min of incubation at 85 °C, in agreement with the thermostability results described for the truncated AnOTA. 9Thermostability assays performed in this study with AnOTA revealed that after prolonged incubation (20 h) at 45 °C, the enzyme retained more than 60% of its initial activity and more than 70% at temperatures lower than 45 °C (Figure 1C).Therefore, using 4MF as a substrate, the pH and temperature effects on AnOTA activity obtained in this study are almost identical to those reported previously for the Nterminal truncated protein.The high optimal temperature shown by AnOTA, together with its thermostability and the fact that the truncated enzyme has been described to be stable for at least two years without loss of activity, 9 represents a great advantage for the potential application of AnOTA in food and feed processing.
The crystal structure of AnOTA 9 demonstrated that it belongs to the superfamily of amidohydrolases, in particular, to subtype I of such enzymes, following a classification that considers seven subtypes according to the structural features of their metal(s)-binding centers. 26In particular, members from the most common subtype I have two metal ions (M α and M β , respectively) that are coordinated by the side chains of six residues, namely, M α is coordinated by two His residues (located in an HXH sequence motif) and an Asp, and M β is coordinated by two His residues.In turn, the two metals are bridged by a carboxylated lysine residue, that originated from a post-translational modification.In the members of subtype II, this latter carboxylated lysine is replaced by a glutamate residue, conserving the other protein ligands.Specifically, within the active site of AnOTA, there are two Zn 2+ ions, which are coordinated by His111, His113, and Asp378 (M α site) and His287 and His307 (M β site), respectively.The carboxylated Lys246 residue acts as a bridge between them.These same structural features are also observed in the recently reported cryo-EM structure of the homologue enzyme ADH3 from Stenotrophomonas acidaminiphila 27 (SaOTA).The participation of the metal-binding center in the catalysis is evidenced in AnOTA by its complete inactivation with the Zn 2+ -specific chelator 1,10-phenanthroline 9 and in ADH3 by the complete lack of activity of single-point mutants of residues that participate in the coordination of the metals; this latter analysis concludes that both metals, M α and M β , are essential for catalysis. 27ithin the metalloprotease realm, this type of a bimetallic Zn 2+ site is defined as cocatalytic, distinct from other sites containing only one Zn 2+ atom known either as catalytic or structural. 28,29In agreement with the crystal structure of AnOTA, 9 the recombinant, full-length form of the enzyme here studied has Zn 2+ atoms as deduced from our atomic absorbance analysis, which in turn discards the presence of Co 2+ that could be derived from the purification protocol.Although the most frequently found metal in cocatalytic sites in metalloproteases is precisely Zn 2+ , there are other cases in which Co 2+ , Mn 2+ , Fe 2+ , and Ni 2+ have also been identified. 29nder this scenario, we have studied the effect of several metals on AnOTA activity both with the EDTA-untreated (AnOTA − ) and the EDTA-treated (AnOTA + ) enzyme.A summary of the obtained results is shown in Figure 2. Regarding the control AnOTA, we noticed that even treatment with 20 mM EDTA, and subsequent dialysis, does not completely abolish its activity, which supports the idea that AnOTA exhibits a high affinity for Zn 2+ , in agreement with our atomic absorbance results.Interestingly, the addition of 1 mM Zn 2+ either to AnOTA − or AnOTA + results in a strong inhibition of its activity.−30 In this regard, previous studies on E. coli aminopeptidase P have revealed a third Zn-binding site, also within the active site of the enzyme, close to the other two catalytic Zn 2+ atoms, 31 whose occupancy causes metal inhibition.Similar results have been obtained for the Mn 2+ inhibition of the aminopeptidase P from Pseudomonas aeruginosa. 32e results obtained with 1 mM Co 2+ when added to AnOTA − can be easily explained in terms of a much less efficient inhibition than that exerted by Zn 2+ , most probably resulting from the low-affinity binding of Co 2+ to a putative third, inhibitory metal-binding site.The activity level observed when Co 2+ is added to AnOTA + would support the latter lowaffinity binding to the third metal-binding site, and importantly, also that Co 2+ would replace the removed Zn 2+ atom(s) from the cocatalytic site.Since, essentially, the same results are obtained with Ni 2+ and Mn 2+ , a similar hypothesis can be proposed for these metals.In contrast to Co 2+ , Ni 2+ , and Mn 2+ , the 100% activity level observed for either AnOTA − or AnOTA + when treated with iron would indicate that this metal binds to the cocatalytic site but not to the third inhibitory site, and therefore no inhibition would be exerted.
As we have indicated above, Yu et al. reported that AnOTA (or amidase 2 sig) exhibits lower Ca 2+ inhibition sensitivity as compared to the N-terminal truncated protein (amidase 2 mat). 24In turn, these authors also showed that the activity of amidase 2 mat was 40% inhibited by 8 mM CaCl 2 , while AnOTA was not appreciably inhibited neither by 20 mM EDTA nor 20 mM Ca 2+ . 24In agreement with these results, we observed that the activities of both AnOTA − or AnOTA + enzymes did not decrease but increased 14 and 34%, respectively, upon preincubation with 8 mM CaCl 2 (Figure 3).Therefore, our results confirmed the low sensitivity of fulllength AnOTA and also its suitability for OTA detoxification in food or feed additives containing a chelator.Citric acid or other organic acids such as propionic acid are widely used as feed additives and can chelate divalent ions; 24 therefore, the use of an enzyme, such as the full-length AnOTA, less sensitive to divalent metal ions and to chelators will be strongly recommended.
Analysis of the Substrate Specificity of AnOTA.As indicated above, determination of the substrate profile of an enzyme to be used in detoxification processes in complex, protein-rich milieus is a must, since degradation of off-target compounds must be avoided.This is particularly a concern for the members of the amidohydrolase superfamily that despite sharing a highly conserved overall protein subunit architecture, they show disparate substrate specificities due to differences in structural elements decorating their substrate-binding sites.Hence, within the closest structural homologues of AnOTA (see Table 1 and Figure S3), the three enzymes isolated from an environmental sample from the Sargasso Sea, Sgx9359b, 33 Sgx9355e, 34 and Sgx9260c, 15 whose subunit structures superimpose almost perfectly to AnOTA, show markedly different substrate specificities, namely, Sgx9359b catalyzes the hydrolysis of L-Xxx-L-Arg dipeptides (and N-acetyl and N-formyl derivatives), Sgx9355e hydrolyzes dipeptides with a C-terminal hydrophobic amino acid (Ile, Leu, Phe, Tyr, Val, Met, and Trp), although also accepts a C-terminal Thr, and Sgx9260c hydrolyzes Gly-L-Pro and L-Ala-L-Pro (and N-acyl derivatives of L-Pro).
Cc2672 from Caulobacter crescentus CB15 hydrolyzes L-Xaa-L-Arg/Lys dipeptides, 33 and the amidohydrolase from Steno-trophomonas sp.CW117 is known to hydrolyze OTA.However, no further data have been reported about its substrate profile.Other structurally more distant homologues of AnOTA such as peptidases belonging to the M38 family of enzymes from the MEROPS peptidase database 35 show a wide spectrum of specificities, such as isoaspartyl dipeptidase, 36 dihydropyrimidinase, 37 guanine deaminase, 38 urease, 39 etc.
As far as we know, no results about the substrate profile of AnOTA have been reported apart from those described by Dobritzsch et al. that revealed hydrolysis of 4MF, L-Phe-L-Tyr, L-Arg-L-Phe, and hippuryl-L-Phe, 9 which supports a preference for an aromatic C-terminal residue in model compounds with one amide bond.To characterize the substrate profile of AnOTA in further detail, we have used a library of 17 Nbenzyloxycarbonyl (or carbobenzyloxy) amino acid derivatives, ten N-acetyl derivatives of L-amino acids, two N-benzoyl-glycyl amino acid derivatives (or hippuryl-amino acids), and two Ldipeptides (Figure S4).However, before proceeding with this set of potential substrates, we first confirmed the hydrolytic activity of AnOTA against OTA, and also OTB, and second, we also tested potential endopeptidase activity against BSA.As expected, we observed complete hydrolysis of OTA and OTB by full-length AnOTA after overnight incubation (Figure 4), but no endopeptidase activity, as deduced by the absence of fragmentation products in an SDS-PAGE gel (data not shown).
The results of the hydrolytic activity of AnOTA against the set of compounds mentioned above reveal that the dipeptide GF is the most efficiently hydrolyzed substrate.Also, only three N-benzyloxycarbonyl derivatives containing one amide bond out of 12 are hydrolyzed: ZF > ZL ∼ ZA (Figure 5), and only four N-benzyloxycarbonyl derivatives containing two amide bonds out of five are hydrolyzed, namely, ZFI > ZAF > ZAL > ZIF (Figure 5).Finally, regarding the hippuryl compounds, only HF is hydrolyzed.From these results, it can be concluded that (i) hydrolysis by AnOTA is exclusively observed against compounds containing a C-terminal Phe or aliphatic amino acid (Ala, Leu, and Ile); (ii) AnOTA hydrolyzes Nbenzyloxycarbonyl derivatives containing either one or two potentially scissile amide bonds; (iii) N-acetyl-L-amino acids are not hydrolyzed by AnOTA, irrespective of the amino acid residue; hence, most probably, AnOTA is not an N-acetyl amidohydrolase; and (iv) compounds with a C-terminal polar or charged side chain are not hydrolyzed.
Regarding the N-benzyloxycarbonyl derivatives containing two potentially scissile amide bonds, it is worth noting that the above results do not permit a response neither to the question of how many amide bonds are hydrolyzed nor to the order of hydrolysis in case both amide bonds are hydrolyzed.Although the absence of endopeptidase for AnOTA suggests that the unique scissile amide bond of these N-benzyloxycarbonyl derivatives is the C-terminal one, we tackle these questions from a structural perspective using molecular docking simulations following the same experimental approach as in our previous studies. 13,40Previous to these studies, and to validate our approach, we proceeded to dock the OTA molecule into the binding site of SaOTA and compared the results with the cryoEM structure of such a complex. 28As can be observed in Figure S5, the obtained results are in excellent agreement with the experimental structure (PDB entry: 8IHS), since the docked OTA molecule superimposes almost perfectly with the experimental one, revealing the same pattern of . 24The activity exhibited by native untreated AnOTA in the absence of additive was defined as 100%.interactions with the amino acid side chains of SaOTA and the Zn 2+ ions.
Molecular docking of OTA into the active site of AnOTA revealed a similar binding mode as in SaOTA (Figure 6A), which entails the following interactions: (i) the C-terminal, Phe aromatic ring of OTA is located in an aliphatic pocket formed by the side chains of Val253, Val332, Ile333, and Leu355; (ii) the negatively charged carboxy moiety would interact favorably with the nearby side chains of His191, His287, and His289; (iii) the OTA carbonyl oxygen of the scissile amide bond would point toward the Zn 2+ (M β site), which is consistent with the catalytic mechanism of hydrolysis of metallocarboxypeptidases; 13,41 and (iv) the N-terminal, 5chloro-8-hydroxy-3-methyl-1-oxo-7-isochromanyl ring of OTA (ISO) would establish aromatic stacking interactions with the imidazole side chain of His113.
Interestingly, docking of OTB, which is the nonchlorinated version of OTA, reveals that its N-terminal 8-hydroxy-3methyl-1-oxo-7-isochromanyl ring is rotated 180°when compared to OTA (Figure 6B), revealing a different binding mode of the N-terminal end.This unexpected result may provide the structural basis underlying the higher efficiency of hydrolysis of OTA vs OTB by AnOTA, as has been reported previously. 24he molecular docking results obtained with the dipeptide GF reveal that this substrate binds as the structurally equivalent C-terminal end of OTA (Figure 6C).Taking into account the structure of this substrate, this result suggests that the built-up of the interactions between the C-terminal end of the substrates and the aliphatic pocket of AnOTA is the main driving force for yielding a catalytically productive binding; in fact, the location of the C-terminal end of substrates in this pocket is a shared feature for all substrates studied here.Thus, as expected, the results obtained with the N-benzyloxycarbonyl-L-amino acid substrates ZA, ZL, and ZF (Figure 6F) reveal binding modes which are also consistent with the binding of OTA.Specifically, the C-terminal amino acid side chain is located in the aliphatic pocket with their N-terminal carbobenzyloxy moieties occupying equivalent positions as that of ISO.Importantly, in all cases, the scissile amide bond occupies the same position with the same orientation.Regarding the N-benzyloxycarbonyl-L-amino acid substrates containing two amide bonds ZAL, ZAF, and ZFI, the docking results indicate that the binding site of AnOTA accommodates these larger molecules (Figure 6I).Whereas the C-terminal amino acid side chain is housed in the aliphatic pocket, the Nterminal aromatic ring of the N-benzyloxycarbonyl moiety would establish favorable stacking interactions with the side chain of Tyr124; finally, the arrangement of the side chain of the intermediate residue of the substrate (Ala in ZAL or ZAF, and Phe in ZFI) would be constrained by the bulky side chain of Met254.Finally, similar conclusions can be deduced from the docking of the substrate HF (Figure 6J): its C-terminal end is situated within the aliphatic pocket, as expected.Interestingly, no hydrolytic activity is observed against the other hippuryl derivative studied here (hippuryl-arginine, HR) which bears a positively charged C-terminal amino acid side chain.As expected, no pose compatible with the catalytic mechanism of hydrolysis is obtained in the docking assays, indicating that the positively charged side chain cannot be stabilized within the aliphatic pocket.These in silico results provide the structural details revealing that AnOTA only hydrolyzes the C-terminal amide bond of the model substrates.In this regard, it has not escaped our notice that since hydrolysis of the C-terminal residue of ZAL, ZAF, or ZFI produces ZA and ZF, respectively, which in turn are substrates for AnOTA, they could be also hydrolyzed.Nevertheless, we consider that the contribution of this second hydrolysis reaction is negligible since ZA and ZF concentrations (resulting from the hydrolysis of ZAL, ZAF, and ZFI, respectively) are much lower that ZAL, ZAF, and ZFI in our experimental conditions.
Hence, our studies provide a plausible scenario about the structural bases of the mycotoxin-and substrate-binding modes supporting the notion that AnOTA is a carboxypeptidase that shows a marked preference for C-terminal Leu, Phe, or Ala residues.The AnOTA binding site is an open structure that permits the binding of large substrates such as ZAL, ZAF, ZFI and ZIF, which in turn explains previous results reporting the hydrolysis of the dipeptide N-benzyloxycarbonyl-glycinephenylalanine, tripeptide N-benzyloxycarbonyl-glycine-phenylalanine-phenylalanine, and pentapeptide N-benzyloxycarbonylglycine-leucine-glycine-phenylalanine. 25 This ability to accommodate a wide range of substrates, or promiscuity, is especially relevant in the case of amidohydrolases which can hydrolyze a wide range of substrates.Therefore, it can be suggested that in OTA-producing fungi, it is likely that the physiological substrate of AnOTA, and close homologues, is ochratoxins.In this sense, it has been described that AnOTA-like enzymes from OTA-producing fungi showed more favorable substratebinding pockets to accommodate OTA and OTB than similar enzymes from other nonproducer fungi. 42he OTA-hydrolyzing activity of AnOTA, and probably that of other homologues from OTA-producing fungi, together with its marked preference for C-terminal Phe, Ala, Leu, and Ile residues of AnOTA represent an advantage for the use of these enzymes in detoxification processes since the production of unwanted byproducts in protein-rich food substrates is minimized, preventing the occurrence of relevant changes in the organoleptic and nutritional characteristics of the food substrate.
Application of AnOTA in Plant-Based Beverages for OTA Detoxification.The emergence of plant-based beverages as a nondairy alternative offers a new option due to environmental and ethical concerns, as well as lactose intolerance and casein allergies.However, raw materials employed in the elaboration of these beverages are susceptible to mycotoxin contamination.For this reason, in this work, we carried out OTA-detoxification assays by AnOTA in oat, soy, and almond beverages.
Plant-based beverages contain proteins, lipids, dietary fiber, minerals, and vitamins.Enzymatic hydrolysis (such as amylase, protease, cellulase, and pectinase) is a specific and effective method to hydrolyze vegetable components to convert some of them into easily absorbed nutrients like glucose, maltose, amino acids, and soluble phenolics.As reported in their labels, the three beverages assayed here possessed a diverse biochemical composition.To know if these plant-based beverages possessed adequate properties for AnOTA activity, their pH was determined.All beverages assayed possessed a close to neutral pH, ranging from a pH of 7.16 in the soy beverage to a pH of 7.84 in the almond beverage.The optimal pH for AnOTA activity is 6.0−6.5 (Figure 1A), although almost 60% of its maximal activity was found at pH 8.0.Concerning temperature, plant-based beverages containing AnOTA were incubated at 20 and 37 °C.Incubation at a temperature close to room temperature (20 °C) was chosen as it is an economically affordable temperature since at this temperature, AnOTA could exert its activity during beverage storage.In spite that the optimal temperature for AnOTA activity is high (65 °C), at 20 and 37 °C, this enzyme exhibited 30 and 40% of its maximal activity (Figure 1B).Therefore, at the pH of the beverage assayed and the incubation temperatures used in the study, AnOTA should remain active (Figure 1A). Figure 7 shows that OTA was fully hydrolyzed in all beverages incubated during 4 h at 20 °C; identical results were obtained at 37 °C incubation (data not shown).These results indicate that AnOTA can be used efficiently to eliminate OTA from plant-food beverages, even at storage conditions.
Activity assays performed in this study revealed that AnOTA does not possess endopeptidase activity on BSA.To confirm the absence of endopeptidase activity in these assays, plantfood beverages were incubated for 18 h in the presence of AnOTA at 37 °C, since at this temperature, enzyme activity is higher than at 20 °C.After incubation, proteins in beverage samples were analyzed on SDS-PAGE gels.Previously, protein content was spectrophotometrically determined.The results indicated that soy beverage possessed the highest protein content (88.19 mg/mL), followed by almond (40.34 mg/mL) and oat (34.68 mg/mL) beverages.As shown in Figure S6, the protein profile of the AnOTA-treated beverages presented identical protein profiles to the nontreated beverages.Most of the protein bands from oat beverage correspond to proteins lower than 45 kDa; however, almond and soy beverages also possess proteins having high molecular size.In any case, AnOTA treatment did not produce an evident variation in the protein profile of the beverage proteins.
In addition to the study of the potential hydrolytic effect of AnOTA on the plant beverages by analyzing the pattern of protein bands in SDS-PAGE, the free amino acid composition of AnOTA-treated and untreated soy beverage was also studied (Figure 8).The obtained results indicate that the amino acid composition does not vary significantly upon addition of AnOTA, suggesting that the unwanted hydrolytic effect on the proteins from the beverage is negligible, both in the absence and in the presence of OTA.More rigorously, we used as an estimation of the overall uncertainty in the values of the relative amount of the amino acids the mean of the standard deviations (s.d.) of the data for the amino acids that cannot derive from the catalytic activity of AnOTA, namely, polar or charged residues (Asp, Glu, Lys, Arg, Ser, and His).The calculated overall s.d. is 0.9.In turn, the s.d.values for Phe and Tyr were 0.8 and 0.6, respectively, revealing that the variation of the values for these two latter amino acids fall within the instrumental uncertainty.Therefore, we conclude that the addition of AnOTA to plant-food beverages, with or without OTA, does not yield unwanted byproducts originated from the hydrolysis of plant proteins.These results support our claim that the OTA detoxification with AnOTA may be a promising  approach to eliminate this mycotoxin effectively, without modifying the beverage characteristics.
Computational Predictions of AnOTA In Vivo Characteristics.Like other proteins, once ingested, enzymes are generally easily broken down into their constituent amino acids that are indistinguishable from other food molecules.In general, ingestion of microbial enzymes is not likely to be of concern regarding food allergies. 43However, allergenicity, antigenicity, and immunogenicity of food proteins are crucial aspects associated with the widespread usage of new foods and additives.As OTA-degrading enzymes have a potential application in the food industry, the potential allergenicity and antigenicity of AnOTA were predicted in silico.In this regard, the prediction of allergenic proteins has become very relevant nowadays due to the use of enzymes in foods, therapeutic compounds, and biopharmaceuticals.Currently, allergen prediction methods are heavily used by the scientific community in designing proteins with low allergenicity.As a first approach to studying the allergenicity of AnOTA, the AllerTOP v.2 and AlgPred programs were used. 19,20The results obtained with both tools indicate that AnOTA is a nonallergenic protein.Here, it is important to remark that the criteria used for each program to define AnOTA as nonallergenic are different, and therefore, the overall results are not redundant.Thus, whereas AllerTOP v.2 predicts protein allergenicity based on known allergenic and nonallergenic protein sequences, AlgPred combines three methods: (i) a BLAST search to identify allergens based on their similarity to known allergens; (ii) prediction of IgE epitopes, and (iii) MOTIF scanning search for conserved motifs in the protein that are also present in allergenic proteins.
Immunoinformatics can also aid in the discovery of immunogenicity prediction.The possibility of an immune response toward AnOTA was evaluated by the VaxiJen program. 21This program is a widely used method to distinguish between immunogens and nonimmunogens among proteins from different origins.In agreement with the above results, VaxiJen assigned a low antigenicity for AnOTA classifying it as a non-antigen.The prediction of epitope binding to major histocompatibility complex MHC class I and II molecules was made with NetMHCcons and NetMHCIIpan, respectively. 22Using the default settings of NetMHCIIpan and considering HLA supertype representatives (the most frequent alleles of HLA A/B), 16 potential epitopes with a length of 15 amino acids were identified in AnOTA as strong binders to MHC class II.Additional 49 weak binders were also identified to MHC class II.The detected number of epitopes for MHC class I was lower in both forms, being identified only 4 strong binders and 8 weak binders.
Finally, we used the PROSPERous program 23 to predict potential substrate cleavage sites for 90 proteases.Of these proteases, 51 were of human origin including aspartic protease, cysteine protease, metalloprotease, and serine protease.According to PROSPERous, AnOTA would be sensitive to multiple proteases as eighty-four cleavage sites were found.Thus, AnOTA may be sensitive to cathepsin K as a cysteine protease, various matrix metalloproteases, and some serine proteases, including elastase, cathepsin G, and glutamyl peptidase I.
Mycotoxin contamination can cause huge economic losses to global agriculture.The enzymatic biodetoxification of mycotoxins is a suitable strategy due to its high efficiency, strong specificity, and no damage to nutrients.However, these advantages are highly dependent on the biochemical characteristics of the selected enzyme.In this study, AnOTA peptidase from A. niger, generally regarded as a safe organism, has been characterized.The pH and temperature profiles exhibited by AnOTA, together with its thermostability, make this enzyme appropriate for its use in technological processes.The enzyme exhibited strong specificity toward substrates containing a Cterminal Phe or aliphatic residue, which is a desirable feature since it minimizes potential damage to nutrients.The structural basis for this specificity has been studied by molecular docking simulations, which have revealed the pattern of interactions between the substrates and the amino acid side chains within the AnOTA active site.Moreover, computationally predicted characteristics indicated that AnOTA is a nonallergenic, nonantigenic, and low immunogenic protein, which can be cleaved by human proteases.All these in vitro and in silico results support the suitability of AnOTA for its use for OTA detoxification in food processing and animal feed.However, further studies are needed to decipher the remaining knowledge gap about AnOTA efficiency in vivo.

Figure 1 .
Figure 1. Biochemical properties of AnOTA from A. niger CBS 513.88.(A) AnOTA pH activity profile.(B) AnOTA temperature activity profile.(C) AnOTA thermal stability after incubation at 20 (orange circles and line), 37 (blue circles and line), 55 (green squares and line), and 65 °C (black squares and line), in MOPS buffer (50 mM, 20 mM NaCl, and pH 7) at indicated times.Inset: activity values at 20 h incubation at different temperatures.The mean value and standard error are shown (n = 3).The percentage of residual activity was calculated by comparing it to the nonincubated enzyme.

Figure 2 .
Figure 2. Effects of divalent cations on untreated (AnOTA − ; light blue) and EDTA-treated (AnOTA + ; dark blue) AnOTA activity.The relative activity values of AnOTA − or AnOTA + are those registered after incubation for 15 min with 1 mM concentration of different divalent cations.The activity of corresponding to AnOTA − in the absence of added divalent cations was defined as 100%.Light-blue bars correspond to AnOTA − and dark ones correspond to AnOTA + .The experiments were performed in triplicate.The mean value and standard error are shown.

Figure 3 .
Figure 3.Effect of 8 and 20 mM CaCl 2 on AnOTA − and AnOTA + .Proteins were preincubated with the corresponding additives during 15 min at room temperature before the reaction with 4MF as a substrate.The error bars represent the standard deviation estimated from three independent assays.Light-brown bars represent AnOTA − (EDTA-untreated AnOTA), and dark-brown bars represent AnOTA + .Green bar represents the activity of amidase 2 mat (truncated form of AnOTA), and the red one represents the activity of amidase 2 sig (AnOTA).These two latter results are from Yu et al. (2015).24The activity exhibited by native untreated AnOTA in the absence of additive was defined as 100%.

Figure 4 .
Figure 4. AnOTA activity on OTA (left) and OTB (right) mycotoxins.The HPLC chromatograms of OTA and OTB (5 μM) incubated at 37 °C for 16 h in the presence of AnOTA are shown in blue lines.Control reactions without AnOTA are shown in orange.The fluorescence wavelengths were 330 nm for excitation and 460 nm for emission.

Figure 5 .
Figure 5. Specificity profile of AnOTA.The height of the bars represents the enzymatic activity level of AnOTA against different substrates, normalized to the maximum observed activity (against glycyl-phenylalanine; GF).Only the results from hydrolyzed substrates are shown.Each color corresponds to a different family of substrates (see Figure S4): blue, N-benzyloxycarbonyl amino acid derivatives with one amide bond; light brown, N-benzyloxycarbonyl amino acid derivatives with two amide bonds; green, hippuryl-amino acids; and gray, L-dipeptides.ZA: N-benzyloxycarbonyl-L-alanine; ZF: N-benzyloxycarbonyl-L-phenylalanine; and ZL: N-benzyloxycarbonyl-L-leucine.ZAF: N-benzyloxycarbonyl-L-alanyl-L-phenylalanine; ZAL: N-benzyloxycarbonyl-L-alanyl-L-leucine; and ZFI: N-benzyloxycarbonyl-L-phenylalanine-L-isoleucine.HF: hippuryl-L-phenylalanine.GF: Lglycyl-phenylalanine.GA: L-glycyl-alanine.The error bars represent the standard deviation estimated from the three independent assays.The percentage of hydrolyzed compounds is indicated on top of each bar (initial concentration: 1 mM).

Figure 7 .
Figure 7. AnOTA activity on OTA containing plant-food beverages.HPLC chromatograms of (A) almond, (B) oat, and (C) soy commercial beverages containing 5 μM OTA, without AnOTA (orange traces) and incubated with AnOTA at 20 °C for 4 h are shown (blue traces).The fluorescence wavelengths were 340 nm for excitation and 460 nm for emission.

Figure 8 .
Figure 8. Free amino acid composition of soy commercial beverages.The relative amount of each amino acid is shown in different conditions: control without AnOTA and OTA (green); with AnOTA (orange); and with AnOTA and OTA (red).The reactions were performed with the soy beverages by adding OTA at 5 μM final concentration (red) and using 5 μg of the AnOTA enzyme (orange and red).Reactions were incubated at 20 °C for 4 h.

Table 1 .
Closest Structural Homologues of AnOTA and Substrate Specificity Protein whose DNA has been isolated from the Sargasso Sea.d Hyd: hydrophobic, C-terminal amino acid.

■ ASSOCIATED CONTENT * sı Supporting Information The
Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jafc.4c02944.Scheme of the hydrolysis of the amide bond of OTA, SDS-PAGE analysis of recombinant AnOTA from A. niger CBS 513.88, structural comparison of the AnOTA subunit with those from its close structural homologues, structures of the substrates used in our study classified into five families and also those of OTA and 4MF, docking of OTA into the cryoEM 3D structure of the AnOTA homologue enzyme ADH3 from S. acidaminiphila (SaOTA) and comparison with the experimentally determined structure of the mycotoxin, and SDS-PAGE analysis of plant-based beverages containing OTA (5 μM) incubated with AnOTA at 37 °C for 18 h (PDF)