Biochemical characterization, cDNA cloning, and molecular modeling of araujiain aII, a papain-like cysteine protease from Araujia angustifolia latex

Araujiain aII, the protease with highest specific activity purified from latex of Araujia angustifolia (Apocynaceae), shows optimum proteolytic activity at alkaline pH, and it is completely inhibited by the irreversible inhibitor of cysteine proteases trans-epoxysucciny-l-leucyl-amido(4-guanidino) butane. It exhibits esterolytic activity on several N-α-Cbz-amino acid p-nitrophenyl esters with a preference for Gln, Ala, and Gly derivatives. Kinetic enzymatic assays were performed with the thiol proteinase substrate p-Glu-Phe-Leu-p-nitroanilide (K m = 0.18 ± 0.03 mM, k cat = 1.078 ± 0.055 s−1, k cat/K m = 5.99 ± 0.57 s−1 mM−l). The enzyme has a pI value above 9.3 and a molecular mass of 23.528 kDa determined by mass spectrometry. cDNA of the peptidase was obtained by reverse transcription-PCR starting from total RNA isolated from latex. The deduced amino acid sequence was confirmed by peptide mass fingerprinting analysis. The N-terminus of the mature protein was determined by automated sequencing using Edman’s degradation and compared with the sequence deduced from cDNA. The full araujiain aII sequence was thus obtained with a total of 213 amino acid residues. The peptidase, as well as other Apocynaceae latex peptidases, is a member of the subfamily C1A of cysteine proteases. The enzyme belongs to the alpha + beta class of proteins, with two disulfide bridges (Cys22–Cys63 and Cys56–Cys95) in the alpha domain, and another one (Cys150–Cys201) in the beta domain, as was suggested by molecular modeling.


AaCPII
Amino acid sequence corresponding to the cysteine protease araujiain aII deduced from cDNA and N-terminal sequence ACN Acetonitrile AMPSO (N-(1,1-dimethyl-2-hydroxyethyl)-3-amino-2hydroxy-3-[(1-hydroxy-2-methylpropan-2-yl) amino] propane-1-sulfonic acid CAPS 3-  Endress and Bruyn (2000)], a native species from South America, commonly known as ''tasi'', ''taso'' or ''doca'' (Bucciarelli et al. 2008) oozes sticky and toxic latex upon tissue damage.Latex is a milky fluid with a complex mixture of constituents, including proteins, vitamins, carbohydrates, lipids, terpenes, alkaloids, starches, oils, tannins, resins, gums and free amino acids.Laticifers not only serve as a repository for natural products but also exhibit unique proteomes.Many latex proteins are thought to have defensive roles in plants, and include proteases, chitinases and pathogenesis-related polypeptides (Hagel et al. 2008;Liggieri et al. 2009).Over 110 latices of different plant families are known to contain at least one proteolytic enzyme.Most of them belong to the cysteine or serine endopeptidase catalytic type (Domsalla and Melzig 2008).Species belonging to the Asclepiadoideae subfamily often contain papain-like cysteine proteases (PLCPs) in latex, which may be integrated in the defense scheme of those plants against a specialist herbivore playing crucial roles in plant-pathogen/pest interactions (Rasmann et al. 2010).During these parasitic interactions, PLCPs act on host pathogen substrates.In this manner proteolysis induces the selection of counteracting inhibitors, non-cleavable substrates and other means to evade proteolysis.Therefore, the interactions of proteases with their substrates and inhibitors can be seen as a molecular battlefield.Intriguingly, both plants and their invaders use PLCPs at these molecular battlefields (Shindo and van der Hoorn 2008).In A. angustifolia, latex is contained within unarticulated ramified laticifers creating a network throughout most of the plant.
From a pharmaceutical and biotechnological point of view, latex is an important source of plant peptidases used in industry due to their property of being active over wide ranges of temperature and pH (Domsalla and Melzig 2008;Liggieri et al. 2009).Applications include the use of these enzymes as meat tenderizers (such as Panol Ò Purified Papain, Liquipanol Ò T100), component of detergent formulations (Khaparde and Singhal 2001), or surfactants synthesis (Morcelle et al. 2009) among others.Papain alone or in combination with other proteolytic enzymes has long been available for diverse medical indications (Beuth 2008;Salas et al. 2008); as an example we can mention Papacarie, a gel available in the market, used for chemomechanical dental caries removal.Latices from Asclepiadoideae subfamily have been used in folk medicine as antiparasitic agents against gastrointestinal nematodes and in wound healing in addition to other ailments.The pharmacological actions of these milky latices have been attributed to the presence of cysteine peptidases (Stepek et al. 2004;Obrego ´n et al. 2009a).Shivaprasad et al. (2009) have shown that the PLCPs present in Asclepias curassavica latex exhibited strong procoagulant action and were found to be specific in their action (thrombin like).This is considered as the basis for the traditional use of plant latex to stop bleeding on fresh cuts.On the other hand, PLCP activity of Calotropis procera latex is involved in the larvicidal action of latex proteins on Aedes aegypti larvae (Ramos et al. 2009).
PLCPs are usually 23-30 kDa in size, and use a catalytic cysteine residue to cleave peptide bonds in protein substrates.This catalytic cysteine is part of a catalytic triad situated in the middle of a cleft that binds the substrate through specific interactions (Shindo and van der Hoorn 2008).PLCPs are included in clan CA, family C1 and subfamily A in the MEROPS classification (http://www.merops.sanger.ac.uk).Subfamily C1A comprises proteases that contain disulfide bridges and accumulate in vesicles, vacuoles, or the apoplast (van der Hoorn 2008).In the current paper we report the biochemical, proteomic and molecular characterization of araujiain aII, the PLCP with higher specific activity, purified from latex of the milkweed A. angustifolia.

Plant material
Fruits of A. angustifolia (Hook.et Arn.) Decaisne were obtained from plants grown in the town of Ringuelet, Province of Buenos Aires, Argentina (Obrego ´n et al. 2006).The plant is an attractive, small, tender vine with small green twining stems, evergreen arrow-shaped leaves and white bell-shaped flowers with the petals reflexed leaving a pink pointed central structure; fruits are green oblong-fusiform follicles (Burkart 1979).Voucher specimens were deposited in the LPE herbarium (Facultad de Ciencias Exactas, Universidad Nacional de La Plata, Argentina).

Crude extract preparation and protein determination
Latex obtained by superficial incisions of fruits, collected in 0.05 M citric-citrate buffer (pH 4.5) containing 5 mM EDTA, was first centrifuged at 5,0009g for 30 min at 4°C.Gums and other insoluble materials were discarded, and the supernatant was ultra centrifuged at 100,0009g for 30 min at 4°C.This new supernatant (''crude extract''), containing soluble proteins, was fractionated and conserved at -20°C for further studies.
Proteins present in the crude extract (CE) were determined by Bradford's method using bovine albumin as standard (Bradford 1976).

Protein purification
Purification was carried out by cation exchange chromatography according to the method of Obrego ´n et al. (2009b).Briefly, 1 ml of the crude extract containing 1.5 mg of protein was loaded onto a Pharmacia XK 16/40 column having AK16 adaptors, packed with SP-Sepharose Fast Flow and equilibrated with 0.055 M Tris-HCl (pH 7.4).Chromatography was developed in an FPLC equipment (Pharmacia, Uppsala, Sweden) by washing with the equilibrating buffer and further elution of the bound material with a step gradient of sodium chloride (0-0.5, 0.5-0.8, and 0.8-2.0M) in the same buffer.Cation exchange chromatography was spectrophotometrically monitored by absorbance measurement at 280 nm.
Caseinolytic activity and pH profile Proteolytic assays were made using casein as substrate.The reaction mixture was prepared by mixing 0.1 ml of enzyme extract with 1.1 ml of 1% casein containing 12 mM cysteine, in a 0.1 M Tris-HCl buffer (pH 8.0).The reaction was performed at 45°C on an Agilent 8453 UV-visible spectrophotometer equipped with a thermostated cell and was stopped 2 min later by the addition of 1.8 ml of 5% TCA.Each test tube was centrifuged at 3,0009g for 30 min and the absorbance of the supernatant measured at 280 nm.An arbitrary enzyme unit (caseinolytic unit, U cas ) was defined as the amount of protease which produces an increment of one absorbance unit per min in the assay conditions (Priolo et al. 1991).
The effect of pH on enzyme activity of the purified protease was measured with casein (pH range 6.4-10.5)using 10 mM sodium salts of the following ''Good'' buffers: MES, MOPS, TAPS, AMPSO and CAPS (Liggieri et al. 2004).
Triplicate measurements were performed for each assay.

Endoesterolytic activity
These assays were performed under optimal conditions according to the method of Silverstein (1974) modified by Obrego ´n et al. ( 2001) with N-a-carbobenzoxy-p-nitrophenyl esters of the following amino acids: Ala, Asn, Asp, Gln, Gly, Ile, Leu, Lys, Phe, Pro, Trp, Tyr and Val.Reaction mixture consisted of 1.8 ml of 0.1 M Tris-HCl buffer (pH 8.0), 0.1 ml of 1 mM substrate in ACN solution and 0.1 ml of purified peptidase.Continuous liberation of p-nitrophenol was followed at 37°C with an Agilent 8453 E UV-visible spectroscopy system at 405 nm every 3 s for 2 min.Triplicate measurements were performed for each assay and control assays were also made.In each case the non-enzymatic hydrolysis for each substrate was determined by replacing the enzyme solution with 0.1 ml of buffer.Enzyme kinetic data were analyzed for the most preferred N-Cbz-aa p-nitrophenyl ester derivative determined in the aforementioned assays; substrate concentrations range from 0.1 to 1 mM in the reaction mixture.

Amidolytic activity and enzyme kinetic
Amidasic activity was measured according to the method of Filippova et al. (1984) modified by Morcelle et al. (2004).The reaction mixture contained 1.5 ml of 0.1 M phosphate buffer pH 6.5, 0.3 M KCl, 10 mM EDTA, 3 mM DTT, 0.18 ml of substrate (stock solution: 4 mM PFLNA in DMSO) and 0.12 ml of enzyme.The p-nitroaniline released at 37°C was spectrophotometrically detected at 410 nm.Triplicate measurements were performed and control assays were also made.An arbitrary enzyme activity unit (U PFLNA ) was defined as the amount of protease that released one micromole of p-nitroaniline per minute in the assay conditions.Kinetic parameters using this substrate were determined for the purified enzyme.K m and k cat were calculated by non-linear regression analysis of Michaelis-Menten equation.

Determination of catalytic type
The action of different cysteine protease inhibitors (50 mM sodium iodoacetate, 50 mM iodoacetamide and 100 lM E-64) was evaluated by incubating the purified enzyme for 30 min at 25°C (Salvesen and Nagase 2001).The residual caseinolytic activity was measured as indicated in ''Caseinolytic activity and pH profile'' (Obrego ´n et al. 2006).

Gradient SDS-PAGE analysis
The active fraction, uninhibited and inhibited with 10 mM sodium iodoacetate, was submitted to denaturing SDS polyacrylamide gel electrophoresis according to Scha ¨gger and von Jagow (1987) with modifications.The electrophoresis was performed in a 10-16% linear gradient of polyacrylamide and 0-9% sucrose linear gradient to amplify resolution.Electrophoresis was run in two constant voltage steps (stacking gel: 30 V, resolution gel: 90 V).Gels were stained with Coomassie Brilliant Blue R-250.Silver staining was also performed to improve visualization of protein bands (O'Connell and Stults 1997).

Mass spectrometry analyses
Matrix-assisted laser desorption ionization-time-of-flight mass spectrometry (MALDI-TOF MS) was used for the determination of endopeptidase molecular mass, as well as its purity degree.MALDI-TOF mass spectra was acquired on a BRUKER Ultraflex spectrometer equipped with a pulsed nitrogen laser (337 nm), in linear positive ion mode, using a 19 kV acceleration voltage.Samples were prepared by mixing equal volumes of a saturated solution of matrix (3,5-dimethoxy-4-hydroxycinnamic acid-sinapinic acid) in 0.1% TFA in water/ACN 2:1, and 1 lM protein solution.

N-terminal sequence
The N-terminal sequence was determined by automated Edman's degradation.Samples were prepared by immobilizing 10-100 pmol of pure protein on polyvinylidene fluoride (PVDF) membrane in Ultrafree-Probind Filter tubes (Millipore Corporation, Billerica, MA, USA).Each sample was centrifuged at 10,000 rpm for 2 min and the PVDF filter was washed with MilliQ water.Finally, we introduced the immobilized protein in the reaction chamber of the Procise protein sequencer (Applied Biosystem, Life Technologies, Carlsbad, CA, USA).
To confirm that the PCR products corresponded to the sequence of the gene sought we conducted a nested PCR reaction (Frohman 1994).Reaction conditions were as specified above using two combinations of internal primers: CAapo 1 -R 1 , and CAapo 2 -R 1 (CAapo 1 5 0 -CCTATCAG AAATCAAGGAAAATGTGGGAGTTGCTGG-3 0 and CAapo 2 5 0 -ATCAAGGAAAATGTGGGAGTTGCTGG-3 0 ).The nested-PCR products were separated by agarose gel (2%) electrophoresis, and selected DNA bands were excised and purified using a DNA extraction kit (QIAEX II Agarose Gel Extraction, QIAGEN GmbH).Finally, the purified DNA fragments were cloned into the pGEM-T Easy vector in heat shock competent E. coli XL1-Blue.Plasmids of selected colonies were purified using the GFX extraction kit Micro Plasmid Prep Kit (GE HealthCare Life Sciences) and sequenced by the Sequencing Services of Veterinary Faculty, Autonomous University of Barcelona (Barcelona, Spain).Chromatograms were analyzed with the software Chromas v2.13 (Technelysium Pty.Ltd.).cDNA sequence analysis, homology modeling and validation of 3-D structure cDNA sequences obtained from selected clones were analyzed with Clustal-W alignment software (Thompson et al. 1994) to obtain a consensus sequence.The consensus sequence was translated in all six reading frames with the Translate Tool software available on the ExPASy server to identify the presence of conserved cysteine endopeptidase elements in the amino acid sequence.At the same time, a PSI-BLAST search was conducted using the NCBI nr (non-redundant) database restricted to the Viridiplanteae Kingdom with the matrix Blosum 62. Distance map was constructed using the Neighbor-Joining Method.Physicochemical properties of araujiain aII were predicted by GPMAW v6.0 (Lighthouse data, DK-5230, Odense M, Denmark) and theoretical values were compared with experimental data.
The 3D model of araujiain aII from A. angustifolia was built by homology modeling using Modeller 9v7 software.Sequence similarity search was performed with the NCBI Basic Local Alignment Search Tool (BLASTprotein) and crystal structures of the closest homologues available in the Brookhaven Protein Data Bank (PDB: http://www.rcsb.org/pdb/home/home.do) were selected.The high score and low E-value were considered.Results yielded by NCBI BLAST revealed papaya proteinase omega with a resolution of 1.80 A ˚(PDB ID: 1PPO) of latex of Carica papaya as a suitable template.Models obtained were evaluated and validated with DOPE assessment score.

Characterization of araujiain aII
Araujiain aII, the most active peptidase isolated from the latex of A. angustifolia, was purified by cation exchange chromatography as previously reported by Obrego ´n et al. (2009b).Homogeneity of this peptidase was analyzed by electrophoresis on polyacrylamide gradient gel (Fig. 1) and mass spectrometry (Fig. 2).Unlike what happens with other peptidases when preparing electrophoresis samples (Priolo et al. 2000;Obrego ´n et al. 2001) the presence of an inhibitor is not required to prevent autodigestion.Autolysis resistance proteases are very useful for industrial or therapeutic applications and frequently proteases are modified by protein engineering to avoid rapid autolysis (Singh et al. 2010).Araujiain aII molecular mass measured by MALDI-TOF was 23.528 kDa (Fig. 2).This value is of the same order of magnitude than those obtained for other Apocynaceae peptidases (Obrego ´n et al. 2009b) and as reviewed by Domsalla and Melzig (2008) all latex cysteine peptidases are in the range from 21 to 29 kDa.
A suspension of pure enzyme with a concentration of 55 lg protein/ml displayed a caseinolytic activity of 0.17 U cas /ml (specific activity: 3.1 U cas /g).The range of pH at which the enzyme had more than 10% activity on the same substrate was 7.5-9.0,while it showed more than 90% of activity in a very narrow pH range, between 8.0 and 8.5 (Fig. 3).These data were consistent with that observed for the crude extract of A. angustifolia latex with a pH optimum between 6.7 and 8.5 (Obrego ´n et al. 2006).The action of cysteine protease inhibitors produced a total and irreversible inhibition of enzyme activity.While b-mercaptoethanol, DTT, and cysteine produced a significant activation of the enzyme confirming the cysteine peptidase catalytic mechanism of araujian aII (data not shown).
As seen in Fig. 4 the highest preferences were shown for Gln and Ala derivatives, followed by Gly, Tyr, Phe and Asp derivatives in decreasing order.Both araujiain aII and asclepain f hydrolyse the same substrates (Trejo et al. 2001).Other peptidases of the Apocynaceae family display different substrate specificity.Asclepain cII and asclepain cI from latex of A. curassavica strongly prefer Asp and Tyr derivatives (Liggieri et al. 2004(Liggieri et al. , 2009)).Morrenain bI exhibits strong preference for Ala and Asp derivatives while morrenain bII prefers Asp and Gly, both enzymes isolated from latex of Morrenia brachystephana (Vairo-Cavalli et al. 2001, 2003).With regard to the enzyme kinetics, araujiain aII had a non-Michaelian behavior with these kinds of substrates.Native peptidase N-terminal amino acid sequence was determined up to 23 aminoacids by EDMAN sequencing: LPDSVDWRDKGVVFPIRRQGKCG. On the other hand, the araujiain aII peptide map is shown in Fig. 5. Using MASCOT search engine it was not possible to identify this enzyme; therefore, it may be considered a novel proteinase.Protein identification and differentiation by PMF has been adopted in our group as an excellent tool to differentiate, in a fast and unequivocal way, proteases with very similar physicochemical and functional properties (Obrego ´n et al. 2009a;Torres et al. 2010).The robustness of the method even allowed the differentiation of the homologue isoenzymes of the latex of A. curassavica (Obrego ´n et al. 2009a).

Peptidase cloning and sequence analysis
Araujiain aII was cloned with the template synthesized from A. angustifolia latex RNA and primers Nt 1-8 and R 0 .Identity of PCR products was confirmed by amplification with nested primers CAapo 1 and R 1 .From nested PCR   1).As shown in Table 1, five experimental tryptic fragments match peptides of the putative sequence generated by simulated digestion (mass error tolerance: 0.6 Da); the protein sequence coverage of peptide fragments was 33% with a fine peptide distribution.The proposed sequence for the mature enzyme is shown in Fig. 6, containing 213 aminoacids and including those residues coming from the N-terminus.
As could be expected, a high conservation degree was observed for those amino acid residues which are essential for catalytic activity (Cys25, His159 and Asn 175 according to mature papain numbering; Beers et al. 2004) and those important for maintaining the tertiary structure.Araujiain aII as well as asclepain cI, asclepain cII (Obrego ´n et al. 2009a), asclepain f (Trejo et al. 2009), philibertain gA and philibertain gB (Sequeiros et al. 2005), all enzymes from latex of Apocynaceae family, presented Fig. 8 Distance map of papain-like cysteine peptidases.Distance model was constructed with the PSI-BLAST tool; tree method: neighbor joining, distance according to Grishin (protein) restricted to 0.85.Peptidase names are indicated in bold followed by plant species in italic and accession numbers are represented in parenthesis.The tree includes peptidases most closely related in sequence to araujiain aII, and papain (the subfamily C1A type example).Green circles monocots, blue circles eudicots.CP cysteine protease, SAG senescence associated gene the catalytic Cys at position 25, while His was located at position 156 (Fig. 6).
Neighbor joining analysis (Fig. 8) showed enzymes from Apocynaceae latex as a separate group.Despite their homology, as was discussed previously, each peptidase displayed unique substrate specificity.
AaCPII was analyzed by GPMAW v6.0 to estimate some physicochemical properties: relative molecular mass (23,390.66Da); 280 nm molar extinction coefficient (48,010 M -1 cm -1 ); and pI with reduced and oxidized thiols (9.36 and 10.05, respectively).The latter is slightly more basic than the experimental pI: 8.9 (Fig. 9).This experimental pI is similar to papain pI (8.75) and less basic than pI of other Apocynaceae peptidases (Dubey and Jagannadham 2003;Sequeiros et al. 2005;Liggieri et al. 2009).The estimated molecular mass of araujiain aII is fairly close to the value obtained by mass spectrometry and of the same order of molecular masses of other peptidases from Apocynaceae (Obrego ´n et al. 2009a, b;Trejo et al. 2009).

Molecular modeling of araujiain aII
Among all the available molecular structures of PLCPs, papaya proteinase omega from latex of C. papaya was identified as the best structural template for modeling the molecular structure of araujiain aII.The root mean square deviation (RMS) of the Ca atoms between the modeled structure and the template structure was within a reasonable range (0.467 A ˚).The modeled overall structure of araujiain aII is shown in Fig. 10.The fold consisted of two domains (lobes) with the catalytic site lying between them.
The location of residues involved in catalysis is shown in Fig. 11.The L domain (N-terminal domain) composed of a helices is stabilized by the presence of two disulfide bonds and included Gln19 and Cys25.The R domain with a b-barrel structure and a disulfide bond contributes to the active site with His156 and Asn176 residues.PLCPs use Cys25 as nucleophile, activated by His 159 in the active site (papain numbering).Besides Cys and His, Gln19 helps in the formation of the electrophilic center that stabilizes the tetrahedral intermediate; while Asn175 is thought to orientate the imidazolium ring of catalytic His (Barrett and Rawlings 2004). Phe140, Trp178 and Trp182 (141, 177 and 181 according to papain numbering, respectively), involved in the hydrophobic pocket of the catalytic site in which the hydrogen bond between Asn176 and His156 is located, are also conserved in araujiain aII.In addition to the Cys25 residue of the active site, six Cys residues are  , 56, 63, 95, 150 and 201, as suggested by modeling probably involved in disulfide bridges(Cys22-Cys63, Cys56-Cys95, Cys150-Cys201).While asclepain f, chymopapain, Carica candamarcensis endopeptidase, and ficain have, in addition to the active site cysteine, a free thiol group (Cys133), AaCPII and papain both lack free sulfhydryl (Trejo et al. 2009).

Table 1
Identification of trysin-digested peptides of araujiain aII with theoretical PMF (Table