Physiological and Molecular Plant Pathology 74 (2009) 27–33

Contents lists available at ScienceDirect

Physiological and Molecular Plant Pathology

journal homepage: www.elsevier.com/locate/pmpp

Inhibition of a Hevea brasiliensis by a Kazal-like protease inhibitor from Phytophthora palmivora

Dutsadee Chinnapun a, Miaoying Tian b, Brad Day b, Nunta Churngchow a,* a Department of Biochemistry, Faculty of Science, Prince of Songkla University, Kanchanawanich Street, Hat-Yai, Songkhla 90112, Thailand b Department of Plant Pathology, Michigan State University, East Lansing, MI 48824, USA article info abstract

Article history: Protease inhibitors have been implicated in virulence of the oomycete plant pathogen Phytophthora Accepted 24 August 2009 infestans. Phytophthora palmivora, the causative agent of ‘‘leaf fall’’ and ‘‘black stripe’’ in the rubber plant (Hevea brasiliensis), belongs to the same genus as P. infestans and likely shares conserved pathogenesis Keywords: mechanism. Based on the sequences of the Kazal-like inhibitor EPI10 from P. infestans and Serine protease inhibitor its ortholog from Phytophthora ramorum, we designed a pair of primers to amplify the potential homolog Kazal family from P. palmivora. A full-length cDNA was isolated using reverse transcription polymerase chain reaction Phytophthora palmivora (RT-PCR) followed by rapid amplification of cDNA ends (RACE), and designated Ppepi10. Ppepi10 encodes Hevea brasiliensis Subtilisin A a 222 amino acid containing three putative Kazal domains, designated Kazal1, Kazal2 and Kazal3. In vitro protein expression and protease inhibition analyses revealed that both rKazal1 and rKazal2 domains inhibited the activity of subtilisin A but neither had an effect on the chymotrypsin and . Moreover, both of them interacted with a 95 kDa protease from H. brasiliensis leaf extracts, suggesting a role for Ppepi10 in pathogenicity through suppression of host plant defenses. Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction such as Neospora caninum which secretes a single domain Kazal inhibitor called NcPI-S [16]. This inhibitor was found to be highly Protease inhibitors are important natural tools for regulating the inhibitory to subtilisin, yet has little or no activity against elastase proteolytic activity of their targets. For example, inhibitors can or chymotrypsin. Serine protease inhibitors EPI1 and EPI10, which block the activity of a protease to regulate signaling mechanisms also belong to the Kazal family have been reported to be secreted by through receptor interactions [2]. By definition, protease inhibitors Phytophthora infestans [24,25]. The two-domain EPI1 and the three- function by impairing the proteolytic activity of target . In domain EPI10 proteins were shown to inhibit and interact with the the case of host–microbe interactions, the functions of host plant pathogenesis-related protein P69B a subtilase of tomato [24,25]. proteases are impacted through the specific recognition and activity Phytophthora palmivora is an oomycete that is the causative of their cognate pathogen inhibitors. Based on the catalytic types of agent of ‘‘leaf fall’’ and ‘‘black stripe’’ in the rubber plant. It attacks the inhibited proteases, protease inhibitors may be classified into at the petioles, causing mature leaves to fall prematurely and attacks least 6 types: cysteine protease inhibitors, serine protease inhibi- the tapping surface resulting in poor latex production. Although it tors, threonine protease inhibitors, aspartic protease inhibitors, is a pathogen of great economic importance in Thailand, little is glutamic protease inhibitors and metalloprotease inhibitors. known about the molecular mechanisms involved in the pathoge- The protease inhibitors directed against serine proteases can be nicity and host specificity of P. palmivora. P. palmivora belongs to divided into at least 20 different families based on sequence simi- the same genus as P. infestans and likely shares some conserved larity, topology, and mechanism of binding [13]. Among these, the pathogenesis mechanisms. Based on previous work by Tian et al. Kazal family is one family of serine protease inhibitors that has [24,25], we sought to identify potential inhibitors of defense previously been characterized in plant–pathogen interactions response in the Hevea brasiliensis–P. palmivora interaction. In short, [24,25]. This family is named after L. Kazal, who discovered the we hypothesized that protease inhibitors from P. palmivora might pancreatic secretory trypsin inhibitors, PSTI, present in all verte- also play a role in the suppression of plant defense responses. To brates [23]. Recently, they have been found in many organisms, this end, we focused on identifying protease inhibitors from P. palmivora with similarity to EPI10 from P. infestans because it was * Corresponding author. Tel.: þ66 74 288261; fax: þ66 74 446656. shown to have specific activity as well as play a role in suppression E-mail address: [email protected] (N. Churngchow). plant defense [24] and it has highly conserved sequences with

0885-5765/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.pmpp.2009.08.005 28 D. Chinnapun et al. / Physiological and Molecular Plant Pathology 74 (2009) 27–33 other Phytophthora species such as Phytophthora ramorum and products from the first RACE-PCR reaction were used as the DNA Phytophthora sojae. template in the nested PCR reactions using the oligonucleotides (50- Previous work has described that the full-length protein, EPI10, TGCCTGGACGTGTACGACCCGGTG-30)and(50-GACCCTCGTTCGAG- from P. infestans has protease inhibitor activity; however, functional TATTCCTTTCCA-30)asnested30-RACE and nested 50-RACE primers, activity of individual Kazal domains has not been fully characterized respectively. The primers for RACE-PCR reactions were designed based [24]. In this study, we describe the isolation, cloning and functional on the obtained partial Ppepi10 sequence. characterization of a Kazal-like extracellular serine protease inhibitor gene from P. palmivora, Ppepi10. This is the first serine 2.5. Cloning and sequencing RACE products protease inhibitor gene that has been cloned and characterized from P. palmivora. The Ppepi10 gene was produced using RT-PCR and The PCR products from RACE-PCR were cloned into the TOPO RACE-PCR, and was identified to contain 3 Kazal domains, desig- PCR 2.1 Vector (Invitrogen) and transformed into Escherichia coli nated Kazal1, Kazal2 and Kazal3. Moreover, we describe the inhibi- TOP10 cells, according to the manufacturer’s instructions. Plasmids tory activity against proteins from rubber leaf using a zymogram for sequencing were extracted using the QIAprep spin miniprep kit buffer system and co-immunoprecipitation to identify the specific (Qiagen). protease targeted by rKazal1 and rKazal2 of rPpEPI10. 2.6. Sequence analyses 2. Materials and methods The signal of Ppepi10 was predicted using SignalP 3.0 2.1. Plant growth and protein extraction [7]. The Kazal domains of Ppepi10 were identified by searching with the InterPro database (http://www.ebi.ac.uk/tool/InterProScan). Rubber plants (RRIM600 cultivar) were grown in a growth Multiple alignments of the Kazal domains from P. palmivora chamber with a photoperiod of 12 h of light and 12 h of dark at 25 C. (Ppepi10), P. infestans (EPI1 and EPI10) [24,25], the crayfish Paci- Leaves (5 g) of a 12 week-old rubber plant were homogenized with fastacus leniusculus (PAPI-1) [9], and the apicomplexan Toxoplasma liquid nitrogen using a mortar and pestle. Total protein was extracted gondii (TgPI1) [18] were conducted using the program CLUSTAL-X. with 10 ml of 100 mM Tris–HCl buffer pH 7.5 followed by precipi- The sequences were obtained from the NCBI database (www.ncbi. tation with 90% ammonium sulfate saturation at 4 C. The pellet was nlm.nih.gov). The P. palmivora sequence described in this paper collected by centrifugation at 12,000 g for 20 min at 4 Candthen was deposited in GenBank under accession no. FJ643536. dissolved in sterile distilled water. The solution was then desalted by loading onto a PD-10 column and eluted with distilled water. Eluted 2.7. Plasmid and bacterial strain used for production of rKazal1, fractions from the PD-10 column were monitored for protein content rKazal2 and rKazal3 proteins at a wavelength of 280 nm. The eluted fractions with high protein content were pooled and analyzed further, as described below. Plasmid pFLAG-Kazal1, pFLAG-Kazal2 and pFLAG-Kazal3 were constructed by cloning the PCR-amplified DNA fragments of Kazal1, 2.2. Phytophthora strain and culture condition Kazal2 and Kazal3, respectively into the EcoRI and KpnIsitesofpFLAG- ATS (Sigma), a vector allowing secreted expression of N-terminal FLAG P. palmivora, isolated from a diseased H. brasiliensis plant was fusion proteins. The oligonucleotides Kazal1-F (50-GCGGAATTCCATC- maintained on potato dextrose agar (PDA) medium at 25 C. For GACGACGACAAGTGCTCAT TC-30), Kazal1-R (50-GGGGTACCCTAGTCTG RNA extraction, P. palmivora was grown in Henninger medium [6] CGGGGCCGCTGG-30), Kazal2-F (50-GGGAATTCCATGTGCCCGGACGC on a rotary shaker at 100 rpm and 25 C for 15 days. TTGCCTG-30), Kaza2-R (50-GGGGTACCCTACGGTGGTCCCGTGTAGCC-30), Kazal3-F (50-GGGAATTCCATGTGCGCTGACATGTTGTGTCC-30) and 2.3. Primers design for P. palmivora epi10 (Ppepi10) Kazal3-R (50-GCGGGTACCTTACAGATTTAAAGTTTGAGAATAGGTC-30) were used to amplify the fragments. The introduced EcoRI and KpnI DNA primers for isolation of P. palmivora epi10 (Ppepi10)were restriction site are underlined. The transformation of pFLAG-Kazal1, designed based on alignments using the program CLUSTAL-X of pFLAG-Kazal2 and pFLAG-Kazal3 into the E. coli strain BL21 was EPI10 from P. infestans (GenBank accession no. AY586282) and performed using electroporator at 2500 volts. PramEPI10 from P. ramorum (Trace identified no. 303447516) [25]. These sequences were obtained from the NCBI databases (www. 2.8. Expression and purification of rKazal1, rKazal2 and rKazal3 ncbi.nlm.nih.gov). The conserved regions of epi10 from P. infestans and P. ramorum were used to create the primers for Ppepi10. Cultures of E. coli BL21 containing pFLAG-Kazal1, pFLAG-Kazal2 and pFLAG-Kazal3 were grown overnight on a rotary shaker at 2.4. RNA isolation, RT-PCR and RACE-PCR 250 rpm and 37 C. Overnight cultures were diluted (1:100) in LB medium containing carbenicillin (50 mg/ml) and incubated on a rotary Mycelium from P. palmivora wasgroundinliquidnitrogen,andtotal shaker at 250 rpm and 37 C until an optical density of OD600 ¼ 0.3, at RNA was extracted using the RNeasy plant mini kit (Qiagen). For RT- which time isopropyl-b-D-thiogalactopyranoside (IPTG) was added to PCR, total RNA was treated with DNA-Free (Ambion, Austin, TX), and a final concentration of 0.4 mM. The cultures were further incubated first-strand cDNAs were synthesized using the SuperScript III reverse overnight on a rotary shaker at 250 rpm and 28 C. The culture transcriptase RT-PCR system (Invitrogen) from 3 mg of total RNA. The supernatants of pFLAG-Kazal1, pFLAG-Kazal2 and pFLAG-Kazal3 were forward primer (50-TTTGGATGCCTCGACGTGTA-30) and degenerate used to purify rKazal1, rKazal2 and rKazal3 with anti-FLAG M2 affinity reverse primer (50-CGGAGCCGCACACAGGRGCATAGTTGTC-30)were gel column (Sigma). After loading each sample, the column was used for RT-PCR. RACE-PCR was employed to obtain the full-length washed with TBS buffer. Then the protein was eluted with 0.1 M Ppepi10 cDNA sequence, using Smart RACE cDNA amplification kit glycine (pH 3.5), and equilibrated to a neutral pH with 20 mlof1MTris following the instructions of the manufacturer (Clontech). The forward (pH 8.0), for each 1 ml eluted fraction. Protein concentration was primer (50-ATGCTACTTGCGCCTTGCGTCTTGC-30)andreverseprimer measured at 280 nm and calculated using an extinction coefficient of (50-CACAGGGGCATAGTTGTCGGGACAC-30) were used respectively for 6147 M1 cm1 for rKazal1 and 7575 M1 cm1 for rKazal2 deter- 30-RACE and 50-RACE in the firstround of RACE-PCR reactions. The PCR mined with the approach of Gill and von Hippel [5]. D. Chinnapun et al. / Physiological and Molecular Plant Pathology 74 (2009) 27–33 29

2.9. SDS-PAGE and western blot analyses was incubated in 1zymogram renaturation buffer, 2.5% (v/v) Triton X-100, for 30 min at room temperature. Then it was equilibrated in SDS-PAGE was performed using 15% (w/v) polyacrylamide gels. 1zymogram developing buffer containing 50 mM Tris: 0.2 M NaCl: After electrophoresis, gels were stained with silver nitrate following 5mMCaCl2 and 0.02% Brij 35 for another 30 min at room tempera- the method of Merril [17], or stained with Coomassie Brilliant Blue ture, followed by a 4 h incubation at 37 C in fresh 1zymogram [20]. A value for the relative molecular mass (Mr) of the protein band developing buffer before staining with 0.5% Coomassie brilliant blue. was estimated by comparison with a BenchMark pre-stained protein After the gel was destained with solution containing methanol: acetic ladder (6–180 kDa; Invitrogen). For Western blot analyses, proteins acid: water, 50:10:40, areas of protease activity appear as clear band were transferred to nitrocellulose membranes using a Mini trans-blot against a blue background where the protease has digested the apparatus. Detection of antigen–antibody complexes was carried out gelatin substrate. with anti-Flag M2-peroxidase (Sigma) and Super Signal West Pico chemiluminescent substrate (Pierce). 2.11. Co-immunoprecipitation

2.10. Protease inhibition assays The FLAG-tagged protein immunoprecipitation kit (Sigma) was used for testing for the co-immunoprecipitation of rKazal1 and Protease inhibition analysis of commercial serine proteases against rKazal2 with protein extracts from rubber leaf. A total of 800 pmol rKazal1 and rKazal2 was performed using the colorimetric Quanti- of purified rKazal1 or rKazal2 were incubated with 200 ml (2 mg) of cleave protease assay kit (Pierce). Twenty pmol of rKazal1 or rKazal2 rubber leaf proteins for 30 min at 25 C. Forty ml of anti-FLAG M2 was incubated with 20 pmol of chymotrypsin (Sigma), subtilisin resin was added and incubated at 4 C, overnight with gentle A (Sigma), or trypsin (Pierce), in a volume of 50 mlfor30minat25C. shaking. The bound protein complexes were eluted with 60 mlof After 30 min, 100 ml of succinylated casein (2 mg/ml in 50 mM Tris FLAG peptide solution (150 ng/ml). The analysis was performed buffer, pH 8) was added and incubated at room temperature for using an 8% gel for SDS-PAGE and silver nitrate staining. 20 min. After that, 50 ml of chromogenic reagent 2,4,6-trini- trobenzenesulfonic acid was added and incubated for 20 min at room 3. Results temperature before measuring protease activity at 450 nm by spectrophotometry. 3.1. Isolation of a full-length Ppepi10 cDNA The BioRad zymogram buffer system was used to test for inhibi- tion of plant protease by rKazal1 and rKazal2. Both 20 and 80 pmol of CLUSTAL-X alignment of sequences of EPI10 from P. infestans rKazal1 or rKazal2 were incubated with 8 ml(0.08mg)ofrubberleaf (GenBank accession no. AY586282) and PramEPI10 from P. ramo- proteins for 30 min at 25 C before mixing with zymogram sample rum (Trace identified no. 303447516) was used to generate the buffer. The 8% zymogram gel was prepared as standard SDS-poly- Ppepi10 primer set. As shown in Fig. 1, the alignment of epi10 acrylamide gel with 0.1% gelatin added in both separating gel and sequences from two Phytophthora species revealed a highly stacking gel. The samples were then loaded onto the gel without conserved basic sequence structure. Two relatively long conserved boiling or addition of reducing reagents. After electrophoresis, the gel stretches were chosen for the forward primer and the degenerate

Fig. 1. Alignment of epi10 sequences from P. infestans and P. ramorum using CLUSTAL-X. The conserved sequences are highlighted in black. The underlined sequences indicated DNA primer sequences used for RT-PCR. F, forward primer; R, reverse primer. 30 D. Chinnapun et al. / Physiological and Molecular Plant Pathology 74 (2009) 27–33 reverse primer (Fig. 1). We obtained the partial Ppepi10 sequence and rKazal3 with FLAG antibody, we found that only rKazal1 and (Fig. 2) by sequencing a cDNA fragment amplified by RT-PCR. rKazal2 were expressed (data not shown). Additional protein The full-length Ppepi10 sequence was obtained using RACE-PCR expression constructs (e.g., pGEX 4T-1; GE Healthcare Life Sciences) as described in Materials and methods. The gene specific primers were made for expressing and purifying Kazal3, however, for both 50-RACE and 30-RACE reactions were designed based on the none expressed the Kazal3 domain, suggesting a possible level of partial Ppepi10 sequence (Fig. 2). The sequences from two RACE-PCR toxicity in E. coli that prevented us from moving forward with this reactions were assembled into the full-length Ppepi10 sequence, construct. Consequently only rKazal1 and rKazal2 were purified by which was deposited at the National Center for Biotechnology immunoaffinity using a gravity column packed with anti-FLAG M2 Information GenBank under accession no. FJ643536. affinity gel. The purified rKazal1 and rKazal2 were eluted from the column and analyzed by SDS-PAGE and staining with Coomassie Brilliant Blue. The apparent molecular weights were identified as 3.2. Ppepi10 encodes a putative extracellular Kazal-like 17 kDa for rKazal1 (Fig. 4A) and 15 kDa for rKazal2 (Fig. 4B). serine protease inhibitor 3.4. rKazal1 and rKazal2 inhibit the serine protease subtilisin A From the full-length Ppepi10 cDNA sequence, we identified an open reading frame of 669 bp that corresponded to a predicted The rKazal1 and rKazal2 proteins were tested for the inhibition translated product of 222 amino acids. SignalP 3.0 analysis of the of several commercially available serine proteases using the color- putative protein identified a 19-amino acid signal peptide with imetric Quanti-cleave protease assay kit. Three serine proteases a significant mean S value of 0.931. Using the InterPro database were used for identifying the inhibitor specificity: chymotrypsin, (http://www.ebi.ac.uk/tool/InterProScan), three domains of Ppepi10 subtilisin A and trypsin. In three independent experiments, we that were similar to Kazal inhibitors (InterPro IPR002350) were found that rKazal1 and rKazal2 proteins did not inhibit chymo- detected (Fig. 3A). trypsin or trypsin. However, both of them inhibited the activity of Multiple sequence alignment of the Kazal domains from P. pal- subtilisin A, with inhibition values of approximately 48% for rKazal1 mivora (PpEPI10; this study), P. infestans (EPI1 and EPI10), the and 26% for rKazal2 (Fig. 5). crayfish P. leniusculus (PAPI-1), and the apicomplexan T. gondii These results indicate that the Kazal1 and Kazal2 domains of (TgPI1) produced by CLUSTAL-X, illustrate the signature sequence Ppepi10 encoded a functional protease inhibitor that specifically of the Kazal-like serine protease inhibitor family (Fig. 3B). The first targets the subtilisin class of serine proteases. Moreover, our data domain of PpEPI10 (PpEPI10a) represented a typical Kazal domain, indicates that the rKazal1 domain has a higher activity than rKazal2. which contains six Cys residues forming 3 disulfide bridges of Cys1–5, Cys2–4 and Cys3–6. For the second (PpEPI10b) and the 3.5. rKazal1 and rKazal2 inhibit a protease from rubber leaf third domains (PpEPI10c), the Cys residues at position 6 and/or 3 were missing, representing atypical Kazal domains which are very rKazal1 and rKazal2 were tested for inhibition of proteases present common in Kazal-like inhibitors of plant pathogenic oomycetes in rubber leaf using the zymogram buffer system. A protein extract [26]. For all three domains, the predicted P1 residues, which play from rubber leaves was incubated with or without rKazal1 and rKa- central roles in determining the specificity of Kazal inhibitors, are zal2,andtheproteaseactivitywasmonitoredusingthegelprotease Asp, as are also found in the Kazal domains of P. infestans EPI1 and assay, described above. Interestingly, we found that 20 pmol of rKazal1 EPI10 (Fig. 3). completely inhibited the protease present in rubber leaf extracts (Fig. 6A); however, rKazal2 required a higher amount (80 pmol) to 3.3. Expression and purification of PpEPI10 Kazal domains inhibit the protease (Fig. 6B). The zymogram gels showed that the rubber leaf extract had at least two proteases, but only one of them rKazal1, rKazal2 and rKazal3 were cloned into the pFLAG-ATS was inhibited by rKazal1 and rKazal2 (Fig. 6). This result demonstrates protein expression vector (Sigma) in order to express and purify that rKazal1 and rKazal2 could inhibit a protease (likely a serine recombinant proteins fused with the FLAG epitope (N-terminal protease) from rubber leaf, and the rKazal1 was more active. fusion). After performing Western blot analyses of rKazal1, rKazal2 3.6. rKazal1 and rKazal2 interact with protease from rubber leaf

rKazal1 and rKazal2 were further tested for co-immunoprecip- itation with rubber tree proteins using FLAG antibody covalently linked to agarose beads. As shown in Fig. 7, both rKazal1 and rKa- zal2 co-precipitated a 95 kDa rubber tree protein, consistent with our results of protease inhibition, described above.

4. Discussion

Plant pathogens produce effector proteins that can either promote infection (virulence proteins) or trigger defense responses (avirulence proteins) in their host plants [4,10,11,15,22]. Several oomycete pathogens produce elicitor proteins as avirulence prod- ucts which in turn induce plant defense responses and pro- grammed cell death referred to as the ‘‘hypersensitive response’’ (HR). Recently, many effector molecules from oomycete pathogens have been reported which function to inhibit plant protein func- tions and in term, presumably shut down plant defense responses. Fig. 2. Partial Ppepi10 sequence obtained from sequencing of the Ppepi10 cDNA frag- ment amplified by RT-PCR. The underlined nucleotides indicate the primer sequences For example, P. sojae secretes glucanase inhibitor proteins (GIPs) used for RACE-PCR. that specifically inhibit the endoglucanase activity of soybeans. D. Chinnapun et al. / Physiological and Molecular Plant Pathology 74 (2009) 27–33 31

Fig. 3. PpEPI10 is predicted to be a member of the Kazal family of serine protease inhibitors. (A) Schematic representation of PpEPI10 structure. SP, signal peptide. The three Kazal domains are designated as PpEPI10a, PpEPI10b and PpEPI10c. The positions of the amino acid residues starting from the N terminus are indicated by numbers. The predicted P1 residues, which play central roles in determining the specificity of Kazal inhibitors, are Asp. C is the cysteine residues and the disulfide linkages predicted based on the structure of other Kazal domains are shown. (B) Sequence alignment of the Kazal inhibitor domains. Protease inhibitor PpEPI10a–c, FJ643536 from P. palmivora, EPI10a–c, AY586282 and EPI1a–b, AY586273 from P. infestans, PAPI-1a–d, CAA56043 from the crayfish Pacifastacus leniusculus and TgPI-1a–d, AF121778 from the apicomplexan T. gondii. The asterisks indicate amino acid residues that define the Kazal family protease inhibitor domains. The arrows indicate the positions of the Cys residues that are missing in PpEPI10b and PpEPI10c. The P6, P5, P4, P3, P2, P1, P10,P20,P30,P140,P150 and P180 are contact positions of Kazal domains to interact with their serine proteases.

Furthermore, GIPs and soybean endoglucanases interact in vivo however, no protease inhibitor has been identified from P. palmivora, during pathogenesis in soybean roots [8]. EPI1 and EPI10 of the thus making the identification of the specific protease inhibitor Kazal family are secreted from P. infestans during pathogenesis and described herein an important advance in understanding the viru- both the two-domain EPI1 and the three-domain EPI10 proteins lence strategies of P. palmivora. were shown to inhibit and interact with the pathogenesis-related Inthepresentstudy,wedescribetheisolationofaprotease protein P69B a subtilase of tomato [24,25]. inhibitor gene from P. palmivora, Ppepi10, and identified it as a member P. palmivora, the causative agent of leaf fall and black stripe in the of the Kazal family of serine protease inhibitors (Fig. 3B). Furthermore, rubber plant (H. brasiliensis) produces the avirulence molecules we showed that Kazal1 and Kazal2 of PpEPI10 encoded functional elicitin, as well as a 75 kDa elicitor. Both of them trigger defense protease inhibitor domains that specifically targeted the subtilisin class reactions, inducing accumulation of scopoletin, peroxidase, phenolic of serine proteases using the colorimetric Quanti-cleave protease assay compounds and local resistance in H. brasiliensis [3].Todate, kit. The rKazal1 and rKazal2 from PpEPI10 inhibited only the activity of subtilisin A but did not inhibit chymotrypsin and trypsin (subtilisin A,

Fig. 4. SDS-PAGE analysis of purified rKazal1 and rKazal2 proteins from anti-FLAG M2 Fig. 5. Protease inhibition assay of rKazal1 and rKazal2 using the colorimetric Quanti- affinity column stained with Coomassie brilliant blue. (A) Lane M indicates protein cleave protease assay kit. Standard error of deviation was calculated from the three standard. Lane ‘‘rKazal1’’ represents purified rKazal1 from the anti-FLAG M2 affinity independent replications. The experiment was repeated three times with similar column. (B) Purified rKazal2-Flag, as described in ‘‘A’’, above. results. 32 D. Chinnapun et al. / Physiological and Molecular Plant Pathology 74 (2009) 27–33

Fig. 6. Protease inhibition assay using the zymogram buffer system. ‘‘Control’’ is the rubber leaf proteins without incubation with rKazal1 or rKazal2. The ‘‘rKazal1 þ plant proteins’’ are rubber leaf proteins incubated with rKazal1. The ‘‘rKazal2 þ plant proteins’’ are rubber tree proteins incubated with rKazal2. (A) The zymogram buffer gel using 20 pmol of rKazal1 or rKazal2 incubated with 8 ml (0.8 mg) of rubber leaf proteins. The arrow indicates the band inhibited by rKazal1. (B) The zymogram buffer gel using 80 pmol of rKazal1 or rKazal2 incubated with 8 ml (0.8 mg) of rubber leaf proteins. The arrow indicates the band inhibited by rKazal1 or rKazal2. chymotrypsin and trypsin, represent three major classes of serine assassin bug Rhodnius prolixus. Rhodniin is composed of two Kazal- proteases) (Fig. 5). type domains, however, only one of which displays the typical activity The Ppepi10 gene shares approximately 72% sequence similarity of Kazal-type inhibitors. Interestingly, the second, ‘‘inactive’’ domain to P. infestans epi10, and like epi10, encodes for a protein containing has been shown to function independently through binding to the three Kazal domains (Fig. 3B). The first domain (PpEPI10a) is similar fibrinogen recognition exosite of thrombin by electrostatic interaction to the Kazal domains from previously identified oomycete inhibi- [1]. In the current study, we demonstrate that both rKazal1 and tors (Fig. 3B). The second domain (PpEPI10b) and the third domain rKazal2 specifically targeted the subtilisin class of serine proteases. (PpEPI10c) are atypical, lacking the third and sixth Cys for PpEPI10b Protein inhibitors of serine proteases are well studied and much is and the sixth Cys for PpEPI10c. Nonetheless, PpEPI10b exhibits the known about the interactions between an inhibitor and its cognate same pattern of abnormal sequences to that of the EPI10b domain , as most of these inhibitors act through a common mecha- of P. infestans. That is, EPI10b from P. infestans also lacked both the nism [14]. In short, this interaction is characterized by a tight asso- third and sixth Cys but retained the other four Cys (Fig. 3B). ciation between the enzyme and substrate (low Km)andthe Inhibitors of the Kazal-type typically exhibit two or more inhibi- hydrolysis reaction is very slow [19]. The higher activity of rKazal1 on tory domains that may be specific for different protease substrates the enzyme activity of subtilisin A shown in Fig. 5 infers that the Ki of [12]. For example, seven Kazal-type domains of ovoinhibitor from rKazal1 is lower than for rKazal2. However, the lack of the third and avian egg white have multiple targets, and at least five active sites: sixth Cys for rKazal2 should not affect the Ki of rKazal2 because the two for trypsin, two for chymotrypsin and one for porcine pancreatic contact residues of Kazal domains to interact with their serine elastase [21]. Another well-characterized member of the Kazal-type proteases are not the third and sixth Cys position. The contact posi- inhibitors is the thrombin inhibitor rhodniin, isolated from the tions of Kazal domains presented by Stephen et al. [23], P6, P5, P4, P3,

Fig. 7. Co-immunoprecipitation of rKazal1 (A) or rKazal2 (B) with rubber leaf proteins using FLAG antibody. The elutes were run on SDS-PAGE gel followed by staining with silver nitrate. The various inputs are shown below the gel image. Lane ‘‘Plant proteins’’ represents the input with rubber leaf proteins alone. Lane ‘‘rKazal1’’ or ‘‘rKazal2’’ represents the input with purified rKazal1 or rKazal2. Lane ‘‘Plant protein þ rKazal1/rKazal2’’ represents the input with the mixture of rubber leaf proteins incubated with rKazal1 or rKazal2. The protein band pulled down with the FLAG antibody in the presence of rubber leaf proteins and rKazal1/rKazal2 is indicated by the arrow. M, protein standard. D. Chinnapun et al. / Physiological and Molecular Plant Pathology 74 (2009) 27–33 33

P2, P1, P10,P20,P30,P140,P150 and P180 were shown in Fig. 3B. [3] Chinnapun D, Churngchow N. Induction of peroxidase, scopoletin, phenolic Nevertheless, the P3 (the second conserved cysteine residue) and compounds and resistance in Hevea brasiliensis by elicitin and a novel protein elicitor purified from Phytophthora palmivora. Physiol Mol Plant Pathol 2008; 0 P15 (a conserved asparagine residue) of contact positions show a low 72:179–87. variation in each Kazal domain then the other 10 contact positions [4] Collmer A, Badel JL, Charkowski AO, Deng WL, Fouts DE, Ramos AR, et al. Pseudomonas syringae Hrp type III secretion system and effector proteins. Proc were usually used for calculation the Ki of Kazal domains [23]. Natl Acad Sci 2000;97:8770–7. The amount of rKazal2 (80 pmol) that completely inhibited [5] Gill SC, von Hippel PH. Calculation of protein extinction coefficients from a protease from rubber leaf was 4 times higher than that for rKazal1 amino acid sequence data. Anal Biochem 1989;182:319–26. (20 pmol) and yet only one band was removed by these two inhibi- [6] Henninger H. Zur Kultur von Phytophthora infestans auf vollsynthetischen Nahrsubstraten. Z Allg Mikrobiol 1963;3:126–35. tors (Fig. 6). We also performed co-immunoprecipitation on the [7] Jannick DB, Henrik N, Gunnar H, Søren B. Improved prediction of signal rubber tree proteins incubated with rKazal1 and rKazal2 to identify : signalP 3.0. J Mol Biol 2004;340:783–95. the rubber tree protease targeted by rKazal1 and rKazal2 from [8] Jocelyn KCR, Kyung SH, Alan GD, Peter A. Molecular cloning and character- ization of glucanase inhibitor proteins: coevolution of a counterdefense rPpEPI10, and found that both rKazal1 and rKazal2 co-precipitated mechanism by plant pathogens. The Plant Cell 2002;14:1–17. with the protein from the rubber leaf. Therefore, it is expected that [9] Johansson MW, Keyser P, Soderhall K. Purification and cDNA cloning of a four- this protein should be a rubber leaf protease (Fig. 7). As a conse- domain Kazal proteinase inhibitor from crayfish blood cells. Eur J Biochem quence, rKazal1 and rKazal2 should interact with the same rubber 1994;223:389–94. [10] Kjemtrup S, Nimchuk Z, Dangl JL. Effector proteins of phytopathogenic leaf protease because the proteins from our co-immunoprecipitation bacteria: bifunctional signals in virulence and host recognition. Curr Opin of rKazal1 and rKazal2 with rubber leaf protein each had a molecular Microbiol 2000;3:73–8. weight of about 95 kDa (Fig. 7) which concurs with the result [11] Knogge W. Fungal infections of plants. Plant Cell 1996;8:1711–22. [12] Kreutzmann P, Schulz A, Standker L, Forssmann WG, Ma¨gert HJ. Recombinant obtained in Fig. 6. We hypothesize that these 95 kDa proteins are production, purification and biochemical characterization of domain 6 of PR-proteins found in H. brasiliensis, and as such, serve a primary role LEKTI: a temporary Kazal-type-related serine proteinase inhibitor. J Chroma- in plant defense. At present, we are pursuing the identification of the togr B Analyt Technol Biomed Life Sci 2004;803:75–81. [13] Krowarsch D, Cierpicki T, Jelen F, Otlewski J. Canonical protein inhibitors of 95 kDa protein(s), however, this advance is limited in the face of a lack serine proteases. Cell Mol Life Sci 2003;60(11):2427–44. of genomic sequence data from H. brasiliensis. [14] Laskowski Jr M, Kato I. Protein inhibitors of proteinases. Annu Rev Biochem In summary, the work presented herein leads us to conclude 1980;49:593–626. [15] Lauge R, De Wit PJ. Fungal avirulence genes: structure and possible functions. that PpEPI10 from P. palmivora is a Kazal-like extracellular serine Fungal Genet Biol 1998;24:285–97. protease inhibitor. It contains three Kazal domains, of which, Kazal1 [16] Meredith TM, Wen CC, Xing WZ, Susannah DB, Vern BC. Neospora caninum and Kazal2 were demonstrated to be important functional domains expresses an unusual single-domain Kazal protease inhibitor that is dis- charged into the parasitophorous vacuole. Int J Parasitol 2004;34:693–701. capable of inhibiting a protease from rubber leaf. [17] Merril CR, Goldman D, Sedman SA, Ebert MH. Ultrasensitive stain for proteins in polyacrylamide gels shows regional variation in cerebrospinal fluid Acknowledgements proteins. Science 1981;211:1437–8. [18] Pszenny V, Angel SO, Duschak VG, Paulino M, Ledesma B, Yabo MI, et al. Molecular cloning, sequencing and expression of a serine proteinase inhibitor This research was carried out in the laboratory of Brad Day at gene from Toxoplasma gondii. Mol Biochem Parasitol 2000;107:241–9. Michigan State University and supported by the Thailand Research [19] Read RJ, Fujinaga M, Sielecki AR, James MNG. Structure of the complex of Fund (TRF) through the Royal Golden Jubilee Ph.D. Program (RGJ- Streptomyces griseus protease B and the third domain of the turkey ovomucoid inhibitor at 1.8-A resolution. Biochemistry 1983;22:4420–33. PHD) to Ms. Dutsadee Chinnapun (PHD/0235/2548), a Graduate [20] Sambrook J, Fritsch EF, Maniatis T. Molecular cloning: a laboratory manual. Research Fund from the Prince of Songkla University, Thailand and 2nd ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory; 1989. in part by a National Science Foundation CAREER Award (B. Day; [21] Saxena MJ, Tayyab S. Protein proteinase inhibitors from avian egg whites. Cell Mol Life Sci 1997;53:13–23. IOS-0641319), as well as support from the Michigan Agricultural [22] Staskawicz BJ, Mudgett MB, Dangl JL, Galan JE. Common and contrasting Experiment Station (MAES). We thank the Research Technology themes of plant and animal diseases. Science 2001;292:2285–9. Support Facility, Michigan State University for help with DNA [23] Stephen ML, Wuyuan L, Qasim MA, Stephen A, Izydor A, Wojciech A, et al. Predicting the reactivity of proteins from their sequence alone: Kazal family sequencing and Prof. Dr. Brian Hodgson, Prince of Songkla Univer- of protein inhibitors of serine proteinases. Proc Natl Acad Sci 2001; sity for revision the manuscript and valuable comments. 98:1410–5. [24] Tian M, Benedetti B, Kamoun S. A second Kazal-like protease inhibitor from Phytophthora infestans inhibits and interacts with the apoplastic pathogenesis- References related protease P69B of tomato. Plant Physiol 2005;138:1785–93. [25] Tian M, Huitema E, Da Cunha L, Torto-Alalibo T, Kamoun SA. Kazal-like [1] Andreas L, Doriano L, Margit B, Robert H, Thomas F, Burkhard K, et al. Two extracellular serine protease inhibitor from Phytophthora infestans targets heads are better than one: crystal structure of the insect derived double the tomato pathogenesis-related protease P69B. J Biol Chem 2004;279: domain Kazal inhibitor rhodniin in complex with thrombin. EMBO J 1995; 26370–7. 14:5149–57. [26] Tian M, Kamoun S. A two disulfide bridge Kazal domain from Phytophthora [2] Bode W, Humer R. Natural protein proteinase inhibitors and their interaction exhibits stable inhibitory activity against serine proteases of the subtilisin with proteinases. Eur J Biochem 1992;204:433–51. family. BMC Biochem 2005;6:15.