Solvent-Stable Pseudomonas aeruginosa PseA ...

Research Article J Mol Microbiol Biotechnol 2008;15:234–243 DOI: 10.1159/000107488

Published online: August 21, 2007

Solvent-Stable Pseudomonas aeruginosa PseA Protease Gene: Identification, Molecular Characterization, Phylogenetic and Bioinformatic Analysis to Study Reasons for Solvent Stability Anshu Gupta a Swatismita Ray b Sanjay Kapoor b S.K. Khare a a

Chemistry Department, Indian Institute of Technology, and b Department of Plant Molecular Biology, University of Delhi South Campus, New Delhi, India

Key Words Cloning Gene identification lasB gene Pseudomonas aeruginosa Solvent-tolerant protease

and Cys-494) and had a number of hydrophobic clusters at the protein surface. These hydrophobic patches (21% of the sequence) and disulfide bonds may possibly be responsible for the solvent-stable nature of the enzyme. Copyright © 2007 S. Karger AG, Basel

Abstract We have previously isolated a solvent-stable protease from a novel solvent-tolerant strain of Pseudomonas aeruginosa (PseA). Here we report cloning and characterization of the gene coding for this solvent-tolerant protease. A homology search of the N-terminal amino acid sequence of the purified PseA protease revealed an exact match to a P. aeruginosa PST-01 protease gene, lasB. The c-DNA sequence of the PST01 protease was used to design primers for the amplification of a 1,494-bp open reading frame encoding a 53.6-kDa, 498amino-acid PseA LasB polypeptide. The deduced PseA LasB protein contained a 23-residue signal peptide (2.6 kDa) followed by a propeptide of 174 residues and a 33-kDa mature product of 301 residues. A phylogenetic analysis placed PseA lasB closest to the known zinc metalloproteases from P. aeruginosa. This gene was also found to contain a conserved HEXXH zinc-binding motif, characteristic of all zinc metallopeptidases. The 3D structure analysis of PseA protease revealed the presence of 7 -helices (36% of the sequence). The molecule was found to have two disulfide bonds (between Cys-227 and Cys-255 and between Cys-467

© 2007 S. Karger AG, Basel 1464–1801/08/0154–0234$24.50/0 Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/mmb

Introduction

We have previously reported on the isolation of a novel solvent-tolerant Pseudomonas aeruginosa strain [Gupta and Khare, 2006]. This strain is able to grow in the presence of a wide range of hydrophobic solvents at large concentrations. It secretes a novel protease, which is catalytically active and remarkably stable in all the solvents above log P 3.2, in as high as 75% solvent concentrations for at least up to 72 h. P. aeruginosa protease was purified to homogeneity by hydrophobic interaction chromatography and found to be a 32-kDa zinc metalloprotease [Gupta et al., 2005]. Its novel properties, being alkaline in nature and fairly stable in all the solvents and surfactants, make it a promising candidate for applications in detergent formulations and peptide synthesis in non-aqueous media. Proteases are earliest-known enzymes and extensively characterized from a variety of sources. However, the S.K. Khare Chemistry Department Indian Institute of Technology, Delhi, Hauz Khas New Delhi 110016 (India) Tel. +91 11 2659 6533, Fax +91 11 2658 1073, E-Mail [email protected]

ability to withstand surfactant and stability in solvent medium is a lesser-known quality of these enzymes. P. aeruginosa has been reported to produce at least four different types of endopeptidases [Kessler et al. 1998; Traidej et al., 2003]: (i) elastase or pseudolysin (LasB) – a 33kDa zinc metallopeptidase [Kessler and Ohman, 2004a; Ogino et al., 2000], (ii) staphylolysin or LasA endopeptidase – a 20-kDa zinc metalloendopeptidase [Kessler et al., 1998; Kessler and Ohman, 2004b], (iii) alkaline protease or aeruginolysin – a 50-kDa zinc metalloendopeptidase [Morihara and Homma, 1985; Okuda et al., 1990; Wallach, 2004] and (iv) protease IV or lysyl endopeptidase – a 26-kDa serine protease encoded by protease IV gene [Caballero et al., 2004; Engel et al., 1998]. Of these, the alkaline protease and elastase (both metalloproteases) have been reported to be the most abundant ones [Parmely, 1993]. Nucleotide sequences of these proteases have been deduced for some strains of P. aeruginosa. However, there is very little information on their stability in solvents and on the structural features that are responsible for this enzyme property. Ogino et al. [2001] attributed the solvent stability to the presence of intermolecular disulfide bonds, between Cys-30 and Cys-58 in case of PST-01 protease. Relative presence of hydrophobic amino acids, which should apparently be critical in interactions with solvents, has not been looked into. The phylogenetic relationships between proteases of solvent-tolerant and nontolerant strains are yet to be established. Such analyses would be vital for understanding the processing, secretion and mechanism of solvent tolerance in proteases as well as for elaborating their applications in non-aqueous enzymology. The present study was undertaken to (i) determine the identity of the protease, (ii) characterize its gene by cloning and sequencing and (iii) undertake in silico analysis to correlate structural features of this protein to its biochemical characteristics. Bioinformatic tools were used to predict the three-dimensional structure of PseA protease and to study the possible reason for the organic solvent stability of the protein. The role of disulfide bridges and a hydrophobic surface are implicated in the organic solvent stability of the protease. Although proteases from different strains of P. aeruginosa have been studied before, this is the first report (to the best of our knowledge) establishing the surface hydrophobicity of solvent-tolerant P. aeruginosa protease. The surface features and the pylogenetic relation of solvent-tolerant P. aeruginosa metalloprotease with other zinc metallopeptidases are the highlights of the present work. Solvent-Stable P. aeruginosa PseA Protease Gene

Results and Discussion

N-Terminal Amino Acid Sequence Analysis The N-terminal amino acid sequence of the purified protease from P. aeruginosa PseA revealed the following 15 amino acid residues: N-Ala-Glu-Ala-Gly-Gly-ProGly-Gly-Asn-Gln-Lys-Ile-Gly-Lys-Tyr-. This sequence was used to search the National Center for Biotechnology Information (NCBI) database to identify the protein by using the Protein-Protein Basic Local Alignment Search Tool program (BLASTP). An exact match with pseudolysin protease gene lasB from P. aeruginosa PST-01 (GenBank accession No. AB029328) [Ogino et al., 2000] was found in the database. It has a 1,494bp open reading frame (ORF). In order to clone the corresponding gene in our isolate, a polymerase chain reaction (PCR)-based approach was used. A cDNA sequence of PST-01 protease was used to design the PCR primers. Amplification and Cloning of the Protease Gene For amplifying the 1,494-bp ORF of the protease gene, the genomic DNA of P. aeruginosa PseA strain was isolated. The 1.69-kb coding region of this gene was PCR amplified by using primers homologous to 5 and 3 untranslated regions (UTR), as described in the Experimental Procedures. The PCR products were resolved on an agarose gel. An approximately 1.7-kb amplified DNA fragment was purified by using a Qiagen gel extraction kit (data not shown). The gel-purified fragment was cloned in pGEM-T Easy vector and transformed in Escherichia coli strain XL-1 Blue MRF’. Amongst approximately 100 transformants of E. coli strain XL-1 Blue MRF’, five white colonies were selected for plasmid isolation. All the clones showed an insert of 1.7 kb along with a 3.0-kb vector band after digestion with EcoRI (data not shown). Since the protease gene was found to have internal SalI and NotI restriction sites, all the plasmids were individually digested with these enzymes. An expected restriction pattern was obtained with both the enzymes confirming the cloning of the protease gene (data not shown). Analysis of the Sequence The plasmids isolated from clones 1 and 2 were used for sequencing. Since the size of the insert was 1.7 kb, four primers (two standard primers, M13 forward and M13 reverse, and two internal primers, IPF and IPR) were used to obtain the complete sequence of both the strands. J Mol Microbiol Biotechnol 2008;15:234–243

235

1 81 161 241 321 401 481 561 641 721 801 881 961 1041 1121 1201 1281 1361 1441 1521 1601 1681

TCTACCCGAAGGACTGATACGGCTGTTCCGATCAGCCCACAAGGCGGCGGTAAGCGTCGGCCGAGTACTTCGGCCTGAAA M K K V S T L D L L F V A I M G V S P A AAACCAGGAGAACTGAACAAGATGAAGAAGGTTTCTACGCTTGACCTGTTGTTCGTTGCGATCATGGGTGTTTCGCCGGC SD A F A A D L I D V S K L P S K A A Q G A P G P V T L Q CGCTTTTGCCGCCGACCTGATCGACGTGTCCAAACTCCCCAGCAAGGCTGCCCAGGGCGCGCCCGGCCCGGTCACCTTGC A A V G A G G A D E L K A I R S T T L P N G K Q V T AAGCCGCGGTCGGCGCCGGCGGTGCCGACGAACTGAAAGCGATCCGCAGCACGACCCTGCCCAACGGCAAGCAGGTCACC R Y E Q F H N G V R V V G E A I T E V K G P G K S V A CGCTACGAGCAATTCCACAACGGCGTACGGGTGGTCGGCGAAGCCATCACCGAAGTCAAGGGTCCCGGCAAGAGCGTGGC A R R S G H F V A N I A A D L P G S T T A A V S A E Q GGCGCGGCGCAGCGGCCATTTCGTCGCCAACATCGCCGCCGACCTGCCGGGCAGCACCACCGCGGCGGTATCCGCCGAGC V L A Q A K S L K A Q G R K T E N D K V E L V I R L AGGTGCTGGCCCAGGCCAAGAGCCTGAAGGCCCAGGGCCGCAAGACCGAGAATGACAAAGTGGAACTGGTGATCCGCCTG G E N N I A Q L V Y N V S Y L I P G E G L S R P H F V GGCGAGAACAACATCGCCCAACTGGTCTACAACGTCTCCTACCTGATTCCCGGCGAGGGACTGTCGCGGCCGCATTTCGT I D A K T G E V L D Q W E G L A H A E A G G P G G N Q CATCGACGCCAAGACCGGTGAAGTGCTCGATCAGTGGGAAGGCCTGGCCCACGCCGAGGCGGGCGGCCCCGGTGGCAACC K I G K Y T Y G S D Y G P L I V N D R C E M D D G N AGAAGATCGGCAAGTACACCTACGGTAGCGACTACGGTCCGCTGATCGTCAACGACCGCTGCGAGATGGACGACGGCAAC V I T V D M N G S T N D S K T T P F R F A C P T N T Y GTCATCACCGTCGACATGAACGGCAGCACCAACGACAGCAAGACCACGCCGTTCCGCTTCGCCTGCCCGACCAACACCTA K Q V N G A Y S P L N D A H F F G G V V F N L Y R D W CAAGCAGGTCAACGGCGCTTATTCGCCACTGAACGACGCGCATTTCTTCGGCGGCGTGGTGTTCAACCTGTACCGGGACT F G T S P L T H K L Y M K V H Y G R S V E N A Y W D GGTTCGGCACCAGCCCGCTGACCCACAAGCTGTACATGAAGGTGCACTACGGGCGCAGCGTGGAGAACGCCTACTGGGAC G T A M L F G D G A T M F Y P L V S L D V A A H E V S GGCACGGCGATGCTCTTCGGCGACGGCGCCACCATGTTCTATCCGCTGGTGTCGCTGGACGTGGCGGCCCACGAGGTCAG H G F T E Q N S G L I Y R G Q S G G M N E A F S D M A CCACGGCTTCACCGAGCAGAACTCCGGGCTGATCTACCGCGGGCAATCCGGCGGAATGAACGAGGCGTTCTCCGACATGG G E A A E F Y M R G K N D F L I G Y D I K K G S G A CCGGCGAGGCCGCCGAGTTCTACATGCGCGGCAAGAACGACTTCCTGATCGGCTACGACATCAAGAAGGGCAGCGGTGCG L R Y M D Q P S R D G R S I D N A S Q Y Y N G I D V H TTGCGCTACATGGACCAGCCCAGCCGCGACGGGCGATCCATCGACAACGCCTCGCAGTACTACAACGGTATCGACGTGCA H S S G V Y N R A F Y L L A N S P G W D T R K A F E V CCACTCCAGCGGCGTGTACAACCGTGCGTTCTACCTGCTGGCCAACTCGCCGGGCTGGGATACCCGCAAGGCCTTCGAGG F V D A N R Y Y W T A T S N Y N S G A C G V I S S A TGTTCGTCGACGCCAACCGCTACTACTGGACCGCCACCAGCAACTACAACAGCGGTGCCTGTGGAGTGATTAGCTCGGCG Q N R N Y S A A D V T R A F S T V G V T C P S A L CAGAACCGCAACTACTCGGCGGCTGACGTCACCCGGGCGTTCAGCACTGTCGGCGTGACCTGCCCGAGCGCGTTGTAAGC TCGGTGGCCCCGGCCGGCACTCCAGGAAGGAATGCCGGCCGGGGCCGCTCAAGCCGTCTTCCGCCAGGAGGACGGCTGCT SR SR TTATGTCGCTT

20 46 73 100 126 153 180 206 233 260 286 313 340 366 393 420 446 473 498

Fig. 1. The nucleotide sequence obtained after the assembly of all sequenced fragments and deduced amino acid sequence (in single letter codes) of lasB gene encoding 53.6-kDa protein from P. aeruginosa PseA. The start codon (ATG) and the stop codon (TAA) are marked by rightward arrow and downward arrow, respectively. Shaded region represents the signal peptide. Boxed region represents the mature protein and rounded box within this region is the conserved Zn-binding motif. The underlined region is the Shine-Dalgarno (SD)-like sequence while dashed lines indicate stacking region (SR). The numbers on the left are nucleotide counts whereas those on the right represent amino acid sequences. Bold amino acids are the N-terminal amino acid sequence of the purified protease determined by the Edman degradation method.

Nucleotide sequence of the cloned DNA fragment of 1,691 bp (fig. 1) contained 64.2% G + C base pairs, which was found to be similar (67.2%) to the genomic DNA of P. aeruginosa [Palleroni, 1984]. The nucleotide sequence of P. aeruginosa PseA genomic DNA exhibited more than 95% homology with the lasB gene of P. aeruginosa PST-01 [Ogino et al., 2000], PAO1 [Bever and Iglewski, 1988], IFO 3455 [Fukushima et al., 1989], IFO 3080, N-10, PA103 [Tanaka et al., 1991] and PA14 (GenBank accession No. NZ_AABQ07000001). The 1,494-bp PseA ORF (between positions 102 and 1,595), showed 98% ho236

J Mol Microbiol Biotechnol 2008;15:234–243

mology to the lasB gene for PST-01 protease. Thus, this protease can be concluded to be of the elastase type of the four protease types of P. aeruginosa, viz. staphylolysin, elastase, alkaline protease and lysine-specific protease. This ORF was found to have a codon bias towards C or G at position 3. Comparison of the nucleotide sequence with the N-terminal protein sequence revealed that the initiation codon ATG was located 589–591 bp upstream of the mature protein-coding sequence (fig. 1). This putative translation initiation site was also preceded by a Shine-Dalgarno-like sequence 11–16 bp upstream of the Gupta /Ray /Kapoor /Khare

ATG codon suggesting that it could be the in vivo translation initiation site. This is in agreement with protease genes of PAO1 and PST-01 reported by Bever and Iglewski [1988] and Ogino et al. [2000], respectively, which also get processed by removal of 197 amino acids from the amino terminal. Leucine was found to be the C-terminal amino acid followed by a termination codon TAA at position 1,596. The existence of this termination codon at this site was further confirmed by the presence of a stacking region at nucleotide positions 1,609–1,644. Leucine as C-terminal amino acid was previously observed by Fukushima et al. [1989] in case of P. aeruginosa IFO 3455. Processing and Secretion of the Protease The 1,494-bp ORF of the PseA gene with a coding capacity of 498 amino acids was calculated to code for a 53.6-kDa protein (fig. 1). The purified protease, however, has previously been reported to be of 35 kDa [Gupta et al., 2005]. We determined the N-terminal amino acid sequence (first 15 amino acids) of purified protease by the Edman degradation method. Comparison of the N-terminal sequence with the predicted ORF revealed that the amino terminus of the mature extracellular purified protease matched with the 198th amino acid of the sequence predicted polypeptide (fig. 1). This indicates that protease is probably synthesized as a 53.6-kDa preproenzyme consisting of 498 amino acids, which is processed to produce a mature enzyme of 33 kDa (301 amino acids). These data suggest that the first 23 amino acid residues of preproenzyme are a putative signal sequence. This sequence has two lysine residues, basic amino acids at positions 2 and 3, and 20 highly hydrophobic amino acids between positions 4 and 23, being in good agreement with the general features of the signal sequences for the periplasmic proteins [von Heijne, 1986]. This 2.4-kDa Nterminal signal peptide would be removed to generate a proenzyme (51.3 kDa) upon passage through the inner membrane into the periplasm, as shown in case of an elastase from P. aeruginosa [Kessler et al., 1998]. From the resultant proenzyme, an 18.2-kDa propeptide of 174 amino acids may be rapidly cleaved off by autoproteolysis to yield a 33-kDa mature enzyme consisting of 301 amino acids (198–498), as described by Kessler and Ohman [2004a]. Kessler et al. [1998] have suggested a role of this propeptide as an inhibitor of the enzyme and its involvement in correct folding and secretion competence. In P. aeruginosa cultures, the mature enzyme (elastase) is secreted as a noncovalent complex with its propeptide, and Solvent-Stable P. aeruginosa PseA Protease Gene

then the propeptide is degraded extracellularly by the mature enzyme itself. Our results confirm that the mature enzyme contains 301 amino acids and has a calculated molecular mass of 33.1 kDa, which agrees well with the molecular mass of the purified protease (35 kDa) measured by SDS-PAGE [Gupta et al., 2005]. Phylogenetic Analysis of the PseA Protease The nucleotide and amino acid sequences of the proteases produced by various isolates of P. aeruginosa and other bacteria have been reported [Black et al., 1990; Cascon et al., 2000; David et al., 1992; Hase and Finkelstein, 1991; Ogino et al., 2000; Rust et al., 1996]. A search of databases by the BLASTP revealed that the P. aeruginosa PseA protease was most similar in sequence to the zinc metalloprotease (elastase) secreted from various P. aeruginosa strains as PST-01 [Ogino et al., 2000], PA14 and then with that of PA103, IFO 3455 and N-10 [Tanaka et al., 1991]. Analysis of the PseA protease (498 amino acids) using PROSITE (Swiss Institute of Bioinformatics) revealed a Zn-binding motif (VAAHEVSHGF) starting at amino acid position 334 (fig. 1). It may indicate that PseA belonged to a group of neutral zinc metallopeptidases [Hase and Finkelstein, 1993; Rawlings and Barrett, 2004]. PseA also showed a high degree of similarity with the primary structure of pseudolysin (elastase from P. aeruginosa). P. aeruginosa PST-01 pseudolysin was characterized by Xray crystallography and shown to contain three zinc ligands at His-140, His-144 and Glu-164. The ligands of calcium ions are carboxyl groups of Asp-136, Glu-172, Glu-175 and Asp-183, carbonyl groups of Leu-185, and one water molecule. The residues of the active site are Glu-141, Tyr-155 and His-223 [Ogino et al., 2000]. The regions from amino acids 332–343 (135–146 of mature enzyme) of PseA protease show 100% similarity to the zinc-binding or active site regions of PST-01 elastase (pseudolysin) and other P. aeruginosa elastases (fig. 2). On the basis of these similarities, His-140, His-144 and Glu-164 could also perform the function of zinc ligands and carboxyl groups of Asp-136, Glu-172, Glu-175 and Asp-183, and carbonyl groups of Leu-185 could engage calcium ions in the mature PseA. Similarly, Glu-141, Tyr155 and His’s-223, which form the active center in pseudolysin, could as well perform the same function in case of PseA. A phylogenetic tree representing known Zn metalloproteases from a number of bacteria demonstrated that P. aeruginosa PseA protease shares homology with sevJ Mol Microbiol Biotechnol 2008;15:234–243

237

LDVAAHEVSHGFTEQNSGLIYRGQSGGMNEAFSDMAGEAAEFYMRGKND----FLIG LDVAAHEVSHGFTEQNSGLIYRGQSGGMNEAFSDMAGEAAEFYMRGKND----FLIG LDVAAHEVSHGFTEQNSGLIYRGQSGGMNEAFSDMAGEAAEFYMRGKND----FLIG LDVAAHEVSHGFTEQNSGLIYRGQSGGMNEAFSDMAGEAAEFYMRGKND----FLIG LDVAAHEVSHGFTEQNSGLIYRGQSGGMNEAFSDMAGEAAEFYMRGKND----FLIG LDVAAHEVSHGFTEQNSGLIYRGQSGGMNEAFSDMAGEAAEFYMRGKND----FLIG NDVVAHELTHGVTQETANLEYKDQSGALNESFSDVFGYFVD----DED-----FLMG NDVVAHEITHGVTQQTANLVYRSQSGALNESFSDVFGYFID----DED-----FLMG INVSAHEVSHGFTEQNSGLVYRDMSGGINEAFSDIAGEAAEFYMRGNVD----YIVG LDVVAHELTHAVTEYTAGLVYQNESGAINEAVSDIMGTVAE----YSVGSNFDWLVG IDVIAHELTHGITQHEAGLIYYGEPGALNESFSDVFGALVKQRVKNQKAEEADWLIG LDVAAHEVSHGFTEQNSGLVYSGQSGGINEAFSDMAGEAAENYMKGSND----WLVG IDVIGHELTHGITQYEAGLQYYGEPGALNESFSDVFGSLVKQKSKNQTAQEADWLIG LGVGGHEVSHGFTEQHSGLEYFGQSGGMNESFSDMAAQAAEYYSVGKNS----WQIG IDVVGHELAHGVTESEAGLIYFQQAGALNESLSDVFGSLVKQFHLKQTADKADWLIG LDVVAHEITHAVTERTAGLVYEYQPGALNESFSDVLGNLVE----NKND--PKWLLG LDVIAHELAHGITEHEAGLIYFRQSGALNESLSDVFGSMVKQYHLGQTAEQADWFIG LDVTAHEMTHGVTQETANLIYENQPGALNESFSDVFGYFND----TED-----WDIG IDVVAHELTHAVTDYTAGLIYQNESGAINEAISDIFGTLVE----FYANKNPDWEIG *Ζ *

*

* Ζ

*

Fig. 2. Comparison of the high homology regions in the primary

structures of zinc metalloproteases from various bacterial strains. The numbers written on both sides of the lines indicate the amino acid position from the start of translation. Hyphens indicate gaps introduced so that the maximum matching may be obtained. The

**

**

*

sa-3080 P. aerugino K sa(D) gino eru P. a D1 (D) FR a-

os gin ru ae P. ) (C ps in

(A

E.r wi ( nia Serr F) ca a tia p rot (F) r ote ama ov or Nosto c ula a c pun (F) c tif or ns me

238

)

Anabaena variabilis (F)


Sta (F phy ) lo

co cc us St. au epid re erm idis us (F)

ly tic us oproteo p. B. therm ms eriu (F) ac t b o u (F) . p Ex ig ilis ss ubt s (F) ce lus my o cil a n B ti ac mo er h T ) (F

eral Zn metalloproteases from P. aeruginosa strains [Rust et al., 1996], Aeromonas spp. [Cascon et al., 2000], Legionella spp. [Black et al., 1990] and Vibrio spp. [David et al., 1992; Hase and Finkelstein, 1991] confirming it to be a member of this family (fig. 3).

*

*Ζ

**

433 433 433 433 433 419 449 448 440 439 294 417 294 476 297 487 297 462 476

* * *

v rio cte

us or

ba io ibr v o ell B) . Bd ) p. ( r sp (B illus s ac te erob a n Bac moa Ther (B)

Legionella pneumophila (G) V ibrio c hloer ae (G A er ) omo ( nas P. G) hy d aer r op ug hila ino saela s ta se (G )

) (G ) O1 (G A -P 01 ) sa T(G no PS 455 gi sa3 ru ino IFO ae ug saP. aer ino ) P. erug s eA (G P. a eruginosa-P P. a PA 14 (G) P. aeruginosa-

the known proteases from various strains. A–G = Trypsin-like serine protease, subtilisin-like serine protease, LasA endopeptidase, alkaline protease, protease-IV (lysil endopeptidase), zinc metalloprotease (elastase close to thermolysin) and zincmetalloprotease (elastase close to pseudolysin), respectively. These data were generated by CLUSTAL X software using the neighbor-joining method. Trypsin has been used as the outgroup.

D---VHHSSGVYNRAFYLL D---VHHSSGVYNRAFYLL D---VHHSSGVYNRAFYLL D---VHHSSGVYNRAFYLL D---VHHSSGVYNRAFYLL D---VHHSSGVYNRAFYLL DNGGVHTNSGIPNKAAYNV DNGGVHTNSGIPNKAAYNT D---VHHSSGVFNRAFYLL DNGGVHTNSGIVNKAAYLL DNGGVHINSGIPNRAFYLA D---VHHSSGVYNRAFYLL DNGGVHINSGIPNHAFYLA D---VHYSSGVYNHLFYIL DNGGVHLNSGIPNRAFYLA DWGGVHINSGIPNKAFYNF DNGGVHLNSGIPNRAFYLT DYGGVHTNSGIPNKAAYNT DNGGVHINSGIINKAAYLI

bold amino acids indicate identity with the P. aeruginosa PseA protease. The boxed amino acids show the residues serving as zinc ligands. Z/* = Residues constructing active centers of pseudolysin and other identical residues in all the zinc metalloproteases including the PseA protease, respectively.

Tr y

Fig. 3. A phylogenic relationship amongst

418 418 418 418 418 404 431 430 425 421 276 402 276 461 279 469 279 444 458

(A) P. a e r u Bo (A) ginos a-P (A rd A1 ) ete 4 lla pe rt us si s

*

384 384 384 384 384 370 394 393 390 384 230 368 230 424 213 428 213 406 421

P. aer ugino sa-C

332 332 332 332 332 318 347 346 338 332 174 316 174 372 157 378 157 359 369

P. aerugino sa-PA103 (E)

P. aeruginosa PseA (DQ350610) P. aeruginosa PST-01 (AB029328) P. aeruginosa PAO1 (AE004791) P. aeruginosa - elastase (M19472) P. aeruginosa IFO3455 (M24531) P. aeruginosa PA14 (NZ_AABQ0700000) Staphylococcus aureus (NC_002953) S. epidermis (NC_002976) Vibrio chloerae (M59466) Exiguobacterium sp. (NZ_AADW01000031) Anabaena variabilis (NZ_AAEA01000001) Aeromonas hydrophila (AF193422) Nostoc punctiforme (NZ_AAAY02000092) Legionella pneumophila (A35265) Serratia proteamaculans (AY819662) Thermoactinomyces sp. (AY280367) Erwinia carotovora (NC_004547) Bacillus subtilis (NC_000964) B. thermoproteolyticus (A20191)

0.1

Structural Aspects of PseA Protease Contributing to Its Solvent-Tolerant Nature The hydrophilicity plot of the deduced amino acid sequence of P. aeruginosa PseA protease was determined (fig. 4) according to the method of Kyte and Doolittle Gupta /Ray /Kapoor /Khare

4.5

Fig. 4. Hydrophilicity plot for PseA prote-

ase according to Kyte and Doolittle [1982]: the profile of the hydropathic index (ordinate) of the amino acid sequence as a function of the residue number (abscissa), which starts from the initiating methionine. Signal peptide, propeptide and mature proteins are marked, respectively. On the plot, a positive peak indicates a probability that the corresponding polypeptide fragment is hydrophilic (a negative peak indicating a probable hydrophobic segment).

0

–4.5 Signal peptide 50 aa

Propeptide

150 aa

Zn-binding region 250 aa

350 aa

Mature peptide

450 aa

Fig. 5. A comparison of 3D structure features of solvent-tolerant and solvent-labile proteases. a 3D structure of the P. aeruginosa

stick models of P. aeruginosa PseA protease, thermolysin (PDB code 1LNF) and A. niger protease (PDB code 1IZD), respectively. The hydrophobic patches present on the protein surface have been given yellow surface. Disulfide bridges are marked by S-S and shown blue. N/C = N-terminal and C-terminal amino acids of proteins, respectively.

Solvent-Stable P. aeruginosa PseA Protease Gene


PseA protease. In the solid ribbon diagram, -helices, -strands, active centers and disulfide bonds are colored red, blue, purple and yellow, respectively. Residues of Zn ligands and that for the Ca atom are shown in green and pink, respectively. b–d Ball and

239

[1982]. It shows that the sequence had hydrophobic inclination. Since this enzyme has previously been shown to be extremely stable in various organic solvents [Gupta et al., 2005; Gupta and Khare, 2006], we believe that the presence of hydrophobic amino acids could contribute towards this property of P. aeruginosa PseA protease. In order to correlate the structural features responsible for solvent-stable function, the amino acid sequence of PseA protease was used to determine its 3D structure. The P. aeruginosa elastase found in the Protein Data Bank (PDB, code 1EZM) was considered as template. The BLAST pairwise alignment between PseA (198- to 495amino-acid region) and 1EZM sequences showed amino acid identities of 98%. The amino acid sequence of PseA protease was submitted to different servers, namely the HMMSTR/ROSETTA server [Bystroff and Shao, 2002], Swiss-Model [Schwede et al., 2003], 3D-JIGSAW [Bates et al., 2001] and CPHmodels 2.0 Server [Lund et al., 2002], to obtain the 3D structural model. The obtained models were superimposed with the template for verification. We found all the models to be highly similar with the template. Figure 5a shows the model with secondary structure topology in evidence. The molecule was found to have 36% -helices, 22% strands and 41% coils, and two disulfide bonds (between Cys-227 and Cys-255 and between Cys467 and Cys-494). In this protein, 31.8% of amino acid residues are hydrophobic, of which 66.3% were localized on the surface of this protein, as shown in yellow (fig. 5b). The two solvent-labile proteases, namely thermolysin and Aspergillus niger protease (PDB code 1LNF and 1IZD, respectively), taken for comparison, contained 30.4 and 30% hydrophobic residues, respectively. Incidentally, of these only 56.2 and 44.3% could be localized on the surface of these proteins, respectively (fig. 5c, d). Ogino et al. [2001] have reported that the presence of disulfide bonds played an important role in the organic solvent stability of the PST-01 protease. Based on the above observations, we hypothesize that along with disulfide bridges, accumulation of hydrophobic amino acids on the surface of the polypeptide might impart solvent tolerance to proteins/enzymes in general in case of solvent-tolerant microbes. The role of disulfide bonds in imparting solvent stability is further supported by chemical modification experiments (table 1). Protease activity was completely inhibited (100% inhibition) by 5 mM dithiothreitol (DTT), suggesting that enzyme must contain S-S bonds as part of its monomeric structure [Kessler et al., 1997]. This effect is consistent with its organic solvent and thermal stability, 240


Table 1. Chemical modification of cysteine and cystine in PseA

Reagent None (control) p-Chloromercuribenzoic acid Iodoacetic acid 1,4-DTT Glutathione

Concentration mM

Activity %

5 5 5 5

100 95 93 0 90

Purified PseA protease was preincubated with various reagents at 25° C for 10 min and residual protease activity determined. The activity in the absence of reagent (control) was taken as 100%. The experiment was carried out in duplicate and the difference between individual sets of readings was