Comprehensive inhibitor profiling of the Proteus mirabilis ...

Biochimie 93 (2011) 1824e1827

Contents lists available at ScienceDirect

Biochimie journal homepage: www.elsevier.com/locate/biochi

Research paper

Comprehensive inhibitor profiling of the Proteus mirabilis metalloprotease virulence factor ZapA (mirabilysin) Louise Carson a, George R. Cathcart a, Christopher J. Scott a, Morley D. Hollenberg b, Brian Walker a, Howard Ceri c, Brendan F. Gilmore a, * a b c

School of Pharmacy, Queens University Belfast, Medical Biology Centre, 97 Lisburn Road, Belfast BT9 7BL, UK Department of Physiology & Pharmacology, University of Calgary, Calgary, AB, Canada The Biofilm Research Group, Department of Biological Sciences, University of Calgary, 2500 University Drive NW, Calgary, AB T2N 1N4, Canada

a r t i c l e i n f o

a b s t r a c t

Article history: Received 25 February 2011 Accepted 24 June 2011 Available online 6 July 2011

In this study we report for the first time the comprehensive inhibitor profiling of the Proteus mirabilis metalloprotease virulence factor ZapA (mirabilysin) using a 160 compound focused library of N-alpha mercaptoamide dipeptides, in order to map the S01 and S02 binding site preferences of this important enzyme. This study has revealed a preference for the aromatic residues tyrosine and tryptophan in P01 and aliphatic residues in P02 . From this library, six compounds were identified which exhibited sub- to low-micromolar Ki values. The most potent inactivator, SHeCO2eYeVeNH2 was capable of preventing ZapA-mediated hydrolysis of heat-denatured IgA, indicating that these inhibitors may be capable of protecting host proteins against ZapA during colonisation and infection. Ó 2011 Elsevier Masson SAS. All rights reserved.

Keywords: Mirabilysin ZapA Inhibitor N-alpha mercaptoamide Protease Virulence

1. Introduction Extracellular secreted metalloproteases are produced by a number of Gram-negative bacterial pathogens, and are known to play important roles in a number of key processes central to successful colonisation and infection of the host [1e3]. Proteus mirabilis, an opportunistic urinary tract pathogen, produces a 54-kDa secreted metalloprotease, mirabilysin, or ZapA [4], belonging to protease family M10 (peptidase M10.057) with a high degree of homology to the serralysin subgroup of this family [5,6]. Initially, ZapA was reported to degrade host immunoglobulins, primarily IgA [4,7,8], however subsequent studies have concluded that IgA is not a natural substrate of this protease, rather, ZapA possesses a broad specificity and is capable of degrading a wide range of host defensive proteins [9,10]. ZapA is specifically upregulated during swarmer cell differentiation [11], another critical virulence process in the pathogenesis of this organism, facilitating migration over surfaces [12]. Recently we have characterised the importance of ZapA as a virulence factor in a rat model of P. mirabilis-induced acute and chronic prostatitis [13]. Inhibition of ZapA may therefore represent a novel anti-virulence

target for therapeutic intervention in P. mirabilis infections of the prostate. We have previously reported the solid-phase synthesis of N-alpha mercaptoamide dipeptides [14], and the profiling of such inhibitors against an extracellular metalloprotease and virulence factor of Pseudomonas aeruginosa, LasB, also known as Pseudomonas elastase [15]. In the present study, we have used an established library of N-alpha mercaptoamide dipeptides to identify effective inhibitors of P. mirabilis ZapA. This inhibitor screen has provided vital information for the design of improved inhibitors and substrates of this metalloprotease virulence factor. 2. Experimental 2.1. Bacterial strains P. mirabilis BB2000, previously characterised [6] was a kind gift from Prof. Robert Belas, Center for Marine Biotechnology, University of Maryland, Baltimore, MD, USA. 2.2. Isolation, purification and identification of ZapA protein

* Corresponding author. Tel.: þ44 (0) 28 90 972 047; fax: þ44 (0) 28 90 247 794. E-mail address: [email protected] (B.F. Gilmore). 0300-9084/$ e see front matter Ó 2011 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.biochi.2011.06.030

Isolation and purification of ZapA were achieved through modification of the method previously described by Loomes et al. (1992)

L. Carson et al. / Biochimie 93 (2011) 1824e1827

[16]. Essentially, P. mirabilis BB2000 was cultured for 24 h in LB broth at 37 C after which 200 ml of the culture medium was centrifuged at 1700 g at 4 C for 1 h. Ammonium sulphate (0.1 M) was added to the decanted supernatants, which were then passed through a 0.45 mM pore filter before purification on HiTrap Phenyl HP columns (GE Healthcare), using the AktaPrime Plus FPLC apparatus (GE Healthcare). A 3 ml column volume was equilibrated with 50 mM sodium phosphate, 0.1 M ammonium sulphate, pH 7.0, before sample loading at a flow rate of 1 ml/min. Columns were then washed with the same buffer, and elution achieved with 50 mM TriseHCl, pH 11.0. Fractions containing purified protein were pooled, dialysed overnight at 4 C against 50 mM TriseHCl, pH 8.0, and stored at 80 C until required. Proteolytic activity was confirmed by an azo-casein assay [16]. The release of azo dye from azocasein was determined at 440 nm using the ELx808 Absorbance Microplate Reader (BioTek Instruments Inc.). One unit of protease was defined as that capable of hydrolysing 1 mg of azocasein in 60 min at 37 C. Protein content was quantified using the BCA Protein Assay (Pierce). Following SDS-PAGE with colloidal Coomassie blue staining, an in-gel tryptic digest was performed on the single protein band observed at 54 kDa. Using ElectroSpray Ionisation Ion-Trap Mass Spectrometry (ThermoFinnigan LCQ Deca), in nanospray mode, the sample was analysed and MS/MS data collected by Xcalibur. MS/MS ion search was performed by Mascot (www. matrixscience.com) [17], using the database NCBInr.

1825

2.3. Synthesis of N-alpha mercaptoamide dipeptide inhibitors The synthesis of the focused library of N-alpha mercaptoamide dipeptide inhibitors was carried out as described previously [15] using a CEM LibertyÔ microwave assisted peptide synthesiser and standard Fmoc solid-phase peptide synthesis protocols on Rink amide resin. The identity and purity of the synthesised inhibitors were confirmed by electrospray mass spectrometry and reverse phase (C 18) HPLC analysis. The general structure of the synthesized inhibitors is shown in Fig. 1A. 2.4. Fluorimetric inhibition assays Each dipeptide inhibitor was screened for activity against purified ZapA using a microtitre-based fluorimetric assay. Hydrolysis of the fluorogenic substrate Aminobenzoyl-Ala-Gly-Leu-Ala-pNitro-Benzyl-Amide (Peptides International) by ZapA was carried out in assay buffer containing 50 mM TriseHCl, 2.5 mM CaCl2, pH 8.0. Changes in fluorescence were monitored by a BMG FLUOstar OPTIMA fluorescence microtitre plate reader. Inhibitors were initially screened at a 25 mM concentration, and those displaying fractional inhibition greater than 0.8 were selected for further investigation over an extended concentration range, in order to determine accurate kinetic parameters (Ki). Ki values were calculated using the Morrison equation for tight binding inhibitors, on GraphPad Prism software (version 5).

Fig. 1. (A) General structure of the N-alpha mercaptoamide dipeptide inhibitor library evaluated in this study. (B) Relative potency of individual compounds displayed as the fractional inhibition of ZapA activity at an inhibitor concentration of 25 mM. The library has been arranged according to the dipeptide amino acid sequences, and colour-coded according to the P01 residue.

1826


2.5. Potential of N-alpha mercaptoamide dipeptide inhibitors to protect biological substrates against ZapA-mediated hydrolysis IgA from human colostrum (SigmaeAldrich) was used either in its native conformation, or denatured by heating to 80 C for 15 min. Assays were carried out at 37 C with gentle agitation in 200 mL buffer (50 mM TriseHCl, 2.5 mM CaCl2, pH 8.0). Controls consisted of 20 mg of IgA protein incubated alone, or with 4 mg of ZapA. Both controls also contained 5 mL DMF (inhibitor vehicle). In order to test the ability of inhibitor to preserve the integrity of IgA, 20 mg of IgA was incubated with 4 mg ZapA and 50 mM SHeCH2eCOeYeVeNH2. Aliquots were withdrawn at regular time intervals, heat inactivated at 100 C for 10 min and subjected to SDS-PAGE and Coomassie blue staining. 3. Results 3.1. Purification and identification of ZapA A 150-fold purification of ZapA was achieved through a one-step purification procedure directly from culture supernatants of P. mirabilis BB2000. Elution from the column resulted in a single peak of UV-absorbing material (280 nm). A typical yield of ZapA protein was 4 mg/l culture, while proteolytic activity was in the region of 40e60 U/mg, as determined by azo-casein assay. Following in-gel tryptic digestion of the purified protein, recovered peptides were analysed by MS/MS and 6 peptides were identified as being derived by the action of trypsin on P. mirabilis ZapA (Accession number AAC33449), these are highlighted in red in Fig. 2B. 3.2. Fluorimetric inhibition assays A number of inhibitor sequences have been identified which display inhibitory activity against ZapA (Fig. 1B). Those sequences displaying fractional inhibition of 0.8 or greater at 25 mM were selected for further characterisation. As shown in Table 1, a number of these inhibitors display Ki values in the low micromolar range, with SHeCH2eCOeYeVeNH2 being the most potent of the series, having a Ki value of 0.75 mM. It is clear from the data presented that in the P01 subsite, there is a strong preference for the large aromatic residues of tryptophan (Trp) and tyrosine (Tyr). 3.3. ZapA-mediated degradation of IgA and preservation of substrate by inhibitor Although it has been reported that ZapA hydrolyses host immunoglobulins, in particular IgA [4,7,8], we observed negligible proteolytic activity of ZapA towards IgA in its native conformation, even over extended periods of time of up to 60 h (Fig. 2C). However, IgA which had been heat inactivated was rapidly hydrolysed (Fig. 2D). The ability of the most potent inhibitor identified in this study, SHeCH2eCOeYeVeNH2 at 50 mM concentration, to protect biological substrates from ZapA hydrolysis in vitro is demonstrated by its ability to prevent the ZapA-mediated hydrolysis of heatdenatured IgA (Fig. 2D, lane 4), highlighting the potential therapeutic application of ZapA inhibitors as an anti-virulence strategy in P. mirabilis infection. 4. Discussion In keeping with the findings of our current study, several other groups have demonstrated that ZapA exhibits negligible proteolytic degradation of native human IgA [9,10]. Once denatured, however, IgA is degraded rapidly by ZapA (Fig. 2C and D). Denatured IgA is

Fig. 2. (A) SDS-PAGE analysis of ZapA as purified using FPLC. Lane 1 shows molecular weight marker (SeeBlueÒ Plus 2 pre-stained standard, Invitrogen). Lanes 2e5 show ZapA containing fractions, eluted from PhenylHP columns showing a single band at 54 kDa. (B) In-gel tryptic digestion of partially purified protein was performed and recovered peptides subjected to MS/MS. MS/MS ion searches using MASCOT identified 6 peptides, highlighted in red, from P. mirabilis ZapA (Accession number AAC33449). (C) SDS-PAGE analysis displaying the inability of ZapA to hydrolyse IgA in its native conformation. Lane 1 shows the molecular weight marker (SeeBlueÒ Plus 2 pre-stained standard, Invitrogen). Lane 2 shows 2 mg IgA alone (negative control). Lane 3 shows 2 mg IgA with 400 mg of ZapA. Lane 4 shows 2 mg IgA with 400 mg of ZapA and 50 mM SHeCH2eCOeYeVeNH2. All samples and control were incubated at 37 C for 60 h. (D) SDS-PAGE analysis displaying the hydrolysis of heat-denatured IgA by ZapA. Lane 1 shows the molecular weight marker (SeeBlueÒ Plus 2 pre-stained standard, Invitrogen). Lane 2 shows 2 mg IgA alone (negative control). Lane 3 shows 2 mg IgA with 400 mg of ZapA. Lane 4 shows 2 mg IgA with 400 mg of ZapA and 50 mM SHeCH2eCOeYeVeNH2. All samples and controls were incubated at 37 C for 15 min. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

unlikely to possess any appreciable immune function, and so ZapAmediated hydrolysis of denatured IgA would not be expected to contribute to pathogenicity of P. mirabilis. Although the specific mechanism of ZapA in pathogenesis is yet to be elucidated, it is Table 1 Ki values (mM) for the most potent inhibitors of ZapA identified from the N-alpha mercaptoamide dipeptide library screen. P01

P02

Ki (mM)

Tyr Trp Trp Tyr Trp Met Tyr

Val Arg Val Trp Lys Lys Lys

0.75 2.09 2.53 3.07 3.97 5.58 13.15


known to be critical for colonisation and establishment of infection [13]. Therefore, the effective inactivation of ZapA may provide protection to the cognate substrates in vivo and may represent a putative anti-infective/anti-virulence strategy. Such anti-virulence strategies are non-destructive to the infecting pathogen, and as such exert much less selective pressure than conventional antimicrobial therapies, possibly reducing the likelihood of resistance [15,18,19]. We have previously reported the inhibitor profiling of P. elastase, LasB, using N-alpha mercaptoamide dipeptides [17]. In this present study we report for the first time the inhibitor profiling of the P. mirabilis metalloprotease, ZapA, using the same focused library of N-alpha mercaptoamide template based inhibitors. Inhibitors of ZapA with Ki values in the low micromolar range are described, with the most potent inhibitor, SHeCH2eCOeYeVeNH2 possessing a Ki of 0.75 mM. Interestingly, this compound was the most potent inactivator of LasB, having a Ki of 0.77 mM [17]. The screening of this focused library has also revealed important information regarding the characteristics of the S01 and S02 subsites of the ZapA active site. A common feature of the most potent inhibitors identified is the presence of Tyr or Trp in the P01 position of the dipeptide mercaptoamide inhibitor. This suggests that the S01 binding pocket of ZapA preferentially accommodates large aromatic residues. Interestingly, however, inhibitors with Phe in the P01 position did not exhibit good inhibitory activity. This may be explained by the fact that, unlike Phe, both Tyr and Trp are capable of participating in hydrogen bonding within S01 , and suggests that while the S01 pocket of the ZapA preferentially accommodates large hydrophobic aromatic groups, the ability of this group to form hydrogen bonds is a driving determinant of specificity. While none of the amino acids were favoured in the P02 position, all those appearing in the most potent inhibitors were the aliphatic residues of valine, arginine and lysine. In keeping with our previous observations with LasB, the P01 residue appears to be the primary determinant of specificity of these compounds. 5. Conclusion It is interesting that many of the trends observed in the screening of this inhibitor library against ZapA metalloprotease of P. mirabilis are mirrored by the trends observed for the LasB metalloprotease of P. aeruginosa, i.e. the preference for large aromatic residues in P01 and aliphatic residues in P02 . Although these two enzymes are from distinct families of metalloproteases (LasB from the Thermolysin family and ZapA from the Serralysin family), both are important virulence factors of their respective pathogens. Since both proteases share similar binding specificities, it is possible that development of metalloprotease inhibitors for specific antimicrobial use would be an effective anti-virulence strategy in a number of bacterial infections. To assess further the potential of protease inhibitors as a useful class of antimicrobial drugs, more work is needed to understand in depth the mechanism by which bacterial

1827

proteases facilitate virulence of their respective pathogens, alongside in vivo studies to investigate the therapeutic application of bacterial protease inhibitors. Acknowledgements The authors gratefully acknowledge Adrienne Healy, QUB Proteomics Facility for assistance in MS/MS protein identification and the support of the Department for Employment and Learning (DEL). References [1] S.I. Miyoshi, S. Shinoda, Bacterial metalloproteases as the toxic factor in infection, J. Toxicol. Toxicon. Rev. 16 (1997) 177e194. [2] C.T. Supuran, A. Scozzafava, B.W. Clare, Bacterial protease inhibitors, Med. Res. Rev. 22 (2002) 329e372. [3] J. Potempa, R.N. Pike, Corruption of innate immunity by bacterial proteases, J. Innate Immun. 1 (2009) 70e87. [4] L.M. Loomes, B.W. Senior, M.A. Kerr, A proteolytic enzyme secreted by Proteus mirabilis degrades immunoglobulins of the immunoglobulin A1 (IgA1), IgA2, and IgG isotypes, Infect. Immun. 58 (1990) 1979e1985. [5] R. Belas, Mirabilysin. in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, second ed. Elsevier Academic Press, London, 2004, pp. 579e581. [6] C. Wassif, D. Cheek, R. Belas, Molecular analysis of a metalloprotease from Proteus mirabilis, J. Bacteriol. 177 (1995) 5790e5798. [7] B.W. Senior, L.M. Loomes, M.A. Kerr, The production and activity in vivo of Proteus mirabilis IgA protease in infections of the urinary tract, J. Med. Microbiol. 35 (1991) 203e207. [8] M.A. Kerr, L.M. Loomes, B.W. Senior, Cleavage of IgG and IgA in vitro and in vivo by the urinary tract pathogen Proteus mirabilis, Adv. Exp. Med. Biol. 371A (1995) 609e611. [9] R. Belas, J. Manos, R. Savanasuthi, Proteus mirabilis ZapA metalloprotease degrades a broad spectrum of substrates, including antimicrobial peptides, Infect. Immun. 72 (2004) 5159e5167. [10] M.A.F. Anéas, F.C.V. Portaro, I. Lebrun, L. Juliano, M.S. Palma, B.L. Fernandes, ZapA, a possible virulence factor from Proteus mirabilis exhibits broad protease substrate specificity, Braz. J. Med. Biol. Res. 34 (2001) 1397e1403. [11] K.E. Walker, S. Moghaddame-Jafari, C.V. Lockatell, D. Johnson, R. Belas, ZapA, the IgA-degrading metalloprotease of Proteus mirabilis, is a virulence factor expressed specifically in swarmer cells, Mol. Microbiol. 32 (1999) 825e836. [12] H. Mobley, R. Belas, Swarming and pathogenicity of Proteus mirabilis in the urinary tract, Trends Microbiol. 3 (1995) 280e284. [13] V. Phan, R. Belas, B.F. Gilmore, H. Ceri, ZapA, a virulence factor on a rat model of Proteus mirabilis-induced acute and chronic prostatitis, Infect. Immun. 76 (2008) 4859e4864. [14] J.F. Lynas, S.L. Martin, B. Walker, A.D. Baxter, J. Bird, R. Bhogal, J.G. Montana, D.A. Owen, Solid-phase synthesis and biological screening of N-alphamercaptoamide template-based matrix metalloprotease inhibitors, Comb. Chem. High Throughput Screen. 3 (2000) 37e41. [15] G.R. Cathcart, B.F. Gilmore, B. Greer, P. Harriott, B. Walker, Inhibitor profiling of the Pseudomonas aeruginosa virulence factor lasB using N-alpha mercatoamide template-based inhibitors, Bioorg. Med. Chem. Lett. 19 (2009) 6230e6232. [16] L.M. Loomes, B.W. Senior, M.A. Kerr, Proteinases of proteus spp.: purification, properties, and detection n urine of infected patients, Infect. Immun. 60 (1992) 2267e2273. [17] D.N. Perkins, D.J. Pappin, D.M. Creasy, J.S. Cottrell, Probability-based protein identification by searching sequence databases using mass spectrometry data, Electrophoresis 20 (1999) 3551e3567. [18] S. Crunkhorn, Antibacterial drugs: new paths to beating bacteria, Nat. Rev. Drug Disc. 7 (2008) 891. [19] T.B. Rasmussen, M. Givskov, Quorum sensing inhibitors: a bargain of effects, Microbiology 152 (2006) 895e904.