PauA: a novel plasminogen activator from Streptococcus uberis

FEMS Microbiology Letters 178 (1999) 27^33

PauA: a novel plasminogen activator from Streptococcus uberis E.L. Rosey b , R.A. Lincoln a , P.N. Ward a , R.J. Yancey Jr b , J.A. Leigh a; * b

a Institute for Animal Health Compton Laboratory, Compton, Newbury, Berks RG20 7NN, UK Central Research Division, P¢zer, Animal Health Biological Discovery, Eastern Point Road, Groton, CT 06340, USA

Received 17 February 1999; received in revised form 27 May 1999; accepted 20 June 1999

Abstract Chromosomal DNA from two geographically distinct isolates of Streptococcus uberis was used to clone the plasminogen activator in an active form in Escherichia coli. The cloned fragments from each strain contained four potential open reading frames (ORFs). That for the plasminogen activator encoded a protein of 286 amino acids (33.4 kDa) which is cleaved between residues 25 and 26 during secretion by S. uberis. The amino acid sequence of the mature protein showed only weak homology (23.5^28%) to streptokinase. The plasminogen activator gene, pauA, in S. uberis was located between two ORFs with high homology to the DNA mismatch repair genes, hexA and hexB, and not on a DNA fragment between the genes encoding an ATP binding cassette transporter protein (abc) and a protein involved in the formation and degradation of guanosine polyphosphates (rel) as is the case for streptokinase in other streptococci. ß 1999 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. Keywords : Plasminogen activator; Cloning; Streptokinase ; Mastitis ; hexA; hexB; Streptococcus uberis

1. Introduction Streptococcus uberis is a common cause of infection and in£ammatory disease in the bovine udder. Mastitis has been recognized as a serious threat to the welfare of the dairy cow and has a severe economic impact on the dairy industry. Infection by S. uberis is not controlled by currently recommended dairy husbandry procedures and alternate methods of control such as vaccination are being sought. The success of such an approach has been hampered by a lack of knowledge of the crucial virulence determinants and their in£uence on pathogenesis (for

* Corresponding author. Fax: +44 (16) 355 77299.

review see [1]). The acquisition of such knowledge will be greatly assisted by the application of molecular and genetic technologies. S. uberis has been shown to speci¢cally activate bovine and ovine plasminogen [2] and this is mediated by an extracellular protein, plasminogen activator uberis (PauA), which has a molecular mass of around 30 kDa [3]. The role of PauA in the pathogenesis of intramammary infection and mastitis has not been fully elucidated. In vitro, PauA has been shown to mediate the acquisition of plasmin at the bacterial surface following growth in media containing plasminogen [4] and peptides derived from bovine casein via the action of plasmin have been shown to satisfy some of the amino acid requirements of this bacterium [5]. Plasminogen is present

0378-1097 / 99 / $20.00 ß 1999 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 1 0 9 7 ( 9 9 ) 0 0 3 3 5 - 3

FEMSLE 8925 11-8-99

28

E.L. Rosey et al. / FEMS Microbiology Letters 178 (1999) 27^33

in bovine milk at a concentration of around 1.5 Wg ml31 [6] and it has been postulated that activation of plasminogen by PauA may facilitate early colonization of the lactating gland by promoting the release of nutrients [2] which are likely to be restricted in milk [7]. In support of this hypothesis, vaccination with PauA, which induced a neutralizing antibody response, reduced the rate of colonization and decreased the incidence of disease following experimental challenge [8]. Many pathogenic streptococci have been shown to activate plasminogen in a species speci¢c fashion [9] through the action of secreted proteins. Streptokinases, the best studied of these, are a group of extracellular proteins with a molecular mass of around 47 kDa and sequence homology which varies between 80 and 98% [10]. Over much of the sequence, the homology is very high and variation is largely con¢ned to a single domain (V1) located between residues 147^218 of the secreted protein [11]. Another plasminogen activating protein (Esk) puri¢ed from S. equisimilis has a molecular mass of 49 kDa and an N-terminal amino acid sequence which di¡ers from that of the classical streptokinase [12]. However, it has since been revealed that chromosomal DNA from the Esk producing strain hybridizes with the streptokinase sequence (skc) and that this gene was located within a region of the chromosome which contained the streptokinase gene in other streptococci [13]. The authors of this latter study concluded that streptokinase and Esk were genetic homologues. In contrast, chromosomal DNA from S. uberis which produces PauA failed to hybridize with the same skc gene probe [13] and the authors concluded that PauA was not a genetic homologue of streptokinase. This communication compares PauA sequences obtained from two geographically distinct isolates of S. uberis with existing streptokinase sequences and describes their location within the chromosome.

2. Materials and methods 2.1. Bacterial strains and storage

in the streptococcal culture collections at the Institute for Animal Health and P¢zer. The bacteria were stored at 320 and 380³C in Todd Hewitt broth containing 25% (v/v) glycerol. 2.2. Extraction of chromosomal DNA from S. uberis S. uberis cells were inoculated into BHI and incubated at 37³C for 24 h. Cells were harvested by centrifugation (7700Ug, 20 min) and the cell pellet was washed in 10 mM Tris-HCl containing 5 mM EDTA (TE, pH 7.8) and ¢nally resuspended in 2 ml TE. Washed cells were incubated at 65³C for 20 min, frozen overnight at 320³C and lysis was initiated by the addition of mutanolysin (250 U), lysozyme (100 mg) and RNAse A (50 Wg) and incubation at 37³C for 1.5 h. Subsequently, proteinase K was added and the incubation continued for a further 3 h at 37³C prior to the addition of SDS to a ¢nal concentration of 2% (w/v) and a further 30-min incubation. The lysate was extracted once with Trissaturated phenol and extracted twice with phenol/ chloroform/isoamyl alcohol (25:24:1). Chromosomal DNA was precipitated by the addition of 2.5 volumes of absolute ethanol and 0.1 volume of sodium acetate (3.0 M, pH 5.2), followed by incubation at 320³C for 24 h. Precipitated DNA was collected by centrifugation (15 000Ug, 15 min at 4³C), rinsed once in 70% (v/v) ethanol, dried, resuspended in TE and quanti¢ed by measuring the absorbance at 260 nm. 2.3. Construction of S. uberis genomic libraries in Escherichia coli Chromosomal DNA from S. uberis (10 Wg) was digested with BglII at 37³C for 8 h, extracted with phenol/chloroform (24:1) and stored at 320³C. The plasmid vector pUC18 was digested with BamHI and dephosphorylated using calf intestinal phosphatase as recommended by the supplier (Promega). Linear dephosphorylated vector (0.5 Wg) was ligated to digested chromosomal DNA (5 Wg) and used to transform E. coli DH5K as recommended by the supplier (Gibco, BRL).

All strains of S. uberis were originally isolated from cases of clinical bovine mastitis and retained

FEMSLE 8925 11-8-99

E.L. Rosey et al. / FEMS Microbiology Letters 178 (1999) 27^33

2.4. Screening of genomic libraries for expression of PauA activity

29

primers ER45 (5P-GAGATTCCTCTCTAGATATCA-3P) and ER46 (5P-GGGCTGCAGATCCGTTAAAAAATGACATTAATAT-3P). The positions of the primers ER45, ER 46, P38 and P39 are indicated in Fig. 1.

A skim milk agar plate assay [2] was used to detect plasminogen activation. Plates consisted of BHI agar supplemented with 1% (w/v) skimmed milk (Oxoid), bovine plasminogen (0.015 U ml31 ), ampicillin (100 Wg ml31 ) and IPTG (0.1 mM). This assay allowed for the di¡erentiation of non-speci¢c protease activity from plasminogen-dependent activity by the omission of plasminogen from control plates. This modi¢cation was used to con¢rm the plasminogendependent clearing of the skimmed milk by putative positive clones. Portions of inserts within the plasmids of two positive clones were sequenced using an automated sequencer (ABI model 377, Applied Biosystems).

2.6. Southern analysis Standard procedures for Southern analysis of chromosomal and ampli¢ed DNA were used [14]. Digoxygenin-labelled probe was prepared by ampli¢cation of DNA encoding mature PauA from S. uberis 0140J using a DIG PCR labelling mix (Boehringer).

3. Results and discussion 3.1. Screening of plasmid libraries for plasminogen activation

2.5. PCR ampli¢cation of chromosomal DNA encoding plasminogen activators

Colonies of transformed E. coli which produced zones of clearing on BHI skimmed milk agar plates containing plasminogen were analyzed to ensure that the observed protease activity was plasminogen-dependent. Two isolates were chosen for further analysis, one isolate (Pz318) contained DNA from S. uberis strain C216 and the other (Pz319) contained DNA from S. uberis strain 95-140. Their respective plasmids were designated pER318 and pER319. Both of these isolates were also capable of producing plasminogen activating activity in the absence of IPTG, suggesting that the pauA promoter was also present within the cloned fragments and that this was functional in E. coli DH5K.

Chromosomal DNA located between streptococcal genes abc and rel was ampli¢ed by PCR using the oligonucleotide primers abc (5P-CCACATCCTGAAGGCCCAAC-3P) and rel (5P-GCGTGAAGCGGACCAATGG-3P) and that encoding mature PauA was ampli¢ed by PCR using oligonucleotide primers P38 (5P-AATAACCGGTTATGATTCCGACTAC3P) and P39 (5P-AAAATTTACTCGAGACTTCCTTTAAGG-3P). In each case, the annealing temperature was 54³C and the MgCl2 concentration was 1.5 mM. Chromosomal DNA encoding the entire PauA open reading frame (ORF) was similarly ampli¢ed at an annealing temperature of 50³C using Table 1 Sequence variation of S. uberis PauA Database accession number

AJ131604 AJ131605 AJ012549 AJ006413 AJ012548

Amino acid residue 48

54

90

99

186

215

217

D D D D N

V V V A V

Q Q Q R R

Q Q Q R R

L L L R R

Q Q Q P P

H H H D D

Six amino acid sequences of mature PauA were aligned using the Pileup program, part of Wisconsin Package Version 9.1, Genetics Computer Group (GCG), Madison, WI, USA. Sequence variations are shown using the single letter code for amino acids with numbering starting from the ¢rst residue (isoleucine) of the secreted gene product.

FEMSLE 8925 11-8-99

30

E.L. Rosey et al. / FEMS Microbiology Letters 178 (1999) 27^33

Fig. 1. Structural organization of the S. uberis pauA region. The transcriptional orientation of ORFs in the pauA region is indicated by arrows. The chromosomal fragments isolated from strains C216 (pER318) and 95^140 (pER319) are shown below a scale line indicating distances (nucleotides) between the major ORFs.

to BglII restriction polymorphism, within or upstream of an ORF with high homology to S. pneumoniae HexA [16]. An ORF designated orf1 which potentially encodes a 130 amino acid protein was identi¢ed adjacent to hexB. The fourth ORF was consistent with that expected for PauA and recently reported by Johnsen et al. [17]. A putative bi-directional transcriptional terminator was identi¢ed in both clones between pauA and hexA. Expression of the full-length PauA (amino acid 1^

3.2. Molecular characterization of PauA producing clones Analysis of pER318 and pER319 demonstrated that both contained cloned inserts with a considerable identity. The cloned fragments contained four ORFs (Fig. 1). The fragments shared one common endpoint within an ORF with high homology to amino acids 1^68 of Streptococcus pneumoniae HexB [15], while exhibiting distinct endpoints, due Table 2 Sequence comparison of PauA with streptokinase 1 1 ^ 2 3 4 5 6 7 8 9 10 11 12 13 14 15

2

3

4

5

6

7

8

9

10

11

12

13

14

15

100 ^

100 100 ^

97.71 97.71 97.71 ^

97.71 97.71 97.71 99.24 ^

98.09 98.09 98.09 99.24 99.24 ^

23.55 23.55 23.55 23.55 24.71 24.32

24.71 24.71 24.71 25.10 24.71 25.10

25.48 25.48 25.48 26.26 25.87 26.26

26.26 26.26 26.26 26.64 26.26 26.64

26.26 26.26 26.26 26.64 26.26 26.64

26.26 26.26 26.26 26.64 26.26 26.64

26.26 26.26 26.26 26.64 26.26 26.64

26.26 26.26 26.26 26.64 26.26 26.64

27.38 27.38 27.38 27.78 28.06 27.78

^

95.66 ^

95.18 95.42 ^

85.54 85.78 84.10 ^

85.54 85.78 84.10 100 ^

85.54 85.78 84.34 100 100 ^

85.30 85.54 84.34 98.80 98.80 98.80 ^

85.30 85.54 84.34 98.80 98.80 98.80 100 ^

78.80 79.76 77.83 89.64 89.64 89.64 88.43 88.43 ^

Amino acid sequences of mature plasminogen activators were aligned and clustered using the Pileup multiple sequence alignment program (Wisconsin Package Version 9.1, Genetics Computer Group (GCG), Madison, WI, USA). Sequence identity scores were generated using the gap program from the same package. The boxed region denotes comparisons of PauA with streptokinase (Sk) variants. Numbers refer to the following sequences (accession number or reference). 1, PauA (AJ131604); 2, PauA (AJ131605) ; 3, PauA (AJ012549); 4, PauA (AJ006413) ; 5, PauA (AJ012548); 6, PauA (AJ131631); 7, Ska; ([20]) ; 8, Ska (X51517) ; 9, Ska (Z48617); 10, Skc (A20006) ; 11, Skc (S46536); 12, Skg (X13400) ; 13, Skc (A04926) ; 14, Ska (K02986) ; 15, Ska (X13399).

FEMSLE 8925 11-8-99

E.L. Rosey et al. / FEMS Microbiology Letters 178 (1999) 27^33

31

286) in E. coli resulted in a marked decrease in the cell growth rate compared to that observed during expression of mature PauA (amino acid 26^286). Analysis of whole cell lysates by Western blotting indicated that although the majority of expressed protein remained full-length (amino acid 1^286), a substantial portion co-migrated with recombinant PauA (amino acid 26^286) and puri¢ed native PauA (Rosey et al., unpublished). This suggests that post-translational processing and removal of the putative signal sequence occurs in E. coli at a sub-optimal rate and that the observed toxicity possibly resulted from interference with the normal secretion apparatus when full-length PauA was overproduced in E. coli. Similar deleterious e¡ects have been noted during expression of skc with an intact signal sequence in E. coli [18] and this e¡ect was also attributed to defective secretion of the heterologous gene product. 3.3. Analysis of the pauA ORF The mature PauA sequences obtained during the present investigation (accession numbers AJ012549 and AJ012548) and those from other studies [17] varied at only seven amino acid residues (Table 1) and consequently showed a high (97.7^100%) degree of homology (Table 2). In contrast, the mature streptokinase protein sequences exhibited greater polymorphism, the sequence homology ranged from 78.8 to 100%(Table 2). The sequences of PauA showed only weak homologies (23.5^28%) to streptokinase sequences (Table 2) and no continuous regions of homology were detected. This is consistent with other observations [13] which showed that a gene probe based on the streptokinase gene (skc) from S. equisimilis H46A failed to hybridize with chromosomal DNA from three strains of S. uberis (including one strain, C216, used in the present study). In addition, the amino acid sequence of PauA showed no signi¢cant homology to the N-terminal amino acid sequence of the plasminogen activator Esk [12]. Variation within the streptokinase sequences has been shown to be con¢ned to restricted regions of the protein and it has been implied that this variation is associated with strains which exhibit di¡erent pathological e¡ects following infection [11]. No such

Fig. 2. Comparison of the loci of S. pyogenes and S. uberis plasminogen activators. (A) Agarose gel electrophoresis of the abc-rel PCR ampli¢cation products (S. pyogenes 0358, lane 1; S. uberis C216, lane 2 and S. uberis 0140J, lane 3) and the pauA ampli¢cation products (S. pyogenes 0358, lane 4; S. uberis C216, lane 5 and S. uberis 0140J, lane 6). Molecular mass markers (lane 7) are as indicated. (B) Southern analysis of genomic DNA from S. pyogenes 0358 (lanes 1 and 4), S. uberis C216 (lanes 2 and 5) and S. uberis 0140J (lanes 3 and 6) using the digoxygenin-labelled pauA coding sequence as probe. Genomic DNA (5 Wg) was digested with HindIII (lanes 1^3) or BamHI (lanes 4^6).

divergence was detected within the PauA sequences. Interestingly, the three PauA sequences from isolates from the UK and Denmark (AJ131604, AJ131605 and AJ012549) were identical (Tables 1 and 2). Similarly, those from two isolates from the USA (AJ012548 and AJ131631) di¡ered by only two amino acids (Table 1). However, these two groups of strains showed consistent di¡erences in four of the seven variant positions (Table 1). Analysis of PauA from a greater number of strains will be required to determine if these subtle variations are consistent with the geographical site of isolation. The PauA sequence from the NCTC strain (AJ006413) showed a greater similarity to those obtained from the isolates from the USA (two amino acid di¡erences in

FEMSLE 8925 11-8-99

32

E.L. Rosey et al. / FEMS Microbiology Letters 178 (1999) 27^33

each case), compared to that obtained from the European isolates (six amino acid di¡erences). However, the geographical origin of this strain (isolated in 1932) is uncertain. 3.4. Location of pauA within the chromosome of S. uberis The location of pauA within the chromosome of S. uberis clearly di¡ered from that of the streptokinase genes (ska, skc, skg) in other streptococci [13,19]. Streptokinase has been located in a region of the chromosome containing dexB, abc, lrp, skc/ska/skg, orf and rel, in that order over an 8.5-kb stretch of DNA. Streptokinase genes have been shown to be in the same chromosomal location in three strains of S. pyogenes, a human Lance¢eld group G streptococcus and the strain of S. equisimilis shown to produce the protein, Esk [13]. However, chromosomal DNA from S. uberis failed to hybridize with skc but did hybridize with probes designed to dexB, abc and rel [13]. These genes were also shown to be in close proximity in S. uberis [13]. In the present investigation, ampli¢cation of DNA from this region of Streptococcus pyogenes, using primers complementary to £anking genes abc and rel [19], generated a product of approximately 3.5 kb, consistent with that predicted [19]. However, a product of only 1.6 kb was obtained from S. uberis (Fig. 2a). This is consistent with the absence of a streptokinase gene within this location in S. uberis. Furthermore, this fragment from S. uberis failed to hybridize with a probe representing the coding sequence of mature PauA (data not shown), thus con¢rming that pauA does not occupy this chromosomal location. Southern analysis of genomic DNA using the same probe demonstrated the presence of a single copy of pauA (containing one HindIII restriction site (Fig. 1) and lacking a BamHI site) within S. uberis and the absence of any homologous sequence within the S. pyogenes genome (Fig. 2b). The terminal sequences of the cloned fragments within pER318 and pER319 showed high homology to hexA and hexB from S. pneumoniae (Fig. 1). The HexA and HexB proteins are highly conserved with MutS and MutL homologues from both bacteria and eukaryotes [16,17]. HexA and B play a role in mismatch DNA repair during transformation and DNA

replication in S. pneumoniae. Mutations within hexA or hexB lead to an increased mutation rate and HexA has also been shown to induce an increase in the spontaneous mutation rate when overexpressed in E. coli [17]. It is not readily apparent whether the chromosomal location of pauA between hexA and hexB may a¡ect regulation or expression of these DNA repair proteins signi¢cantly. It is tempting to speculate, however, that the location of orf1 and pauA at this chromosomal region may provide insight into possible evolutionary events leading to acquisition of plasminogen activating capabilities in S. uberis. Earlier characterization suggested that PauA was distinct from streptokinase by virtue of its molecular size and substrate speci¢city [2,3]. The data in this communication have extended the evidence for this by demonstration of the distinct nature of the encoded plasminogen activator, the coding sequence and the chromosomal location of pauA. A number of streptococci capable of infecting a mammalian host are also capable of speci¢cally activating plasminogen from that host [9]. In one such case, S. uberis and activation of bovine plasminogen by PauA, it is interesting to speculate that this property may have arisen as a result of an independent evolutionary event which was distinct from the acquisition of streptokinase and streptokinase-like molecules.

References [1] Leigh, J.A. (1999) Streptococcus uberis: A permanent barrier to the control of bovine mastitis ? Vet. J. 157, 225^238. [2] Leigh, J.A. (1993) Activation of bovine plasminogen by Streptococcus uberis. FEMS Microbiol. Lett. 114, 67^72. [3] Leigh, J.A. (1994) Puri¢cation of a plasminogen activator from Streptococcus uberis. FEMS Microbiol. Lett. 118, 153^ 158. [4] Leigh, J.A. and Lincoln, R.A. (1997) Streptococcus uberis acquires plasmin activity following growth in the presence of bovine plasminogen through the action of its speci¢c plasminogen activator. FEMS Microbiol. Lett. 154, 123^129. [5] Kitt, A.J. and Leigh, J.A. (1997) The auxotrophic nature of Streptococcus uberis: The acquisition of essential amino acids from plasmin derived casein peptides. Adv. Exp. Med. Biol. 418, 647^650. [6] Benfeldt, C., Larsen, L.B., Rasmussen, J.T., Andreasen, P.A. and Petersen, T.E. (1995) Isolation and characterization of plasminogen and plasmin from bovine milk. Int. Dairy J. 5, 577^592.

FEMSLE 8925 11-8-99

E.L. Rosey et al. / FEMS Microbiology Letters 178 (1999) 27^33 [7] Aston, J.W. (1975) Amino acids in milk. Their determination by gas-liquid chromatography and their variation due to mastitic infection. Aust. J. Dairy Technol. 30, 55^59. [8] Leigh, J.A., Finch, J.M., Field, T.R., Real, N.C., Winter, A., Walton, A.W. and Hodgkinson, S.M. (1999) Vaccination with the plasminogen activator from Streptococcus uberis induces an inhibitory response and protects against experimental infection in the dairy cow. Vaccine 17, 851^857. [9] McCoy, H.E., Broder, C.C. and Lottenberg, R. (1991) Streptokinases produced by pathogenic group C streptococci demonstrate species speci¢c plasminogen activation. J. Infect. Dis. 164, 515^521. [10] Malke, H. (1993) Polymorphism of the streptokinase gene : implications for the pathogenesis of post streptococcal glomerulonephritis. Zent.bl. Bakt. 278, 246^257. [11] Huang, T.-T., Malke, H. and Ferretti, J.J. (1989) Heterogeneity of the streptokinase gene in group A streptococci. Infect. Immun. 57, 502^506. [12] Nowicki, S.T., Minning-wenz, D., Johnston, K.H. and Lottenberg, R. (1994) Characterization of a novel streptokinase produced by Streptococcus equisimilis of non-human origin. Thromb. Haemost. 72 (4), 595^603. [13] Frank, C., Steiner, K. and Malke, H. (1995) Conservation of the organization of the streptokinase gene region among pathogenic streptococci. Med. Microbiol. Immunol. 184, 139^146. [14] Sambrook, J., Fritsch, E.F. and Maniatis, T (1989) Molecular

[15]

[16]

[17]

[18]

[19]

[20]

33

Cloning : a Laboratory Manual, 2nd edn., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Prudhomme, M., Martin, B., Mejean, V. and Claverys, J.-P. (1989) Nucleotide sequence of the Streptococcus pneumoniae hexB mismatch repair gene : Homology of HexB to MutL of Salmonella typhimurium and to PMS1 of Saccharomyces cerevisiae. J. Bacteriol. 171, 5332^5338. Prudhomme, M., Mejean, V., Martin, B. and Claverys, J.-P. (1991) Mismatch repair genes of Streptococcus pneumoniae : HexA confers a mutator phenotype in Escherichia coli by negative complementation. J. Bacteriol. 173, 7196^7203. Johnsen, L., Poulsen, K., Killian, M. and Petersen, T.E. (1999) Puri¢cation and cloning of a streptokinase from Streptococcus uberis. Infect. Immun. 67, 1072^1078. Muller, J., Reinert, H. and Malke, H. (1989) Streptokinase mutation relieving Escherichia coli K-12 (prlA4) of detriments caused by the wild-type skc gene. J. Bacteriol. 171, 2202^ 2208. Mechold, U., Steiner, K., Vetterman, S. and Malke, H. (1993) Genetic organization of the streptokinase region of the S. equisimilis H46A chromosome. Mol. Gen. Genet. 241, 129^ 140. Ohkuni, H., Todome, Y., Suzuki, H., Mizuse, M., Kotani, N., Horiuchi, K., Shikama, N., Tsugita, A. and Johnston, K.H. (1992) Immunochemical studies and complete amino acid sequence of the streptokinase from Streptococcus pyogenes (group A) M type 12 strain A374. Infect. Immun. 60, 278^283.

FEMSLE 8925 11-8-99