System. Appl. Microbiol. 23, 325-329 (2000) SYSTEI\IL4TIC AND © Urban & Fischer Verlag _htt-,-p_:llw_w_w_.ur_ba_n_fis_ch_er_.de-.:./jo_u_rn_als_/s_am ____________ APPLIED MICROBIOLOGY

Characterization of Acid Phosphatase Activities in the Equine Pathogen Streptococcus equi

ANDREA HAMILTON, DEAN HARRINGTON, and lAIN C. SUTCLIFFE

School of Sciences, University of Sunderland, UK

Received June 16,2000

Summary

Acid phosphatases hydrolyse phosphomonoesters at acidic pH in a variety of physiological contexts. The recently defined class C family of acid phosphatases includes the 32 kDa LppC lipoprotein of Strep­tococcus equisimilis. To define further the distribution of acid phosphatases in the genus Streptococcus we have examined the equine pathogens Streptococcus equi subsp. equi and Streptococcus equi subsp. zooepidemicus. Whole cell assays indicated that these organisms possess two acid phosphatases with ac­tivity optima at pH 5.0 and pH 6.0-6.5 and that only the former of these was, like LppC, resistant to EDTA. Western blotting with a polyclonal anti-LppC antiserum revealed the presence of a cross-reactive 32 kDa protein in both organisms. The cross-reactive protein in S. equi was shown to be a surface acces­sible lipoprotein as its processing was inhibited by the antibiotic globomycin and it was released from whole cells by treatment with trypsin. The presence of DNA sequences homologous to the S. equisimilis IppC gene were confirmed by PCR. These data strongly suggest that Streptococcus equi subsp. equi and Streptococcus equi subsp. zooepidemicus produce a lipoprotein acid phosphatase homologous to LppC of S. equisimilis.

Key words: Acid phosphatase- Lipoprotein - Streptococcus - Strangles

Introduction

Acid phosphatases are a ubiquitous and diverse group of enzymes which hydrolyse phosphomonoesters at acidic pH and are thought to participate in a wide variety of functions including nutrient acquisition, signal transduc­tion, metabolic regulation and energy conversion (VIN­CENT et al., 1992; ROSSOLINI et al., 1998). They may also contribute to the virulence of some bacterial pathogens (ROSSOLINI et al., 1998). Recently, a new family of bacte­rial non-specific acid phosphatases, designated class C, has been described (ROSSOLINI et al., 1998; THALLER et al., 1998). Members of this class include the OlpA puta­tive lipoprotein of Chryseobacterium meningosepticum (THALLER et al., 1998), the P4 outer membrane lipopro­tein of Haemophilus inf/uenzae (REILLY et al., 1999) and the LppC lipoprotein of Streptococcus equisimilis (GASE et al., 1997; MALKE, 1998). A lipoprotein that is antigeni­cally related to LppC was also identified in Streptococcus pyogenes (GASE et al., 1997) and a gene encoding LppA, a protein highly homologous to LppC (82% sequence iden­tity over 284 amino acids), has been identified through the S. pyogenes genome project (MALKE, 1998).

To explore further the distribution of lipoprotein acid phosphatases in the genus Streptococcus we have exam­ined the equine pathogen Streptococcus equi subsp. equi

and the closely related Streptococcus equi subsp. zooepi­demicus (herein S. zooepidemicus). Streptococcus equi is of major significance as the causative agent of strangles, a highly contagious and debilitating disease characterised by pharyngeal constriction which is a consequence of lymph node swelling (often accompanied by abscessa­tion) in the horses upper respiratory tract (CHANTER, 1997). Streptococcus zooepidemicus is a significant cause of equine lower airway disease, foal pneumonia, en­dometritis and abortion (CHANTER, 1997).

Materials and Methods

Bacterial Strains and Growth The type strains S. equi NCTC 9682T and S. zooepidemicus

7023T were obtained from the National Collection of Type Cul­tures (Reading, UK). Clinical isolates of S. equi (strains 1026, 1742,4047) and S. zooepidemicus (strains 3682, 461 and K3)

Abbreviations: NBT-BCIP - Nitro Blue Tetrazolium/5-Bromo-4-Chloro-3-Indolyl Phosphate; PBS - phosphate-buffered saline; SDS-PAGE - sodium dodecyl sulphate polyacrylamide gel elec­trophoresis

0723-2020100/23/03-325 $ 15.0010

326 A. HAMILTON et a!.

were kindly supplied by Dr Neil CHANTER (Animal Health Trust, Newmarket, UK). Laboratory stocks were maintained on Brain Heart Infusion agar. Broth cultures in either Brain Heart Infusion or Todd Hewitt media were grown without agitation at 37°C.

For growth of S. equi in the presence of the antibiotic globomycin, an overnight culture was diluted 1:20 into fresh 20 mL broths then incubated until early log phase. Globomycin (10 mg/mL in ethanol) was added to one culture to give a final concentration of 120 pg/mL and a control culture received an equivalent volume of ethanol alone. Growth was continued for 4.5 hours, then the cells were harvested by centrifugation (2800 x g, 10 min., 4 0c) and washed twice with phosphate­buffered saline (PBS).

Acid phosphatase assays Bacteria from 20 mL Todd Hewitt broth cultures were har­

vested by centrifugation and washed twice with PBS. Whole cell acid phosphatase activity was assayed in citrate-chloride buffers using p-nitrophenylphosphate as substrate, essentially as de­scribed by MALKE (1998). Released p-nitrophenol was measured at 405 nm using a Dynatech MRX plate-reader. Initial assays were performed at pH 5.0 and subsequent assays were performed across a range of pH values from 3.0-7.5. The effect of enzyme inhibitors was assayed by including EDTA (15 mM), L-(+) tartar­ic acid (16.7 mM) and 2 % Triton X-I 00 in the assay buffers.

Preparation of cell extracts, electrophoresis and immuno­blotting

Bacteria from 20 mL Todd Hewitt broth cultures were har­vested by centrifugation and washed with PBS. Cell extracts were prepared by boiling the washed cell pellets for 10 minutes in 130 pL sodium dodecyl sulphate polyacrylamide gel elec­trophoresis (SDS-PAGE) loading buffer followed by removal of cell debris by microcentrifugation. Mild detergent extracts were also prepared by solubilising whole cells with 1 % w/v octyl-glu­coside at 4 °C followed by centrifugation to remove cell debris, extensive dialysis to remove the detergent and concentration by freeze-drying. Dried extracts were prepared for electrophoresis by boiling in SDS-PAGE loading buffer as above.

Washed cell suspensions were also incubated in the presence or absence of trypsin (200 units/mL, 1 hr, 37°C) and the re­leased proteins in the supernatants recovered following removal of the cells by centrifugation. Supernatants were prepared for SDS-PAGE as described above.

Samples were electrophoresed through 13 % resolving gels (120 V; 75 minutes; BioRad mini-Protean II system) and semi­dry blotted onto nitrocellulose membranes. The blots were blocked overnight with 5% w/v skimmed milk in PBS-0.05% w/v Tween, then probed for two hours with polyclonal rabbit anti-LppC antiserum (GASE et a!', 1997) at 1/1,000 dilution. Blots were washed extensively with PBS-Tween and incubated with alkaline phosphatase-conjugated anti-rabbit IgG (Sigma, 1/30,000 dilution). After extensive washing, blots were devel­oped with the alkaline phosphatase substrate Nitro Blue Tetra­zolium/ 5-Bromo-4-Chloro-3-Indolyl Phosphate (NBT-BCIP; Zymed Laboratories Inc.).

Substrate SDS-PAGE analysis (zymography) of acid phos­phatase activity was carried out using a minor modification of the method of THALLER et al. (1995). Briefly, cell extracts were subjected to SDS-PAGE as described above, then the separated proteins were renatured by incubation for 3-4 hours in several changes of renaturation buffer (100 mM Tris-HCI pH 7.0, con­taining 2 mM MgS04' 0.05 mM ZnCI2 and 2% v/v Triton X-100). Following incubation for 1 hour at 37°C in several changes of equilibration buffer (100 mM sodium acetate, pH 5.0), acid phosphatase activity was detected by incubation

at 37°C for up to 16 hours in freshly prepared equilibration buffer containing 0.25 mM BCIP and 0.25 mM NBT.

Polymerase Chain Reaction (PCR) Conserved regions identified from the alignment of the S.

equisimilis LppC sequence with that of LppA from S. pyogenes (amino acids 129-135 and 249-255 respectively in the LppC protein; MALKE, 1998) were used to design the degenerate PCR primers SEZLPPC 1 U and SEZLPPC 2L. Primer sequences were as follows: SEZLPPC lU, 5'-TGGGTNCARAARAARGARGC-3' and SEZLPPC 2L, 5'-CCRTACATNGGRTTNGGRAA-3'.

The cells from 300 pi aliquots of overnight cultures of each bacterial strain were pelleted and resuspended in 100 pi of 10 mM Tris-HCI pH 8.0-1 mM EDTA buffer. The resuspended cells were boiled for 10 min, centrifuged to remove cell debris and the resulting supernatant was recovered as a source of ge­nomic DNA.

An aliquot (10 pi) of genomic DNA was added to each PCR reaction mixture which contained 50 pmols of each PCR primer (SEZLPPC lU and SEZLPPC 2L), 200 pM each of the de­oxynucleotide triphosphates, 3 mM magnesium chloride and 1 unit of Taq polymerase (Promega, UK) in a total volume of 50 pI. PCR was performed using the following thermal cycling profile: 94°C for 5 min (1 cycle); 94 °C for 30 sec, 45°C for 45 sec, 57°C for 30 sec (35 cycles); 72 °C for 7 min (1 cycle). Following PCR, 10 pi of each reaction mixture was elec­trophoresed on a 1.5% (w/v) agarose gel and amplification products were visualised under UV light.

The PCR product generated from S. equi 9682T template DNA was purified, cloned into the pGEM-T cloning vector (Promega, UK) according to manufacturers instructions and the resultant recombinant plasmids were used to transform compe­tent E. coli ]MI09 cells. DNA obtained from transformed cells harbouring putative recombinant plasmids was analysed by PCR using the SEZLPPC 1 U and SEZLPPC 2L primers as de­scribed above. Cloned inserts in plasmids extracted from trans­formants giving an amplification product of the appropriate size were sequenced using an ABI Prism 377 DNA sequencer. The sequence obtained has been deposited in the EMBL nu­cleotide sequence database under the accession number A]403972.

Results

Acid phosphatase activity of s. equi

Using a whole cell assay acid phosphatase activity of S. equi was detected over the pH range 4.0 to 7.0 (Figure 1). Maximal acid phosphatase activity centered around pH 5.0 but there also appeared to be a second activity with an optimum at ca. pH 6.0-6.5. Comparable results were obtained with whole cells of S. zooepidemicus (data not shown). Acid phosphatase activity at pH 5.0 was readily detectable during all phases of growth of both or­gamsms.

To determine whether these two optima represented distinct activities, acid phosphatase was assayed in the presence of EDTA, an inhibitor of class B but not class A acid phosphatases (ROSSOLINI et aI., 1998). Whereas the peak of acid phosphatase activity at pH 5.0 was clearly resistant to EDTA, it appeared that the activity at ca. pH 6.0-6.5 was sensitive to EDTA (Fig. 1), consistent with the presence of distinct acid phosphatases. Further assays revealed that both acid phosphatase activities were resis-

Fig. 1. Acid phosphatase activity in S. equi. Enzyme activity was measured as described in METH­ODS in the presence ["j or ab­sence [.] of EDTA. Results shown are a representative experi­ment from quadruplicate repeats.

E c: It) 0 "<t .... res Q) u c: res ..c ... 0 1/1 ..c

<C

0.5

0.45

0.4

0.35

0.3

0.25

0.2

0.15

0.1

0.05

0

3 4

tant to the enzyme inhibitors tartrate and Triton X-I 00 (data not shown).

Screening for acid phosphatase lipoproteins by Western blotting

To investigate the nature of the acid phosphatases of S. equi and S. zooepidemicus cell extracts were examined

Fig. 2. Western blot analysis of S. equi and S. zooepidemicus using a rabbit polyclonal antibody to anti-LppC of S. equisimilis. Cell extracts were separated by SDS­PAGE, blotted onto nitrocellulose and probed with anti-LppC. Ex­tracts were as follows: Lane 1, Streptococcus pneumoniae R36A; Lane 2, S. equi; Lane 3, S. equi grown m the presence of globomycin; Lane 4, octyl-gluco­side extract from S. equi; Lane 5, octyl-glucoside extract from S. zooepidemicus; Lane 6, molecular weight standards (kDa).

1 2

4.5 5

Acid Phosphatase of Streptococcus equi 327

5.5

pH

6 6.5 7 7.5

by Western blotting using a rabbit polyclonal antibody to LppC of S. equisimilis. As shown in Figure 2 a single cross-reacting protein of ca. 32 kDa was detected in each extract, whereas no cross-reaction was observed in a cell extract of Streptococcus pneumoniae. Cross-reacting proteins of 32 kDa were also detected in extracts from S. equi strains 1026, 1742, 4047 and S. zooepidemicus strains 3682, 461 and K3 (data not shown) and were de-

3 4 5 6

218.0 125.0

78.0

44.0

31 .9

17.5

7.7

328 A. HAMILTON et al.

1 2 3

44.0-

31 .9-

17.5-

Fig. 3. Western blot analysis of proteins released from S. equi by treatment with trypsin. S. equi cells were incubated in the presence or absence of trypsin and the released proteins separated by SDS­PAGE, blotted onto nitrocellulose and probed with anti-LppC anti­body. Lane 1, SDS extract of untreated S. equi cells; Lane 2, super­natant from trypsin-treated cells; Lane 3, supernatant from cells in­cubated without trypsin. Positions of the molecular weight stan­dards (kDa) are indicated on the left.

tected following mild detergent extraction with octyl-glu­coside (Fig. 2, lanes 4 and 5). Zymographic analysis of octyl-glucoside extracts from S. equi 9682T and strain 4047 revealed low level acid phosphatase activity at the same molecular weight as the cross-reacting proteins (data not shown).

Growth of S. equi NCTC 9682T in the presence of globomycin, an antibiotic which specifically inhibits lipoprotein signal peptide processing (INUKAI et al., 1978), interfered with the processing of the 32 kDa pro­tein resulting in the appearance of an additional cross-re­acting band at ca. 34 kDa (Fig. 2 lane 3). These data strongly suggest that the 32 kDa protein detected in S. equi and S. zooepidemicus is a lipoprotein electrophoret­ically and antigenic ally similar to the 32.4 kDa LppC acid phosphatase of S. equisimilis. To confirm that in S. equi this lipoprotein is surface exposed whole bacterial cells were subjected to mild digestion with trypsin. As shown in Figure 3, truncated forms of the anti-LppC cross-reactive protein were released from cells of S. equi 9682T in the presence but not the absence of trypsin.

PCR analysis to identify an lppC homologue in S. equi

Degenerate primers designed from alignment of the translated sequences of S. equisimilis LppC and S. pyo­genes LppA amplified a PCR product of 360 bp from S. equi and S. zooepidemicus genomic DNA. The size of these PCR products is consistent with the predicted size of the corresponding internal sequence of the IppC gene. The S. equi PCR product was cloned and sequenced and

yielded a translated sequence of 113 amino acids fol­lowing removal of primer sequences. Homologues of this translated sequence were identified by a BLAST search (ALTSCHUL et al., 1997) using the NCB! server (http:// www.ncbi.nlm.nih.gov/BLAST). The translat­ed S. equi sequence showed 71 % identity to LppC of S. equisimilis (811113 amino acids) and 46% identity (52/111 amino acids) to OlpA, the class C acid phos­phatase of C. meningosepticum (THALLER et al., 1998). Furthermore, a BLAST search of the S. pyo­genes genome (http//www.genome.ou.edu/strep. html) revealed that the translated S. equi sequence was 72 % identical to LppA of S. pyogenes (821113 amino acids).

Discussion

In this study we have examined the acid phos­phatase activity of S. equi subsp. equi and of S. equi subsp. zooepidemicus. A whole cell enzyme assay re­vealed readily detectable, apparently constitutive acid phosphatase activity in both organisms. Analysis of the pH optimum for acid phosphatase activity sug­gested that both organisms possess more than one acid phosphatase, with the major activity optimal at pH 5.0 (Fig. 1). This acid phosphatase activity was resistant to EDTA, tartrate and Triton X-100 and thus resembles the LppC acid phosphatase of S. equi­similis which also has a pH optimum at 5.0 (MALKE, 1998). Moreover, Western blotting with a polyclonal anti-LppC antibody revealed that both S. equi and S. zooepidemicus produced a single cross-reacting anti­gen of ca. 32 kDa (Fig. 2), consistent with the mole­cular weight of LppC (GASE et al., 1997; MALKE, 1998). Low level acid phosphatase activity could also be detected zymographically at this molecular weight using NBT-BCIP as substrate. Difficulty in detecting LppC acid phosphatase activity by zymography has previously been noted: in contrast to the enzyme ex­pressed in Escherichia coli ]M109, the native S. equi­similis enzyme was not detectable zymographically with BCIP as substrate (MALKE, 1998). In this study, the use of NBT as an enhancer in the development buffer was found to increase the sensitivity of the zy­mography (data not shown). Since only the activity attributed to LppC was detected zymographically, it is possible that the EDTA-sensitive activity (Fig.1) is a class B acid phosphatase as these enzymes are unable to dephosphorylate BCIP and are sensitive to inhibi­tion by EDTA (ROSSOLINI et al., 1998).

Processing of the S. equi phosphatase was sensitive to the action of globomycin, consistent with lipopro­tein modification, as has been demonstrated for LppC and the closely related LppA homologue of S. pyogenes (GASE et al., 1997). The surface localisation of the S. equi phosphatase was confirmed by mild trypsinolysis, which released truncated proteins that cross-reacted with the anti-LppC antibody (Fig. 3). Trypsin cleaves polypeptides adjacent to lysine and

arginine residues and it is interesting to note that align­ment of the streptococcal LppC and LppA protein se­quences reveals conservation of several potential trypsin cleavage sites in the 40 amino acid region immediately adjacent to their lipid-modified N-termini (MALKE, 1998). It can thus be speculated that at least two of these sites are also present in the S. equi homologue resulting in the release by trypsin of the slightly truncated prod­ucts observed (Fig. 3).

To provide genetic evidence that S. equi possesses a gene encoding a homologue of LppC we used degenerate PCR to amplify a product from S. equi and S. zooepi­demicus genomic DNA. This yielded a translated S. equi sequence which displayed high levels of identity with LppC and LppA (811113 and 82/113 identical amino acids respectively). The amino acid sequence was also significantly homologous (521111 amino acids identical) to OlpA of C. meningosepticum, the archetypal bacterial class C acid phosphatase lipoprotein described by ROSSOLINI and co-workers (ROSSOLINI et aL, 1998; THALLER et aL, 1998). It is notable that the deduced S. equi amino acid sequence also contains the domain B motif characteristic of members of the 'DDDD' super­family of phosphohydrolases (THALLER et aL, 1998). These data strongly suggest that the cross-reacting anti­gen detected by Western blotting in S. equi is indeed a homologue of LppC from S. equisimilis. It is anticipated that the full gene sequence for the S. equi LppC homo­logue will soon become available through the recently initiated S. equi strain 4047 genome project (http://www. sanger.ac. uk/Projects/S _equi/).

These data suggest that lipoprotein acid phosphatases of the LppC type are produced in several members of the genus Streptococcus. We have also been able to detect e1ectrophoretically similar antigens cross-reacting with the anti-LppC antibody in whole cell extracts of Strepto­coccus gordonii DL1 and two strains of Streptococcus agalactiae (data not shown). However, whole cell ex­tracts of Streptococcus pneumoniae (Fig. 2) and Strepto­coccus mutans (data not shown) lacked cross-reactive antigens, consistent with the absence of LppC homo­logues in the unfinished genomes of these organisms as determined by BLAST searches. Cumulatively, these data suggest a discontinuous distribution of lipoprotein acid phosphatases within the genus Streptococcus that war­rants further investigation. Furthermore, our data sug­gest that there is heterogeneity within the bacterial class C acid phosphatases as the S. equi and S. equisimilis ac­tivities are both able to hydrolyse BCIP and are resistant to EDTA, whereas the P4 lipoprotein of H. inf/uenzae is EDTA-sensitive and unable to hydrolyse BCIP (REILLY et aL,1999).

Lipoprotein modification provides a versatile mecha­nism by which diverse proteins may be anchored within the Gram-positive cell envelope (SUTCLIFFE and RUSSELL, 1995) and the class C acid phosphatases represent a novel functional class of Gram-positive lipoproteins (MALKE, 1998). As acid phosphatases may play multiple

Acid Phosphatase of Streptococcus equi 329

roles in bacterial physiology, the precise significance of the cell surface localisation of these enzymes has yet to be determined. Moreover, further studies are now needed to clarify if the acid phosphatases described here for S. equi and S. zooepidemicus contribute to the virulence of these important equine pathogens.

The authors are grateful to Dr Horst MALKE (Institute for Molecular Biology, Jena University, Germany) for supplying the anti-LppC antibody, to Dr Neil CHANTER for supplying clinical isolates of S. equi and to Dr Masatoshi INUKAI (Sankyo Chemical Co., Tokyo, Japan) who kindly supplied the globomycin. DNA sequencing was carried out at the University of Newcastle Central Facility for Molecular Biology.

References

ALTSCHUL, S. E, MADDEN, T. L., SCHAFER, A. A., ZHANG, ] ., ZHANG, Z., MILLER, W. and LIPMAN, D. ] .: Gapped BLAST and PSI-BLAST: a new generation of protein database search programmes. Nucleic Acids Res. 25,3389-3402 (1997) .

CHANTER, N.: Streptococci and enterococci as animal patho­gens. J. Appl. Microbiol. Symp. Suppl. 83, 100S-109S (1997).

GASE, K., LIU, G., BRUCKMANN, A., STETNER, K., OZEGOWSKI, ]., MALKE, H.: The IppC gene of Streptococcus equisimilis en­codes a lipoprotein that is homologous to the e (P4) outer membrane protein from Haemophilus in{luenzae. Med. Mi­crobiol. Immunol. 186, 63-73 (1997).

MALKE, H.: Cytoplasmic membrane lipoprotein LppC of Strep­tococcus equisimilis functions as an acid phosphatase. Appl. Env. Microbiol. 64,2439-2442 (1998).

REILLY, T. j., CHANCE, D. L., SMTTH, A. L.: Outer membrane lipoprotein e (P4) of Haemophilus influenzae is a novel phos­phomonoesterase. J. Bacteriol. 181,6797-6805 (1999).

ROSSOLINI, C. M., SCHTPPA, S., RICCIO, M. L., BERLUlTI, E, MACASKIE, L. E., THALLER, M. c.: Bacterial nonspecific acid phosphohydrolases: physiology, evolution and use as tools in microbial biotechnology. Cell. Mol. Life Sci. 54, 833-850 (1998).

SUTCLIFFE, I. c., RUSSELL, R. R. B.: Lipoproteins of Gram-posi­tive bacteria. J. Bacteriol. 177, 1123-1128 (1995).

THALLER, M. c., BERLUTTI, E, SCHIPPA, S., lORI, P., PASSARIELLO, c., ROSSOLlNI, C. M.: Heterogeneous patterns of acid phos­phatases containing low-molecular-mass polypeptides in members of the family Enterobacteriaceae. Int.]. Syst. Bacte­riol. 45, 255-261 (1995) .

THALLER, M. c., SCHIPPA, S., ROSSOLlNI, C. M.: Conserved se­quence motifs among the bacterial, eukaryotic and archeal phosphatases that define a new phosphohydrolase superfam­ily. Prot. Sci. 7, 1647-1652 (1998).

VINCENT, ]. B., CROWDER, M.W., AVERILL, B.A.: Hydrolysis of phosphate monoesters: a biological problem with multiple chemical solutions. Trends Biochem. Sci. 17: 105-110 (1992).

Address for correspondence: lAIN SUTCLIFFE, Fleming Building, School of Sciences, The Uni­versity of Sunderland, Sunderland SR2 3SD, UK Telephone: + 441915152995; Fax: + 44191515 3747 e-mail: iain.sutcliffe@sunderland.ac.uk