virulence of protein a-deficient and alpha-toxin-deficient ... · 3104 patel et al. table 1....

Vol. 55, No. 12INFECTION AND IMMUNITY, Dec. 1987, p. 3103-31100019-9567/87/123103-08$02.00/0Copyright © 1987, American Society for Microbiology

Virulence of Protein A-Deficient and Alpha-Toxin-Deficient Mutantsof Staphylococcus aureus Isolated by Allele ReplacementARVIND H. PATEL,' PETER NOWLAN,2 EDWARD D. WEAVERS,3 AND TIMOTHY FOSTER'*

Microbiology Department, Moyne Institute,' and Wellcome Research Laboratories,2 Trinity College, University of Dublin,Dublin 2, and Veterinary Research Laboratories, Abbottstown, Dublin 15,3 Ireland

Received 17 June 1987/Accepted 14 September 1987

The gene coding for protein A (spa) of Staphylococcus aureus 8325-4 has been inactivated by substituting partof the spa coding sequence for a DNA fragment specifying resistance to ethidium bromide. The in vitro-constructed spa::EtBrr substitution mutation was introduced into the S. aureus chromosome by recombi-national allele replacement. Southern blot hybridization showed that the in vitro-constructed mutation was

present in the chromosomal spa locus. We have previously reported the inactivation of the alpha-toxin gene

(hly) by allele replacement with an in vitro-constructed hly::Emr (erythromycin resistance) mutation (M.O'Reilly, J. C. S. de Azavedo, S. Kennedy, and T. J. Foster, Microb. Pathogen. 1:125-138, 1986). A doubleSpa- Hly- mutant was constructed by transduction. The virulence of Spa- and Hly- mutants was tested byexperimental infection of mice. When subcutaneous injections were given, Hly- mutants formed a flat,darkened lesion, whereas Hly+ strains caused a raised, cream lesion. Alpha-toxin was shown to be a majorfactor in forming subcutaneous lesions and in causing the death of mice injected intraperitoneally. Spa-mutants were slightly less virulent than their Spa' counterparts, which suggests that protein A is also a

virulence factor of S. aureus.

Staphylococcus aureus is a pathogenic bacterium whichcauses a variety of infections in humans, including endocar-ditis, osteomyelitis, wound sepsis, skin abcesses, and theeconomically important disease of dairy cattle, mastitis (fora review, see reference 12). The organism can produce anarray of potential virulence factors, such as alpha-, beta-,gamma- and delta-toxins, leucocidin, coagulase, and proteinA. However, apart from alpha-lysin (1, 20, 46) and coagulase(20), direct evidence that any of these factors is required forthe pathogenesis of staphylococcal infections is lacking.

Protein A is a cell wall-associated protein of S. aureus (14,15, 23, 40). It has the unusual property of binding to the Fcregion of immunoglobulins from different mammalian spe-cies. Molecular cloning and DNA sequencing have shownthat protein A from strain 8325-4 is composed of fiverepeated units, each of which is an immunoglobulin G(IgG)-binding domain (25, 28, 43). In addition, the carboxy-terminal region is required for the linkage of protein A to thecell envelope of S. aureus (15).

Protein A is synthesized mainly in the logarithmic phase ofgrowth (4, 29). In contrast, many other exoproteins (e.g.,alpha-toxin) are expressed predominantly during the station-ary phase of growth (4, 32, 33, 35, 36). There is evidence thatthe protein A gene is controlled negatively and that alpha-toxin is controlled positively, possibly at the transcriptionallevel, by a regulatory gene called agr (4, 15, 30, 35, 36, 44).

It is probable that protein A has a role in the pathogenesisof staphylococcal infections. It produces a variety of biolog-ical effects when administered to laboratory animals or whentested in vitro (14, 23). These include hypersensitivity reac-tions, release of histamine from basophils, activation ofcomplement, and derepression of the opsonizing activity ofserum. One approach to establishing the importance of sucha factor in virulence is to examine the ability of mutants tocause experimental infections (42). However, previous vir-

* Corresponding author.

ulence studies with protein A-deficient mutants isolated bychemical mutagenesis have given ambiguous results (20, 45).To assess the relevance of protein A in virulence and to

avoid the problems of chemical mutagenesis, we have iso-lated a protein A-deficient mutant of S. aureus 8325-4 byallele replacement site-directed mutagenesis. The virulenceof this mutant and a previously isolated mutant deficient inalpha-toxin (32) was tested in experimental infections ofmice.

MATERIALS AND METHODS

Bacterial strains, plasmids, and bacteriophages. The bacte-rial strains are shown in Table 1, and the plasmids are shownin Table 2. The replacement vector XL47.1 (24) was used toclone the spa gene. The staphylococcal phages 80a and 85,which were used to transduce plasmids and chromosomalmarkers, respectively, were obtained from H. Pomeroy(Microbiology Department, Trinity College, Dublin) as wasthe lysogenic converting phage 42E. S. aureus strains werelysogenized with phage 42E as described previously (32) andconsequently lost the ability to produce beta-toxin (9, 48).Nomenclature and abbreviations. The structural genes for

protein A and alpha- and beta-hemolysins are called spa,hly, and hlb, respectively. The ability to express theseproteins is referred to as Spa', Hly+, and Hlb+. Invitro-constructed mutations in the spa and hly genes aredesignated Aspa: :EtBrr and hly: :Emr. EtBrr and Emr refer toresistance to ethidium bromide (EtBr) and to erythromycin,respectively.

Bacteriological media. Escherichia coli strains were grownin L broth and L agar (11, 30), whereas S. aureus wasroutinely cultured in Trypticase soy broth and agar (BBLMicrobiology Systems, Cockeysville, Md.), and peptonewater (Oxoid Ltd., London, England) was used forresuspending cells before inoculation. A phages were prop-agated in X base and top agar (11, 27). Bernheimer-Schwartzbroth (3) was used for culturing S. aureus for measuring the

3103

on Novem

ber 2, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

http://iai.asm.org/

3104 PATEL ET AL.

TABLE 1. Bacterial strains

Strain Genotype Relevant properties Source orreference

S. aureus8325-4 spa+ hly+ hlb+ NCTC 8325 cured of prophages; plasmid-free 31RN4220 spa+ hly+ hlb+ Mutant of 8325-4 capable of accepting shuttle plasmids 21

propagated in E. coliDU1090 spa' hly421::Emr hlb+ Defective in alpha-hemolysin; carries plasmid pTS01 32DU5723 Aspa-600::EtBrr hly+ hlb+ Protein A-deficient mutant of 8325-4; carries plasmid pEC1 This studyDU5724 Aspa-600::EtBrr hly421::Emr hlb+ DU1090 x DU5723; carries plasmid pEC1 This studyDU5719 spa+ hly+ hlb Phage 42E lysogen of 8325-4; plasmid-free This studyDU5721 Aspa-600::EtBrr hly+ hlb Phage 42E lysogen of DU5723; plasmid-free This studyDU5722 Aspa-600::EtBrr hly-421::Emr hlb Phage 42E lysogen of DU5724; plasmid-free This studyDU5720 spa' hly421::Emr hlb Phage 42E lysogen of DU1090; plasmid-free This studyDU5729-5733 Spontaneous EtBrr mutants of 8325-4 This studyDU5734-5738 Spontaneous CTABr mutants of 8325-4 This study

E. coliC600 lac thr leu thi tonA hspR hspM 2TB1 ara thi rpsL Alac-proxjii (+80dlacl Host for detecting chimeric pUC plasmids 49

AM15) hspR

expression of extracellular products in culture supernatants. sively with bovine albumin (Sigma), rabbit anti-protein ABernheimer-Schwartz agar incorporating 1% dog serum was serum, and peroxidase-conjugated swine anti-rabbit serumused for detecting protein A production. Egg yolk medium by the procedure described previously (33, 38). Expression(Oxoid) and toluidine blue DNA (TBD) agar (39) were used of Spa by E. coli colonies was detected by the colonyfor estimating lipase and DNase production, respectively. immunoblotting procedure of Helfman et al. (16).

Chemicals, antibiotics, and enzymes. Laboratory chemicals Immunochemical analysis of Spa produced in E. coli and S.were obtained from Sigma Chemical Co. (St. Louis, Mo.) or aureus. Proteins produced in Xspa lysates of E. coli werewere the best grade available from BDH (Poole, England). prepared as described previously (33). Protein A expressionAmpicillin was a gift from Beecham Laboratories (Brent- by S. aureus was determined after growth for 18 h inford, United Kingdom). Chloramphenicol, erythromycin, Bernheimer-Schwartz broth. Cell-bound protein A was de-EtBr, 5-bromo-4-chloro-3-indolyl-,3-D-galactopyranoside tected as described previously (18). Proteins were subjected(X-Gal), and cetyltrimethylammonium bromide (CTAB) to sodium dodecyl sulfate-polyacrylamide gel electrophore-were purchased from Sigma. Restriction enzymes and T4 sis (22) and analyzed by Western immunoblotting (6).DNA ligase were obtained from the Boehringer Corporation Expression of S. aureus extracellular products. Protein Aand were used according to the instructions of the manufac- production was detected as a precipitin ring surroundingturer. colonies stab-inoculated on Bernheimer-Schwartz agar in-

Antisera. Rabbit anti-protein A serum was purchased from corporating 1% dog serum. Production of alpha-, beta-,Sigma. It was adsorbed with a concentrated extract of E. coli delta-, and gamma-hemolysins was tested as described pre-C600 before use. Peroxidase-conjugated swine anti-rabbit viously (32). Lipase and protease were detected by theglobulin was obtained from Dakopatts (Glostrup, Denmark). appearance of clear zones around colonies growing on egg

Transformation. Plasmids were transformed into cells of yolk agar after incubation at 37°C for 24 h. DNase activityE. coli C600, and TB1 was made competent with CaCl2 was indicated by a pink zone around wells in TBD agartreatment (26) and into protoplasts of S. aureus (8, 13, 32). loaded with 20 ,ul of culture supernatant and incubated at

Plaque and colony immunoblotting. X plaques producing 37°C for 18 h. Coagulase expression was tested by mixingprotein A were detected by laying 82-mm-diameter nitrocel- culture supernatants (0.75 ml) with 0.25 ml of neat rabbitlulose disks (Schleicher & Schuell, Inc., Keene, N.H.) onto plasma and incubating for 18 h at 37°C.phage overlay plates. The filters were incubated succes- MIC determinations. Narrow-interval MIC determinations

TABLE 2. Plasmids

Plasmid Host Markers Other relevant properties Source orreference

pACYC184 E. coli Cmr Tcr Cloning vector 7pUC18 E. coli Apr Cloning vector 49pSPA721 E. coli Tcr Spa' 4.3-kb EcoRI fragment from Aspa cloned in pACYC184 This studypEBP962 E. coli Apr pUC18 with deleted spa gene This studypEBM1 E. coli Apr EtBrr 2-kb BglII EtBr' fragment from pCW59-Ebr in Bcll site of This study

pEBP962; Aspa::EtBrrpE194 S. aureus Emr Naturally occurring plasmid 17pC221 S. aureus Cmr Naturally occurring plasmid 5pEC1 S. aureus Cmr Replication origin (1.14-kb Taql fragment) of pE194 and Cmr This study

determinant (1.16-kb TaqI fragment) of pC221pCW59-Ebr S. aureus Tcr EtBr 2-kb BglIl fragment from pA118 cloned in pCW59 19pEBM15 Apr Emr EtBrr Shuttle plasmid; pEBM1 linked to pE194; carries Aspa::EtBrr This study

INFECT. IMMUN.

on Novem


.org/D

ownloaded from

http://iai.asm.org/

VIRULENCE OF Spa- AND Hly- S. AUREUS MUTANTS

with EtBr and CTAB were performed as described previ-ously (10).

Molecular cloning and hybridization. A library of S. aureus

8325-4 DNA was constructed in XL47.1 as described previ-ously (33), except that the genomic DNA was cleaved withSau3AI and the A replacement vector was cut with BamHI.Subcloning and restriction mapping experiments were car-

ried out by standard procedures (26). DNA probes werelabeled by nick translation in vitro (26) with 20 ,uCi of[oc-32P]dATP (Amersham Corp., Bucks, United Kingdom).Southern blot hybridization (41) was carried out as describedpreviously (26).

Construction of pEC1. About 1 ,ug of a 1.14-kilobase (kb)TaqI fragment carrying the origin of replication of plasmidpE194 (17) was ligated with an equal amount of a 1.16-kbTaqI fragment carrying the Cmr determinant of pC221 (5).Cmr transformants of S. aureus RN4220 were isolated, andone was shown to have the desired structure.

Construction of the Aspa::EtBrr mutation in vitro. Samples(20 ,ug) of pSPA721 DNA were cleaved with PstI and Bcll,then separately cleaved with BclI and EcoRI, and fraction-ated on 1.2% agarose gels. A 0.6-kb PstI-BclI fragment anda 0.9-kb BclI-EcoRI fragment were isolated from the gels byelectroelution (26), ligated with 0.5 jig of PstI- and EcoRI-cleaved pUC18 DNA, and transformed into E. coli TB1.White colonies appearing on L agar plates incorporatingampicillin (100 ,ug/ml) and X-Gal were screened for plasmidscarrying both fragments. One plasmid with the Aspa muta-tion (pEBP962; Fig. 1) was kept for further study. A BglIIfragment from pCW59-Ebr, which expresses resistance toEtBr (19) was introduced into the Bcll site of pEBP962 toform pEBM1 and create the Aspa::EtBrr mutation (Fig. 1).Recombinants were detected by replica plating Apr trans-formants onto agar incorporating EtBr (1,000 ,Ig/ml). Ashuttle plasmid (pEBM15; Fig. 1) was then constructed byligating PstI-cleaved pEBM1 DNA with the S. aureus plas-mid pE194 cut with the same enzyme. Recombinants weredetected by replica plating Apr transformants onto agarincorporating erythromycin (100 p.g/ml).

Inactivation of the spa gene of S. aureus 8325-4. The shuttleplasmid pEBM15 carrying the insertionally inactivated spagene was transformed into protoplasts of S. aureus RN4220.The plasmid was then transduced into 8325-4 along withpEC1 to construct a double-plasmid strain. EtBrr Emr Cmrcolonies were purified and inoculated into broth containingchloramphenicol. After growth to saturation, the cultureswere diluted 10-4 into fresh broth containing chloram-phenicol and grown to saturation again. This was repeateduntil the cells had gone through 200 generations. At thisstage, the culture was completely devoid of pEBM15 (therewere no Emr cells). Dilutions were plated on agar incorpo-rating 7 ,ug of EtBr per ml, and resistant colonies weretoothpick stab-inoculated onto dog serum agar.

Virulence tests in mice. S. aureus cultures were grown for18 h in 50 ml of Trypticase soy broth in 250-ml flasks shakingat 150 rpm. Cells were harvested by centrifugation (10,000 x

g for 10 min), washed twice in peptone water, and finallyresuspended in 1/10 the original culture volume of peptonewater. The densities of the cell suspensions were adjusted toan A580 of 45 + 3 (2 x 1010 CFU/ml) (the absorbance was

measured after diluting 1/100 in peptone water). The suspen-sions were diluted in saline, and 0.1-ml volumes were

injected subcutaneously in female Sha Sha mice aged be-tween 8 and 15 weeks. After 24 h, the mice were killed andthe sizes of the lesions (if any) were measured with calipers.The dimensions of the lesions were measured at the widest

E !'E.V~~~~,P '1 ~~ T Eh

E PUC a E11-EwucISP-2E pEBP962

12E u e B7f_t

P m PB/Bg J9 W3 Bg E

E w iC,, E pEBM15E~~~~~E

X X EjsCS

5

FIG. 1. Construction of a mutation in spa by allele replacement.The steps involved in the construction of the Aspa::EtBrr mutationin vitro and replacement of the wild-type spa gene in strain 8325-4 byrecombination are shown. The shaded regions of the thick horizon-tal lines correspond to regions of homology between the in vitro-constructed spa mutation and the spa gene in the chromosome of S.aureus. Steps 1 through 3 were performed in E. coli, whereas steps4 to 5 occurred in S. aureus. E, EcoRI; B, BclI; P, PstI; V, EcoRV;H, HindIII; T, TaqI; Bg, BglII; Ebr, EtBrr; CS, chromosome of S.aureus. The dashed line indicates the BclI fragment deleted fromspa.

and narrowest point, and the approximate area was esti-mated by multiplying the two values. The lesion areasgenerated by the same inoculum of any given pair of organ-isms were ranked and analyzed according to the Mann-Whitney U-test (47). Sample animals were immersed insaline containing 10% Formalin, and lesions were examinedby routine histological procedures.

Bacteria prepared in the same way were also injectedintraperitoneally in female Laca mice aged 8 to 9 weeks. Theanimals were observed at intervals for 48 h, and the time ofdeath was recorded.

Histological examination of skin lesions. Animals werekilled 24 h postinoculation by cervical dislocation. Sites forsectioning were marked, and then whole animals were fixedin 10% Formalin for 48 h. Lesions were removed, trimmed,and embedded in paraffin wax. Sections were cut andprocessed by routine histological procedures. They werestained by the method of Gram and with hematoxylin andeosin.

RESULTSMolecular cloning and mapping of the spa gene. A Aspa

phage was detected in a gene library of S. aureus 8325-4

VOL. 55, 1987 3105

on Novem


.org/D

ownloaded from

http://iai.asm.org/

3106 PATEL ET AL.

DNA cloned in XL47.1 by plaque immunoblotting. The spagene was subcloned into pACYC184 on a 4.3-kb EcoRIfragment to form pSPA721 (Fig. 1). A restriction map of thisEcoRI fragment (Fig. 1) agreed with those published byLofdahl et al. (25) and Uhlen et al. (43).

Isolation of a Aspa::EtBr' mutation. Spa- mutants of S.aureus 8325-4 were isolated by allele replacement site-directed mutagenesis by using a strategy first used in thisorganism by O'Reilly et al. (32) for isolating mutationsde'fective in alpha-toxin. A 1.2-kb BclI fragment which isinternal to the spa gene and which codes for the fiveIgG-binding domains and part of the cell wall attachmentregion of the protein (15, 25, 28, 43) was deleted frompSPA721 in vitro. Then a 2-kb BglII fragment carrying anEtBrr determinant from pCW59-Ebr which specifies low-level resistance to EtBr and CTAB (Table 3) was insertedinto the Bcll site of pEBP962 to form the Aspa::EtBrrmutation on plasmid pEBM1 (Fig. 1). A shuttle plasmid(pEBM15 EtBrr Emr Apr) was constructed (Fig. 1) andtransferred into S. aureus 8325-4, along with an incompatibleplasmid pECI (Cmr).

After growth for 200 generations in broth containingchloramphenicol (to maintain selection for pEC1), the.spa::EtBrr plasmid pEBMiS had been completely elimi-nated from the population. Cells were plated on EtBr andCTAB agar, and the resultant colonies were tested for Spaproduction on dog serum agar. It should be noted thatcontrol experiments showed that spontaneous EtBrr andCTABr mutants of S. aureus arose at a frequency of about 5X 10-5. DNA dot-blot hybridization was used to identify thedesired recombinants with the Bcll fragment from within spa(no hybridization) and the 2-kb BglII fragment of pCW59-Ebr(positive hybridization) as probes. One mutant, DU5723,was kept for further analysis.

Several spontaneous EtBrr and CTABr mutants werepurified, and the resistance of these mutants to EtBr andCTAB compared with those of a strain harboring pEBM15(which carries the EtBrr determinant from pCW59-Ebr) andthe chromosomal spa::EtBrr recombinant (Table 3). Theplasmid-derived EtBrr determinant expressed a lower levelof resistance to both EtBr and'CTAB when it was integratedin the spa locus in the chromosome. The spontaneousmutants exhibited a variety of resistance phenotypes. Gen-erally, the mutants selected on CTAB expressed crossresistance to EtBr, whereas those selected on EtBr were notresistant to CTAB. The levels of EtBrr and CTABr werehigher than that expressed by the spa::EtBrr recombinant.

TABLE 3. Resistance to EtBr and CTAB

MIC (,ug/ml)Strain

EtBr CTAB

8325-4 1.0 1.08325-4(pEBM15) 22 10DU5723 Aspa::EtBrr 4.0 5.0DU5729 10 3.0DU5730 7.0 1.0DU5731 14 2.0DU5732 4.0 1.0DU5733 4.0 1.0DU5734 9.0 5.0DU5735 9.0 5.0DU5736 9.0 3.0DU5737 9.0 3.0DU5738 9.0 2.0

1 2 3 4 5 6

1 69.3

5.94.7

3.73.3

2.6

1.9

1.2

FIG. 2. Southern blot hybridization of S. aureus Aspa. Chromo-somal DNA from S. aureus strain 8325-4 (lanes 2, 4, and 6) andDU5723 (A&spa::EtBr') (lanes 1, 3, and 5) was cleaved with EcoRIand fractionated on a 1% agarose gel. The probes were the 1.2-kbBcll fragment from within spa (lanes 1 and 2), the 2-kb BglIIfragment from pCW59-Ebr (lanes 3 and 4), and the EcoRI fragmentcarrying spa' from pSPA721 (lanes 5 and 6).

Characterization of S. aureus 8325-4Aspa::EtBrr. Southernblot hybridization was performed to verify that the Spa-mutant had the predicted structural changes in the spa locus.Chromosomal DNA from the mutant and the parental Spa'strain was cleaved with EcoRI and' probed with the 1.2-kbBclI fragment (which is intragenic to spa), the 2-kb BglIIfragment (which carries the inserted EtBrr determinant), andthe 4.3-kb EcoRI fragment from pSPA721 (which containsthe intact spa gene).A 4.3-kb fragment in 8325-4 hybridized when chromo-

somal DNA cleaved with EcoRI was probed with the 1.2-kbBclI fragment from the spa gene (Fig. 2, lane 2), whereas themutant failed to hybridize (Fig. 2, lane 1). In contrast,probing with the inserted 2-kb BglII fragment revealed a5.1-kb EcoRI fragment in mutant strain DU5723, whereasthe parental strain failed to hybridize (Fig. 2, lane's 3 and 4).A fragment of 5.1 kb in the mutant also hybridized to the4.3-kb spa'probe (Fig. 2, lane 5), whereas the homologousfragment hybridized in the wild-type strain (Fig. 2, lane 6).This data is consistent with the Spa- phenotype beingcaused by recombination of the in vitro-constructedAspa::EtBrr mutation into the genome of 8325-4 in place ofthe wild-type spa allele.Western blot analysis showed that the mutant strain

DU5723 lacked detectable protein A (Fig. 3, lanes 1 and 2),whereas the parental strain 8325-4 expressed both cell-freeand cell wall-bound forms (Fig. 3, lanes 3 and'4).The mutant strain produced similar levels of alpha-, beta-,

delta-, and gamma toxins and coagulase, lipase, and DNaseas the parental strain (data not shown). Furthermore, thestrains had identical phage susceptibility patterns. Thus, theSpa- mutant does not appear to be pleiotropic.

Construction of a protein A- and alpha-toxin-defectivemutant of S. aureus 8325-4. The alpha-hemolysin gene of8325-4 has been previously inactivated by insertion of anEmr marker by allele replacement mutagenesis (32). The

INFECT. IMMUN.

on Novem


.org/D

ownloaded from

http://iai.asm.org/


1 2 3 4

9467

43

30

20

14

FIG. 3. Western imnmunoblotting of S. aureus extracts. S. aureus

8325-4 spa' (lanes 3 and 4) and DU5723 Aspa::EtBrr (lanes 1 and 2)were analyzed. Protein in cell walls (lanes 1 and 3) and in culturesupernatants (lanes 2 and 4) was extracted and fractionated bysodium dodecyl sulfate-polyacrylamide gel electrophoresis andtransferred onto nitrocellulose. Filters were incubated with rabbitanti-protein A antibodies and then with peroxidase-conjugatedswine anti-rabbit IgG serum. Proteins of known molecular weight (inthousands) are indicated.

hly::Emr locus was transduced from strain DU1090 intostrain DU5723 to construct the doubly defective mutantstrain DU5724. Before use in virulence tests, the strainswere cured of plasmids and then lysogenized with phage 42Ewhich prevents expression of beta-hemolysin by insertioninto the hlb gene (9). The latter manipulation was performedto allow our results to be compared with those of Jonsson etal. (20), who used strain SA113, a derivative of strain 8325which is lysogenic for the staphylokinase converting phage.413 and which consequently does not produce beta-toxin.Virulence tests with S. aureus mutants. Female Sha Sha

mice were injected subcutaneously with strains DU5719,DU5720, DU5721, and DU5722. Lesions were examinedafter 24 h. The mice were sacrificed, and the area of thelesions was measured. Tissue sections were examined his-tologically.

Strains which produce alpha-toxin caused a raised, creamlesion with an outer reddened area (Fig. 4). In contrast,alpha-toxin-negative strains caused a lesion which had a

blackened, sunken center (Fig. 4). This was surrounded by a

raised, cream zone and an outer, reddened area. Micro-scopic examination of stained sections of infected areas

revealed an acute seropurulent inflammatory reaction. Inmice infected with the Hly+ strain, DU5719 neutrophilinfiltration is confined mainly to the periphery of the lesion(Fig. SA.) Few neutrophils are present in the center and are

rarely seen in close contact with bacterial cells. In miceinfected with the Hly- strain, DU5720, neutrophils are

present in large numbers (Fig. 5B). The region containingbacteria is surrounded by neutrophils, a high proportion ofwhich are necrotic.The alpha-toxin-producing strains were at least 10-fold

more virulent than those which lacked the toxin. Thus 2 x108 CFU of the Hly+ Spa' strain DU5719 caused lesionswith a mean area of 866 mm2, whereas the otherwise isogenicHly- strain DU5720 caused lesions with a mean area of 170

A

_<.

.t.i1 t

B

FIG. 4. Skin lesions in Sha Sha miice. (A) Mouse injected sub-cutaneously with S. aureus DU5719 (8325-4 hlb; 108 CFU). (B)Mouse injected with DU5720 (8325-4 hly::Emr hlb; 2 x 109 CFU).The photographs were taken 24 h postinoculation. The arrowsindicate the lesions.

mm2 (Table 4). A similar effect was seen when the Hly+Spa- strain DU5721 was compared with the Hly- Spa-strain DU5722 (Table 4). This shows that alpha-toxin is amajor determinant of subcutaneous abscesses in mice.The data (Table 4) also suggest that protein A contributes

to the size of the lesions caused by both Hly+ and Hly-strains. The differences between Spa' and Spa- strains wereless pronounced than those seen with alpha-toxin. Theabsence of Spa did not affect the nature of the lesion, only itssize.The virulence of the mutant strains was also assessed by

measuring the survival of Laca mice injected intraperitone-ally. At least fourfold more Hly- bacteria were required tokill mice at the same rate as the otherwise isogenic Hly+strains (Table 5). This result was obtained with both theSpa' and Spa- strains. In addition, the Spa- mutants weremarginally but reproducibly less virulent than their Spa'counterparts, although some of the comparisons were notsignificant at the 5% level (Table 5).To confirm this result, pools of 12 mice were injected

TABLE 4. Subcutaneous lesions in Sha Sha micecaused by S. aureus mutants

Inoculum Mean lesion sizea (mm2)(CFU) Hly+ Spa' Hly' Spa- Hly- Spa' Hly- Spa

2 x 109 170.6 342 x 108 866.2 272.8 29.1 02 x 107 63.1 7.81X 107 10.6 0

a Each number is the mean area of lesions on six independently inoculatedanimals. Each pairwise comparison was statistically significant by the Mann-Whitney U-test (P = 5%).

VOL. 55, 1987 3107

on Novem


.org/D

ownloaded from

http://iai.asm.org/

3108 PATEL ET AL.

^r 8 f4. t R. { ;..

.,, 4z;#Jt.8W, {.. <,X4. .," .. .

1;s f ........... .j. 1. ,d .:._Ws sg f p*

t,@..," s :¢' .t 0' Uf 8

w .. @, .'. jj iw .' p ;> Mrr. .. -wS ... e .j f jp. p ... r .. .. ., _¢Ss .t s t *r ts .W s *

.,ff ? Q , '8 ^

, .i, @ ^ t* s.>

2. *sz! tri

A.- *.

e, .0

i. b

., "IF , -V 'EL-J I ;0, Al -:. _'4W.

FIG. 5. Skin lesions in Sha Sha mice. Lesions formed as described in the legend to Fig. 4 were sectioned and stained with hemotoxylinand eosin. Photomicrographs were taken at high power. Arrows indicate area of bacterial proliferation (B) and neutrophils (A) in the lesioncaused by DU5719 and bacterial cells plus neutrophils (C) in the lesion caused by DU5720.

intraperitoneally with 5 x 108 bacteria (Hly+ Spa' and Hly+Spa-) or with 2 x 109 bacteria (Hly- Spa' and Hly- Spa-)and the time of death was recorded at four hourly intervals.In both tests, the Spa' strain killed the mice significantlymore quickly (P = 5%) than the otherwise isogenic Spa-strain.

DISCUSSION

This paper describes the isolation and characterization ofa site-specific substitution mutation in the chromosomal spa

gene of S. aureus strain 8325-4. The sequences coding for thefive IgG-binding domains in Spa were deleted and werereplaced by a fragment specifying resistance to EtBr andCTAB. Allele replacement by recombination was used toinactivate the chromosomal spa gene by a plasmid incom-patibility technique we have previously used to inactivatethe hly gene of S. aureus (32). However, the EtBrr markerwas not ideal for this purpose, because the resistance levelwas reduced substantially when the determinant was inte-grated in the chromosome, possibly due to a reduction in

TABLE 5. Survival of Laca mice after intraperitoneal injection with S. aureus mutants

Inoculum Times of death (h) of individual animals"(CFU) Hly+ Spa' Hly+ Spa- Hly- Spa' Hly Spa-2 x 109 10, 10, 10, 22, 27, 27 27, 27, 27, 29, 44b1 x 109 22, 27, 27, 29, S, S 27, 29, 44, 44, 48C5 x 108 6, 6 6, 6, 6, 6 10, 10, 10, 10, 10, Sb S, S S, S, S, S S, S, S, S, S

2.5 x 108 6, 6, 10, 10, S, S S, S, S, S, S, SC1.25 x108 10,S, S, S, S, S S, S, S, S, S, SC

a Animals were observed at the following times (in hours) postinfection: 6, 10, 22, 27, 29, 44, and 48 h. Each value given represents one animal. S indicatesthat the animal survived the 48-h observation period.

b Significant at P = 5%.c Not significant at P = 5%.

INFECT. IMMUN.

on Novem


.org/D

ownloaded from

http://iai.asm.org/


gene copy number. It was difficult to detect the desiredrecombinants among spontaneous EtBrr and CTAB' mutantsthat occurred on selective plates. We have recently inacti-vated the spa gene with an aminoglycoside resistance deter-minant. This is a much better selective marker which shouldpermit the spa locus to be mapped in the 8325 chromosomeby protoplast fusion and transformation in the same way asthe hly locus (34).

Previous virulence studies with alpha-toxin mutants iso-lated by allele replacement suggested that the toxin contrib-uted to the virulence of S. aureus in mice (32), an observa-tion which we have confirmed here with subcutaneous andperitonitis infections. This is not surprising because alpha-toxin is a potent exotoxin of S. aureus and has been stronglyimplicated as a virulence factor by other studies (1, 20, 46).

In addition to the quantitative differences in virulencebetween Hly+ and Hly- strains measured in the subcutane-ous infections, the lesions obtained were substantially dif-ferent both superficially (Fig. 4) and histologically (Fig. 5).The absence of neutrophils in the vicinity of bacteria inlesions caused by Hly+ strains is consistent with the obser-vation that purified alpha-toxin inhibits neutrophil chemo-taxis in vitro (37).We also demonstrate that protein A makes a contribution

to both types of infection. Spa- mutants formed smaller skinlesions than their Spa' counterparts, as well as being slightlyless virulent when injected intraperitoneally. It will beinteresting to test our mutants in other experimental infec-tions of mice and other animal species and also to examinethe ability of the mutant bacteria to grow in diffusionchambers or in granuloma pouches.These results contradict a previous report that most Spa-

mutants of strain SA113 were not less virulent for mice (20).One Spa- mutant which did have lowered virulence wassubsequently also shown to be deficient in a protein thatbinds fibronectin (45). Our finding that Spa appears to be avirulence factor for staphylococcal infections of mice couldbe due to the use of different animal models or to differencesbetween the bacterial strains used.Although allele replacement is the most satisfactory

method for isolating mutations affecting virulence, somedegree of caution must be exercised when interpretingvirulence experiments. First, the drug resistance insertmight affect the expression of other genes due to polarity.This is unlikely for spa, since the gene is flanked bysequences resembling transcription terminators (43), sug-gesting that it is monocistronic. However, it is not known ifhly is monocistronic or if it is expressed as part of a largertranscription unit.

Second, it is conceivable that the inserted drug resistancemarker is responsible for the reduction in virulence. The bestway to control for this is to isolate a deletion in the chromo-somal gene. However, attempts to isolate a deletion in hly byallele replacement by using colony hybridization to detectthe recombinant have failed, possibly due to the low fre-quency of this type of recombination event (ca. 10-7, unpub-lished data). Virulence could be measured in a strain carry-ing the resistance marker on a plasmid, but controls wouldhave to be performed to ensure that the plasmid was notinterfering with virulence.

Finally, it is possible that the small reduction in virulenceof Spa- mutants is not directly due to loss of protein Aexpression but to a slower growth rate. We have measuredthe doubling times of the four strains during early exponen-tial growth and no significant differences were found (unpub-lished data).

ACKNOWLEDGMENTSWe would like to acknowledge the financial support of the

Wellcome Trust and the National Board for Science and Technol-ogy, Dublin.We also thank Keith Dyke for sending plasmid pCW59-Ebr,

Harriet Pomeroy for staphylococcal bacteriophages, Harriet Pom-eroy, Paul O'Toole, and Mary O'Reilly for helpful discussions, andTheresa Hogan for enthusiastic technical assistance.

LITERATURE CITED1. Adlam, C., P. D. Ward, A. C. McCartney, J. P. Arbuthnott, and

C. M. Thorley. 1977. Effect of immunization with highly purifiedalpha- and beta-toxins on staphylococcal mastitis in rabbits.Infect. Immun. 17:250-256.

2. Appleyard, R. K. 1954. Segregation of new lysogenic typesduring growth of a doubly lysogenic strain derived from Esch-erichia coli K12. Genetics 39:440-452.

3. Berheimer, A. W., and L. L. Schwartz. 1963. Isolation andcomposition of staphylococcal alpha toxin. J. Gen. Microbiol.30:455-468.

4. Bjorklind, A., and S. Arvidson. 1980. Mutants of Staphylococ-cus aureus affected in the regulation of exoprotein synthesis.FEMS Microbiol. Lett. 7:203-206.

5. Brenner, D. G., and W. V. Shaw. 1985. The use of syntheticoligonucleotides with universal templates for rapid DNA se-quencing: results with staphylococcal replicon pC221. EMBO J.4:561-568.

6. Burnette, W. N. 1981. "Western blotting": electrophoretictransfer of protein from sodium dodecyl sulphate polyacryl-amide gels to unmodified nitrocellulose and radiographic de-tection with antibody and radioiodinated protein A. Anal.Biochem. 112:195-203.

7. Chang, A. C. Y., and S. N. Cohen. 1978. Construction andcharacterization of amplifiable multicopy DNA cloning vehiclesderived from the plSA cryptic mini plasmid. J. Bacteriol.134:1141-1156.

8. Chang, S., and S. N. Cohen. 1979. High frequency transforma-tion of B. subtilis protoplasts by plasmid deoxyribonucleic acid.Mol. Gen. Genet. 168:111-119.

9. Coleman, D. C., J. P. Arbuthnott, H. M. Pomeroy, and T. H.Birkbeck. Cloning and expression in Escherichia coli and Staph-ylococcus aureus of the beta-lysin determinant from Staphylo-coccus aureus: evidence that bacteriophage conversion of beta-lysin activity is caused by insertional inactivation of thebeta-lysin determinant. Microb. Pathogen. 1:549-564.

10. Coleman, D. C., H. Pomeroy, J. K. Estridge, C. T. Keane, M. T.Cafferkey, R. Hone, and T. J. Foster. 1985. Susceptibility toantimicrobial agents and analysis of plasmids in gentamicin- andmethicillin-resistant Staphylococcus aureus from Dublin hospi-tals. J. Med. Microbiol. 20:157-167.

11. Davis, R. W., D. Botstein, and J. R. Roth. 1980. Advancedbacterial genetics. Cold Spring Harbor Laboratory, Cold SpringHarbor, N.Y.

12. Easmon, C. S. F., and C. Adlam (ed.). 1983. Staphylococci andstaphylococcal infections, vols. 1 and 2. Academic Press, Inc.(London), Ltd., London.

13. Fairweather, N., S. Kennedy, T. J. Foster, M. Kehoe, and G.Dougan. 1983. Expression of a cloned Staphylococcus aureusa-hemolysin determinant in Bacillus subtilis and Staphylococ-cus aureus. Infect. Immun. 41:1112-1117.

14. Fosgren, A., V. Ghetie, R. Lindmark, and J. Sjoquist. 1983.Protein A and its exploitation, p. 429-478. In C. S. F. Easmonand C. Adlam (ed.), Staphylococci and staphylococcal infec-tions, vol. 2. Academic Press, Inc. (London), Ltd., London.

15. Guss, B., M. Uhlen, B. Nilsson, M. Lindberg, J. Sjoquist, and J.Sjodahl. 1984. Region X, the cell-wall-attachment part ofstaphylococcal protein A. Eur. J. Biochem. 138:413-420.

16. Helfman, D. M., J. R. Feramisco, J. C. Fiddes, G. P. Thomas,and S. H. Hughes. 1983. Identification of clones that encodechicken tropomyosin by direct immunological screening of acDNA expression library. Proc. Natl. Acad. Sci. USA 80:31-35.

17. Horinouchi, S., and B. Weisblum. 1982. Nucleotide sequenceand functional map of pE194, a plasmid that specifies inducible

VOL. 55, 1987 3109

on Novem


.org/D

ownloaded from

http://iai.asm.org/

3110 PATEL ET AL.

resistance to macrolide, lincosamide, and streptogramin type Bantibiotics. J. Bacteriol. 150:804-814.

18. Janzon, L., S. Lofdahl, and S. Arvidson. 1986. Evidence for acoordinate transcriptional control of alpha-toxin and protein Asynthesis in Staphylococcus aureus. FEMS Microbiol. Lett.33:193-198.

19. Jones, I. G., and M. Midgley. 1985. Expression of a plasmidborne ethidium resistance determinant from Staphyloc occusaureus in Escherichia coli: evidence for an efflux system. FEMSMicrobiol. Lett. 28:355-358.

20. Jonsson, P., M. Lindberg, I. Haraldsson, and T. Wadstrom.1985. Virulence of Staphylococcus aureus in a mouse mastitismodel: studies of alpha hemolysin, coagulase, and protein A aspossible virulence determinants with protoplast fusion and genecloning. Infect. Immun. 49:765-769.

21. Kreiswirth, B. N., S. Lofdahl, M. J. Betley, M. O'Reilly, P. M.Shlievert, M. S. Bergdoll, and R. P. Novick. 1983. The toxicshock syndrome exotoxin structural gene is not detectablytransmitted by a prophage. Nature (London) 305:709-712.

22. Laemmli, U. K. 1970. Cleavage of the structural protein duringthe assembly of the head of bacteriophage T4. Nature (London)227:680-685.

23. Langone, J. J. 1982. Protein A of Staphylococcus aureus andrelated immunoglobulin receptors produced by streptococci andpneumococci. Adv. Immunol. 32:157-252.

24. Loenen, W. A. M., and W. J. Brammar. 1980. A bacteriophagevector for cloning large restriction fragments made with severalrestriction enzymes. Gene 10:249-259.

25. Lofdahl, S., B. Guss, M. Uhlen, L. Philipson, and M. Lindberg.1983. Gene for staphylococcal protein A. Proc. Natl. Acad. Sci.USA 80:697-701.

26. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecularcloning: a laboratory manual. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

27. Miller, J. H. 1972. Experiments in molecular genetics. ColdSpring Harbor Laboratory, Cold Spring Harbor, N.Y.

28. Moks, T., L. Abrahmsen, B. Nilsson, U. Hellman, J. Sjoquist,and M. Uhlen. Staphylococcal protein A consists of five IgG-binding domains. Eur. J. Biochem. 156:637-643.

29. Movitz, J. 1974. A study on the biosynthesis of protein A inStaphylococcus aureus. Eur. J. Biochem. 48:131-136.

30. Nordstrom, K., and M. Lindberg. 1978. Effects of streptomycinand novobiocin on Staphylococcus aureus gene expression. J.Bacteriol. 133:614-620.

31. Novick, R. P. 1963. Properties of a cryptic high-frequency trans-ducing phage in Staphylococcus aureus. Virology 33:155-166.

32. O'Reilly, M., J. C. S. de Azavedo, S. Kennedy, and T. J. Foster.1986. Inactivation of the alpha-haemolysin gene of Staphylococ-clis aureus by site-directed mutagenesis and studies on theexpression of its haemolysins. Microb. Pathogen. 1:125-138.

33. O'Toole, P. W., and T. J. Foster. 1986. Molecular cloning andexpression of the epidermolytic toxin A gene of Staphylococcusaureus. Microb. Pathogen. 1:583-594.

34. Pattee, P. A. 1986. Chromosomal map location of the alpha-hemolysin structural gene in Staphylococcus aureus NCTC

8325. Infect. Immun. 54:593-596.35. Recsei, P., B. Kreiswirth, M. O'Reilly, P. Schlievert, A. Gruss,

and R. P. Novick. 1985. Regulation of exoprotein gene expres-sion by agr. In The staphylococci, p. 701-706. Gustav Fischer-Verlag, Stuttgart, Federal Republic of Germany.

36. Recsei, P., B. Kreiswirth, M. O'Reilly, P. Schlievert, A. Gruss,and R. P. Novick. 1986. Regulation of exoprotein gene expres-sion in Staphylococcus aureus by agr. Mol. Gen. Genet.202:58-61.

37. Russell, R. J., P. C. Wilkinson, R. J. Mclnroy, S. McKay, A. C.McCartney, and J. P. Arbuthnott. 1976. Effects of staphylococ-cal products on locomotion and chemotaxis of human bloodneutrophils and monocytes. J. Med. Microbiol. 8:433-449.

38. Russell, R. R. B., D. Coleman, and G. Dougan. 1985. Expressionof a gene for glucan-binding protein from Streptococcus mutansin Escherichia coli. J. Gen. Microbiol. 131:295-299.

39. Shortle, D. 1983. A genetic system for analysis of staphylococ-cal nuclease. Gene 22:181-189.

40. Sjoquist, J., J. Movitz, I. B. Johansson, and H. Hjelm. 1972.Localization of protein A in the bacteria. Eur. J. Biochem.30:190-194.

41. Southern, E. M. 1975. Detection of specific sequences amongDNA fragments separated by gel electrophoresis. J. Mol. Biol.98:503-517.

42. Sparling, P. F. 1983. Applications of genetics to studies ofbacterial virulence. Philos. Trans. R. Soc. Lond. B. 303:199-207.

43. Uhlen, M., B. Guss, B. Nilsson, S. Gatenbeck, L. Philipson, andM. Lindberg. 1984. Complete sequence of the staphylococcalgene encoding protein A. A gene evolved through multipleduplications. J. Biol. Chem. 259:1695-1702.

44. Uhlen, M., B. Guss, B. Nilsson, F. Gotz, and M. Lindberg. 1984.Expression of the gene encoding protein A in Staphylococcusaureus and coagulase-negative staphylococci. J. Bacteriol.159:713-719.

45. Wadstrom, T., L. Switalski, P. Speziale, K. Rubin, C. Ryden, G.Froman, A. Faris, M. Lindberg, and M. Hook. 1985. Binding ofmicrobial pathogens to connective tissue fibronectin: an earlystep in localized and invasive infections, p. 193-207. In Thepathogenesis of bacterial infections. Bayer Symposium VIII.Springer-Verlag, Berlin.

46. Ward, P. D., C. Adlam, A. C. McCartney, J. P. Arbuthnott, andC. M. Thorley. 1979. A histopathological study of the effects ofhighly purified staphylococcal alpha and beta toxins on thelactating mammary gland and skin of the rabbit. J. Comp.Pathol. 89:169-177.

47. Wardlaw, A. C. 1985. Practical statistics for experimentalbiologists. John Wiley & Sons Ltd. Chichester, United King-dom.

48. Winkler, K. C., J. de Wart, and C. Grootsen. 1965. Lysogenicconversion of staphylococci to loss of P-toxin. J. Gen. Micro-biol. 39:321-333.

49. Yanisch-Perron, C., J. Vieira, and J. Messing. 1985. ImprovedM13 phage cloning vectors and host strains: nucleotide se-quences of M13mpl8 and pUC19 vectors. Gene 33:103-119.

INFECT. IMMUN.

on Novem


.org/D

ownloaded from

http://iai.asm.org/

virulence of protein a-deficient and alpha-toxin-deficient ... · 3104 patel et al. table 1....

Documents