Staphylococcus aureus menD and hemB Mutants Are as Infective as the Parent Strains, but the Menadione Biosynthetic Mutant Persists within the Kidney

JID Staphylococcus aureus menD and hemB Mutants Are as Infective as the Parent Strains, but the Menadione Biosynthetic Mutant Persists within the Kidney Donna M. Bates 2 Christof von Eiff 3 Peter J. McNamara 2 Georg Peters 3 Michael R. Yeaman 0 4 Arnold S. Bayer 0 4 Richard A. Proctor 1 2 0 Department of Medicine, University of California-Los Angeles (UCLA) Medical School and Harbor-UCLA Medical Center 1 Medicine, University of Wisconsin Medical School , Madison 2 Medical Microbiology and Immunology 3 Department of Medical Microbiology, University of Mu ̈nster Medical School , Mu ̈nster , Germany 4 Research and Education Institute , Torrance, California Small colony variants (SCVs) of Staphylococcus aureus were generated via mutations in menD or hemB, yielding menadione and hemin auxotrophs, respectively, and studied in the rabbit endocarditis model. No differences in the 95% infectious dose occurred between strains with regard to seeding heart valves (∼106 cfu) or other target organs. No differences were observed between the response of the hemB mutant to oxacillin therapy and that of the parent strain in any target tissues, and significant reductions in bacterial densities were seen in all tissues (compared with untreated controls). In contrast, oxacillin therapy did not significantly reduce bacterial densities of the menD mutant in either kidney or spleen and significantly reduced densities within vegetations. These data show that SCVs are able to colonize multiple tissues in vivo and that the menD mutation provides the organism with a survival advantage during antimicrobial therapy, compared with its parent strain, in selected target tissues. - Staphylococcus aureus small colony variants (SCVs) are found in patients with persistent, antibiotic-resistant, and recurrent infections [1–7] and have been recognized as an emerging problem [8]. One of the hallmark clinical features of S. aureus SCVs is their ability to persist despite aggressive antimicrobial therapy; some clinical infections have persisted for 150 years [1]. The frequency of isolation of SCVs from human infections has been found to be as high as 40% in patients with osteomyelitis [8], 1% in blood cultures when the sample was plated directly onto solid agar [9], and 130% in bronchial secretions or throat swabs from patients with cystic fibrosis [5]. Many of the clinical SCV isolates are defective in electron transport [1–4, 6–8]. SCVs from clinical infections are commonly auxotrophic for menadione and hemin, which are key cofactors in the formation of menaquinone and cytochromes, respectively, and are thus important components of the electron transport chain [1–3]. The alterations in electron transport in such SCVs result in multiple phenotypic changes, including slow growth, increased MICs of several antibiotics [2, 10], and reduced a-toxin production [1–4, 6–8]. Because a-toxin production by S. aureus that enter into host cells (e.g., endothelial cells) results in host cell lysis or induction of intracellular apoptosis, the lack of a-toxin production in SCVs would allow such bacteria to persist within these anatomic niches. These events may also protect intracellular SCVs from both the host defense system and antimicrobial exposures and may represent an important nidus for recurrent infections. In laboratory-generated SCVs, intracellular persistence has been attributed to an ability of such organisms to reside within mammalian cells without lysing them, as a result of reduced amounts of a-toxin production. However, no data directly comparing a parent and a genetically well-defined isogenic variant have been published that characterize the role of SCVs in vivo. In addition, naturally occurring clinical SCV isolates have a tendency to exhibit an unstable phenotype, which makes them unsuitable for study in relevant in vivo models. Therefore, we have generated stable SCVs via mutations in menD or hemB, yielding menadione and hemin auxotrophs, respectively, which reflect 2 of the most common auxotrophs among naturally occurring SCVs from clinical infections [1–9]. To study the relative infectivity and antibiotic-response profiles of the 2 parent-mutant strain sets, we used the rabbit endocarditis model, which is an acute multitarget tissue infection model. The aims of the present study were 3-fold: (1) to create parent-isogenic SCV strain sets that represent the 2 most common auxotrophies among clinical SCVs (i.e., menadione and hemin); (2) to compare in vivo the ability of the parent strains and the SCVs to infect damaged heart valves and to cause infective endocarditis and metastatic infection; and (3) to evaluate the impact of the 2 SCV genotypes on the organism’s capacity to resist antimicrobial therapy in the experimental endocarditis model. MATERIALS AND METHODS Growth conditions. For routine culture, Escherichia coli strains were grown in Luria-Bertani broth, and S. aureus strains were grown in trypticase soy broth in a shaking water bath with rotary agitation (at 250 rpm) for 3–8 h at 37 C, stopping the incubation when the medium became cloudy, which indicated that sufficient organisms were present for harvesting. Antibiotic-resistant E. coli strains were grown in medium containing ampicillin (100 mg/mL), and antibiotic-resistant S. aureus strains were grown in medium containing tetracycline (5 mg/mL) or erythromycin (5 mg/mL), as appropriate. Plasmid and strain construction. S. aureus 8325-4 containing an antibiotic resistance cassette within the chromosomal copy of menD was constructed in several steps (table 1). First, a 3.8-kb fragment containing the menD gene was generated by polymerase chain reaction (PCR) amplification of S. aureus RN6390 chromosomal DNA 285 bp upstream of menF and 535 bp downstream of menD, using primers flanking regions upstream of menF (5 -GAATTCGCTGTACCAACTAAAACGGGTAC-3 ) and downstream (5 -GAATTCGGCGATTGAGACAATCGTTGTTG-3 ) of menD, respectively. DNA sequences containing the BamHI restriction site, added to the 5 end of each primer, are underlined. The menFD PCR product was ttail cloned into pCRII( ), a vector prepared to facilitate ligation of PCR products that have overhanging 3 -thymidine residues to base pair with complementary terminal bases of the amplified DNA, generating pJM625. The BamHI fragment containing menFD from pJM625 was cloned into the BamHI site of pJM377, a derivative of pBluescript SK( ) lacking the EcoRI site, generating pJM682. The menD gene in pJM682 was disrupted by cloning the ermC into a unique EcoRI site within menD. A 1.9-kb PCR fragment containing the ermC gene was obtained by PCR amplification of pHB201 (kindly provided by Prof. B. Weisblum, University of Wisconsin, Madison), using primers containing EcoRI restriction sites on the 5 ends: 5 -TGGCTTATTGGCATCCTGGC-3 and 5 -TCGTGCGCTCTCCTGTTCC-3 . The ermC PCR product was then t-tail cloned into pCRII( ), generating pJM624. A 1.9kb EcoRI fragment containing the ermC gene was cloned from pJM624 into the unique EcoRI restriction site within menD, generating pJM664. The PstI fragment from pJM664, containing menFD::ermC, was cloned into similar sites in the temperaturesensitive shuttle vector pSPT181(ts) [14], generating pJM678. Plasmid DNA derived from E. coli strains was introduced into the S. aureus restriction minus strain, RN4220, by electroporation, as described elsewhere [15], before introduction to other strains of S. aureus. S. aureus 8325-4 containing a chromosomal insertion of ermC within menD was obtained by allelic exchange, using plasmid pJM678. The conditions used for plasmid integration and cointegrate resolution were as described elsewhere [14]. A pSPT181(ts) derivative containing the wild-type menFD sequence (pDB23) was obtained by cloning the 3.9-kb BamHI fragment from pJM625 into the BamHI site of pSPT181(ts). Plasmid pDB23 was used to restore the menD::ermC allele in S. aureus DB24 to the wild-type menD sequence by allelic exchange [14]. The hemB mutant has been described elsewhere [11]. F11 phage transductions for the menD mutant were performed as described elsewhere [12]. The mutation in hemB was repaired by complementing hemB in trans with pCE12. DNA isolation and Southern blot analysis. Chromosomal DNA was isolated from S. aureus strains as described elsewhere [16]. Plasmid DNA was obtained from S. aureus strains and purified with a Qiagen kit, with modifications described elsewhere [12]. Southern blot analysis was performed as described elsewhere [12]. MICs and kill curves. The oxacillin MICs for the study strains were determined by the broth dilution technique, fol Phage Genotype Source/reference 8325 cured of phages F11, F12, and F13 Restriction minus 8325-4 8325-4 hemB::ermB I10 complemented in trans with hemB on pCE12 8325-4 menD::ermC DB24 with menD restored by allelic exchange F F80lacZDM15 D(lacZYA-argF) U196 endA1 recA1 hsdR17(rk mk ) deoR thi-1 supE44 l gyrA96 relA1 phoA F {lacIq Tn10(TetR)} mcrA D(mrr-hsdRMS-mcrBC) F80lacZDM15 DlacX74 deoR recA1 araD139 D(ara-leu)7697 galU galK rpsL (StrR) endA1 nupG recA1 endA1 gyrA1 thi-1 hdsR17 supE44 relA1 lac [F proAB lacIqZDM15 Tn10 (TetR)] E. coli TOP10F pJM625 E. coli TOP10F pJM624 E. coli DH5a pJM678 E. coli XL1-Blue pJM664 E. coli DH5a pJM682 Generalized transducing phage Invitrogen Life Technologies Invitrogen Life Technologies Stratagene [12] [12] Invitrogen Life Technologies Present study Present study [13] Present study Present study NOTE. NCTC, National Collection of Type Cultures. lowing NCCLS protocols, using trypticase soy broth and final inocula of 105 and 107 cfu. To confirm the MICs, an agar dilution assay was carried out in parallel. For the parent strain, the oxacillin MICs were 0.25 and 0.5 mg/mL, respectively. For the hemB and menD mutants, MICs were 0.5 mg/mL at both inocula. There were no significant differences in MICs obtained by broth dilution or agar dilution assays (data not shown). The effects of oxacillin on the parent strains, mutants, and restored mutants, as well as a time kill curve for the menD mutant, are shown in figure 1. Susceptibility to thrombin-induced platelet microbicidal protein 1 (tPMP-1). The cationic antimicrobial peptide tPMP-1 was prepared from rabbit platelets, and its bioactivity (mg/mL) was determined by our standard assay techniques, as described elsewhere [17]. To determine the in vitro tPMP-1 susceptibility phenotype for the 5 study strains, we used a previously published microtiter assay, in which tPMP-1 was tested at 2 mg/mL, at a final staphylococcal inoculum of 103 cfu [17]. Two independent runs were performed on separate days. As in previous studies, the breakpoint for in vitro resistance of grampositive bacteria to tPMP-1 was defined as 50% survival after 2-h peptide exposures [18]. Experimental endocarditis model. A rabbit model of endocarditis [19] was used in these investigations. In brief, anesthetized rabbits underwent transcarotid artery–to–left ventricle catheterization to induce sterile vegetations on the aortic valve. Endocarditis was induced by intravenous bacterial challenge 24 h later. To determine the ID95 for all 5 of the S. aureus study strains in this model, 4 or 5 catheterized animals were challenged with a range of inocula for each construct that encompassed the ID95 for most S. aureus strains (i.e., 2 104–2 106 cfu). Rabbits were killed 48 h after challenge, and cardiac vegetations, kidneys, and spleen were aseptically removed, homogenized, and processed for quantitative culture, as described elsewhere [19–22]. For the formal antibiotic treatment study, experimental endocarditis was induced in catheterized animals by challenge with the ID95 inoculum. Animals were randomly assigned at 48 h after induction of endocarditis to receive either no therapy (untreated controls) or oxacillin for 3 days (50 mg/kg 3 times daily intramuscularly). This oxacillin dosing strategy readily achieved serum levels well in excess of the MICs for all the strains used in the present study [19–22]. Oxacillin was selected for several reasons: (1) it is highly effective in treating experimental endocarditis caused by susceptible S. aureus strains; (2) it is commonly used for the treatment of S. aureus endocarditis; and (3) like other b-lactams, it penetrates phagocytic cells Untreated samples Oxacillin-treated samples Bacterial density, mean log10 cfu/g Bacterial density, mean log10 cfu/g Tissue, strain a P .05 was considered to be statistically significant. poorly [23–25], thus allowing intracellular bacterial persistence within 1 target tissue. At 24 h after the last dose of oxacillin, animals were killed, and their target tissues were processed as described above for quantitative cultures. Statistical analysis. Bacterial densities in target tissues from untreated controls and oxacillin-treated animals were compared using Student’s t test for unpaired samples. P .05 was considered to be statistically significant. The parent S. aureus strain was susceptible to tPMP-1 (mean survival SD, 30% 7%). The hemB mutant was modestly less susceptible to tPMP-1 (mean survival SD, 40% 7%), although the level of susceptibility was below the breakpoint for resistance [17]. Complementation of hemB resulted in a reduction of the susceptibility profile to levels near those of the parent strain (mean survival SD, 27% 6%). In contrast, the menD mutant was highly resistant to tPMP-1 (mean survival SD, 71% 15%); complementation of menD restored the susceptibility profile to levels near those of the parent strain (mean survival SD, 21% 4%). The overall infectivity of each strain was determined by quantification of the ID50, ID90, and ID95. The challenge inocula for all strains used for inducing endocarditis and metastatic infection in kidneys and spleen in 50%, 90%, and 95% of rabbits were as follows: ID50, 2 104 cfu; ID90, 2 105 cfu; and ID95, 2 106 cfu. For the untreated parent strains, as well as the hemB mutant and hemB-complemented strain, bacterial densities in vegetations, kidneys, and spleen were not significantly different (table 2). For the menD mutant, bacterial densities in vegetations, kidneys, and spleen of untreated controls were substantially lower than those for the parent strain. Complementation of the menD mutant restored virulence; bacterial densities of this mutant were equivalent to those of the parent strain. During oxacillin therapy, bacterial densities were significantly reduced in all target tissues of animals infected with the parent strain, the hemB mutant, and the hemB-complemented strain, compared with untreated controls (table 2). In contrast, oxacillin therapy significantly reduced bacterial densities of the menD mutant in vegetations but not in kidneys and spleen. Reversal of the menD mutation restored the responsiveness of this strain to oxacillin therapy in the kidney and spleen to levels similar to those of the parent strain. DISCUSSION It is increasingly recognized that S. aureus SCVs play a role in serious human infections [1–9, 26–28]. They have been implicated as a cause of persistence, recurrence, and antibiotic resistance in lung, bone, skin, wound, and foreign body infections [1–9, 26–28]. Most recently, infections caused by S. aureus SCVs were found to be associated with dramatically increased (nearly 6-fold) mortality rates in critically ill patients in an intensive care unit in Brazil [29]. We and others have shown that S. aureus SCVs can readily enter and reside in endothelial cells [5, 11, 30]. This intracellular persistence has been associated with reduced a-toxin production [4, 5, 11, 30], which enables SCVs to reside within the endothelial cell milieu without lysing the host cells, a normal target for a-toxin [31]. Moreover, this intracellular environment protects the SCVs from antibiotic-induced killing by many conventional antistaphylococcal antibiotics (e.g., b-lactams and aminoglycosides) because of their lack of penetration into cells [2, 23–25, 32]. In addition, extracellular SCVs also are relatively resistant to several such antimicrobial agents (e.g., aminoglycosides), because an intact electron transport system that generates a threshold transmembrane potential is critical for their bacterial uptake [2, 8, 32]. Furthermore, SCVs are highly likely to escape, by 1 of at least 2 methods, being killed by host defense systems that are active at sites of endothelial damage. First, many SCVs are intrinsically resistant to killing by platelet-secreted cationic staphylocidal peptides (e.g., tPMP1) that are induced by agonists generated at such endothelial sites [32]. Second, the reduced production of a-toxin, a potent platelet peptide secretagogue, by SCVs mitigates elaboration of such peptides [19]. Taken together, these data suggested that the in vivo relevance of the SCV phenotype could be adjudicated and quantified in the experimental model of infective endocarditis (as used in the present study), in which endothelial cells, platelets, and metastatic sites of infection play a major pathogenic role [19]. Several interesting observations emerged from the studies cited above. First, there were no differences in the capacity of either the hemB or menD mutants to induce experimental endocarditis, in comparison with the parent and complemented strains (i.e., the ID95 was virtually identical for all strains). Although the ability of SCVs to produce a-toxin, a known virulence factor [31], is impaired, SCVs exhibit an enhanced expression of fibrinogen-binding and fibronectin-binding adhesins [33]. Each of these adhesins has been shown to play an important role in the pathogenesis of endocarditis [34–36]. Thus, although the mutants may grow more slowly and produce less a-toxin than the parent strains, this may be balanced by enhanced adhesion to damaged valves, which may explain why the infectivity of the mutant strains in vivo is similar to that of the parent strains. One notable difference between the 2 SCV phenotype mutants was in achievable bacterial densities in kidneys and spleen, in comparison with the parent strain. For the hemB mutant, there were no differences in kidney or spleen densities, compared with those in animals infected with the parent strain. We postulate that this is likely related to the fact that each organ is probably replete with hemin derived from the embolic infarcts that occur during the course of experimental endocarditis in these organs, which circumvents the hemB knockout-induced defect in the cytochrome system [1–4, 8, 11, 30]. In contrast, bacterial densities achieved by the menD mutant in the same target tissues were significantly lower than those achieved by the parent strain. This probably reflects the in vivo retention of the SCV phenotype and the attendant slow growth rates of the menD mutant. Of note, reconstitution of the menD mutation restored the capacity of this altered strain to achieve bacterial densities in these target organs to levels similar to those of the parent strain. These data are consistent with clinical reports that hemin auxotrophic SCVs are most frequently isolated from relatively avascular sites (e.g., sputum and osteomyelitic bone), whereas menadione biosynthetic mutants have been recovered from a broader range of host tissues. The response of menD and hemB mutants to oxacillin therapy in the endocarditis model further emphasized the relative differences in survival advantage between hemB and menD mutants in vivo in diverse anatomic niches. For the hemB mutant, the capacity of oxacillin to clear the organisms from all target tissues was virtually identical to that observed for the parent and complemented strains. This may, again, underscore the likely availability of hemin in these niches in experimental endocarditis, where hemorrhagic necrosis will occur within the infected emboli [37]. In addition, the relative susceptibility of the hemB mutant to tPMP-1 (compared with that of the tPMP1–resistant menD mutant) may have contributed to the oxacillin-induced clearance of this strain from all target tissues [38]. In contrast, the presence of the menD mutation substantially reduced the ability of oxacillin to eradicate such SCVs from kidneys and spleen but not from vegetations. This disparity in tissue-specific SCV clearance is not well understood. As noted above, the resistance of the menD mutant to tPMP1 may play a role in the reduced clearance of this construct from kidneys and spleen by oxacillin. However, this would not adequately explain the ability of oxacillin to efficiently clear such SCVs from vegetations. Most bacteria within a vegetation are trapped extracellularly within the relatively acellular platelet-fibrin matrix, whereas tissues infected by microemboli are replete with cells, which may offer an opportunity for intracellular invasion by SCVs. This may allow a larger percentage of the organisms to enter host cells, providing a protective environment. Because intracellular levels of menadione are 100fold lower than the levels needed to reverse the SCV phenotype in menadione biosynthetic mutants [39], this may be an additional explanation for the resistance of the menD mutant to antibiotic therapy within the spleen and kidneys, compared with vegetations. Collectively, these data emphasize that the menD and the hemB mutants can be virulent, despite their slow growth rates and reduced a-toxin production. We have reported elsewhere that the hemB mutant can produce as much tissue damage as the parent strain in the mouse arthritis model [40] and that SCVs may contribute significantly to the lung pathology in patients with cystic fibrosis [4, 5]. Thus, although we have found that the menD mutant is able to persist in this animal model of endocarditis, several of the original paradigms about the virulence of SCVs have changed. Initially, SCV isolates were primarily found in patients with relatively mild, but persistent, infections. However, menadione and hemin auxotrophic SCVs are proving to be quite virulent, both in experimental models and in human infection. Acknowledgment We thank Yin Li Chai for excellent technical assistance. References 1. Proctor RA , van Langevelde P , Kristjansson M , Maslow JN , Arbeit RD . Persistent and relapsing infections associated with small-colony variants of Staphylococcus aureus . Clin Infect Dis 1995 ; 20 : 95 - 102 . 2. Proctor RA . Bacterial energetics and antimicrobial resistance . Drug Resist Updat 1998 ; 1 : 227 - 35 . 3. Proctor RA . Respiration and small colony variants of Staphylococcal aureus . In: Fischetti V, Novick R , Ferretti J , Portnoy D , Rood J, eds. Gram-positive pathogens: staphylococci . Washington, DC: American Society for Microbiology Press, 2000 : 345 - 50 . 4. Kahl B , Herrmann M , Everding AS , et al. Persistent infection with small colony variant strains of Staphylococcus aureus in patients with cystic fibrosis . J Infect Dis 1998 ; 177 : 1023 - 9 . 5. Kahl BC , Duebbers A , Lubritz G , et al. Persistence of small colony variants (SCV) and/or normal S . aureus in respiratory secretions of cystic fibrosis (CF) patients: results of a 6 year prospective study [abstract D-41] . In: Program and abstracts of the 101st general meeting of the American Society for Microbiology (Orlando) . Washington, DC: American Society for Microbiology, 2001 : 283 . 6. von Eiff C , Bettin D , Proctor RA , et al. Recovery of small colony variants of Staphylococcus aureus following gentamicin bead placement for osteomyelitis . Clin Infect Dis 1997 ; 25 : 1250 - 1 . 7. von Eiff C , Vaudaux P , Kahl BC , et al. Bloodstream infections caused by small-colony variants of coagulase negative staphylococci following pacemaker implantation . Clin Infect Dis 1999 ; 29 : 932 - 4 . 8. Proctor RA , Bates DM , McNamara PJ . Electron transport-deficient Staphylococcus aureus small-colony variants as emerging pathogens . In: Craig W, ed. Emerging infections. Vol 5. Washington, DC: American Society for Microbiology Press, 2001 : 95 - 110 . 9. Acar JF , Goldstein FN , Lagrange P. Human infections caused by thiamine- or menadione-requiring Staphylococcus aureus . J Clin Microbiol 1978 ; 8 : 142 - 7 . 10. Baumert N , von Eiff C , Schaaff F , Peters G , Proctor RA , Sahl H-G. Physiology and antibiotic susceptibility of Staphylococcus aureus small colony variants . Microb Drug Resist 2002 ; 8 : 253 - 60 . 11. von Eiff C , Heilmann C , Proctor RA , Woltz C , Peters G , Go¨tz F. A site directed Staphylococcus aureus hemB mutant is a small colony variant which persists intracellularly . J Bacteriol 1997 ; 179 : 4706 - 12 . 12. McNamara PJ , Milligan-Monroe KC , Khalili S , Proctor RA . Identification, cloning, and initial characterization of rot, a locus encoding a repressor of extracellular toxins in Staphylococcus aureus . J Bacteriol 2000 ; 182 : 3197 - 203 . 13. Byeon WH , Weisblum B. Replication genes of plasmid pE194-cop and repF: transcripts and encoded proteins . J Bacteriol 1990 ; 172 : 5892 - 900 . 14. Janzon L , Arvidson S. The role of the delta-lysin gene (hld) in the regulation of virulence genes by the accessory gene regulator (agr) in Staphylococcus aureus . EMBO J 1990 ; 9 : 1391 - 9 . 15. Kraemer GR , Iandolo JJ. High-frequency transformation of Staphylococcus aureus by electroporation . Curr Microbiol 1990 ; 21 : 373 - 6 . 16. Dyer DW , Iandolo JJ . Rapid isolation of DNA from Staphylococcus aureus . Appl Environ Microbiol 1983 ; 46 : 283 - 5 . 17. Bayer AS , Cheng D , Yeaman MR , et al. In vitro resistance to thrombin-induced platelet microbicial protein among clinical bacteremic isolates of Staphylococcus aureus correlates with an endovascular infection source . Antimicrob Agents Chemother 1998 ; 42 : 3169 - 72 . 18. Wu T , Yeaman MR , Bayer AS . In vitro resistance to platelet microbicidal protein correlates with endocarditis source among bacteremic staphylococcal and streptococcal isolates . Antimicrob Agents Chemother 1994 ; 38 : 729 - 32 . 19. Bayer AS , Ramos MD , Menzies BE , Yeaman MR , Shen AJ , Cheung AL. Hyperproduction of alpha-toxin by Staphylococcus aureus results in paradoxically reduced virulence in experimental endocarditis: a host defense role for platelet microbicidal proteins . Infect Immun 1997 ; 65 : 4652 - 60 . 20. Egert J , Carrizosa J , Kaye D , Kobasa WD . Comparison of methicillin, nafcillin, and oxacillin in therapy of Staphylococcus aureus endocarditis in rabbits . J Lab Clin Med 1977 ; 89 : 1262 - 8 . 21. Cars O , Henning C , Holm SE . Penetration of ampicillin and dicloxacillin into tissue cage fluid in rabbits: relation to serum and tissue protein binding . Scand J Infect Dis 1981 ; 13 : 69 - 74 . 22. Cars O. Tissue distribution of beta-lactam antibiotics: experimental studies in rabbits . Scand J Infect Dis Suppl 1981 ; 27 : 1 - 48 . 23. Craven N , Anderson JC . Antibiotic activity against intraleukocytic Staphylococcus aureus in vitro and in experimental mastitis in mice . Am J Vet Res 1983 ; 44 : 709 - 12 . 24. Roesler J , Hockertz S , Vogt B , Lohmann-Matthes ML . Staphylococci surviving intracellularly in phagocytes from patients suffering from chronic granulomatous disease are killed in vitro by antibiotics encapsulated in liposomes . J Clin Invest 1991 ; 88 : 1224 - 9 . 25. Frank MO , Sullivan GW , Carper HT , Mandell GL . In vitro demonstration of transport and delivery of antibiotics by polymorphonuclear leukocytes . Antimicrob Agents Chemother 1992 ; 36 : 2584 - 8 . 26. Seifert H , von Eiff C , Fatkenheuer G . Fatal case due to methicillinresistant Staphylococcus aureus small colony variants in an AIDS patient . Emerg Infect Dis 1999 ; 5 : 450 - 3 . 27. Abele-Horn M , Schupfner B , Emmerling P , Waldner H , Goring H. Persistent wound infection after herniotomy associated with smallcolony variants of Staphylococcus aureus . Infection 2000 ; 28 : 53 - 4 . 28. Spearman P , Lakey D , Jotte S , Chernowitz A , Claycomb S , Stratton C. Sternoclavicular joint septic arthritis with small-colony variant Staphylococcus aureus . Diagn Microbiol Infect Dis 1996 ; 26 : 13 - 15 . 29. Salgado DR , Pinto M , Bozza FA , Sampaio J , Leite JP . Outbreak with small colony variants of methicillin-resistant Staphylococcus aureus in an ICU [abstract K-1226] . In: Program and abstracts of the Interscience Conference on Antimicrobial Agents and Chemotherapy (Chicago) . Washington, DC: American Society for Microbiology, 2001 : 416 . 30. Balwit JM , van Langevelde P , Vann JM , Proctor RA . Gentamicin-resistant menadione and hemin auxotrophic Staphylococcus aureus persist within cultured endothelial cells . J Infect Dis 1994 ; 170 : 1033 - 7 . 31. Vann JM , Proctor RA . Cytotoxic effects of ingested Staphylococcus aureus on bovine endothelial cells: role of S. aureus a-hemolysin . Microbial Pathogenesis 1988 ; 4 : 443 - 53 . 32. Koo S-P , Bayer AS , Sahl H-G , Proctor RA , Yeaman MR . Staphylocidal action of thrombin-induced platelet microbicidal protein (tPMP) is not solely dependent on transmembrane potential (DW) . Infect Immun 1996 ; 64 : 1070 - 4 . 33. Vaudaux P , Francois P , Bisognano C , et al. Increased expression of clumping factor and fibronectin-binding proteins by small colony variants of Staphylococcus aureus . Infect Immun 2002 ; 70 : 5428 - 37 . 34. Moreillon P , Entenza JM , Francioli P , et al. Role of Staphylococcus aureus coagulase and clumping factor in pathogenesis of experimental endocarditis . Infect Immun 1995 ; 63 : 4738 - 43 . 35. Que YA , Francois P , Haefliger JA , Entenza JM , Vaudaux P , Moreillon P. Reassessing the role of Staphylococcus aureus clumping factor and fibronectin-binding protein by expression in Lactococcus lactis . Infect Immun 2001 ; 69 : 6296 - 302 . 36. Kuypers JM , Proctor RA . A fibronectin low-binding mutant of Staphylococcus aureus shows reduced adherence to traumatized rat heart valves . Infect Immun 1989 ; 57 : 2306 - 12 . 37. Padmanaban G , Venkateswar V , Rangarajan PN . Haem as a multifunctional regulator . Trends Biochem Sci 1989 ; 14 : 492 - 6 . 38. Dhawan VK , Bayer AS , Yeaman MR . Thrombin-induced platelet microbicidal protein susceptibility phenotype influences the outcome of oxacillin prophylaxis and therapy of experimental Staphylococcus aureus endocarditis . Antimicrob Agents Chemother 2000 ; 44 : 3206 - 9 . 39. Shearer MJ , McCarthy PT , Crampton OE , Mattock MB . The assessment of human vitamin K status from tissue measurements . In: Suttie JW, ed. Current advances in vitamin K research. New York : Elsevier, 1987 : 437 - 52 . 40. Jonsson I-M , von Eiff C , Proctor RA , Peters G , Ryden C , Tarkowski A. Virulence of a hemB mutant displaying the phenotype of a Staphylococcus aureus small colony variant in a murine model of septic arthritis . Microb Pathog 2003 ; 34 : 73 - 9 .