JOURNAL 01 CLINICAL MICROBIOLOGY, Aug. 1978, p. 142-147 Vol. 8, No. 2 0095-1 137/78/0008-0142$02.00/0 Copyright © 1978 American Society for Microbiology Printed in U.S.A. Human Infections Caused by Thiamine- or Menadione- Requiring Staphylococcus aureus J. F. ACAR,' F. W. GOLDSTEIN,' AND P. LAGRANGE2 Hôpital Saint-Joseph, Université Pierre et Marie Curie,' and Institut Pasteur,2 Paris, France Received for publication 21 February 1978 Stable dwarf forms of Staphylococcus aureus have been identified in clinical specimens as the sole or predominant isolate in eight cases. These organisms have been shown to be menadione or thiamine dependent, i.e., cultivation in the presence of one of these agents has permitted growth of colonies which appear typical of S. aureus. In vitro resistance to aminoglycosides was overcome by cultivation in the presence of menadione or thiamine. Menadione- or thiamine- requiring S. aureus can be considered as causative agents in severe human infections. Special care must be taken if they are to be identified in pathological specimens. Their antibiotic sensitivity testing should be done comparatively on supplemented and nonsupplemented media. Staphylococcus aureus growing as dwarf col- onies. S. aureus was identified on the basis of micro- onies on usual media have been documented by scopic morphology, coagulase, phosphatase, deoxyri- many investigators after in vitro exposure of the bonuclease, catalase production, and biochemical tests bacteria to adverse environmental conditions according to Baird-Parker (3). S. aureus 209 P (Insti- tut Pasteur) was used as control. such as chemicals, antibiotics, or aging (4, 8-13, Supplementation of dwarf colony variants. 16, 19, 21-24, 26-29, 37, 39-41, 43-45). In some Dwarf colonies were subcultured by using one or more cases, these organisms have been shown to have of the following media: TSA, TSA with 5% horse increased nutritional requirements, requiring blood, chocolate agar with 1% IsoVitaleX (Baltimore supplementation with carbon dioxide, hemin, Biological Laboratories), Mueller Hinton (MH) agar menadione, thiamine, or pantothenate (6, 10, 12, with disks containing 5 jug of thiamine-hydrochloride 13, 18-20, 22, 23, 26-28, 31-36, 38-41). Isolation (Hoffman-LaRoche) or 1 gg of menadione bisulfite (E. of such strains from clinical material was first Merk AG). Incubation was carried out for 18 and 48 h reported by Hale in 1951 (20). Thiamine-requir- aerobically, aerobically with 5% CO2 supplement, and ing strains isolated et al. as the anaerobically (H2-CO2, GasPak). by Sompolinsky Quantitative supplementation was performed on causative agent of bovine mastitis were never MH agar containing either thiamine or menadione at isolated from human subjects (33-35). Menadi- a 10-fold concentration ranging from 0.01 to 100 ,tg/ml. one-requiring strains similar to those that were The colony size was measured after 18 and 48 h on selected in vitro by Sasarman et al. (27, 28) have each plate. been isolated by Borderon and Horodniceanu Antibiotic sensitivity testing. Disk susceptibility from a clinical specimen (6). testing was performed on MH agar by the ICS method The present study concerns eight strains of S. (15). Dwarf colony variants and reverse mutants of aureus which were isolated as dwarf colonies each strain were tested separately. Zones of inhibition from clinical specimens and which require thia- for kanamycin, gentamicin, erythromycin, lincomycin, mine or to and vancomycin were read after 24 and 48 h of incu- menadione grow normally. The prob- bation at 37°C. A separate test was performed at 30°C lems related to the detection of these strains in with an oxacillin disk to detect methicillin-resistant a clinical laboratory and the methods for testing strains (2). their antibiotic sensitivity are presented here. MIC. Minimal inhibitory concentrations (MICs) of AND the same antibiotics were determined by an agar di- MATERIALS METHODS lution method by using MH agar supplemented with Bacterial strains. From June, 1975, to May, 1976, 5 ,ug of thiamine per ml or 1 ,ug of menadione per ml. among 1,110 strains of S. aureus isolated from clinical Standardized bacterial inocula (102 to 103 colony-form- material, 15 isolates were observed on initial culture ing units per spot) were deposited on the surface of to grow as dwarf colonies on our routine media, Tryp- the agar plates with an automatic Steers replicator ticase soy agar (TSA; Baltimore Biological Laborato- device. Dwarf colony variants and the reverse mutants ries) and TSA with 5% horse blood. Eight of these were tested simultaneously. Results were read after 24 isolates were stable on serial subculture with these and 48 h of incubation. same media and required either thiamine or menadi- Effect of supplementing substances on anti- one to grow as normal-appearing staphylococcal col- biotie sensitivity. (i) Paper strip method. Two 142 VOL. 8, 1978 THIAMINE- OR MENADIONE-REQUIRING S. AUREUS 143 filter paper strips were placed at right angles on an MH agar seeded with 10i colony-forming units per ml by the technique of Bonifas (5) and Dye (14). One paper contained the supplementing substance, and the other contained one of the antibiotics tested by MIC. Concentrations of solutions used for filter paper im- pregnation were (in pg/ml): thiamine, 400, menadione, 100; oxacillin, 100; kanamycin, 200-, gentamicin, 200, erythromycin, 200e, lincomycin, 200; and vancomycin, 400. Checkerboard method. MICs of kanamycin for a thiamine- and a menadione-requiring strain were de- termined by a checkerboard arrangement as described previously (1). Thiamine or menadione at concentra- tions ranging from 0.01 to 100 ug/mI were tested with kanamycin at concentrations ranging from 0.1 to 100 ug/mI in MH broth. Results of growth after 24 and 48 h of incubation at 37°C were drawn as an isobologram on a logarithmic scale. RESULTS Bacterial strain and supplementation. Three menadione- and five thiamine-requiring S. aureus were isolated from specimens obtained from human subjects (Table 1); in most of the cases, dwarf colony variants were recovered as predominant bacteria in mixed cultures with the reverse mutants (Fig. 1). All the strains had normal microscopic morphology and produced coagulase, deoxyribonuclease, and catalase. FIG. 1. Menadione-requiring S. aureus, TSA, 48-h Menadione-requiring strains grew very poorly growth. Some reverse mutants are noted together with after 48 h on all of the nonsupplemented media, dwarf colonies. except on the plate containing the l-,g menadi- one disk where satellite growth with normal- sized colonies occurred around the disk (Fig. 2). Reverse mutants growing normally could be ob- served. Quantitative supplementation tests showed adequate colonial morphology with menadione concentrations ranging from 0.1 to 2 fg/ml. At concentrations higher than 10 gg/ml, growth was inhibited. Thiamine-requiring strains grew poorly on MH agar but had normal appearance on choco- late agar containing IsoVitaleX. On TSA and TS blood agar, intermediary-sized colonies could be observed, probably because these media con- tain a small amount of thia.nne. Reverse mu- tants were isolated from six of the eight strains tested, and three of these strains showed the TABLE 1. Menadione- or thiamine-requiring S. aureus isolated frompatients FIG. 2. Supplementation test, MH agar, 18-h Oitgin No. of iso- Supplement- growth. Satellite growth of menadione-requiring lates ing substance strain around a disk containing I pg of menadione. Blood i Menadione Osteomyelitis 1 Menadione phenomenon of autosatelliting. All of the pri- Subcutaneous abscess i Menadione mary isolates had a heterogenous appearance, Cerebrospinal fluid i Thiamine with a range from normal to tiny variants being Blood 2 Thiamine observed (Fig. 3). Osteomyelitis 2 Thiamine Satellite growth around a disk containing 5 144 ACAR, GOLDSTEIN, AND LAGRANGE J. CLIN. MICROBIOL. ,g of thiamine was very similar to that observed dwarf colonies were lower but did not reach the with a menadione disk. Quantitative supplemen- normal level of the reverse mutants. tation showed adequate colonial morphology at Except for one strain which was highly resist- concentrations ranging from 0.05 to 25 ,ug of ant to erythromycin, all of the strains were in- thiamine. Growth in 5% C02 in air or under hibited by 0.2 ,g of erythromycin per ml, 0.5 ,g anaerobic conditions did not significantly en- of lincomycin per ml, and 1 ,ug of vancomycin hance growth of any thianine- or menadione- per ml. No significant differences relating to requrng strains. colonial morphology or media supplementation Antibiotic sensitivity testing. All of the were observed with these antibiotics. strains tested were sensitive by disk sensitivity Paper strip method. Results obtained by testing to oxacillin, lincomycin, and vancomycin. this method showed that the more a strain is One strain was highly resistant to erythromycin. supplemented with required metabolites, the Results of MIC determinations are presented in greater is its sensitivity to kanamycin and gen- Table 2. Oxacillin inhibited all of the strains at tamicin (Fig. 4). The checkerboard method con- 0.5 ,g/ml. Dwarf colony variants were more firmed this qualitative observation. MICs ofkan- susceptible to oxacillin when tested on nonsup- amycin were higher when thiamine or menadi- plemented media. On media supplemented with one concentrations were decreased. The ratio of thiamine or menadione, dwarf colonies and their MIC on supplemented media to MIC on nonsup- reverse mutants had similar MICs. plemented media was 1/16 for the menadione- Kanamycin and gentamicin inhibited all of requiring strain and 1/4 for the thiamine-requir- the reverse mutants at 1 mg/ml (except for one ing strain. The graphic expression of the MICs strain which was highly resistant to kanamycin). obtained by the checkerboard method is similar In each case, dwarf colony variants had higher to the figure obtained by the paper strip tech- MICs to kanamycin and gentamicin when tested nique (Fig.