JOURNAL OF CLINICAL MICROBIOLOGY, Jan. 1994, p. 82-86 Vol. 32, No. 10095-1137/94/$04.00+0Copyright © 1994, American Society for Microbiology

Plasmid DNA Fingerprinting of Acinetobacter Species Otherthan Acinetobacter baumannii

H. SEIFERT,* A. SCHULZE, R. BAGINSKI, AND G. PULVERER

Institut fiir Medizinische Mikrobiologie und Hygiene der Universitat Koin, 50935 Cologne, Germany

Received 14 June 1993/Returned for modification 7 September 1993/Accepted 12 October 1993

During the last years Acinetobacter species have emerged as clinically significant pathogens. Most infectionsare nosocomially acquired and mainly due to Acinetobacter baumannii. Little is known about the epidemiologyand clinical significance of unnamed Acinetobacter species 3 (the second most often encountered member of thegenus Acinetobacter) and other Acinetobacter species such as A. johnsonii, A. junii, and A. lwoffli. Seventy-fiveclinical isolates ofAcinetobacter species other than A. baumannii (Acinetobacter species 3, n = 37; A.johnsonii,n = 20; A.junii, n = 8; A. iwoffli, n = 10) recovered from 66 patients over a period of 12 months were analyzedby plasmid DNA fingerprinting. Plasmids were found in 84.4%o ofAcinetobacter species 3 isolates and in all A.johnsonii, A. junii, and A. iwoffi isolates. Strains harbored up to 15 plasmids each. Almost every isolate gavea unique plasmid pattern. With one exception, identical plasmid profiles were detected only in correspondingisolates recovered from blood cultures and intravascular catheters from a given patient. Plasmid DNAfingerprinting proved to be useful for typing Acinetobacter species other than A. baumannii. There was noevidence of patient-to-patient transmission or hospital outbreaks due to these species. This finding is in contrastto the results obtained in studies of the hospital epidemiology ofA. baumannii.

In recent years Acinetobacter baumannii has been in-creasingly recognized as an important nosocomial pathogen(4, 10, 17). This organism has been implicated as the cause ofa wide spectrum of infectious diseases such as pneumonia(10), meningitis (6), bacteremia (4, 25), and device-relatedinfections (4, 27). Outbreaks of infections have been re-ported from neonatal intensive care units (25), medical andsurgical intensive care units (4, 17), and burn units (30) andhave been associated with respiratory equipment (10), hu-midifiers (15), pressure transducers (4), patients' mattresses(30), and contaminated gloves (23).For epidemiological studies, various typing methods, such

as antibiotic resistance typing (1, 19), bacteriocin typing (3),biotyping (8, 9, 19), phage typing (9, 16, 26), serotyping (32,33), cell envelope protein typing (1, 9, 11, 12), plasmid typing(1, 13, 17, 23), ribotyping (11, 14) and restriction fragmentlength polymorphisms of chromosomal DNA determined bypulsed-field gel electrophoresis (2), have been developed.As A. baumannii, formerly classified as A. calcoaceticus

subspecies anitratus, has been frequently shown to be themost prevalent species among Acinetobacter strains (8, 16,20, 33) and was predominantly responsible for nosocomialinfection and hospital outbreaks (4, 10, 17, 23), epidemiolog-ical studies have focused mainly on A. baumannii. Little isknown about the epidemiology and clinical significance ofAcinetobacter species other than A. baumannii, previouslyclassified as A. calcoaceticus subspecies iwoffii. The occur-rence of bloodstream infections, mainly in three hospitals,due to Acinetobacter species other than A. baumannii asidentified on the basis of the latest terminology (7, 8)prompted us to investigate these isolates with regard to theirpossible epidemiologic relationship. The purpose of thisstudy was to find out whether plasmid typing of unnamedAcinetobacter species 3 (the second most prevalent memberof the genus Acinetobacter), A. johnsonii, A. junii, and A.

* Corresponding author.

iwoffii would be useful for epidemiological investigations ofinfections due to these organisms.

(Part of this work has been presented previously [29a].)

MATERUILS AND METHODS

Bacterial strains. During a 12-month survey a total of 584Acinetobacter isolates from 420 patients were consecutivelyrecovered from different clinical specimens submitted to ourinstitution. Details of these isolates have been presentedelsewhere (28). Isolates were identified according to thesimplified identification scheme described by Bouvet andGrimont (8) including growth at 37, 41, and 44°C; productionof acid from glucose; gelatin hydrolysis; and assimilation of14 different carbon sources. Identification of isolates at thegenus level was confirmed by the transformation assay ofJuni (21). The following reference strains served as controls:Acinetobacter species 3 strain CIP 70-29, A. johnsonii ATCC17909, A. junii ATCC 17908, and A. iwoffli ATCC 15309.Reference strains were obtained through the courtesy ofP. J. M. Bouvet and P. A. D. Grimont (Service Enterobac-teries, Institut Pasteur, Paris, France). One hundred fifty-eight isolates were identified as species other than A. bau-mannii. The most common species were unnamedAcinetobacter species 3 (n = 55), A. johnsonii (n = 29), A.Iwoffii (n = 21), and A. junii (n = 11). Of these, all isolatesfrom relevant clinical sources, blood cultures, catheter tips,and cerebrospinal fluid, were included in this study. Otherisolates were included if they were recovered from the samehospital or ward. Thus, a total of 75 isolates were selected:37, Acinetobacter species 3; 20, A. johnsonii; 10, A. iwoffii;and 8,A. junii. The origins of the isolates and the distributionof the species are shown in Table 1. All isolates were storedin glycerol broth at -70°C until further use.

Susceptibility testing. MICs of selected antimicrobialagents were determined by a microtiter broth dilutionmethod (MicroScan MIC Plus Type MK Dried Panels;Baxter Healthcare Corp., West Sacramento, Calif.) and bythe disk diffusion method on Mueller-Hinton agar.

82

on Septem

ber 20, 2018 by guesthttp://jcm

.asm.org/

Dow

nloaded from

PLASMID DNA FINGERPRINTING OF ACINETOBACTER SPECIES 83

TABLE 1. Origins of Acinetobacter isolates and distribution of species

No. of isolates from: Total no. of Total no. ofOrganism

Blood culture Central venous catheter Sputum Wound swab Cerebrospinal fluid isolates patients

Acinetobacter species 3 11 13 6 5 2 37 33A. johnsonii 15 4 1 0 0 20 16A. junii 4 2 0 2 0 8 8A. iwoffii 8 1 0 0 1 10 9

Total 38 20 7 7 3 75 66

Plasmid isolation. Plasmid DNA was prepared as de-scribed by Hartstein et al. (17), with minor modifications. Inbrief, isolates were grown on brain heart infusion agar platesat 30°C for 24 h. Cells from one quarter of the plate-growthfrom one-half of the plate when A. johnsonii, A. junWi or A.iwoffii was used-were suspended in 3 ml of 2.5 M NaCl-10mM EDTA (pH 8.0). After centrifugation, cells were resus-pended in 900 RI of 20% sucrose-50 mM Tris-10 mM EDTA(pH 8.0) and 200 ,ul of lysozyme (10 mg/ml). After incubationat 37°C for 30 min, 1 ml of 2.5 M NaCl-10 mM EDTA (pH8.0) and a lysis solution containing 0.5 ml of 0.5% mixedalkyltrimethylammonium bromide (ATAB) and 0.5 ml 1%Triton X-100 was added. The resulting lysate was incubatedin a water bath at 56°C for 15 min and then centrifuged at33,000 x g for 45 min at 20°C in a Sorvall centrifuge. Fivemilliliters of 0.5% ATAB was added to the supernatant.Following incubation on ice for 10 min and centrifugation at2,500 x g for 15 min at 4°C, the pellet was dissolved in 0.25ml of 2.5 M NaCl-10 mM EDTA (pH 8.0) and 0.5 ml of 10mM Tris-1 mM EDTA (pH 8.0). After addition of 3 ,u ofRNase (500 ,ug/ml; Boehringer GmbH, Mannheim, Germa-ny), the solutions were incubated at 37°C for 30 min. Theremainder of the procedure was performed in Eppendorftubes. Protein was extracted with phenol-chloroform-isoamyl alcohol (25:24:1), and plasmid DNA was precipi-tated with an equal volume of ice-cold isopropanol. Theprecipitate was collected by centrifugation, dried undervacuum, and then dissolved in 60 ,ul of sterile distilled water.Electrophoresis was performed on a horizontal gel contain-ing 0.8% agarose in TAE buffer (0.04 M Tris-acetate, 0.001M EDTA [pH 8.2]). The sample (20 ,ul) and 5 ,ul of runningdye were loaded onto the gel and run for 18 h at 30 V. Gelswere stained with ethidium bromide and photographed undera UV lamp. Molecular weights were estimated by compari-son with molecular size markers. A plasmid type was definedas any plasmid pattern which varied from another patternwith regard to the number and size of plasmid bands.Isolates were considered similar if one of the comparedpatterns contained one or two additional bands. Isolateswere run in duplicate on different gels.

RESULTS

Antibiotic resistance patterns. Analysis of the antibioticresistance patterns of Acinetobacter species 3 isolates re-vealed only three different patterns. Susceptibility pattern Aaccounted for 35 (95%) of the 37 isolates tested. Isolateswere susceptible to amoxicillin-clavulanate, gentamicin, to-bramycin, amikacin, trimethoprim-sulfamethoxazole, tetra-cycline, ciprofloxacin, and imipenem and resistant to ampi-cillin, mezlocillin, piperacillin, cefazolin, and cefuroxime;resistance was variable with regard to broad-spectrum ceph-alosporins. Only two isolates showed greater susceptibility

to penicillins (pattern B) or penicillins and older cephalos-porins (pattern C), respectively.With only a few exceptions, isolates identified as A.

johnsonii, A. junii, orA. Iwoffii were susceptible to ampicil-lin, amoxicillin-clavulanate, broad-spectrum cephalospor-ins, the aminoglycosides, trimethoprim-sulfamethoxazole,tetracycline, ciprofloxacin, and imipenem. Resistance wasvariable with respect to mezlocillin, piperacillin, and oldercephalosporins.

Plasmid profiles. Twenty-seven different plasmid profileswere observed among the 37Acinetobacter species 3 isolatesobtained from 33 patients. Plasmids were found in 32 (86.5%)of 37 isolates, and no plasmids were found in 5 isolates(13.5%). If apparently identical isolates were eliminated,plasmids were detected in 27 (84.4%) of 32 strains. Therewas a median number of four plasmids in each strain, with arange of 0 to 10 plasmids. The sizes of the plasmids were inthe range of 2 to >30 kb. Representative plasmid profiles areshown in Fig. 1.

Plasmid profiles were identical, with one exception, onlyin isolates recovered from the blood culture and intravascu-lar catheter or cerebrospinal fluid from a given patient (Fig.1, lanes 3 to 8, patients B to D). Plasmid profiles were alsoidentical in two isolates obtained from two different patientsin the same hospital. Profiles were similar for two isolatesrecovered from a blood culture and a central venous catheterin a patient with device-related infection; the latter isolateshowed an additional plasmid with a low molecular weight(Fig. 1, lanes 1 and 2, patient A).

Plasmids were found in all A. johnsonii isolates tested;each strain harbored a median number of 10 plasmids (range,3 to 15). Sixteen different profiles were detected in 20isolates obtained from 16 patients; identical profiles werefound only in those isolates that were recovered from theblood culture and intravascular catheter of the same patients(Fig. 2, lanes 1 to 8, patients A to D).

Similar results were obtained with A. junii (data notshown). All isolates harbored plasmids, with an average ofsix plasmids in each strain (range, 2 to 15). Eight differentplasmid profiles were detected among the eight isolatestested.Among the 10 A. Iwoffli isolates tested, nine different

plasmid profiles were found (data not shown). All strainsharbored plasmids, with a median number of nine plasmidsin each strain (range, 4 to 16). Only two isolates from a bloodculture and a peripheral venous catheter of a patient withcatheter-related sepsis were shown to have identical plasmidprofiles.

Reproducibility. For comparison and reproducibility test-ing, one strain was retested on each gel. In addition, fourAcinetobacter species 3 strains and two A. johnsonii strainswere maintained at room temperature for up to eight weeksand subcultured weekly. Longitudinal reproducibility of

VOL. 32, 1994

on Septem

ber 20, 2018 by guesthttp://jcm

.asm.org/

Dow

nloaded from

84 SEIFERT ET AL.

C h r

K b

i3 1

9.4

6. 5

4. 3

2 3

2.0

FIG. 1. Plasmid DNA fingerprints ofAcinetobacter species 3 strains. Lanes: 1 and 2, 3 and 4, 5 and 6, blood culture and central venouscatheter isolates, respectively, from patients A to C; 7 and 8, blood culture and cerebrospinal fluid isolates from patient D; 9 to 19, isolatesrecovered from different patients, demonstrating highly variable profiles; 20, reference strain Acinetobacter species 3 CIP 70-29; Ref.,reference molecular size markers (indicated on the right). Chr, chromosomal DNA.

plasmid profiles was studied by running the isolates side byside on the same gel with corresponding isolates that werekept frozen. Plasmid profiles were identical; no loss ofplasmid bands was found.

DISCUSSION

BecauseAcinetobacter species are ubiquitous in the envi-ronment and are commensals on human skin and mucousmembranes (24, 31), they are common contaminants and

there is difficulty in determining whether the isolation ofAcinetobacter species indicates infection or merely contam-ination. The diagnosis of infection with Acinetobacter spe-cies often depends on the isolation of the same strain fromthe lesion, blood, intravascular device, urine, cerebrospinalfluid, or peritoneal dialysate. Epidemiologic typing is there-fore applied in the diagnosis of a single patient. Transmissionof a strain between patients or staff is another level at whichtyping can elucidate the epidemiology. Examples of this typeof transmission exist for A. baumannii and Acinetobacter

K b

2 3.1

9.4

6.5

4.3

Ch r

2.32.0

FIG. 2. Plasmid DNA fingerprints ofA. johnsonii strains. Lanes: 1 and 2, 3 and 4, 5 and 6, 7 and 8, blood culture and intravascular catheterisolates, respectively, from patients A to D; 9 to 18 and 21, isolates recovered from different patients, demonstrating highly variable profiles;19, reference strain A. johnsonii ATCC 17909; Ref. and 20, molecular size markers (indicated on the left). Chr, chromosomal DNA.

J. CLIN. MICROBIOL.

on Septem

ber 20, 2018 by guesthttp://jcm

.asm.org/

Dow

nloaded from

PLASMID DNA FINGERPRINTING OF ACINETOBACTER SPECIES 85

species 3 (11, 18), but it is currently not known if patient-to-patient transmission or hospital outbreaks occur with otherAcinetobacter species.

Various methods for epidemiological typing of Acineto-bacter strains have been developed and have recently beenreviewed (1, 5). A biotyping system has been introduced byBouvet and Grimont (8) and has proved useful in thedelineation of hospital outbreaks. This system, however, hasbeen developed only for A. baumannii and Acinetobacterspecies 3; but, whereas 19 different biotypes ofA. baumanniiwere identified, only 3 biotypes were observed among Aci-netobacter species 3 strains (9). The bacteriophage typingsystem was based on the use of two complementary sets ofbacteriophages. This typing system has been used only in asingle laboratory in the world. Typeability among A. bau-mannii strains was about 75% (9, 26), whereas, when strainsother than A. baumannii were considered, 68% of strainswere untypeable (9). Antibiogram typing was used in severalstudies as an epidemiological marker (1, 16, 19) and may bevaluable when there are marked similarities or dissimilaritiesin resistance patterns. This method has not been used forAcinetobacter strains other than A. baumannii, but ourresults indicate that antibiotic resistance typing of Acineto-bacter species 3, A. johnsonii, A. junii, and A. Iwoffli hasonly poor discriminative power and cannot be used as areliable marker in epidemiological investigations. Antibio-grams of these isolates revealed that organisms with identi-cal susceptibilities belonged to as many as 25 differentplasmid types. In two recent studies a serotyping system hasbeen designed forA. baumannii and Acinetobacter species 3with polyclonal rabbit immune sera; 20 and 13 serovars,respectively, have been identified (32, 33). This systemseems promising for the study of outbreaks of nosocomialinfection but is rather labor intensive.The typing procedures mentioned so far involved the use

of conventional systems based mainly on phenotype charac-teristics of the bacteria. Some of these techniques, such asphage typing and serotyping, involve the use of specificreagents that are not readily available and often difficult todevelop. The application of modern molecular biology haspermitted new approaches for epidemiologic typing of Aci-netobacter species. These methods are being applied to anever-increasing number of organisms. Conflicting resultswere obtained by using the electrophoretic patterns of cellenvelope proteins. Whereas several studies showed that cellenvelope protein patterns were a valuable tool in typingAcinetobacter strains and allowed differentiation amongstrains associated with hospital outbreaks (9, 11, 12), Giam-manco et al. (16) failed to show any diversity not only withinbut also between some of the biotypes of A. baumannii.Strains identified as species other than A. baumannii mostlyshowed unique profiles, but epidemiologically well-definedstrains have only rarely been studied. Recently, however,two hospital outbreaks due to Acinetobacter species 3 havebeen described and characterized by cell envelope proteinprofiles. All isolates gave identical patterns (11, 18).

Plasmid profile typing and plasmid fingerprinting wereshown to be useful methods of epidemiological typing ofvarious bacterial species (22). Few studies only have beenperformed with Acinetobacter species. Analysis of plasmidprofiles of A. baumannii strains proved to be helpful indelineating nosocomial outbreaks (17, 23). Gerner-Smidt (13)studied the plasmid content of 93 clinical Acinetobacterstrains; unfortunately, strain identification was based onolder taxonomic schemes of the genus. The investigatorreported that 81% of the strains harbored plasmids.

We studied the value of plasmid profile typing of 75Acinetobacter isolates identified as species other than A.baumannii. A total of 84.4% of Acinetobacter species 3isolates and allA. johnsonii, A. junji, and A. Iwoffii isolates,contained plasmids. With one exception, identical profileswere found only in corresponding blood culture and intra-vascular catheter or cerebrospinal fluid isolates from a givenpatient, thus providing evidence that Acinetobacter speciesother than A. baumannii may be involved in device-relatedinfections. This was recently demonstrated by us for A.johnsonii (29). Restriction endonuclease digestion of plasmidDNA may increase the discriminatory power of plasmidanalysis (17), in particular for those strains that contain onlya few plasmids with similar sizes. Digestion of plasmids wasnot considered necessary in our study because the highnumber of plasmid bands in the strains investigated facili-tates interpretation and comparison of plasmid profiles andcontributes to the discriminative power of this typingmethod. The demonstration of such diversity among plasmidpatterns supports the assumption that isolates with differentplasmid profiles represent different strains and that plasmiddifferences do not reflect only acquisition or loss of R-plas-mids in a highly selective environment. Reproducibility ofplasmid profiles was excellent, even after prolonged storageof strains at room temperature.Although our primary intention was to assess the diversity

and reproducibility of plasmid types inAcinetobacter strainsother than A. baumannii, the data we have accumulatedsuggest some specific epidemiologic conclusions. Patient-to-patient transmission, acquisition of strains from a commonsource, or outbreaks of infections seem to be rather unusualwith these species. We have no conclusive proof for this,because colonization in patients was not monitored prospec-tively so that a possible hospital outbreak might have goneundetected. However, we evaluated all isolates that wereconsecutively recovered at our institute during the studyperiod. With one exception, identical or similar profiles inisolates obtained from two different patients were not de-tected, though all isolates were obtained from the samegeographical area and the majority of isolates were obtainedfrom only three different hospitals. It is therefore unlikelythat we might have missed an outbreak of infection orcolonization. No clustering of type by ward or time wasobserved. All but two patients were treated in normal wardsand not in intensive care units. Though all infections werenosocomially acquired, these cases may represent infectionsby strains resident in the patients even before hospitaliza-tion. These findings are in contrast to the results obtained inepidemiological studies of A. baumannii, as patient-to-pa-tient transmission or common-source outbreaks in intensivecare units have often been reported for A. baumannii (2, 4,17, 23). Similar data were reported by Traub (33), who didnot find hospital outbreaks or nosocomial cross-infectionsdue to Acinetobacter species 3. However, nosocomialspread of Acinetobacter species 3 may occur, as was re-cently demonstrated for the first time in a medical and aneonatal intensive care unit (11, 18).No single typing system has generally been accepted, and

most investigators advocate the use of at least two differentmethods in epidemiological studies ofAcinetobacter species(1, 5, 9, 11, 19). With strains belonging to species other thanA. baumannii, species identification and plasmid typingseem to be a good approach in epidemiological investiga-tions, whereas antimicrobial susceptibility testing would notbe able to distinguish among most organisms.

VOL. 32, 1994

on Septem

ber 20, 2018 by guesthttp://jcm

.asm.org/

Dow

nloaded from

86 SEIFERT ET AL.

REFERENCES1. Alexander, M., M. Rahman, M. Taylor, and W. C. Noble. 1988.A study of the value of electrophoretic and other techniques fortyping Acinetobacter calcoaceticus. J. Hosp. Infect. 12:273-287.

2. Allardet-Servent, A., N. Bouziges, M. J. Carles-Nurit, G. Bourg,A. Gouby, and M. Ramuz. 1989. Use of low-frequency-cleavagerestriction endonucleases for DNA analysis in epidemiologicalinvestigations of nosocomial bacterial infections. J. Clin. Micro-biol. 27:2057-2061.

3. Andrews, H. J. 1986. Acinetobacter bacteriocin typing. J. Hosp.Infect. 7:743-751.

4. Beck-Sague, C. M., W. R. Jarvis, J. H. Brook, D. H. Culver, A.Potts, E. Gay, B. W. Shotts, B. Hill, R. L. Anderson, and M. P.Weinstein. 1990. Epidemic bacteremia due to Acinetobacterbaumannii in five intensive care units. Am. J. Epidemiol.132:723-733.

5. Bergogne-Berezin, E., and M. L. Joly-Guillou. 1991. Hospitalinfection with Acinetobacter spp.: an increasing problem. J.Hosp. Infect. 18(Suppl A):250-255.

6. Berk, S. L., and W. R. McCabe. 1981. Meningitis caused byAcinetobacter calcoaceticus var. anitratus. A specific hazard inneurosurgical patients. Arch. Neurol. 38:95-98.

7. Bouvet, P. J. M., and P. A. D. Grimont. 1986. Taxonomy of thegenus Acinetobacter with the recognition of Acinetobacterbaumannii sp. nov., Acinetobacter haemolyticus sp. nov., Aci-netobacterjohnsonii sp. nov., and Acinetobacterjunii sp. nov.and emended description of Acinetobacter calcoaceticus andAcinetobacter Iwoffii. Int. J. Syst. Bacteriol. 36:228-240.

8. Bouvet, P. J. M., and P. A. D. Grimont. 1987. Identification andbiotyping of clinical isolates of Acinetobacter. Ann. Inst. Pas-teur Microbiol. 138:569-578.

9. Bouvet P. J. M., S. Jeanjean, J.-F. Vieu, and L. DUkshoorn.1990. Species, biotype, and bacteriophage type determinationscompared with cell envelope protein profiles for typing Acine-tobacter strains. J. Clin. Microbiol. 28:170-176.

10. Cefai, C., J. Richards, F. K. Gould, and P. McPeake. 1990. Anoutbreak of Acinetobacter respiratory tract infection resultingfrom incomplete disinfection of ventilatory equipment. J. Hosp.Infect. 15:177-182.

11. DUkshoorn, L., H. M. Aucken, P. Gerner-Smidt, M. E.Kaufmann, J. Ursing, and T. L. Pitt. 1993. Correlation of typingmethods for Acinetobacter isolates from hospital outbreaks. J.Clin. Microbiol. 31:702-705.

12. Dikshoorn, L., M. F. Michel, and J. E. Degener. 1987. Cellenvelope protein profiles of Acinetobacter calcoaceticus strainsisolated in hospitals. J. Med. Microbiol. 23:313-319.

13. Gerner-Smidt, P. 1989. Frequency of plasmids in strains ofAcinetobacter calcoaceticus. J. Hosp. Infect. 14:23-28.

14. Gerner-Smidt, P. 1992. Ribotyping of the Acinetobacter cal-coaceticus-Acinetobacter baumannii complex. J. Clin. Micro-biol. 30:2680-2685.

15. Gervich, D. H., and C. S. Grout. 1985. An outbreak of nosoco-mial Acinetobacter infections from humidifiers. Am. J. Infect.Control 13:210-215.

16. Giammanco, A., J. F. Vieu, P. J. Bouvet, A. Sarzana, and A.Sinatra. 1989. A comparative assay of epidemiological markersfor Acinetobacter strains isolated in a hospital. Int. J. Microbiol.272:231-241.

17. Hartstein, A. I., V. H. Morthland, J. W. Rourke, J. Freeman, S.

Garber, R. Sykes, and A. L. Rashad. 1990. Plasmid DNAfingerprinting of Acinetobacter calcoaceticus subspecies anitra-tus from intubated and mechanically ventilated patients. Infect.Control Hosp. Epidemiol. 11:531-538.

18. Horrevorts, A., I. Tjernberg, and L. Djkshoorn. 1993. Nosoco-mial spread of Acinetobacter DNA-group 3 in a neonatal inten-sive care unit. Program Abstr. 6th Eur. Congr. Clin. Microbiol.Infect. Dis., abstr. 584.

19. Joly-Guillou, M. L., E. Bergogne-Berezin, and J.-F. Vieu. 1990.A study of the relationships between antibiotic resistance phe-notypes, phage-typing and biotyping of 117 clinical isolates ofAcinetobacter spp. J. Hosp. Infect. 16:49-58.

20. Joly-Guillou, M. L., E. Bergogne-Berezin, and J.-F. Vieu. 1990.Epidemiologie et resistance aux antibiotiques des Acinetobacteren milieu hospitalier. Presse Med. 19:357-61.

21. Juni, E. 1972. Interspecies transformation of Acinetobacter;genetic evidence for a ubiquitous genus. J. Bacteriol. 112:917-931.

22. Mayer, L. W. 1988. Use of plasmid profiles in epidemiologicsurveillance of disease outbreaks and in tracing the transmissionof antibiotic resistance. Clin. Microbiol. Rev. 1:228-243.

23. Patterson, J. E., J. Vecchio, E. L. Pantelick, P. Farrel, D.Mazon, M. J. Zervos, and W. J. Hierholzer. 1991. Association ofcontaminated gloves with transmission of Acinetobacter cal-coaceticus var. anitratus in an intensive care unit. Am. J. Med.91:479483.

24. Rosenthal, S., and I. B. Tager. 1975. Prevalence of gramnegative rods in the normal pharyngeal flora. Ann. Intern. Med.83:355-357.

25. Sakata, H., K. Fujita, S. Maruyama, H. Kakehashi, Y. Mori, andH. Yoshioka. 1989. Acinetobacter calcoaceticus biovar anitratussepticaemia in a neonatal intensive care unit: epidemiology andcontrol. J. Hosp. Infect. 14:15-22.

26. Santos-Ferreira, M. O., J.-F. Vieu, and B. Klein. 1984. Phage-types and susceptibility to 26 antibiotics of nosocomial strains ofAcinetobacter isolated in Portugal. J. Int. Med. Res. 12:364-368.

27. Seifert, H., and R. Baginski. 1992. The clinical significance ofAcinetobacter baumannii in blood cultures. Int. J. Microbiol.277:210-218.

28. Seifert, H., R. Baginski, A. Schulze, and G. Pulverer. 1993. Thedistribution of Acinetobacter species in clinical culture materi-als. Int. J. Microbiol. 279:544-552.

29. Seifert, H., A. Strate, A. Schulze, and G. Pulverer. 1993.Catheter-related bacteremia due to Acinetobacter johnsonii:report of 13 cases. Clin. Infect. Dis. 17:632-636.

29a.Seifert, H., A. Schulze, and R. Baginski. 1993. Plasmid DNAfingerprinting of Acinetobacter species 3 strains. Program Ab-str. 6th Eur. Congr. Clin. Microbiol. Infect. Dis., abstr. 113.

30. Sherertz, R. J., and M. L. Sullivan. 1985. An outbreak ofinfections with Acinetobacter calcoaceticus in burn patients:contamination of patients' mattresses. J. Infect. Dis. 151:252-258.

31. Taplin, D., G. Rebell, and N. Zaiab. 1963. The human skin as asource of Mima-Herella infections. JAMA 186:166-168.

32. Traub, W. H. 1989. Acinetobacter baumannii serotyping fordelineation of outbreaks of nosocomial cross-infection. J. Clin.Microbiol. 27:2713-2716.

33. Traub, W. H. 1990. Serotyping of clinical isolates of Acineto-bacter: serovars of genospecies 3. Int. J. Microbiol. 273:12-23.

J. CLIN. MICROBIOL.

on Septem

ber 20, 2018 by guesthttp://jcm

.asm.org/

Dow

nloaded from