Evaluation of phenotypic and genotypic markers for clinical strains of Acinetobacter baumannii
Adriana S. Limansky1, María I. Zamboni1, 2, María C. Guardati2, Gustavo Rossignol1, 2, Eleonora Campos3, Alejandro M. Viale1
1 Instituto
de Biología Molecular y Celular de Rosario (IBR, CONICET), Facultad de Ciencias
Bioquímicas y Farmacéuticas, Universidad Nacional de Rosario;
2 Hospital de Emergencias
Clemente Alvarez (HECA), Departamento Bioquímico Municipal, Secretaría de Salud
Pública, Rosario;
3 Instituto
de Biotecnología CICVyA, INTA, Castelar, Buenos Aires.
Postal address: Dra. Adriana S. Limansky, Suipacha 531, 2000 Rosario, Argentina. Fax: 54 341 4390465. E-mail: alimansk@infovia.com.ar
Abstract
Acinetobacter baumannii is an
important opportunistic pathogen that is rapidly evolving toward multidrug resistance
and is involved in various nosocomial infections that are often severe. It
strongly prompts the epidemiological study of A. baumannii infections.
However, there is no a generally accepted typing scheme. Different genotypic
and phenotypic procedures were evaluated for the characterization of clinical
isolates of A. baumannii isolated from a community hospital of Rosario,
Argentina (Hospital de Emergencias Clemente Alvarez, HECA), during a period of
four years. These included PCR with degenerate oligonucleotide primers
(DO-PCR), repetitive extragenic palindromic-PCR (REP-PCR), pulsed-field gel
electrophoresis (PFGE), and antibiotyping. Amongst individual methods,
DO-PCR and PFGE were found the most suitable methods to discriminate A.
baumannii clinical isolates [discriminatory indexes (D) of 0.98 and 0.96,
respectively]. On the other hand, both antibiotyping and REP-PCR were much less
discriminatory (D: 0.86 and 0.77, respectively). The combination of
antibiotyping with any of the above genotypic procedures was found to largely
increase D. In particular, the combination of DO-PCR and antibiotyping provided
the best discriminatory method for epidemiological studies of A. baumannii.Combination of the different genotypic and phenotypic procedures allowed
the inference of genetic relationships and dissemination of multidrug-resistant A. baumannii clones in HECA in the period 1994-1999. One particular
strain, which showed sensibility to carbapenems, was found widely distributed
in this hospital during 1994-1996. A different strain, showing additional
resistance to carbapenems, rapidly disseminated in HECA in coincidence with the
introduction of imipenem therapy in 1997.
Key words: Molecular typing; Nosocomial infections; Acinetobacter baumannii; Genotypic markers.
Resumen
Evaluación de marcadores fenotípicos
y genotípicos para cepas clínicas de Acinetobacter baumannii. Acinetobacter baumannii es un importante patógeno oportunista. Este
microorganismo adquiere con facilidad resistencia a antimicrobianos,
involucrándose en infecciones nosocomiales generalmente graves. Estas
características promueven el análisis epidemiológico de las infecciones
provocadas por el mismo. Sin embargo, no hay aún un esquema de tipificación generalmente
aceptado para este patógeno. Hemos evaluado en este trabajo diferentes
procedimientos fenotípicos y genotípicos para la caracterización de
aislamientos clínicos de A. baumannii aislados en un Hospital Público de
Rosario (Hospital de Emergencias Clemente Alvarez, HECA), durante un período de
cuatro años. Estos incluyeron PCR con oligonucleótidos degenerados (OD-PCR),
PCR empleando cebadores homólogos a secuencias palindrómicas extragénicas
repetitivas (REP-PCR), electroforesis en geles de agarosa con campo pulsado
(PFGE) y ensayo de susceptibilidad a antimicrobianos. OD-PCR y PFGE, entre los
métodos individuales, fueron los métodos de mayor poder discriminatorio (índice
discriminatorio, D, de 0.98 y 0.96; respectivamente). Por otra parte, el antibiotipo
y REP-PCR presentaron menor discriminación (D: 0.86 y 0.77; respectivamente).
La combinación del antibiotipo con cada uno de los procedimientos genotípicos
mencionados originó un aumento importante en los índices discriminatorios de
cada método. En particular, la combinación de OD-PCR y antibiotipo constituyó
la mejor metodología para el estudio epidemiológico de A. baumannii.
Así, la combinación de los procedimientos feno- y genotípicos mencionados
permitió inferir las relaciones genéticas y la diseminación de clones de A.
baumannii multirresistentes en el HECA en el período 1994-99. Una cepa
particular, sensible a imipenem, estuvo ampliamente diseminada en el hospital
durante 1994-1996. Por otra parte, un clon diferente, con resistencia adicional
a carbapenemes, se diseminó rápidamente en el hospital en 1997, en coincidencia
con la introducción de imipenem como terapia antibiótica.
Palabras clave: Tipificación molecular; Infecciones nosocomiales; Acinetobacter baumannii; Marcadores genotípicos.
During the last years, Acinetobacter baumannii has
increasingly become known as an agent of nosocomial infection1. They cause a wide range of clinical
complications, such as pneumonia, septicemia, urinary tract infection, wound
infection, and meningitis, especially in inmunocompromised patients1. This situation is aggravated by the
rapid selection of multiple-drug resistant strains of this bacterium as a
non-desired side effect of antimicrobial treatments2. The challenge posed by this emerging pathogen demands,
among other responses, the use of discriminatory and reproducible typing
methods to accurately characterize strains responsible of possible outbreaks,
evaluate their persistence, and identify their routes of transmission3. A variety of typing methods have been
tested for this purpose, including antibiotyping, plasmid profile, ribotyping,
pulsed-field gel electrophoresis (PFGE), and PCR fingerprinting4-10. Yet, no generally accepted typing
strategy has emerged for Acinetobacter strains from these studies,
although PCR fingerprinting is generally employed for typing procedures due to
its simplicity and rapidity11-13.
Still, a comparative evaluation of different techniques is always recommended
in the sense that it identifies the particular methodology or methodologies
most suited for epidemiological purposes7-9.
We have evaluated different methods aimed to characterize A.
baumannii isolates from a community hospital of Rosario, Argentina,
including PCR fingerprinting, repetitive extragenic palindromic-PCR,
pulsed-field gel electrophoresis, and antibiotyping. A combination of these
procedures was also employed for the inference of clonal relationships amongst
multidrug-resistant A. baumannii strains causing a variety of infections
during 1994-1999 in the above nosocomial institution.
Materials and Methods
Bacterial isolates
Bacterial isolates were obtained from
colonized or infected patients during November 1994 through March 1999 in
Hospital of Emergencies Clemente Alvarez (HECA), Rosario, Argentina. HECA is a
public, 150-bed tertiary-care teaching hospital with an average attendance of
around 35000 patients per year. The intensive-care unit of this hospital is
constituted by a 10-bed facility, admitting an average of around 700 patients
per year.
Identification of bacterial isolates as A. baumannii was
performed by the API 20NE Identification System (bioMérieux, Lyon,
France) and by their ability to grow at 37°C, 41°C, and 44°C4. Relevant data of the 42 isolates
employed in this study are provided in Tables 1 and 2.
TABLE 1.– Characterization of Acinetobacter baumannii isolates
included in this study
*ITU:
Intensive Therapy Unit; OPD: Out-patients Department
†CSF: cerebrospinal fluid; BAL: bronchoalveolar
lavage; AF: ascitic fluid
‡ ATB:
antibiotype (see Materials and Methods).
§ ND:
No determined.
TABLE 2.– Antimicrobial susceptibility patterns of Acinetobacter
baumannii strains
* Criteria for antibiotype (Atb)
characterization were defined in Materials and Methods.
† Isolates were considered
multidrug-resistant (MR) when they expressed resistance to at least three of
the antimicrobial groups defined (see Materials and Methods), and susceptible
(S) when they showed susceptibility to at least three groups (in both cases
with the exclusion of the carbapenems). A given isolate was considered
resistant to a particular antimicrobial group when it expressed resistance to
at least half of the antimicrobials that composed the group. MRImpI or MRImpR strains correspond to multidrug-resistant
strains with intermediate or resistant category for imipenem, respectively14.
‡ QN: quinolones. MER: meropenem; IMP:
imipenem; AMS: ampicillin-sulbactam; CAZ: ceftazidime; CTX: cefotaxime; PIP:
piperacillin; PIP-Taz: piperacillin-tazobactam; GEN: gentamicin; AKN: amikacin;
CIP: ciprofloxacin; TMS: trimethoprim-sulfamethoxazol; S: susceptible; R:
resistant; I: intermediate; V: variable14.
Antibiotype patterns
The susceptibility of A. baumannii isolates to different antimicrobial agents was assessed by the disk diffusion method on Mueller-Hinton agar14 (Table 2). In total, 11 antimicrobials were tested, which were classified in 5 different groups as follows: group 1 comprised meropenem and imipenem; group 2, other b-lactams including ampicillin-sulbactam, ceftazidime, cefota-xime, piperacillin, and piperacillin-tazobactam; group 3, aminoglucosides including gentamicin and amikacin; group 4, ciprofloxacin; and group 5, trimethoprim-sulfamethoxazole. Bacterial isolates were assigned to different antibiotypes (I-VIII, Table 2) on the basis of the following criteria: two isolates were considered to belong to a different antibiotype when the diameters of the inhibition zones corresponding to four or more of the above antimicrobials showed differences higher than 3 mm (data not shown).
Isolation of Acinetobacter baumannii DNA
Bacterial DNA was obtained from cells treated with lysozyme/proteinase K, followed by phenol extraction to remove contaminating proteins15. DNA concentrations were determined spectrophotometrically15.
PCR fingerprinting
A PCR assay employing degenerate oligonucleotide
primers (DO-PCR) was employed, following essentially previously described
methods16. A. baumannii
isolates were considered different when their amplification profiles differed
in more than two bands. Differences in band intensity at a given position
between two given isolates were not considered significant for their
differentiation.
REP-PCR reactions were performed essentially as previously
described9. A difference of 1
or 2 bands was not considered for discrimination purposes, providing that all
other bands migrated at equivalent positions9.
Pulsed-field gel electrophoresis
Bacterial DNA was prepared and digested with ApaI restriction endonuclease essentially as described17. The resulting fragments were separated in a 1% agarose gel employing a CHEF-DR II Bio-Rad apparatus (Bio-Rad Laboratories, Richmond, CA) at 4.5 V/cm by using the following two intervals of ramped pulse times: 1 to 10 s for 12 h and 10 a 100 s for 16 h. PFGE patterns were interpreted following Tenover et al18. Two given isolates were considered different if their PGFE patterns differed in at least four bands18.
Data acquisition and analysis.
DNA fingerprints generated by the different methods were examined visually, and a numeric profile for each strain was constructed on the basis of the presence or absence of an amplification band at a given position19. A similarity matrix was calculated from these data following the procedure of Nei and Li20. In turn, this matrix was used to generate a dendrogram according to the neighbor-joining method21.
Discriminatory index
A numerical index of discrimination (D)22 was used to evaluate the discriminatory power of each individual method or their different combinations.
Results
As noted above, 8 different antibiotypes
were defined amongst the 42 A. baumannii isolates analyzed in this work
(Tables 1 and 2). Five of them were found amongst the 37 multidrug-resistant
(MR) isolates (i.e., I, II, III, V, and VIII), and 3 amongst the 5
susceptible (S) isolates (i.e., IV, VI, and VII).
The use of DO-PCR analysis disclosed the presence of twelve
distinct patterns among the same isolates (A-L, Fig. 1 and Table 1).
Interestingly, the 5 S strains showed 5 different band profiles (C, E, G, I,
and K, respectively). A lower degree of diversity was observed amongst the 37
MR isolates, which showed the presence of seven different patterns (A, B, D, F,
H, J, and L). It is worth noting at this stage that 17 out of the 19 MR isolates
recovered in HECA during the period 1994-1996 (i.e., 89%) belonged to
DO-PCR type A, whilst only 1 strain was found to possess profile B during the
same period (Table 1). On the other hand, 10 out of the 18 (i.e., 56%)
MR strains isolated in the period 1997-1999 belonged to genotype B, while the
number of isolates retaining profile A had been reduced to only 3 (i.e.,
17%) of the analyzed sample during the same period (Table 1). These results
suggested that different clones disseminated in HECA during each particular
period.
Fig. 1.– Characterization of A. baumannii clinical
isolates by DO-PCR. All procedures are described in Materials and Methods. The
following isolates of A. baumannii were used (see Table 1 for details):
Lane 1, 23395 (A); lane 2, 24214 (B); lane 3, 5922 (D); lane 4, 29345 (F); lane
5, 29697 (J); lane 6, 28871 (H); lane 7, 825 (L); lane 8, 2339 (C); lane 9,
25365 (G); lane 10, 29533 (I); lane 11, 29721 (K); lane 12, 24419 (E). The
pattern assigned to each isolate is indicated between brackets. In some cases,
the amplification patterns correspond to an isolate representative of the clon
(Table 1). MR: multiresistant isolates; S: susceptible isolates. The position
of the size markers (EcoRI/HindIII-digested lambda DNA) are
indicated in the left margin.
The analysis of the same bacterial sample by REP-PCR indicated the existence of eight different patterns (A’ to H’, Fig. 2 and Table 1). Again, all 5 S isolates were clearly differentiated by this procedure (patterns C’, D’, E’, G’, and H’, respectively, Table 1). As above, the number of different genotypes was much reduced for MR isolates (Table 1). However, REP-PCR showed the presence of only three distinct profiles amongst the MR strains (A’, B’, and F’, Table 1), contrasting the results of DO-PCR (see above).
Fig. 2.– Characterization of A. baumannii clinical
isolates by REP-PCR. All procedures are described in Materials and Methods. The
following isolates were used (see Table 1 for details): Lane 1, 23395 (pattern
A’); lane 2, 28871 (B’); lane 3, 28616 (F’); lane 4, isolate 2339 (C’); lane 5,
isolate 24419 (D’); lane 6, isolate 25365 (E’); lane 7, 29533 (G’); lane 8,
29721 (H’). For details see the legend to Fig. 1.
The same strains were further analyzed by PFGE. By this procedure, eleven different patterns were disclosed (A" to K", Fig. 3 and Table 1). Again, 5 of them corresponded to S isolates (C’", F’", H’", I’", and J’", respectively, Table 1), whereas six differentiated patterns were observed amongst the MR isolates (A’", B’", D’", E’", G’", and K’", Table 1).
Fig. 3.– Characterization of A. baumannii clinical
isolates by PFGE. The following isolates were used: lane 1, 5922 (pattern D");
lane 2, 29697 (A") ; lane 3, 24474 (G"); lane 4, 28871 (B"); lane 5, 24214
(E"); lane 6, 825 (K"); lane 7, 2339 (C"); lane 8, 24419 (F"); lane 9, 29533
(I"); lane 10, 29721 (J"); lane 11, 25365 (H"). For other details see the
legend to Fig. 1.
The number of types and discriminatory indexes calculated for each particular procedure and for their combination are summarized in Table 3. As seen in the table, OD-PCR and PFGE showed the highest discriminatory capabilities (D of 0.98 and 0.96, respectively). On the other hand, antibiotyping and REP-PCR were much less discriminatory (D of 0.86 and 0.77; respectively). Interestingly, the combination of antibiotyping with any of these genotypic procedures largely increased discriminatory indexes (Table 3).
TABLE 3.– Discriminatory indexes of the phenotypic and
genotypic techniques used in this study*
*D
estimations were done employing A. baumannii
isolates 8808, 2339, 5922, 23395, 24214, 24419, 24474, 25365, 28616, 28871,
29533, 29697, 29721 and 825 (see Table 1).
A dendrogram was constructed on the basis of the combined data of DO-PCR, REP-PCR and PFGE for 14 representative A. baumannii strains recovered in the period November 1994 through March 1999 (Fig. 4). As seen in the figure, the S isolates appeared clearly differentiated from the MR strains, composing a separate and highly divergent group. In turn, the more compact clustering of the MR strains suggests a closer genetic relationships. In particular, a cluster of highly related isolates was noteworthy in that it contained strains that showed different levels of resistance to imipenem (Fig. 4, see also Table 2).
Fig. 4.– Dendrogram indicating genetic relationships between
relevant A. baumannii strains analyzed in this study. The dendrogram was
inferred from a distance matrix constructed from the combined data of DO-PCR,
REP-PCR and PFGE of fourteen bacterial isolates showing representative patterns.
For details see Materials and Methods.
Discussion
Since 1994, significant increases in the
recovery of multiresistant A. baumannii clinical strains was noted in
HECA, particularly amongst immunocompromised and intensive-care unit patients.
This resulted in considerable overuse of imipenem, to which the organisms were
uniformly susceptible. In 1997, carbapenem-resistant strains emerged and
rapidly disseminated, prompting us to conduct this epidemiological
investigation. Initial screenings were done by determining antibiotic
resistance patterns (Tables 1 and 2), in an attempt to identify the
dissemination of particular strains. However, attempts to perform
epidemiological studies soon required the use of more discriminatory markers
than those provided by antibiotyping. We therefore evaluated different
genotypic procedures, including DO-PCR, REP-PCR and PFGE, in order to analize
their usefulness to characterize A. baumannii isolates obtained in the
period from November 1994 through March 1999.
In our experience, DO-PCR performed as the most discriminatory
amongst all tested methods (D=0.98), even slightly higher than the
most-elaborated PFGE technique (D=0.96). On the other hand, the performance of
REP-PCR was much poorer (D=0.77), even lower than antibiotyping (D=0.86).
Interestingly, the combination of antibiotyping with any of the above genotypic
methodologies increased discriminatory capability (Table 3). In this sense, the
best discrimination was provided by the combination of DO-PCR with
antibiotyping (D=1, Table 3). This indicates that antibiotype may still provide
a valuable tool for epidemiological purposes.
The combination of all genotypic markers was employed to
determine the population structure of A. baumannii during this period
(Fig. 4). As expected, MR strains showed much less genetic variability than S
strains, most probably reflecting the selective pressure caused by
antimicrobial therapy. In addition, this analysis suggested that a particular
epidemic MR A. baumannii strain disseminated in HECA during the period
1994-1996 (clone A, Fig. 1 and Table 1). This clone was replaced since 1997 by
an emerging new clone with additional resistance to imipenem (clone B), a
situation that was coincident with the introduction of this antibiotic therapy.
An infection control strategy that included strict compliance with cross
transmission prevention protocols and restrictions in the use of carbapenem was
implemented after this study, resulting in a marked reduction in the incidence
of infection and colonization by MR A. baumannii strains.
Acknowledgements: This work was supported in part by grants from Subsecretaria de Investigación y Tecnología, Ministerio de Salud de la Nación Argentina; Agencia Nacional de Promoción Científica y Tecnológica (ANPCyT), and Municipalidad de Rosario. AV is a Career Investigator of the Consejo Nacional de Investigaciones Científicas y Técnicas, Argentina.
References
1. Bergogne-Bérézin E, Towner KJ. Acinetobacter
spp. as nosocomial pathogens: microbiological, clinical and epidemiological
features. Clin Microbiol Rev 1996; 9: 148-65.
2. Amyes SGB, Young H-K. Mechanisms of antibiotic
resistance in Acinetobacter spp.-genetics of resistance. In:
Bergogne-Berezin E, Jolly-Guillou ML, Towner KJ, (eds). Acinetobacter:
Microbiology, Epidemiology, Infections, Management. New York, N.Y: CRC Press.
1996, p 185-223.
3. Maslow JN, Mulligan ME, Arbeit RD. Molecular
epidemiology: Application of contemporary techniques to the typing of
microorganisms. Clin Infect Dis 1993; 17: 153-64.
4. Bouvet PJM, Grimont PA. Identification and biotyping of
clinical isolates of Acinetobacter. Ann Inst Pasteur/Microbiol
1987;138: 569-78.
5. Dijkshoorn L, Aucken HM, Gerner-Smidt P, Kaufmann ME,
Ursing J, Pitt TR. Correlation of typing methods for Acinetobacter
isolates from hospital outbreaks. J Clin Microbiol 1993; 31: 702-5.
6. Gouby A, Carles-Nurit MJ, Bouziges N, Bourg G, Mesnard
R, Bouvet PJ M. Use of pulsed-field gel electrophoresis for investigation of
hospital outbreaks of Acinetobacter baumannii. J Clin Microbiol
1992; 30: 1588-91.
7. Graser Y, Klare I, Halle E, et al. Epidemiological
study of an Acinetobacter baumannii outbreak by using polymerase chain
reaction fingerprinting. J Clin Microbiol 1993; 31: 2417-20.
8. Marcos MA, Jimenez de Anta MT, Vila J. Correlation of
six methods for typing nosocomial isolates of Acinetobacter baumannii. J
Med Microbiol 1995; 42: 328-35.
9. Reboli AC, Houston ED, Monteforte JS, Wood CA, Hamill
RJ. Discrimination of epidemic and sporadic isolates of Acinetobacter
baumannii by repetitive element PCR-mediated DNA fingerprinting. J Clin
Microbiol 1994; 32: 2635-40.
10. Seifert H, Gerner-Smidt P. Comparison of ribotyping and
pulse-field gel electrophoresis for molecular typing of Acinetobacter
isolates. J Clin Microbiol 1995; 33:1402-7.
11. Van Belkum A. DNA fingerprinting of medically important
microorganisms by use of PCR. Clin Microbiol Rev 1994; 7: 174-84.
12. Welsh J, McClelland M. Fingerprinting genomes using PCR
with arbitrary primers. Nucleic Acids Res 1990; 18: 7213-8.
13. Williams GKJ, Kubelik AR, Livak KJ, Rafalski JA, Tingey
SV. DNA polymorphisms amplified by arbitrary primers are useful as
genetic markers. Nucleic Acids Res 1990; 18: 6531-4.
14. National Committee for Clinical Laboratory Standards.
Performance standards for antimicrobial disk susceptibility tests. Approved
standard M2-A6. National Committee for Clinical Laboratory Standards, Wayne, P.
A. 1997.
15. Sambrook J, Fritsch E F, Maniatis T. Molecular Cloning.
A Laboratory Manual, 2nd ed.
Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1989.
16. Limansky A, Viale A. Can composition and structural
features of oligonucleotides contribute to their wide-scale applicability as
random PCR primers in mapping bacterial genome diversity?. J Microbiol Methods 2002; 50: 291-7.
17. Ausubel FM, Brent R, Kingston RE, et al. Current
Protocols in Molecular Biology. John Wiley and Sons; 1998.
18. Tenover FC, Arbeit RD, Goering RV, et al. Interpreting
chromosomal DNA restriction patterns produced by Pulsed-Field Gel
Electrophoresis: criteria for bacterial strain typing. J Clin Microbiol
1995; 33: 2233-9.
19. Woods TC, Helsel LO, Swaminathan B, et al.
Characterization of Neisseria meningitidis serogroup C by multilocus
enzime electrophoresis and ribosomal DNA restriction profiles (ribotyping). J
Clin Microbiol 1992; 30: 132-7.
20. Lamboy WF. Computing genetic similarity coefficients
from RAPD data: the effects of artifacts. PCR Meth Applic 1994; 4:
31-7.
21. Felsenstein J. PHYLIP: phylogeny inference package, v.
3.5c. Seattle, Washington: University of Washington. 1993.
22. Hunter PR, Gaston MA. Numerical index of the
discriminatory ability of typing systems: an application of Simpson’s index of
diversity. J Clin Microbiol 1988; 26: 2465-6.
Recibido: 26-01-2004
Aceptado: 2-04-2004