A Simple Method for Direct Detection and Discrimination of A. baumannii in Tracheal Aspirates without Culture Isolation

Currently, available phenotyping and commercial methods report A. baumannii only as Acinetobacter calcoaceticus-baumannii complex (ACB) and do not identify individual members of the complex. This is a single blind study aimed to evaluate certain commonly used species-specific genetic markers namely, Intergenic Transcribed Spacer region in 16S rRNA of A. baumannii (Ab-ITS) and gyrB, for identification of ACB members. These molecular targets were first validated on clinical isolates (n = 200) and subsequently on uncultured tracheal aspirates (n = 172). Among the clinical isolates, 183/200 (91.5%) were positive for Ab-ITS. The clinical isolates 17 (17/200) which are failed to amplify in Ab-ITS PCR were subsequently di-agnosed by gyrB PCR as A. calcoaceticus (n = 2), A. pitti (n = 6) and A. nosocomialis (n = 9) but not A. baumannii. Among the tracheal aspirates, 62 samples were reported as sterile in


Introduction
Acinetobacter baumannii causes a vast array of infections including ventilator associated pneumonia (VAP), secondary meningitis, urinary tract infections, septicemia and other severe infectious complications in the critical care (CCU) patients [1] [2]. Clinical microbiology laboratories usually report A. baumannii as "ACB complex". This includes six individual species namely A. baumannii, A. calcoaceticus, A. pitti, and A. nosocomialis, A. seifertii, A. lactucae [3]. These members have epidemiological and clinically relevant differences. They exhibit different drug susceptibility profiles though they are genetically and phenotypically very similar. In particular, A. baumannii is responsible for a large proportion of nosocomial infections. There has been a tremendous increase in antibiotic resistance among A. baumannii isolates in the last two decades [4].
Members of ACB complex pose limitations in phenotyping and conventional microbiological identification [5] [6] [7] [8]. Therefore, it would be desirable to apply molecular methods to identify the individual species of the ACB complex [7] [9]. Also, it would be ideal to be able to detect and report the species directly rapidly from the clinical sample without waiting for culture isolation and characterization.
Molecular methods which are used for identification of Acinetobacter spp. include ribotyping, polymerase chain reaction (PCR) hybridization with speciesspecific probes, methods based on the 16S -23S rRNA intergenic spacer region, tRNA spacer fingerprinting, Pulse Field Gel Electrophoresis (PFGE), methods based on gyrB and rpoB sequences, PCR-RFLP and AFLP analysis, Amplified Ribosomal DNA Restriction Analysis (ARDRA) and Random Amplified Polymorphic DNA (RAPD) typing [10] [11]. Amplified Ribosomal DNA Restriction Analysis (ARDRA) is rapid and more practical for species identification [12].
The main disadvantage of these typing methods is low reproducibility, especially in terms of global molecular epidemiology [7]. Sequence-based typing, such as single or multi locus or whole-genome sequencing, provides more reliable data for comparison from a geographical point of view or globally. However, the high cost of sequence-based typing prevents implementation by developing countries [6] [13]. Failure to identify the A. baumannii in ACB complex may adversely impact both diagnosis and infection control. Thus, there is an unmet need for a rapid and affordable diagnostic tool to identify A. baumannii in clinical isolates in general and particularly from uncultured clinical specimens. A promising and robust molecular target for identification of Acinetobacter spp. would be the polymorphisms within the rRNA intergenic spacer (ITS) sequences. These rRNA genes are present in the same order (i.e. 16S, 23S, and 5S) in their transcriptional units and well conserved throughout most bacterial species [14]. Among the available genetic targets, intergenic spacer in A. baumannii (Ab-ITS) holds-out as a good genetic marker for the species identification as it offers low intra species variation and high degrees of interspecies divergence [1] [9]. A simple sample processing protocol and PCR were optimized for detection of A. baumannii in uncultured tracheal aspirates. The internal transcribed spacer region of the ribosomal RNA of A. baumannii, called Ab-ITS, was used for identification of A. baumannii species. Results of this Ab-ITS-PCR were evaluated against conventional clinical microbiology phenotyping methods and VITEK 2 system. We also present in this article our experience of using gyrB based target sequences for differentiation of individual members of ACB complex.

Materials and Methods
This is a prospective study conducted from November 2014 to March 2015, at Gleneagles Global Hospitals.

Specimen Collection
Samples were collected from two tertiary care health centers in Hyderabad for this study.

Genotyping of A. baumannii by Ab-ITS PCR
PCR assay was performed with the genomic DNA in 20 μl total reaction volume.
The PCR reaction mixture contained 20 pmol of each primer, 200 μmol of dNTPs, 1 U of Taq DNA polymerase (KAPA Taq DNA Polymerase, KAPA Biosystems Inc), 2 μl 10X buffer, 1.65 mM MgCl 2 and 100 ng of genomic DNA lysate. The primers and PCR conditions used in this study are given in Table 1.
The amplicons of ITS region and gyrB were resolved and analyzed on 2% agarose gel after ethidium bromide staining and documented in a gel documentation system (UVITEC, UK). The tracheal aspirate's phenotyping results were de-coded after the genotyping was completed.

Phenotyping and AST
The 200 ACB isolates which belonged to Category 1 were identified in the clinical microbiology laboratory using both manual (biochemical) and automated methods and results interpreted as per BSAC guidelines (2015) [16].

Ab-ITS PCR Results
Category 1 (n = 200) ACB isolates 183 (91.5%) isolates were positive for Ab-ITS and a typical pattern of amplicons in agarose gel electrophoresis is represented in Figure 2. Results of identification and drug sensitivity profiles of ACB isolates are listed in Table 2.

Genotyping of ACB Complex by gyrB PCR
Several isolates (17/200) of Category 1 tested negative for Ab-ITS PCR. Since gyrB PCR has been successfully applied to discriminate and identify members of ACB complex [9] [10] [18]. We first optimized the amplification conditions for the four members of ACB complex: they were A. calcoaceticus, A. pitti and A. nosocomialis in addition to A. baumannii and interestingly all the seventeen of them were identified as A. calcoaceticus (n = 2), A. pitti (n = 6) and A. nosocomialis Advances in Infectious Diseases   (n = 9). The results are documented in Figure 3. Four of the tracheal isolates were reported as ACB complex but when we screened them by Ab-ITS PCR, they did not contain the Ab-ITS target sequence (n = 4). These tracheal aspirates were processed as shown in scheme Figure 1 and subjected to gyrB-PCR with the primer sets for the above mentioned four species of ACB complex and analyzed by agarose gel electrophoresis to identify the species (Figure 4). Two of them were identified as A. nosocomialis (n = 2) and we could not confirm the Acinetobacter species of the other two samples.

Statistical Analysis of the Performance of PCR Assays
MedCalc software (MedCalc Statistical Software version 15.6.1, MedCalc Software,    Figure  5).
When we evaluated the performance of the genetic marker Ab-ITS against the reference method VITEK-2, 44(44/110) tracheal aspirates were positive by both PCR and VITEK-2, whereas 31 samples were "false positive" in PCR genotyping. Thus Ab-ITS PCR showed greater sensitivity (100%) and specificity (53.03%) compared to VITEK-2 for identification of Acinetobacter baumannii ( Figure 6).

Discussion
A. baumannii has emerged as one of the most notorious pathogens in health care facilities globally, targeting the critically ill hospitalized patients with its dreadful drug resistance profile. With an MDR rate of 92.1% and no new drug on the horizon, A. baumannii infections have become a real challenge for critical care and general physicians [7] [19] [20] [21]. Reporting time for identification of A. baumannii in clinical samples directly impacts the outcome of critical care. In this study we have focused on the identification of A. baumannii by evaluating a commonly used species-specific genetic marker and subsequently using a simple sample processing protocol for tracheal aspirates coupled to PCR for direct detection   remarkably well on these samples and was able to identify A. baumannii correctly. Ab-ITS seems to be a more reliable and sensitive genetic marker for identification of A. baumannii. Sequences of the ITS regions have been found to have low levels of intra-species variation and high levels of interspecies divergence [9] [10]. We used the primer sequences of Chen et al. (2007) and it was interesting to note that these primers proved absolutely specific to A. baumannii [10]. Including species specific gyrB primers along with Ab-ITS primers would be an ideal way to make differential diagnosis of ACB complex directly from uncultured clinical samples. Acinetobacter genus consists of about 33 species and six of them, grouped as ACB, are more frequently associated with human diseases.
They are A. baumannii, Acinetobacter 13TU and Acinetobacter genome species 3 were frequently found in several outbreaks [7]. Commercial systems like API 20NE and VITEK-2 are increasingly being used for identification of species but these systems fail to discriminate closely related species [22]. The six members of ACB complex are very closely related and difficult to differentiate by conventional phenotyping or as was shown in this study even by automated systems like VITEK 2 and are missed sometimes even by PCR. Individual members of ACB complex should be identified correctly from A. baumannii as their drug sensitivity profiles are different [5]. Ab-ITS PCR is a reliable, accurate method and can be applied directly on uncultured tracheal aspirates for rapid identification of A. baumannii. The sample processing protocol used in this study is derived from the original protocol published decades ago for processing a variety of clinical samples for the detection of M. tuberculosis and therefore, it should work well with any type of clinical specimen [19]. However, since the microbial population would be different in each type of specimen, our protocol is currently being validated on other types of clinical specimen. The combination of genetic markers used in this study, provides powerful information not only on identification of A. baumannii but also on probable outbreak potential and allows, to some extent, prediction of carbapenem resistance.
The superiority of our genotyping approach compared to VITEK-2 is evident from the sensitivity profile and helps significantly in the differential diagnosis of individual members of ACB complex. Further, our protocol can be applied to both clinical isolates and for direct detection from uncultured clinical samples.
This is certainly an advantage as results could be intimated to clinician the same day to make quick therapeutic decisions. Sixty-two tracheal aspirates were reported as sterile by VITEK-2 (mainstay in many advanced clinical microbiology laboratories) while twenty-five of them found to harbor Ab-ITS target sequence specific for A. baumannii. Availability of multiple copies of 16S-23S rRNA in bacterial cells ensures superior sensitivity of PCR compared to culture. It is possible that these patients apparently had A. baumannii infection and the antimicrobial therapy killed them. Such information from the "so called sterile sam- ples" will be very useful for transmission, antibiotic therapy and molecular epidemiology. Our method has not yet been validated on A. seifertii, A. lactucae which are also members of ACB complex and further we have demonstrated our method only on clinical isolates and tracheal aspirates. We are hopeful that the sample processing protocol would work equally well with all types of clinical specimen as it was originally reported [15]. The in-house sample processing and PCR methods for clinical specimen can be easily adopted for rapid detection of A. baumannii much before culture results are obtained. Implementation of this simple and economical protocol would help the clinician to make evidence based therapeutic decision on the same day.

Conclusion
This study showed that the 16S -23S rRNA intergenic spacer region possesses high discriminatory and better sensitivity to identify A. baumannii to the species level among ACB complex compared to conventional microbiology techniques.
In this study, a comparison between a well-established species-specific genetic marker, Ab-ITS and a commonly used microbial identification system VITEK 2, was made. A convenient and rapid sample processing method was evaluated, for identification of individual species of ACB complex and their drug resistance profiles. The results are available in about 6 -8 h, enabling the clinicians in infection control. Implementation of this protocol would help not only in control of hospital acquired infections but also improve the outcome of such infections.