In comparison with many bacterial pathogens, the global population of
Achromobacter has been poorly characterised. Several studies have reported the variation of A. xylosoxidans isolates and suggested that bacteria could spread out in various environments (Amoureux et al., 2012, 2013). This study focused on applied multiple typing approaches to help discriminate A. xylosoxidans from other Achromobacter and to investigate the clustering of A. xylosoxidans from different geographical background.
The combination of DNA fingerprint typing and multilocus gene typing was applied to investigate species segregation and the epidemiology of A. xylosoxidans. In this study, RAPD fingerprinting and MLST were selected as strain typing methods because these methods are generally used in this species (Kaur et al., 2009; Magni et al., 2010; Spilker, Vandamme & LiPuma, 2012; Vandamme et al., 2013; Trancassini et al., 2014). Both RAPD fingerprinting and MLST analysis illustrated the ability to differentiate between A. xylosoxidans and non-xylosoxidans Achromobacter species, except that L1 was grouped with non-
xylosoxidans by RAPD (Figure 3.5). Considering only A. xylosoxidans, both typing methods could not group A. xylosoxidans isolates by their countries of origin (Figure 3.5 and Figure 3.6).
By observing epidemiology of the isolates, the presence of similar genotypes of strains in different patients was observed in this study. Nevertheless, both the
RAPD typing and MLST demonstrated multiple genotypes of A. xylosoxidans, especially Thai strains (Figure 3.4). Both RAPD and MLST demonstrated multiple strain types of clinical isolates and the strains from both British and Thailand were not separated by geography.
As shown in Figure 3.5 and Figure 3.6, there were two pairs of genetically similar strains; L11-L15 and L8-L17. For Thai strains, clonal strains were unsurprisingly found in strain R1, R5 and R14. British strain L11 and L15 were obtained from the same patient (two months apart), who had bronchiectasis, a disease with abnormally widened lung bronchi. Thai strain R1, R5 and R14 were also collected from the same patient (two months apart) who had pneumonia. These evidences represented long-termed colonisation of A. xylosoxidans, at least, over two months in a single patient with respiratory tract problems during intubation (Table 2.1). Interestingly, L8 and L17 were isolated from a non-Cystic Fibrosis patient with chronic obstructive pulmonary disease and a Cystic Fibrosis patient, respectively. It is still argued that how those patients be infected by similar strain type. It can be assumed that these patients contacted with the same environment so that they were infected with genetically-related strain of A. xylosoxidans.
The diversity of A. xylosoxidans isolates demonstrates that the majority of infected patients acquire the pathogens from the environment. There is a possibility of persistent colonisation in patients with respiratory diseases (Kanellopoulou et al., 2004). Furthermore, these evidences suggest no geographic specificity of A. xylosoxidans with respect to the distribution of clonal types. However, larger scale studies with samples from different clinical, environmental and geographical sources are needed to better understand and characterise the global population of this pathogen. This chapter demonstrated the determination of methods for identification of A. xylosoxidans. Molecular- based approaches, including 16S rDNA and MLST, showed relationships between isolates. However, this did not explain entire genetic relationship between these isolates. This emphasised that the construction of whole genomic relationship was required.
3.5. Conclusion and future work
The main purposes of this study were to demonstrate the ability of multiple bacterial identification approaches to identify emerging pathogens and to apply strain typing of clinical isolates of A. xylosoxidans in order to investigate geographical associations. Due to the fact that A. xylosoxidans is one of the most important emerging pathogens that cause severe diseases in Cystic Fibrosis patients and other immune-compromised patients, precise identification is required for epidemiological study and clinical work-up in order to deliver an appropriate treatment to the patients.
Figure 3.7: The proposed algorithm for A. xylosoxidans identification from non-fermenting Gram-negative. Dashed boxes indicate approaches that are affected by the lack of advanced facilities such as MALDI-TOF MS and sequencing facility.
Non-fermenting Gram-negative bacteria
Biochemical reaction test
MALDI-TOF MS Supportive approach Genus Achromobacter Achromobacter xylosoxidans 16S rDNA sequencing Diversity assessment RAPD MLST
With respect to reference labs, the identifications in the reference labs are merely based on biochemical (conventional) tests. MALDI-TOF MS is used in particular samples, such as samples that cannot be clearly identified by means of conventional methods. A number of identification methods were compared in this study. On the basis of the results presented here, a pragmatic algorithm for A. xylosoxidans identification is suggested (Figure 3.7).
Starting with non-fermenting Gram-negative bacteria from MacConkey agar, biochemical reaction-based methods, such as conventional tests and API20 NE can be used as screening tests due to their good performance in genus identification. For species identification, MALDI-TOF MS can be a main method to assign species to the isolates. The gene sequencing of 16S rDNA can be a supportive information for species identification. Finally, the intra-species discrimination can be conducted by additional methods, including MLST analysis and RAPD, which offer the potential to analyse within-species diversity.
Theorically, utilising a single approach that can give a highly reliable result is the best practice in clinical and diagnostic microbiology but the ideal approach is still unavailable. Therefore, using multiple approaches is necessary for current situation because results from several methods can reciprocally support one another to define a final result. However, some identification approachs, such as MALDI-TOF and DNA sequencing, cannot be used in a routine identification practice because some developing countries, such as Thailand, cannot afford those facilities for a routine purpose (They can afford for research purpose). Therefore, phenotypic identification remains a routine diagnostic practice in developing countries as the method still perform well in the identification of common pathogens, such as S. aereus, E. coli, and S. pneumoniae. Although the time-to-result of the phenotypic is approximately, at least, 48 hours, compared to its of MALDI-TOF which takes around 36 hours, the treatment is not affected because the treatment for infections, mostly, relies on antibiotic susceptibility of the pathogens.
However, there is still in need for a better and more robust identification method for the healthcare unit where MALDI-TOF MS and sequencing facility are
unavailable because rapid and accurate diagnostic provides the opportunity to establish empirical treatment, especially for particular pathogens having intrinsic resistance. Likewise, correct species identification is important for microbiological research and epidemiology. Whole genome sequencing of all
Achromobacter isolates will be conducted to achieve species identification and to pursue futher study.
Chapter 4
Pan-genome analysis of Achromobacter xylosoxidans
4.1. Introduction
The previous analysis in Chapter 3 suggests that phylogenetic trees built using the 16S rDNA gene (Figure 3.4) and MLST genes (Figure 3.7) allow for
Achromobacter species identification. However, sub-species level identification has not been clearly elucidated by those analyses. Further analysis of whole genome sequence data would enable a higher resolution of core phylogenomic analysis to be carried out, define the genome structure and pan-genome of A. xylosoxidans, and allow us to elucidate the molecular relationships between isolates.