Population structure of clinical Pseudomonas aeruginosa from West and Central African countries

Abstract

Background: Pseudomonas aeruginosa (PA) has a non-clonal, epidemic population with a few widely distributed and frequently encountered sequence types (STs) called 'high-risk clusters'. Clinical P. aeruginosa (clinPA) has been studied in all inhabited continents excepted in Africa, where a very few isolates have been analyzed. Here, we characterized a collection of clinPA isolates from four countries of West and Central Africa.

Methodology: 184 non-redundant isolates of clinPA from hospitals of Senegal, Ivory Coast, Nigeria, and Central African Republic were genotyped by MLST. We assessed their resistance level to antibiotics by agar diffusion and identified the extended-spectrum β-lactamases (ESBLs) and metallo-β-lactamases (MBLs) by sequencing. The population structure of the species was determined by a nucleotide-based analysis of the entire PA MLST database and further localized on the phylogenetic tree (i) the sequence types (STs) of the present collection, (ii) the STs by continents, (iii) ESBL- and MBL-producing STs from the MLST database.

Principal findings: We found 80 distinct STs, of which 24 had no relationship with any known STs. 'High-risk' international clonal complexes (CC155, CC244, CC235) were frequently found in West and Central Africa. The five VIM-2-producing isolates belonged to CC233 and CC244. GES-1 and GES-9 enzymes were produced by one CC235 and one ST1469 isolate, respectively. We showed the spread of 'high-risk' international clonal complexes, often described as multidrug-resistant on other continents, with a fully susceptible phenotype. The MBL- and ESBL-producing STs were scattered throughout the phylogenetic tree and our data suggest a poor association between a continent and a specific phylogroup.

Conclusions: ESBL- and MBL-encoding genes are borne by both successful international clonal complexes and distinct local STs in clinPA of West and Central Africa. Furthermore, our data suggest that the spread of a ST could be either due to its antibiotic resistance or to features independent from the resistance to antibiotics.

Figure 1. Map of Africa showing the four cities included in the study, the number of Pseudomonas aeruginosa clinical isolates, and the localization of ESBL- and MBL-producing isolates.

Figure 2. Distribution of the STs and CCs of the 184 clinical isolates of Pseudomonas aeruginosa isolated in West and Central Africa on a dendrogram built with the data all known STs (n = 1595).

STs without any isolates producing ESBL or MBL were represented with red spots. STs represented by ≥1 isolate producing ESBL or MBL were represented with green spots. STs of reference isolates are represented with blue diamonds. The dendrogram is based on the alignment of the concatenated sequences of the acsA, aroE, guaA, mutL, nuoD, ppsA and trpE genes (forming an artificial 2,882-bp sequence) of the 1595 STs of the P. aeruginosa MLST database in July 2013 . See Material and methods section for details.

Figure 3. Distribution of the STs and CCs of the Pseudomonas aeruginosa containing isolates producing either extended-spectrum β-lactamase or metallo-β-lactamase defined in the following references , , , on a dendrogram built with the data all known STs (n = 1595).

ESBL- and MBL-producers were represented with red spots. See Figure 2 legend for the tree construction details. Pink zones surround STs that belong to the same clonal complex.

Figure 4. Distribution of the STs of clinical isolates of Pseudomonas aeruginosa on a dendrogram built with the data all known STs (n = 1595), broken down by their continent of origin.

The geographical origin of each ST was extracted from the P. aeruginosa MLST database in July 2013 .