Quinolone Resistance of Actinobacillus pleuropneumoniae Revealed through Genome and Transcriptome Analyses

Abstract

Actinobacillus pleuropneumoniae is a pathogen that infects pigs and poses a serious threat to the pig industry. The emergence of quinolone-resistant strains of A.pleuropneumoniae further limits the choice of treatment. However, the mechanisms behind quinolone resistance in A.pleuropneumoniae remain unclear. The genomes of a ciprofloxacin-resistant strain, A. pleuropneumoniae SC1810 and its isogenic drug-sensitive counterpart were sequenced and analyzed using various bioinformatics tools, revealing 559 differentially expressed genes. The biological membrane, plasmid-mediated quinolone resistance genes and quinolone resistance-determining region were detected. Upregulated expression of efflux pump genes led to ciprofloxacin resistance. The expression of two porins, OmpP2B and LamB, was significantly downregulated in the mutant. Three nonsynonymous mutations in the mutant strain disrupted the water-metal ion bridge, subsequently reducing the affinity of the quinolone-enzyme complex for metal ions and leading to cross-resistance to multiple quinolones. The mechanism of quinolone resistance in A. pleuropneumoniae may involve inhibition of expression of the outer membrane protein genes ompP2B and lamB to decrease drug influx, overexpression of AcrB in the efflux pump to enhance its drug-pumping ability, and mutation in the quinolone resistance-determining region to weaken the binding of the remaining drugs. These findings will provide new potential targets for treatment.

Keywords: Actinobacillus pleuropneumoniae; efflux pump; porin; quinolone resistance; quinolone resistance-determining region; transcriptome.

Figure 1

Chromosome genome and plasmid of APP SC1810. (a) Chromosome circle map (b) Plasmid circle map. From the inside out, the first circle represents the scale; second circle represents the GCSkew; third circle represents the GC content; fourth and seventh circles represent the fifth and sixth circles of COG; to which each coding sequence belongs, representing the position of coding sequence, tRNA and rRNA on the genome.

Figure 2

RNA-sequencing analysis. (a) GO enriched analysis. Number on the bar chart indicates the number of genes enriched under the term (b) and KEGG pathway enriched analysis for up-regulated and down-regulated genes.

Figure 3

Antibiofilm activity of CPF on SC1810. (a) Biofilm biomass of wild-type and mutant strain (* p < 0.05). (b) Effects of CPF on biofilm formation (12 h with drug, n = 3) and (c) eradication at different concentrations (12 h without drug, 12 h with drug, n = 3), respectively. The data are shown as the absorbance at 570 nm (A570 nm) of residual biofilm and associated error bars denote the standard error of the mean. (d) Analysis of differentially expressed genes (DEGs) related to pili formation between wild-type strains and mutants. (e) Relative transcription levels of apfA, apfC and comEA determined using qRT-PCR. Data are expressed as the mean and standard deviation (SD) of 3 experiments (* p < 0.05).

Figure 4

Analysis of differential expression of porins. (a) Volcano plots marked with DEGs of porins. (b) Each row represents the relative expression of single transcript, and each column represents a sample. Colors represent the log₂-transformed gene expression level, with red and blue representing high and low expression levels, respectively. (c) Molecular phylogenetic analysis by maximum likelihood method and motif prediction. (i) The evolutionary history was inferred by using the maximum likelihood method based on the Tamura 3-parameter model. The percentage of trees in which the associated taxa clustered together is shown next to the branches. (ii) Motifs prediction. The same color or number represents the same conservative site. (d) Relative transcription levels of ompP2A, ompP2B and lamB. Data are expressed as mean and standard deviation (SD) of 3 experiments (* p < 0.05).

Figure 5

Analysis of differential expression genes of efflux pump (a–c) and prediction of crystal structure of CPF-bound AcrB (d–g). (a) Volcano plots marked with DEGs of pumps. (b) Each row represents the relative expression of single transcript, and each column represents a sample. Colors represent the log₂-transformed gene expression level, with red and blue representing high and low expression levels, respectively. (c) Relative transcription levels of acrB. The data are expressed as the mean and standard deviation (SD) of 3 experiments (* p < 0.05). (d) Whole trimer structure of AcrB (ribbon model) bound with CPF (electron density map). (e) Close-up view of the CPF-binding site. Electron density of CPF (orange mesh) overlapped with stick models of LMNG (grey). (f) View of the bound CPF (grey), shown in the cut view of the surface model of the distal pocket in the binding monomer. The red indicates higher electron cloud density (gain electrons), whereas blue represents lower electron cloud density (electron loss). (g) 2D representation of the interaction between CPF and AcrB was drawn using LigPlot+ [31].

Figure 6

Structural comparison of GyrA S83F/D87N and ParC E89K mutation between wild and mutant strains. (a) Quinolone and gyrase binding in the wild-type strain is facilitated by a water-metal ion bridge, which forms hydrogen bonds between the water molecules and Ser 83/Asp87 amino acid. (b) Hydrogen bonds are missing in the mutant APP strain, decreasing the binding affinity of quinolone. (c) Quinolone and topoisomerase IV binding in the wild-type strain occurs through a water-metal ion bridge, where forms a hydrogen bond between the quinolone and Ser 85/Glu89 amino acid that act as anchor points to enzyme. (d) Hydrogen bond link with 89 amino acid is missing, weakening the binding affinity of quinolone.

Figure 7

Mechanism of quinolone resistance of A. pleuropneumoniae. (a) Expression of major porins ompP2B and LamB was decreased, which reduced the drug concentration entering the periplasm. (b) Overexpression of multidrug resistance efflux pump reduced the intracellular drug concentraTable 1. RNA sequencing analysis. (a) Pearson correlation of all sample. (b) Heatmap showing gene expression patterns. (c) Volcano plots shows the difference of expression level between the wild and mutant strains.