Multi-Drug Resistance in Bacterial Genomes-A Comprehensive Bioinformatic Analysis

Abstract

Antimicrobial resistance is presently one of the greatest threats to public health. The excessive and indiscriminate use of antibiotics imposes a continuous selective pressure that triggers the emergence of multi-drug resistance. We performed a large-scale analysis of closed bacterial genomes to identify multi-drug resistance considering the ResFinder antimicrobial classes. We found that more than 95% of the genomes harbor genes associated with resistance to disinfectants, glycopeptides, macrolides, and tetracyclines. On average, each genome encodes resistance to more than nine different classes of antimicrobial drugs. We found higher-than-expected co-occurrences of resistance genes in both plasmids and chromosomes for several classes of antibiotic resistance, including classes categorized as critical according to the World Health Organization (WHO). As a result of antibiotic-resistant priority pathogens, higher-than-expected co-occurrences appear in plasmids, increasing the potential for resistance dissemination. For the first time, co-occurrences of antibiotic resistance have been investigated for priority pathogens as defined by the WHO. For critically important pathogens, co-occurrences appear in plasmids, not in chromosomes, suggesting that the resistances may be epidemic and probably recent. These results hint at the need for new approaches to treating infections caused by critically important bacteria.

Keywords: antibiotic resistance; co-selection; evolution; genomics; multi-drug resistance; plasmid.

Figure 1

Proteins associated with different numbers of antimicrobial classes. Bars represent the number of proteins associated with x resistance classes. The x-axis indicates the number of resistance classes, and the y-axis indicates the number of proteins.

Figure 2

Number of antimicrobial classes per genome. Blue bars represent counts considering scenario 1 (specific proteins) and red bars represent counts considering scenario 2 (all proteins). The x-axis indicates the number of antimicrobial classes, and the y-axis indicates the number of genomes.

Figure 3

Co-occurrence of antimicrobial classes in chromosomes and plasmids. The panels correspond to: (a) scenario 1 (including proteins conferring resistance to one class of antibiotics) in chromosome; (b) scenario 2 (including proteins conferring resistance to all classes of antibiotics) in chromosome; (c) scenario 1 in plasmid; and (d) scenario 2 in plasmid. Red and blue circles represent cases where the observed co-occurrence is higher and lower than expected, respectively (binomial test, p < 0.001). Darker circles represent observed/expected values further away from zero. Gray circles represent cases where co-occurrence is not significantly different from the expected value.

Figure 4

Co-occurrences of antibiotic resistance classes in antibiotic-resistant priority pathogens, considering specific proteins (including proteins conferring resistance to one class of antibiotics). Priorities refer to the categorization of priorities according to the WHO (critical, high, and medium). The panels correspond to: (a) priority 1 on chromosomes; (b) priority 1 on plasmids; (c) priority 2 on chromosomes; (d) priority 2 on plasmids; (e) priority 3 on chromosomes; and (f) priority 3 on plasmids. Red and blue circles represent cases where the observed co-occurrence is higher and lower than expected, respectively (binomial test, p < 0.001). Darker circles represent observed/expected values farther from zero. Gray circles represent cases where co-occurrence is not significantly different from the expected value.

Figure 4

Co-occurrences of antibiotic resistance classes in antibiotic-resistant priority pathogens, considering specific proteins (including proteins conferring resistance to one class of antibiotics). Priorities refer to the categorization of priorities according to the WHO (critical, high, and medium). The panels correspond to: (a) priority 1 on chromosomes; (b) priority 1 on plasmids; (c) priority 2 on chromosomes; (d) priority 2 on plasmids; (e) priority 3 on chromosomes; and (f) priority 3 on plasmids. Red and blue circles represent cases where the observed co-occurrence is higher and lower than expected, respectively (binomial test, p < 0.001). Darker circles represent observed/expected values farther from zero. Gray circles represent cases where co-occurrence is not significantly different from the expected value.