Reactive Oxygen Species-Related Ceftazidime Resistance Is Caused by the Pyruvate Cycle Perturbation and Reverted by Fe3 + in Edwardsiella tarda

Abstract

Reactive oxygen species (ROS) are related to antibiotic resistance and have been reported in bacteria. However, whether ROS contribute to ceftazidime resistance and plays a role in ceftazidime-mediated killing is unknown. The present study showed lower ROS production in ceftazidime-resistant Edwardsiella tarda (LTB4-R _CAZ ) than that in LTB4-sensitive E. tarda (LTB4-S), two isogenic E. tarda LTB4 strains, which was related to bacterial viability in the presence of ceftazidime. Consistently, ROS promoter Fe³⁺ and inhibitor thiourea elevated and reduced the ceftazidime-mediated killing, respectively. Further investigation indicated that the reduction of ROS is related to inactivation of the pyruvate cycle, which provides sources for ROS biosynthesis, but not superoxide dismutase (SOD) and catalase (CAT), which degrade ROS. Interestingly, Fe³⁺ promoted the P cycle, increased ROS biosynthesis, and thereby promoted ceftazidime-mediated killing. The Fe³⁺-induced potentiation is generalizable to cephalosporins and clinically isolated multidrug-resistant pathogens. These results show that ROS play a role in bacterial resistance and sensitivity to ceftazidime. More importantly, the present study reveals a previously unknown mechanism that Fe³⁺ elevates ROS production via promoting the P cycle.

Keywords: Edwardsiella tarda; antibiotic resistance; ceftazidime; reactive oxygen species; the pyruvate cycle.

FIGURE 1

Ceftazidime resistance and reactive oxygen species (ROS) level in LTB4-R_CAZ. (A) Minimum inhibitory concentration (MIC) of LTB4-S and LTB4-R_CAZ. (B) Killing efficiency of ceftazidime to LTB4-S and LTB4-R_CAZ. (C) ROS of LTB4-S and LTB4-R_CAZ. Results (B,C) are displayed as mean ± SEM, as determined by two-tailed Student’s t-test. Four biological repeats are carried out. **P < 0.01.

FIGURE 2

The role of ROS in ceftazidime-mediated killing. (A) ROS was quantified in LTB4-R_CAZ in the absence or presence of ceftazidime plus Fe³⁺, H₂O₂, or/and thiourea as indicated by fluorescence. (B) Percent survival of LTB4-R_CAZ in the presence of ceftazidime and the indicated concentrations of Fe³⁺. (C) Percent survival of LTB4-R_CAZ in the presence or absence of Fe³⁺ plus the indicated concentrations of ceftazidime. (D) ROS level in the presence or absence of ROS promoter and inhibitor plus ceftazidime. Results are displayed as mean ± SEM, as determined by two-tailed Student’s t-test. Four biological repeats are carried out. * < 0.05 and ** < 0.01.

FIGURE 3

The P cycle, NADH, and ATP in LTB4-S and LTB4-R_CAZ. (A) Sketch diagram describing the effect of the P cycle on ROS. (B) Real-time quantitative PCR (qRT-PCR) for the expression of the P cycle-related genes in LTB4-R_CAZ compared to LTB4-S. (C) A global view for transcription level of the P cycle genes. Red and blue indicate upregulated and downregulated expression, respectively. (D) The activity of PDH, KGDH, SDH, and MDH in the P cycle. NADH (E), membrane potential (F), and ATP (G) concentrations in LTB4-S and LTB4-R_CAZ. The activity of superoxide dismutase (SOD) (H) and catalase (CAT) (I) in LTB4-S and LTB4-R_CAZ. Results (B,D–I) are displayed as mean ± SEM, as determined by two-tailed Student’s t-test. Four biological repeats are carried out. **P < 0.01.

FIGURE 4

Differential metabolomics of LTB4-R_CAZ in response to Fe³⁺. (A) Heat map showing differential abundance of metabolites. Yellow and blue indicate increase and decrease of metabolites relative to the median metabolite level of the control, respectively (see color scale). (B) Pathway enrichment of varied metabolites in LTB4-R_CAZ. (C) Integrative analysis of metabolites in significantly enriched pathways. Yellow and blue indicate increased and decreased metabolites, respectively. (D) PCA of LTB4-S and LTB4-R_CAZ. Each dot represents the technical replicate analysis of samples in the plot. (E) S-plot generated from OPLS-DA. Predictive component p [1] and correlation p (corr) [1] differentiate LTB4-R_CAZ from LTB4-S. Dot represents metabolites, and candidate biomarkers are highlighted in blue.

FIGURE 5

The effect of Fe³⁺ on the P cycle. (A) qRT-PCR for expression of the P cycle genes in the presence of Fe³⁺. (B) A global view for transcription level of the P cycle genes. Red and blue indicate upregulated and downregulated expression, respectively. The activity of enzymes of the P cycle (C), NADH level (D), membrane potential (E), and ATP (F) of LTB4-R_CAZ in the presence or absence of Fe³⁺. (G) The activity of SOD and CAT in the presence or absence of ceftazidime or/and Fe³⁺. Results (A,C–G) are displayed as mean ± SEM, as determined by two-tailed Student’s t-test. Four biological repeats are carried out. *p < 0.05 and **p < 0.01.

FIGURE 6

Promotion of Fe³⁺ to the P cycle. (A) ROS of LTB4-R_CAZ in the presence of Fe³⁺ and CAZ plus inhibitors of the P cycle. (B) Percent survival of LTB4-R_CAZ in the presence of Fe³⁺ and CAZ plus inhibitors of the P cycle. (C) The activity of PDH, KGDH, SDH, and MDH in the presence of Fe³⁺. Results are displayed as mean ± SEM, as determined by two-tailed Student’s t-test. Four biological repeats are carried out. *p < 0.05 and **p < 0.01.

FIGURE 7

Percent survival of clinically isolated pathogens in the presence of Fe³⁺ and ceftazidime, cefoperazone, or cefazolin. (A) E. tarda. (B) E. coli. (C) K. pneumoniae. (D) P. aeruginosa.

FIGURE 8

Sketch diagram describing a mechanism by which Fe³⁺ promotes ROS level.