Combination of gallium citrate and levofloxacin induces a distinct metabolome profile and enhances growth inhibition of multidrug-resistant Mycobacterium tuberculosis compared to linezolid

Abstract

Introduction: Tuberculosis (TB) treatment typically involves a tailored combination of four antibiotics based on the drug resistance profile of the infecting strain. The increasing drug resistance of Mycobacterium tuberculosis (Mtb) requires the development of novel antibiotics to ensure effective treatment regimens. Gallium (Ga) is being explored as a repurposed drug against TB due to its ability to inhibit Mtb growth and disrupt iron metabolism. Given the potential interactions between Ga and established antibiotics, we investigated how a combination of Ga with levofloxacin (Lfx) or linezolid (Lzd) affects the growth and metabolome of a multidrug-resistant (MDR) Mtb clinical strain.

Methods: Mtb was cultured using a BACTEC 960 system with concentrations of Ga ranging from 125 to 1,000 μM and with 250 to 500 μM of Ga combined with 0.125 mg/L of Lfx or Lzd. For metabolome analysis, the antibacterials were used at concentrations that inhibited the growth of bacteria without causing cell death. Metabolites were extracted from Mtb cells and analyzed using chromatography-mass spectrometry.

Results: The MDR Mtb strain exhibited a dose-dependent response to Ga. Notably, the enhancement in growth inhibition was statistically significant for the Ga/Lfx combination compared to Ga alone, while no such significance was observed for Ga/Lzd. Moreover, exposure to Ga/Lfx or Ga/Lzd resulted in distinct metabolite profiles. Ga treatment increased the level of aconitate, fumarate, and glucose in the cells, suggesting the inhibition of iron-dependent aconitase and fumarate hydratase, as well as disruption of the pentose phosphate pathway. The levels of glucose, succinic acid, citric acid, and hexadecanoic acid followed a similar pattern in cells exposed to Ga and Ga/Lfx at 500 μM Ga but exhibited different trends at 250 μM Ga.

Discussion: In the presence of Lfx, the Mtb metabolome changes induced by Ga are more pronounced compared to those observed with Lzd. Lfx affects nucleic acids and transcription, which may enhance Ga-dependent growth inhibition by preventing the metabolic redirection that bacteria typically use to bypass iron-dependent enzymes.

Keywords: Mycobacterium tuberculosis; central metabolism; drug resistance; drug–drug interaction; gallium; levofloxacin; linezolid; metabolome.

Figure 1

Influence of Ga and its combination with Lfx or Lzd on Mtb growth and metabolome; (A) growth inhibition of Mtb MDR strain in MGIT™ medium with Ga, Lfx or Lzd, presented as percent inhibition of growth in standard MGIT™ medium; experimental conditions labeled as follows: MGIT™ medium with citrate (cit); Ga citrate (Ga); Lfx with citrate (Lfx), Ga citrate with Lfx (Ga/Lfx); Ga citrate with Lzd (Ga/Lzd), Lzd with citrate (Lzd); Lfx with Lzd and citrate (Lfx/Lzd). Ga was used at 250 and 500 μM. Citrate was used at 1.25 and 2.5 mM (as controls to Ga 250 and 500 μM), while Lfx or/and Lzd were used at 0,125 mg/L. Statistical significance after false discovery rate correction: ns >0.05, * p ≤ 0.05, ** p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001; (B) scores plot of an unsupervised PCA model summarizing all GC–MS metabolomics data; R2X [1] = 0.482; R2X [2] = 0.21; Ellipse: Hotelling’s T2 (95%); the model shows that samples (metabolomes of Mtb grown at different experimental conditions) cluster according to two concentrations of Ga and/or citrate; Ga was used at 250 and 500 μM, citrate was used at 1.25 and 2.5 mM (as controls to Ga 250 and 500 μM), while Lfx or/and Lzd were used at 0,125 mg/L; the experimental conditions are labeled as in A with the addition of standard MGIT™ medium (MGIT); color code of B corresponds to A; the loadings plot of this model showing metabolites responsible for clustering is provided as a Supplementary Figure 2; (C) scores plot of a supervised OPLS-DA model built with GC–MS metabolomics data; the model illustrates differences between three experimental conditions at two concentrations of Ga: Ga (Ga); Ga with Lfx (Ga/Lfx); Ga with Lzd (Ga/Lzd); the model is scaled proportionally to R2X; R2X [1] = 0,289; R2X [2] = 0,241; Ellipse: Hotelling’s T2 (95%); CV-ANOVA p = 0.0001; the model identifies two concentration-dependent subclusters in samples with Ga/Lfx: pink – gallium 250 and 500 μM; green-blue – Ga/Lfx 250 and 500 μM; green-yellow – Ga/Lzd 250 and 500 μM; the loadings plot of this model showing metabolites responsible for clustering is provided as a Supplementary Figure 4.

Figure 2

Influence of Ga 500 μM and its combination with Lfx or Lzd on the metabolome of Mtb; (A) scores plot of a supervised OPLS-DA model, scaled proportionally to R2X; R2X [1] = 0.347; R2X [2] = 0.0845; Ellipse: Hotelling’s T2 (95%); CV-ANOVA p = 0,001; the model shows the separation of samples (Mtb metabolomes) based on experimental conditions: pink – Ga 500 μM; green-blue – Ga/Lfx 500 μM; green-yellow – Ga/Lzd 500 μM; the model identifies Ga/Lzd as a group distinct from Ga/Lfx and Ga alone; Ga/Lfx and Ga alone show overlapping metabolite profile; (B) the loadings plot of this model showing metabolites responsible for clustering; colored areas on the plot are used for visualization and matching the scores plot groups to the loadings plot.

Figure 3

Two patterns of metabolite expression in the presence of Ga/Lfx 250 and 500 μM compared to other experimental conditions; (A) metabolites that increase their level in Ga/Lfx 500 μM compared to Ga/Lfx 250 μM; (B) metabolites that decrease their level in Ga/Lfx 500 μM compared to Ga/Lfx 250 μM; both in A and B panel, the peak areas of metabolites measured in cells exposed to Ga/Lfx 500 μM follow the levels of peak areas of metabolites measured in cells exposed to Ga 500 μM.

Figure 4

Effect of Ga (250 and 500 μM) and its combination with Lfx or Lzd on the Mtb metabolome analyzed by GC–MS Volcano plot analysis identifies the most significant metabolites (marked with annotations) for every growth condition based on their p-value and fold change (FC); (A) metabolome of cells exposed to Ga compared to Ga/Lfx 250 μM; (B) metabolome of cells exposed to Ga compared to Ga/Lfx 500 μM; (C) metabolome of cells exposed to Ga compared to Ga/Lzd 250 μM; (D) metabolome of cells exposed to Ga compared to Ga/Lzd 500 μM pink – Ga 250 and 500 μM; green-blue – Ga/Lfx 250 and 500 μM; green-yellow – Ga/Lzd 250 and 500 μM; detailed information on all the metabolites used in the volcano plot analysis with FC and p values can be found in Supplementary Table 2.

Figure 5

Effect of Ga and Lfx on the Mtb metabolome analyzed by LC–MS Volcano plot analysis identifies the most significant metabolites (marked with annotations) for every growth condition based on their p-value and fold change (FC); (A) metabolome of cells exposed to Ga compared to citrate control; (B) metabolome of cells exposed to Ga compared to Lfx; (C) metabolome of cells exposed to citrate compared to Lfx; pink – gallium; green – Ga/Lfx; purple – citrate.