Report

Subcellular antibiotic visualization reveals a dynamic drug reservoir in infected macrophages

See allHide authors and affiliations

Science  28 Jun 2019:
Vol. 364, Issue 6447, pp. 1279-1282
DOI: 10.1126/science.aat9689

Lipid droplets help anti-TB drug efficacy

Improving chemotherapies against intracellular pathogens requires understanding how antibiotic distribution within infected cells affects efficacy. Greenwood et al. developed an approach to visualize antibiotics in human macrophages infected with the tubercle bacillus (see the Perspective by Smith and Aldridge). They showed that the antitubercular (anti-TB) drug bedaquiline accumulated in host lipid droplets. Lipid droplets seemed to act as an antibiotic reservoir that could be transferred to bacteria during host lipid consumption. Indeed, alterations in host lipid droplet content affected the anti-TB activity of bedaquiline against intracellular bacilli.

Science, this issue p. 1279; see also p. 1234

Abstract

Tuberculosis, caused by the intracellular pathogen Mycobacterium tuberculosis, remains the world’s deadliest infectious disease. Sterilizing chemotherapy requires at least 6 months of multidrug therapy. Difficulty visualizing the subcellular localization of antibiotics in infected host cells means that it is unclear whether antibiotics penetrate all mycobacteria-containing compartments in the cell. Here, we combined correlated light, electron, and ion microscopy to image the distribution of bedaquiline in infected human macrophages at submicrometer resolution. Bedaquiline accumulated primarily in host cell lipid droplets, but heterogeneously in mycobacteria within a variety of intracellular compartments. Furthermore, lipid droplets did not sequester antibiotic but constituted a transferable reservoir that enhanced antibacterial efficacy. Thus, strong lipid binding facilitated drug trafficking by host organelles to an intracellular target during antimicrobial treatment.

Mycobacterium tuberculosis (Mtb) can persist in multiple intracellular niches within human macrophages (1). Given that total host cell accumulation does not necessarily correlate with antibiotic efficacy against intracellular pathogens, we hypothesized that efficacy is modulated by partitioning of compounds within the host cell (2). The antitubercular antibiotic bedaquiline (BDQ) is highly lipophilic (3, 4), a trait associated with permeability through tissue at the expense of nonspecific binding and host sequestration (57). We tested whether efficacy is linked to the accumulation of the drug in intracellular compartments.

To characterize the intracellular distribution of BDQ, we infected human monocyte-derived macrophages (hMDMs) with Mtb for 48 hours, giving bacteria time to enter multiple subcellular compartments (1). Macrophages were treated with 2.5 mg/liter BDQ for 24 hours, fixed, and imaged by correlative electron microscopy (EM) and ion microscopy (IM) (8) (fig. S1A). Mtb exhibited a strong 31P signal, likely corresponding to bacterial DNA or polyphosphates. Uninfected macrophages did not have 31P foci outside of the nucleus (fig. S1B).

Because BDQ contains a bromine atom, we could determine its localization by measuring the intensity of the 79Br signal (Fig. 1, A to C). Control macrophages treated only with the solvent carrier did not contain 79Br, confirming that the signal was specific (fig. S1C). Some of the 79Br signal is derived from primary metabolites of BDQ, which are also active against Mtb (9). BDQ accumulated heterogeneously in Mtb within macrophages, even between neighboring bacteria (Fig. 1A). BDQ was found in Mtb in a variety of intracellular environments, including a membranous vacuole and a lysing necrotic macrophage (Fig. 1, A and B), a known mycobacterial niche (1).

Fig. 1 BDQ accumulates in host LDs and Mtb.

(A to C) Mtb-infected hMDM treated with 2.5 mg/liter BDQ. The EM image is overlaid with 79Br and 31P signals. Scale bar, 2 μm. (D) (Left) Maximum projection of macrophages infected with Mtb-RFP (red fluorescent protein) and treated with 2.5 mg/liter BDQ. LDs stained with BODIPY appear yellow because of spectral overlap. (Right) LD staining and 79Br signal on EM. Scale bar, 5 μm. (E) BDQ in macrophages treated with 2.5 mg/liter BDQ. Data are shown as means from four to six technical replicates from two to three donors. (F) EM with 79Br signal on mitochondria (*). Scale bar, 2 μm. (G) Normalized 79Br (BDQ) signal by area. Data are shown as mean intensity per object with three biological replicates ± SE. P values are from Wilcoxon test (n = 6 to 221 objects). Values on the y axis are shown on a square-root scale.

We also observed Mtb interacting with host lipid droplets (LDs), as previously reported (10), and found the LDs to be highly enriched with antibiotic (Fig. 1C). No characteristic ion signal exists for LDs, so to confirm the organelle identity, we stained an infected sample with the neutral lipid dye BODIPY 493/503 and imaged it with correlative light, electron, and ion microscopy (CLEIM, Fig. 1D). Live-cell imaging before fixation showed that the bacteria were intracellular before antibiotic treatment, confirming that they absorbed BDQ from within the host cell (movie S1). Liquid chromatography–mass spectrometry (LC-MS) quantification of BDQ from unfixed macrophages treated with pradigastat (11), an inhibitor of diacyglycerol O-acyltransferase 1, to inhibit LD formation or with oleate to induce it confirmed LD as the primary reservoir of intracellular BDQ (Fig. 1E). This validated the CLEIM protocol in preserving the antibiotic distribution during processing.

To determine the relative abundance of BDQ in cellular structures, we normalized the 79Br signal to the cytosolic 12C14N signal (corresponding to protein content) to provide relative quantifications of BDQ enrichment per bacterium. A weak BDQ signal was also detected from other organelles, particularly mitochondria, reflecting reports that BDQ inhibits mammalian adenosine triphosphate (ATP) synthase (12) (Fig. 1, F and G). A similar distribution of BDQ was observed at a much lower concentration of antibiotic (0.04 mg/liter); however, at this level, the 79Br signal was only marginally detectable (fig. S1D).

Lipid-laden foamy macrophages are a hallmark of tuberculosis pathogenesis (13). Because LDs were the primary site of BDQ accumulation in the host, we investigated the interactions between LDs and Mtb in human macrophages. As expected, Mtb exposure induced host LD proliferation (10, 14, 15), even in uninfected “bystander” macrophages (Fig. 2A and fig. S2A). LC-MS analysis found that 123 triacylglycerides (TAGs), together with two cholesterol esters and four ceramides, were more abundant in infected macrophages. This included many TAGs containing odd-chain fatty acids, which are associated with Mtb virulence (16) (Fig. 2B and fig. S3). Confocal microscopy revealed extreme heterogeneity in both bacterial and LD burdens (Fig. 2C). Because intracellular mycobacteria consume host LDs as a carbon source (17, 18), we investigated the temporal progression of LD induction and consumption. Live-cell imaging showed that, as Mtb grows intracellularly, LD proliferation outweighs consumption for the first ~48 hours of infection, before consumption eventually reduces the number of LDs (Fig. 2D; fig. S2, D and E; and movies S2 and S3).

Fig. 2 Mtb induces host LD accumulation and consumption in hMDM.

(A) LD area per macrophage, with dots showing means. (B) Fold changes of significantly (P < 0.05) altered lipids relative to uninfected controls (n = 6 technical replicates). (C) Mtb versus LD area 96 hours after infection. Values on the axes are shown on a square-root scale. Scale bar, 10 μm. (D) Mean LD and Mtb burden measured by area, averaged from 38 infected macrophages. LD normalized to uninfected average with a starting point of 1. Curve LOESS was fitted with a 95% confidence interval. (E) Mtb-infected, hMDM-stained antibody against PLIN2. Nuclear stain: DAPI. (F) Maximum projection from live-cell video of Mtb-infected macrophages with LD diameter shown on the right.

Transmission EM (TEM) showed extensive physical contacts between LDs and Mtb or tight Mtb-containing vacuoles (fig. S2C). We also found that the LD surface protein Perilipin 2 (PLIN2), known to associate with intracellular M. marinum (19), labeled 3.8% of Mtb 96 hours after infection (95% confidence interval 1.86 to 6.59, n = 4 donors) (Fig. 2E). These variable levels of interaction may have contributed to the heterogeneity of BDQ accumulation. Mtb secretes lipases (20, 21) and we hypothesized that proximity between LDs and Mtb caused degradation of LDs. Indeed, some LDs in contact with Mtb shrank over time, whereas LDs in other parts of the cell remained the same diameter (Fig. 2F and movie S4).

We next investigated whether LDs sequestered or transferred BDQ to Mtb. We preloaded macrophages with antibiotic and then infected them with Mtb. The distribution of BDQ was indistinguishable from that of samples treated after infection (Fig. 3, A and B), thus BDQ can be transferred from a host reservoir to Mtb. Cotreatment with pradigastat, which reduces LD levels, significantly reduced BDQ abundance in Mtb and inhibited preloading (fig. S2B and Fig. 3, A and B). Therefore, LDs accumulate a BDQ pool that can be transferred to bacteria, although they are not absolutely required for BDQ access to Mtb. The induction of LD formation by the addition of the fatty acid oleate did not prevent BDQ transfer to Mtb (Fig. 3, A and B). LD may thus act more as an accessible reservoir than as a sequestrator of BDQ.

Fig. 3 BDQ is transferred from host LDs to Mtb.

(A) (Top-left) hMDM was treated with 2.5 mg/liter BDQ for 24 hours and then infected for 24 hours; (top right) 10 mg/liter pradigastat for 48 hours of infection and then 24 hours with pradigastat and 2.5 mg/liter BDQ; (bottom left) 2.5 mg/liter BDQ and 10 mg/liter pradigastat for 24 hours before infection for 24 hours; (bottom right) 400 μM oleate and then infected for 48 hours of infection followed by 24 hours of treatment with 2.5 mg/liter BDQ. (B) Quantification of (A). Data show mean signal from 187 to 640 bacteria per condition with three biological replicates. P values were derived from linear regression. (C) 79Br signal from hMDM infected and treated with 2.5 mg/liter BDQ for 24 hours. A total of 221 to 643 bacteria were analyzed per condition with three biological replicates. P values are as in (B). (D) 79Br signal from hMDM infected and treated with 2.5 mg/liter BDQ for 24 hours. In “PFA,” macrophages were infected with prekilled Mtb or 40 mg/liter verapamil added to the BDQ. A total of 221 to 350 bacteria were analyzed per condition with three biological replicates. P values are as in (B).

Damage to the phagosomal membrane and escape into the cytosol is an important feature of Mtb virulence (22). Therefore, we compared macrophages infected with wild-type Mtb and the ΔRD1 strain, which lacks the ESX-1 type VII secretion system required for cytosolic access (Fig. 3C). Both strains showed a similar BDQ signal, thus BDQ can be transferred to phagosomal bacteria. Furthermore, we infected macrophages with formaldehyde-killed Mtb to test whether BDQ uptake was actively controlled by bacteria. Killed bacteria showed a nonstatistically significant upward trend in BDQ signal (Fig. 3D). We also treated macrophages with the proposed efflux-pump inhibitor verapamil (23), which slightly reduced BDQ accumulation; however, its mechanism of action has recently been called into question (24) (Fig. 3D).

Although LDs were not essential for BDQ transfer to Mtb, we hypothesized that they may enhance antibiotic efficacy. In this case, BDQ will accumulate over time in the host macrophage LDs, increasing the effective BDQ dose for bacteria that later consume them. To study marginal changes in BDQ efficacy, we developed two intracellular antibiotic activity assays in which Mtb replication was measured by high-content confocal microscopy (fig. S4A). BDQ concentrations were selected on the basis of their ability to partially limit Mtb growth (fig. S4, B and C). Inhibitors targeting lipid metabolism enzymes, as well as low-density lipoproteins, were screened for their ability to modulate macrophage LD burden. Only DGAT-1 inhibitors and oleate were found to be effective and nontoxic (fig. S4, D to H). A multiplicity of infection of 1 was used to avoid host cell necrosis, which would expose Mtb to extracellular BDQ.

Inhibition of LD formation by pradigastat, A922500, and T863 reduced antibiotic efficacy in macrophages treated with 2.5 mg/liter BDQ, whereas induction with oleate significantly increased BDQ efficacy (Fig. 4, A and B), therefore LDs facilitated antibiotic effectiveness. Preloading of macrophages with BDQ before infection results in a similar level of inhibition, indicating an accessible intracellular BDQ reservoir (Fig. 4, A and C). BDQ has a physiological terminal half-life of 6 months (25). Therefore, BDQ stored in LDs may be accessible well after the patient has stopped treatment. Furthermore, oleate induction of LDs enhanced the efficacy of preloaded BDQ at a marginally effective concentration (0.25 mg/liter), suggesting that these macrophages contained a larger antibiotic reservoir. Conversely, pretreatment with pradigastat reduced the efficacy of 2.5 mg/liter BDQ. By separating macrophages by LD and Mtb burden at the single-cell level, we observed that BDQ disproportionately depleted the high-LD/high-Mtb subpopulation compared with low-LD/high-Mtb macrophages (fig. S4I). This pattern held when the LD burden in the population was skewed with oleate or DGAT-1 inhibitors. Thus, BDQ selectively targets Mtb in foamy macrophages.

Fig. 4 LDs enhance BDQ efficacy against intracellular Mtb.

(A) Experimental timelines with hours shown in parentheses. (B) Bacterial area per macrophage at 96 hours normalized to BDQ-untreated controls for each condition. Data shown are means ± SEM from 12,937 to 17,180 infected macrophages per LD modulator from three donors. P values were calculated from linear regression. Values on the y axis are shown on a square-root scale. It can be seen that A922500 overlaps with T863. ***P < 0.001, **P < 0.01, *P < 0.05. (C) Bacterial area normalized to BDQ-untreated controls. P values are as in (B). A total of 38,795 to 48,466 infected macrophages were analyzed per LD modulator from five monocyte donors. Values on the y axis are shown on a square-root scale.

We propose a model for the transfer of lipophilic antibiotics to intracellular pathogens in which drugs accumulate in LDs and are transferred to the pathogen as LDs are consumed. Many intracellular pathogens interact with host LDs (26), and these results will inspire a reevaluation of the parameters used to screen drug candidates and the importance of subcellular pharmacokinetics to understanding antibiotic efficacy.

Supplementary Materials

science.sciencemag.org/content/364/6447/1279/suppl/DC1

Materials and Methods

Figs. S1 to S4

References (2732)

Movies S1 to S4

References and Notes

Acknowledgments: We thank S. Horswell (Francis Crick Institute) for assisting with statistics. Funding: This work was supported by the Francis Crick Institute (to M.G.G.), which receives core funding from Cancer Research UK (FC001092), the Medical Research Council (FC001092), and the Wellcome Trust (FC001092), and by a UWA Research Collaboration Award (to M.G.G. and H.J.). H.J. is supported by a Discovery Early Career Researcher Award from the Australian Research Council. D.J.G. is supported by a CASE studentship from BBSRC in partnership with GlaxoSmithKline. Author contributions: M.G.G. and A.W. conceived the project. M.G.G., A.W., D.J.G., and H.J. designed the experiments. D.J.G. performed infections, fluorescence microscopy, and image analysis. M.R.G.R. performed EM sample preparation and TEM with assistance from L.M.C. H.J. and S.H. performed EM and ion microscopy. M.S.D.S. performed lipidomics and BDQ analysis by LC-MS with guidance from J.I.M. D.J.G. created the figures and wrote the manuscript with input from M.G.G. All authors provided feedback on the manuscript. Competing interests: None. Data and materials availability: Lipidomics data are in curation at https://www.ebi.ac.uk/metabolights (study number: MTBLS959). All other data needed to evaluate the conclusions of the study are present in the paper or the supplementary materials.
View Abstract

Stay Connected to Science

Subjects

Navigate This Article