Ethical declaration
Ethical approval for the use of peripheral blood samples from human donors was granted by the Swedish Ethical Review Authority (registration number 2018/1498-31/3). All human studies were carried out in accordance with the guidelines and policies of Karolinska Institutet and EU legislation.
Human samples
Buffy coats were ordered from anonymous blood donors at Karolinska Universitetslaboratoriet. On the same day as sample collection, PBMCs were isolated by density gradient centrifugation over Lymphoprep (StemCell Technologies), washed with RPMI-1640 (Cytiva HyClone) with 10% FBS, and cryopreservated at -80oC in FBS with 10% DMSO (Sigma)50.
Cell viability assays
Cells were seeded in 96-well plates ranging from 2,000 to 15,000 cells per well depending on different cell lines and treated with antibiotics in serial dilutions for 72 hours (120 hours for the preliminary assays) in a humidified incubator (37°C, 5% CO2) in 100 µl medium (DMEM with 10% FBS, 2.05 mM L-glutamine (Sigma) for Hek293 and Hela; RPMI-1640 with 10% FBS, 2.05 mM L-glutamine (Sigma), and 55 µM 2-ME (Gibco) for Jurkat T cells and PBMCs). Based on preliminary tests, the number of cells per well was optimized to 2,000 for Hela and Hek293, 6,000 for Jurkat T cells, and 15,000 for PBMCs. The stocks of antibiotics were prepared by dissolving in DMSO. As control, the same amount of DMSO was added to the non-treated groups. Serial drug dilutions were prepared in the corresponding medium to provide a total of 9 concentrations up to 300 µM for tetracyclines and up to 900 µM for additional bacterial protein synthesis inhibitors. After drug treatment, 10% CCK-8 (Cell Counting Kit-8, Sigma Aldrich) (10 µl per well) of the total volume was added to the cells and incubated for 1–4 hours depending on cell lines (1 hour for Hela and Hek293; 3 hours for Jurkat T cells; 4 hours for PBMCs) before the evaluation of the effect of antibiotics on cell viability. The number of viable cells was calculated by measuring the absorbance at 450 nm using a microplate reader and normalized to the non-treated control. Results were visually confirmed under the light microscope. Each treatment was performed in technical repeats. Dose-response curves were plotted, mean ± SEM was shown, and half maximal inhibitory concentration (IC50) values were calculated in GraphPad Prism.
De novo mitochondrial translation assay
Jurkat T cells were seeded in 6-well plates at a density of 1 million cells/ml and cultured with different concentrations of antibiotics for 18 h in RPMI-1640 medium (with 10% FBS, 2.05 mM L-glutamine (Sigma), and 55 µM 2-ME (Gibco)). After centrifugation, cells were incubated twice for 5 min at 37°C in Cys-/Met-free medium (DMEM, high glucose, no glutamine, no methionine, no cysteine, supplemented with 10% dialyzed fetal bovine serum, 1× GlutaMax, and sodium pyruvate), followed by 20 min incubation at 37°C in Cys-/Met-free medium supplemented with 100 µg/ml emetine to inhibit cytosolic translation (only for mitochondrial translation groups). Subsequently, 1 ml Cys-/Met-free medium supplemented with 200 (for cytosolic translation) or 400µCi (for mitochondrial translation) of EasyTag EXPRESS [35S] protein labeling mix (methionine and cysteine) (Perkin Elmer) were added to each sample and incubated for 20 min at 37°C, 5% CO2. Following labeling, cells were washed three times with 1 ml PBS, harvested by centrifugation (400 × g, 10min, 4°C), and stored at − 20°C. Cells were lysed by resuspension in 30 µl PBS supplemented with a Complete EDTA-free protease inhibitor cocktail and 50 U Benzonase nuclease (Sigma) and by application of one freeze-thaw cycle. Protein contents were determined by Pierce BCA assay (Thermo Fisher). 1 × NuPAGE LDS sample buffer (Thermo Fisher) was added to 30 µg lysate for mitochondrial translation and 10 µg lysate for cytosolic translation, respectively, and separated on NuPAGE 12% Bis-Tris gels (Thermo Fisher). Coomassie staining was performed using Imperial Protein Stain (Thermo Fisher) according to manufacturer suggestions. The gel was fixed in fixing solution (20% methanol, 7% acetic acid, 3% glycerol) for 1 h at RT and vacuum-dried at 65°C for 2 h. The resultant gel was exposed to storage Phosphor screens (Fujifilm) and visualized with Typhoon FLA 7000 Phosphorimager (GE Healthcare).
Western blotting for OXPHOS complexes
Human PBMCs and Jurkat T cells were seeded in 6-well plates at a density of 1 million cells/ml and incubated with or without antibiotics in RPMI-1640 medium (with 10% FBS, 2.05 mM L-glutamine (Sigma), and 55 µM 2-ME (Gibco)) for 6 days. After centrifugation, cells were lysed in RIPA buffer containing protease and phosphatase inhibitors (cOmplete and PhosSTOP, respectively, Roche). Protein concentrations were determined by BCA (Pierce, Thermo Fisher). Protein samples were resuspended with 1× NuPAGE LDS sample buffer (Thermo Fisher) supplemented with 100 mM dithiothreitol, heated for 10 min at 75°C, and separated on NuPAGE 4–12% Bis-Tris mini gels (Thermo Fisher) using NuPAGE 1× MES (Thermo Fisher) running buffer, and transferred to PVDF membranes (0.45 µm, Immobilon-P, Sigma) using the iBlot2 system (Invitrogen). Non-specific binding was blocked with TBST containing 5% non-fat milk, followed by overnight incubation at 4°C with antibodies against Total OXPHOS Human WB Antibody Cocktail (Abcam, Cambridge, UK), followed by HRP-conjugated secondary antibodies for 1 h at room temperature. β-actin, HSP60, and GAPDH (Cell Signaling Technology) were used as loading control. Proteins were detected using Clarity Western ECL Substrate (Bio-Rad, 170–5061).
Extracellular Metabolic Flux Assay
The Seahorse XFe96 Analyzer (Seahorse Bioscience) was used to measure oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) in human PBMCs. 1.5x106 cells/well were seeded in 24-well plates and stimulated using plate-bound anti-CD3 antibody (2 µg/ml, clone OKT3, BD Biosciences), soluble anti-CD28 antibody (0.5 µg/ml, clone CD28.2, Biolegend) and IL-2 (10ng/ml, Proteintech) in RPMI-1640 (with 10% FBS, 2.05 mM L-glutamine (Sigma), and 55 µM 2-ME (Gibco)). Cells were treated with or without 5 and 10 µM Tigecycline for 6 days with a half medium change after 3 days. After treatment, cells were washed with Seahorse XF RPMI assay medium, and then 200,000 cells per well were seeded (in triplicate) in 40 µl assay medium in XF 96-well cell culture microplate coated with poly-D-lysine (Sigma). The plate was centrifuged at 300 g for 5 sec with no brake, rotated 180 ̊, and centrifuged again for 5 sec at 300 g. After centrifugation, 140 µl of assay medium was added per well and the plate was left to stabilize in a 37°C, non-CO2 incubator for 40 min. During seahorse, wells were sequentially injected with compounds to achieve final concentrations of 1,264 µM oligomycin (Sigma); 2µM FCCP (Sigma); 0.5 µM rotenone (Sigma) together with 0.5 µM antimycin A (Sigma). OCR and ECAR were measured for each well three times, every three minutes, before and after each injection. OCR was normalized to protein concentration using a BCA Protein Assay kit (ThermoFisher Scientific) conducted according to the manufacturer’s instructions.
In vitro stimulation of human PBMCs
After thawing frozen PBMCs, cells were washed with RPMI complemented with 10% FBS, 2.05 mM L-glutamine (Sigma), and 55 µM 2-ME (Gibco) and rested at 37oC (5% CO2) for 30 min. 250,000 cells per well were seeded in 96-well U-bottom plates with/without Human T-Activator CD3/CD28 beads (Invitrogen) at a 1:1 ratio and 10 ng/ml IL-2 (Proteintech) in RPMI-1640 medium mentioned above. At the same time, antibiotics were added. For the short-term activation assay, cells were harvested 18 hours post-stimulation and stained with Aqua fixable live/dead dye (Invitrogen) and surface markers. For the proliferation assays, cells were stained with CFSE (eBioscience) or Cell Trace Violet (ThermoFisher Scientific) dyes before seeding and stimulation, according to the manufacturer’s recommendations. After incubation for 6 days with a half medium (complete, containing antibiotics and IL-2) change after 3 days, cells were harvested via centrifugation, stained with live/dead dye and surface markers for analysis by flow cytometry. The Foxp3/Transcription Factor Staining Buffer Set (Invitrogen) was used to fix cells according to the manufacturer’s instructions. Flow cytometry was carried out on a BD Celesta (BD Biosciences) instrument and analyzed in FlowJo (TreeStar) software.
Human CD4+ T cell sorting and stimulation
Thawed PBMCs were washed with RPMI complemented with 10% FBS, 2.05 mM L-glutamine (Sigma), and 55 µM 2-ME (Gibco) before resting at 37oC (5% CO2) for 30 min. After counting, cells were resuspended in FACS buffer at a density of 5 x 107 cells/ml. CD4+ T cells were isolated using CD4+ T cell negative isolation kits (Miltenyi Biotech) and enriched using CD4+ T cell Enrichment kits (Miltenyi Biotech). Purified CD4+ T cells were counted and stained with CTV dye and live/dead dye. Subsequently, surface makers were stained at 4oC in FACS buffer (DPBS + 2% FBS). The following panels were used to classify human T CD4+ cells from the live lymphocyte gate: CD4+ T naïve: Aqua−CD4+CD45RA+ CD27+; CD4+ T memory: Aqua−CD4+CD45RA−CD27+.
After staining, cells were washed and resuspended and sorted using a FACSAria Fusion (BD Biosciences). Sorted cells were collected in PBS complemented with 2% FBS at 4oC. Cells were then harvested via centrifugation and resuspended in RPMI-1640 (+ 10% FBS, 2.05 mM L-glutamine (Sigma), and 55 µM 2-ME (Gibco)) before counting. Sorted CD4+ subsets were counted and 85,000 cells per well seeded onto 96-well U-bottom plates with/without Human T-Activator CD3/CD28 beads (Invitrogen) at a 1:2.8 cell:bead ratio and 10ng/ml IL-2 (Proteintech) in RPMI-1640 (+ 10% FBS, 2.05 mM L-glutamine (Sigma), and 55 µM 2-ME (Gibco)). At the same time, tigecycline in serial concentrations (from 2.5 to 10 µM) was added. With a half medium change after 3 days, cells were harvested after 6 days, and stained and fixed before analysis by flow cytometry. Antibodies used in the study are presented in Table 2.
Table 2
Antibodies |
Total OXPHOS Human WB Antibody Cocktail | Abcam | Ab110411 |
HSP60 | Enzo Lifesciences | AB1-SPA-807-E |
b-actin | Abcam | Ab8224 |
GAPDH | Abcam | Ab8245 |
HRP secondary rabbit | GE Healthcare | NA9340V |
HRP secondary mouse | GE Healthcare | NA9310V |
Human CD3 | BD Biosciences | Clone 3D12 |
Human CD4 | BD Biosciences | Clone RPA-T4 |
Human CD8 | BD Biosciences | Clone RPA-T8 |
Human CD45RA | BD Biosciences | Clone HI100 |
Human CD27 | BD Biosciences | Clone M-T271 |
Human CD25 | BD Biosciences | Clone 2A3 |
Statistical analyses
Statistical analyses were carried out in Prism 9 (GraphPad). Differences between groups were analyzed by a Student's t-test or one-way ANOVA with Tukey's multiple comparisons test.
Isolation of mitochondria and purification of mitochondrial monosomes
Jurkat T cells were grown in RPM1-1640 medium (+ 10% FBS, 2.05 mM L-glutamine (Sigma), and 55 µM 2-ME (Gibco)) in a vented flask shaking at 120 rpm at 37°C under 5% CO2. The culture was scaled up by splitting at a cell density of 1.6 × 106 cells/ml. A final volume of 1.5-liter cells at a density of 1 × 106 cells/ml was harvested by centrifugation at 1000 x g for 10 min at 4°C. After being washed with cold PBS buffer, the cell pellet was resuspended in the cold hypotonic MSE buffer (with 0.6 M mannitol, 10 mM Tris–HCl pH 7.4, 1 mM EDTA, 0.1% BSA), and ruptured on ice by a semi-automatic homogenizer (Schuett-biotech). The lysate was clarified by centrifugation at 400 × g and 4°C for 10 min. The pellet was resuspended and subsequently homogenized. After 3 cycles of homogenization-centrifugation, the cell lysates were combined and the mitochondria were pelleted by additional centrifugation at 11,000 × g and 4°C for 10 min. The crude mitochondria were loaded onto the sucrose cushion (1.0 M and 1.5 M sucrose in, 20 mM Tris–HCl pH 7.4, 1 mM EDTA) and centrifuged for 1 hr at 77,000 × g (25,000 rpm) in a SW41 Ti rotor (Beckman Coulter). The band formed by the mitochondria in the middle between 1 and 1.5 M sucrose was collected carefully and resuspended in 10 mM Tris–HCl pH 7.4 in a 1:1 ratio, The pure mitochondrial pellet was collected after centrifugation at 11,000 × g and 4°C for 15 min and then resuspended in mitochondrial freezing buffer (300 mM trehalose, 10 mM Tris–HCl pH 7.4, 10 mM KCl, 0.1% BSA, 1 mM EDTA), flash-frozen in liquid N2 and stored at − 80°C.
The purified mitochondria were thawed and lysed by incubating at 4°C for 30 min in the lysis buffer (25 mM HEPES–KOH pH 7.5, 20 mM Mg(OAc)2, 50 mM KCl, 2% (vol/vol) Triton X-100, 2 mM Dithiothreitol (DTT), 1× cOmplete EDTA-free protease inhibitor cocktail (Roche), 40 U/µl RNase inhibitor (Invitrogen)). The mitochondrial lysate was centrifuged at 19,000 × g (13000 rpm) for 12 min at 4°C, and subsequently overlayed on top of a 10–30% sucrose gradient in the ribosome buffer (25 mM HEPES/KOH pH 7.5, 50 mM KCl, 20 mM Mg(OAc)2, 2 mM DTT). After centrifugation for 21 hr at 54,331 × g (21,000 rpm) in a SW41 Ti rotor (Beckman Coulter), the gradients were fractionated with a Biocomp Fractionator. Fractions corresponding to the monosomes were pooled and pelleted at 135,520 × g (55,000 rpm) for 16 hr at 4°C using a TLA55 rotor (Beckman Coulter). The pellet was gently washed 3 times and dissolved in ribosome buffer. The solution was kept on ice for 15 min and the soluble mitoribosomes were collected by centrifugation at 20,000 x g for 15 hr at 4°C to get rid of aggregation. The concentration of the purified monosomes was quantified by nanodrop.
CryoEM data collection and analysis
In vitro reconstitution of the mitoribosomal monosome-tigecycline complex was performed using 30 µM of tigecycline incubated with 100 nM of monosome in a ratio of 1:300. Holey carbon grids (Quantifoil R2/2, copper, 300 mesh) coated with a layer of continuous carbon (~ 3 nm thickness) were subjected to a glow discharge of 25 mA for 120 s. The sample was applied to the grids at 4°C with 100% humidity and incubated for 30 s using a Vitrobot MKIV (Thermo Fisher Scientific), followed by 3 s blotting with blot force 3 and plunge-freezing in liquified ethane. The dataset was acquired on Titan Krios G3i transmission electron microscope (Thermo Fisher Scientific) operated at 300 kV in the Karolinska Institutet’s 3D-EM facility using a slit width of 20 eV with GIF quantum energy filter (Gatan). For imaging, a K3 detector (Gatan) was used capturing micrographs at magnification of 165kX yielding a pixel size of 0.505 Å. A dose of 45 electrons per Å2 in 50 frames was used with defocus values ranging from − 0.4 to − 1.6 µm. Motion correction followed by CTF estimation, Fourier cropping (to 1.01 Å/px), particle picking, and extraction in 512-pixel boxes (size threshold 300 Å, distance threshold 20 Å, using the pre-trained BoxNet2Mask_20180918 model) were performed on the fly using Warp51. Only particles from micrographs with an estimated resolution of 5 Å were retained for further processing. Detailed parameters are given in the Supplementary Table 1.
A total of 12,972 micrographs from 2 datasets were selected based on an estimated resolution cut-off of 5 Å and defocus below 2 µm as estimated by Warp. A total of 519,237 picked particles from Warp were imported to CryoSPARC (v4.2.1)52 to perform further processing. 2D classification was carried out followed by ab initio reconstructions of clean or good classes with high-resolution features and junk classes. These ab initio reconstructions were used for the heterogeneous refinements of all picked particles. After several round of heterogeneous refinements, 266,550 clean particles comprised the mitoribosomal monosome with high-resolution features were retained and used for further processing. Homogenous refinement of these clean particle stacks was performed which yielded a resolution of 2.4 Å. A recent model of human mitoribosomal monosome (PDB:7QI432) purified from the Hek293 T cells bound to A-tRNA, P-tRNA, and mRNA was fitted in our reconstruction of the monosome, which also comprises these features with additional densities in both mtSSU and mtLSU that could be attributed to the antibiotic tigecycline. Mask targeting the P-tRNA was generated to perform 3D variability analysis, and subsequently classified into three particle clusters with each representing different features of the monosome. Particles with clear density for the P-tRNA were subjected to a second round of 3D variability analysis using the A-tRNA as a mask. We could eventually obtain 3 major classes, class 1 with the majority of the particles (‘Empty’: 130,443 particles) lacked densities for the A-site and P-site tRNA on the monosome, class 2 (‘P-tRNA only’: 87,076 particles) lacked the occupancy for the A-tRNA and is represented by the monosome bound to P-tRNA only and mRNA, and class 3 (‘A- and P-tRNA’: 36,974 particles) contained monosome bound A-tRNA, P-tRNA, and mRNA. Each class/particle set was subjected to a homogenous refinement which yielded a corresponding reconstruction at 2.5 Å, 2.7 Å, and 2.8 Å resolution, respectively. To improve the local resolution of the tigecycline-bound regions on the monosome, 3D refined classes were subjected to CTF refinement (global and local refinement). Furthermore, two masks covering the region of the tigecycline-binding sites on the mtSSU (SSU-head) and mtLSU (LSU-body) were prepared (Supplementary Fig. 2), and local-masked 3D refinement was performed. Reported resolutions are based on gold-standard, applying the 0.143 criterion on the FSC between the reconstructed half-maps (Supplementary Fig. 2). Maps underwent local-resolution filtering, superposed to the consensus map, and combined via Phenix53 for model building and refinement.
Model building and refinement
Model building of the tigecycline-bound monosome was carried out using Coot54. The starting model for the monosome was Protein Data Bank (PDB) ID 7QI432. This model was fitted as a rigid body into the map of the ‘P-tRNA only’ monosome class, and further adjustments were made manually. Three active sites for tigecycline were identified on the monosome, which agreed with the density: one at the mtSSU and two in the mtLSU. Water molecules were picked by Coot automatically around the tigecycline-binding region, and adjusted manually. Metal ions, cofactors (2Fe-S), and modifications were placed based on map densities. Geometrical restraints of modified residues and ligands were calculated by Grade Web Server (http://grade.globalphasing.org) or obtained from the CCP4 library55. Hydrogens were added to all molecules except water by REFMAC556. The final model was refined using the composite map via Phenix.real_space_refine v1.1853. The refined model was validated with MolProbity57,58 and the Phenix suite50. Model statistics are listed in Table 1. UCSF ChimeraX 1.6.159was used to make the figures.