High Throughput Screen for Escherichia coli Twin Arginine Translocation (Tat) Inhibitors

Abstract

The twin arginine translocation (Tat) pathway transports fully-folded and assembled proteins in bacteria, archaea and plant thylakoids. The Tat pathway contributes to the virulence of numerous bacterial pathogens that cause disease in humans, cattle and poultry. Thus, the Tat pathway has the potential to be a novel therapeutic target. Deciphering the Tat protein transport mechanism has been challenging since the active translocon only assembles transiently in the presence of substrate and a proton motive force. To identify inhibitors of Tat transport that could be used as biochemical tools and possibly as drug development leads, we developed a high throughput screen (HTS) to assay the effects of compounds in chemical libraries against protein export by the Escherichia coli Tat pathway. The primary screen is a live cell assay based on a fluorescent Tat substrate that becomes degraded in the cytoplasm when Tat transport is inhibited. Consequently, low fluorescence in the presence of a putative Tat inhibitor was scored as a hit. Two diverse chemical libraries were screened, yielding average Z'-factors of 0.74 and 0.44, and hit rates of ~0.5% and 0.04%, respectively. Hits were evaluated by a series of secondary screens. Electric field gradient (Δψ) measurements were particularly important since the bacterial Tat transport requires a Δψ. Seven low IC50 hits were eliminated by Δψ assays, suggesting ionophore activity. As Δψ collapse is generally toxic to animal cells and efficient membrane permeability is generally favored during the selection of library compounds, these results suggest that secondary screening of hits against electrochemical effects should be done early during hit validation. Though none of the short-listed compounds inhibited Tat transport directly, the screening and follow-up assays developed provide a roadmap to pursue Tat transport inhibitors.

Fig 1. Design of the HTS.

(A) The fluorescent Tat precursor protein spTorA-mCherry-SsrA. The N-terminal TorA signal peptide (spTorA) targets the fluorescent protein mCherry (PDB: 2H5W) to the Tat machinery for transport to the periplasm, and the C-terminal SsrA-tag promotes cytoplasmic degradation of the protein. (B) Fluorescence spectra of LB media alone (black) and with cells producing spTorA-GFP-SsrA in the Tat⁺⁺ background (green) (EX = 485 nm). Tat⁺⁺ denotes strain MC4100(DE3) in which TatABC is overproduced from the pTatABC-Duet1 plasmid (see Methods). (C) Fluorescence spectra of LB media alone (black) and with cells producing spTorA-mCherry-SsrA (red) under the indicated conditions (EX = 587 nm). NaSCN collapses the Δψ across the inner membrane. For (B) and (C), cells with (Tat⁺⁺) or without (ΔTat) TatABC were induced with the indicated proteins and then incubated for 12 h at 4°C (ΔTat denotes strain MC4100ΔTatABCDE; see Methods for more details). (D) Design of the HTS assay. The SsrA tag on spTorA-mCherry-SsrA promotes degradation of cytoplasmically-localized spTorA-mCherry-SsrA, and thus, periplasmically-localized mature Tat cargo (mCherry-SsrA) is the dominant contributor to the total mCherry cellular fluorescence. In the ΔTat background, or in the presence of 30 mM NaSCN (positive control; simulated hit), the total cellular fluorescence is lower, as shown in (C). Due to the competition between transport of mCherry to the periplasm by the Tat machinery and degradation by the ClpXP/ClpAP protease system, low fluorescence in the presence of a putative Tat inhibitor was considered a hit. spTorA-mCherry-H6 (replacing the SsrA tag with a 6xHis-tag) is a control protein that is not degraded in the cytoplasm. OM = outer membrane; IM = inner membrane.

Fig 2. Optimization of the Total mCherry Fluorescence Signal and Validation of Substrate Design Properties.

(A) Effect of spTorA-mCherry-SsrA induction levels on the total mCherry cellular fluorescence in wild type (Tat⁺; MC4100(DE3)) and TatABC overproduced (Tat⁺⁺) backgrounds. The production of spTorA-mCherry-SsrA was induced for 8 h at 25°C with the indicated concentrations of arabinose and the total mCherry cellular fluorescence was determined (n = 3; EX = 587 nm, EM = 610 nm). TatABC was induced with 1 mM IPTG. (B) Cell fractionation. Cytoplasmic (C) and periplasmic (P) fractions [60] of the indicated strains were analyzed after 8 h induction by SDS-PAGE and immunoblotting using anti-mCherry antibodies. For the top three gels, the cells overproduced spTorA-mCherry-SsrA. For the bottom gel, the SsrA tag on spTorA-mCherry-SsrA was replaced with a 6xHis-tag (yielding spTorA-mCherry-H6). Precursor (p) and mature (m) proteins are indicated. The * and ** identify what appears to be C-terminally truncated products. (C) Fluorescence microscopy of the strains in (B). The mCherry proteins were induced at 25°C for 15 h with 2 mM arabinose. Cells were grown in fresh media for an additional 5 h with no arabinose before imaging. More than 95% of cells showed a clear and dominating periplasmic localization of mCherry in the Tat⁺⁺ strain versus ~70% for the Tat⁺ strain. No periplasmic mCherry was observed in the ΔTat strain, confirming that the spTorA-mCherry-SsrA protein was indeed targeted to and transported by the Tat system. Replacing the SsrA tag with a 6xHis-tag resulted in an increase in the amount of cytoplasmic mCherry (compare with Fig 2B), confirming that the SsrA tag promotes cytoplasmic degradation of the spTorA-mCherry-SsrA protein. Visually, about 80% of cells showed noticeable cytoplasmic localization of spTorA-mCherry-H6. Bar = 1.3 μm. (D) Effect of TatABC pre-induction on the total cellular fluorescence of mCherry. TatABC was induced 3 h prior to or simultaneously with spTorA-mCherry-SsrA (100 mM arabinose; n = 3). (E) Effect of induction time on the total cellular fluorescence of mCherry. The spTorA-mCherry-SsrA protein was produced in the Tat⁺⁺ background and the total mCherry cellular fluorescence was obtained as in (A) (n = 3).

Fig 3. Stability of the mCherry Fluorescence Signal.

(A) Stability of the total cellular fluorescence after transfer to 4°C. Cells producing spTorA-mCherry-SsrA under Tat⁺ or Tat⁺⁺ conditions were incubated at 25°C for 8 h, and then transferred to 4°C to inhibit growth. The total mCherry cellular fluorescence was monitored periodically (n = 3). (B) Stability of the total cellular fluorescence after glucose addition. The total cellular fluorescence was quantified as in (A), except that spTorA-mCherry-SsrA production was repressed by addition of 0.5% glucose after 8 h of growth. Growth was continued at 25°C (n = 3). These data demonstrate that the intensity of the mCherry fluorescence signal can be maintained at a constant value for long time periods, thereby providing consistency when reading many HTS plates.

Fig 4. Effect of Oxidative Phosphorylation Inhibitors on In Vivo Tat Transport and Sensitivity of Fluorescence Detection on Plates.

(A) Effect of oxidative phosphorylation inhibitors on cell growth and mCherry fluorescence under Tat⁺⁺ conditions. Cells producing spTorA-mCherry-SsrA under Tat⁺⁺ conditions were grown at 25°C for 8 h in the presence of the indicated respiratory inhibitors. After overnight incubation at 4°C, optical densities (upper panel; A₅₀₀) and total mCherry cellular fluorescence intensities (lower panel; EX = 587 nm, EM = 610 nm) were measured (n = 3). Nigericin (N; 50 μM) is a H⁺/K⁺ exchanger, which collapses the ΔpH gradient. Valinomycin (V; 50 μM) is a K⁺ ionophore, which collapses the Δψ. NaSCN (SCN; 50 mM) collapses the Δψ. NaN₃ (N₃; 0.1%) inhibits the cytochrome bo₃ terminal oxidase. (B) Effect of NaSCN concentration on cell density and total mCherry cellular fluorescence. Measurements were made after incubating cells overnight at 4°C (n = 3). (C) Calibration and sensitivity of mCherry fluorescence detection in 384-well plates. Overnight cultures of E. coli MC4100(DE3) cells were diluted to A₅₀₀ = 0.25, 0.2, 0.15, 0.10 and 0.05, and aliquoted into 384-well plates. The spTorA-mCherry-SsrA and TatABC proteins were induced together or neither were induced, as indicated. The cells were incubated at 25°C for 8 hours, and stored overnight at 4°C. Optical densities (upper panel) and mCherry fluorescence intensities (lower panel) were determined with a BMG Polarstar Omega plate reader (at 500 nm for absorbance; EX = 584 nm, EM = 620 nm for fluorescence). For comparison, the absorbance and fluorescence intensities from purified spTorA-mCherry-H6 (30, 25, 20, 15, 10 and 5 nM) are also shown (n = 3 plates, 16 wells/plate). (D) Simulated hits on 384-well plates. Using the approach described in (C), cells (A₅₀₀ = 0.15) were grown on plates under the indicated conditions, and then fluorescence intensities were measured. spTorA-mCherry-SsrA was used for conditions II-IV. spTorA-mCherry-H6 was used for a minus SsrA control (I; 100%). The SsrA tag decreased the mCherry fluorescence intensity (II; 63±14%), consistent with a decrease in cytoplasmic mCherry concentration due to SsrA-dependent degradation. A further decrease in mCherry fluorescence intensity was observed in Tat-deficient cells (ΔTat) (III; 41±6%) and in the presence of 30 mM NaSCN (IV; 36±14%), consistent with enhanced cytoplasmic degradation due to inhibition of export to the periplasm (n = 3 plates, 16 wells/plate).

Fig 5. Sample Data from the Local HTS, Z'-factor Summary and Hit Summary.

(A) Optical density (left panel; A₅₀₀) and fluorescence intensity (right panel; EX = 584 nm; EM = 620 nm) values from a typical local HTS plate obtained under Tat⁺⁺ conditions. DMSO only (columns 2 and 23) and 30 mM NaSCN (columns 1 and 24) controls are identified as − and +, respectively. (B) Graphical representation of the optical density (black) and fluorescence intensity (red) values for each well of the plate shown in (A). A potential Tat inhibitor (well G-14) and growth inhibitor (well J-22) are identified by red and black arrows, respectively. These intensity signatures are detectable in the raw data images in (A). (C) The Z'-factor for each plate from the local HTS. (D) Fluorescence intensity distribution of compounds with the intensity signatures expected for a Tat inhibitor (see text). The percent fluorescence decrease is normalized to the 30 mM NaSCN positive control (100%) and the DMSO control (0%).

Fig 6. Sample Dose Response Data for Hits from the Local HTS.

(A) Optical density (left panel) and fluorescence intensity (right panel) values for a typical local HTS dose response plate obtained under Tat⁺⁺ conditions. DMSO only (columns 2 and 23) and 30 mM NaSCN (columns 1 and 24) controls are identified as − and +, respectively. Compound concentrations are given at the top of each column. Note that there are two different compounds per row. Dose response plates were run in duplicate. (B) Graphical representation of the values for the plate shown in (A).

Fig 7. The BIPDeC HTS and Z'-factor Summary.

(A) Activity results for the 337,881 compounds screened in the BIPDeC HTS. Active, inconclusive, and inactive activity scores are defined in the text. (B) The Z'-factor for each plate from the BIPDeC HTS.

Fig 8. Effect of Hits from the Local HTS on In Vitro Tat Transport.

(A) Tat transport assays of pre-SufI using NADH to generate a Δψ. Inverted membrane vesicles (IMVs; A₂₈₀ = 5.0), the fluorescent substrate pre-SufI-IAC^Atto565 (90 nM), and NADH (4 mM) were incubated at 37°C for 30 min with DMSO (control) or with hit compound (20 μM). Samples were treated with the protease proteinase K, run on SDS-PAGE, and analyzed by in-gel fluorescence imaging. Assays were run in duplicate. Band intensity corresponds to the amount of protein transported into the IMV lumen. The following 7 compounds survived this screen: 1^JS, 51^JS, 55^JS, 111^JS, 181^JS, 202^JS, and 140^JS (boxed in red). (B) Tat transport assays of pre-SufI using ATP to generate a Δψ. All 7 compounds surviving the screen in (A) were tested using a Tat transport assay performed identically to those in (A) except that ATP (1 mM) was used in place of NADH. Compound 55^JS was eliminated by this assay—the other 6 compounds survived (boxed in red). (C) Chemical structure of compound 55^JS, N-(4,7-Dioxo-4,7-dihydro-2,1,3-benzoxadiazol-5-yl)acetamide. The quinone structure suggests NADH dehydrogenase as a possible target, leading to inhibition of Tat transport when using NADH as the energy source (A), but not when ATP is used (B).

Fig 9. Effect of Hits from the BIPDeC HTS on In Vitro Tat Transport.

(A) & (B) Tat transport assays as in (A) and (B) of Fig 7, respectively, using hit compounds from the BIPDeC HTS. Hit compound concentrations of 100 μM were used to maximize Tat transport inhibition, as the effects of a lower concentration (20 μM) were not very pronounced (data not shown). Compounds 18^NIH and 21^NIH survived the NADH-dependent assay (A) and were tested in the ATP-dependent assay (B). Both of these compounds survived the ATP-dependent assay. Surviving compounds at each stage are boxed in red.

Fig 10. Summary of the HTS's and Hit Analysis by Follow-up Screens.

Surviving and eliminated compounds from the local (left) and BIPDeC (right) HTS's are denoted by black and red numbers, respectively. The number of eliminated compounds at each step is shown to the right of the arrows and the number of surviving compounds are shown on the left in bold. Active (123) and inconclusive (1263*) compounds after the primary BIPDeC HTS were evaluated for chemical intractability and then retested in dose response and cell growth assays.