Fig 1.
Strategy for identifying adduction targets of nitro drugs by click chemistry.
An alkyne labeled 5-NI compound enters trophozoites as an inactive prodrug and is reduced to reactive intermediates, which form covalently linked adducts with multiple target molecules. Cell extracts are prepared and reacted with azido-biotin by the click reaction. The adducted and now biotin-labeled molecules can be detected by immunoblotting or ELISA, or can be isolated by streptavidin-affinity purification. Purified proteins are digested in situ with trypsin, and the resulting peptides are analyzed by LC-MS/MS.
Fig 2.
Synthesis and antigiardial activity of Mz-alkyne.
A. Structure of metronidazole (Mz). B. Scheme of two-step synthesis of an alkyne-labeled derivative of Mz (Mz-alkyne). C. In vitro activity of Mz and Mz-alkyne against Mz-sensitive (MzS) and Mz-resistant (MzR) congenic strains of G. lamblia WB, as determined by EC50 assay. Results are shown as mean + SD (n = 3 independent experiments; *p<0.01 vs MzS cells). D. In vivo efficacy of Mz-alkyne against G. lamblia WB in a suckling mouse model of giardiasis. Starting one day after infection, mice were treated orally with a daily dose of 10 mg/kg Mz-alkyne for three days, or given PBS as a control, and trophozoites were enumerated in the small intestine. Each data point represents one animal; horizontal lines show the geometric means (*p<0.05 vs control). The dashed line depicts the assay sensitivity.
Fig 3.
Biochemical detection of 5-NI drug adduction in G. lamblia.
A. Trophozoites of G. lamblia WB were treated for 2 h with Mz-alkyne (Mz-alk) or Mz (both at ~30 x of their respective EC50), or with DMSO as a solvent control. After gel electrophoresis of cell extracts, adducted proteins were analyzed by immunoblotting using in situ click chemistry for coupling with azido-biotin followed by staining with HRP-labeled anti-biotin and visualization by chemiluminescence. Total protein was stained with Coomassie Blue. B,C. Trophozoites were treated with 100 μM (B) or the indicated concentrations (C) of Mz-alkyne for the indicated periods (B) or 2 h (C). Cells were washed and lysed, and the extracts were reacted with azido-biotin by the click reaction and coated onto 96-well plates. Bound biotin-labeled proteins were detected with streptavidin-HRP followed by incubation with the substrates TMB and H2O2. Data are mean ± SD (n = 3 replicates/group). D. Mz-sensitive (MzS) and congenic Mz-resistant (MzR) lines of G. lamblia WB were incubated for 2 h with the indicated concentrations of Mz-alkyne, and cell extracts were analyzed by immunoblotting for adducted proteins as done in panel A. E. Trophozoites attached to glass coverslips were treated for 15 min with 100 μM Mz-alkyne or left untreated, fixed with methanol, and permeabilized with Triton X-100. In situ click reaction was performed with azido-biotin followed by staining with Alexa 488-labeled anti-biotin (IFA, green), and counterstaining with DAPI (IFA, blue). Differential interference contrast (DIC) microscopy was used to visualize the entire cells (bar, 10 μm). F. Suckling mice infected with G. lamblia WB for 4 days were treated with 100 mg/kg of Mz-alkyne for 2 h or left untreated. Frozen sections of the small intestine were stained by in situ click chemistry with azido-biotin followed by Alexa 488-labeled anti-biotin (green). Trophozoites were detected with an antibody against a G. lamblia-specific protein, GASP-180 (red). DAPI was used for counterstaining (blue). Insets show higher magnifications of representative trophozoites.
Table 1.
Proteins adducted by Mz-alkyne in G. lamblia.
Fig 4.
Confirmation of nitro drug adduction and target protein characteristics.
Trophozoites of G. lamblia WB were treated for 2 h with 100 μM (A) or the indicated concentrations (B) of Mz-alkyne (Mz-alk) or solvent (DMSO) alone, and cell lysates were prepared and reacted with azido-biotin using the click reaction. Biotin-labeled proteins were purified by streptavidin affinity chromatography, and identified by in situ trypsin digestion and subsequent LC-MS/MS analysis. A, B. Whole cell extracts (Total), streptavidin affinity-precipitated proteins (PPT), and the remaining supernatants after precipitation (SN) were analyzed by immunoblotting with specific antibodies against enolase, OCT, or ADI. Total protein was detected by Coomassie staining. To control for the efficiency of affinity precipitation, immunoblots were stained in parallel with HRP-conjugated streptavidin. C. Adducted proteins (closed bars) and all others (non-adducted) proteins (open bars) predicted from the G. lamblia WB genome were analyzed for expression levels, as determined by SAGE tags in prior work [43] (left panel), or cysteine content (right panel).
Fig 5.
Functional impact of 5-NI adduction of ADI.
(A) Proposed reaction mechanisms and structures of Mz-alkyne adducts on protein targets. (B) Example MS/MS spectra of the ADI peptide, 279-MHLDCTFSVLHDK-291, obtained by tryptic digestion of Mz-alkyne adducted ADI-HA. On cysteine 283, three y+2 ions and one b+2 ion with an added mass of 147 were observed that constitute evidence for 5-amino-4-thioether (C8H9N3) adduction, and one b+2 ion with an added mass of 163 that provides evidence for 5-sulfinamide (C8H9N3O) adduction. The predicted structures of these cysteine adducts are shown on the right. (C) Scheme of G. lamblia ADI with known active site residues labeled above and cysteines labeled below. (D) Homology model of ADI protein structure and adduction sites. Residues lining the active site pocket are shown in blue, the catalytic cysteine 424 in yellow, and adducted cysteines in red. Images on top show the whole protein from different angles. One of the images was optically sectioned (dashed arrows) and rotated 90° around the indicated horizontal axis to reveal a top view of the section. The inset shows a magnification of the active site pocket. The adducted cysteine 283 is located at the entrance to the pocket. (E) ADI activity in lysates of G. lamblia trophozoites treated with the indicated concentrations of Mz-alkyne (mean ± SD, n = 3 experiments). (F) Activity of recombinant ADI produced as wild-type (WT) or with mutations of cysteine 283 to alanine (C283A) or cysteine 442 to alanine (C442A) after 2 h treatment with (+) or without (-) Mz-alkyne under cell-free conditions in the presence of dithionite. (G) Activity of WT and mutant (C283A) ADI after 2 h treatment with iodoacetamide. Activities in G and H are expressed relative to solvent controls (mean +/- SD, n = 3; *p<0.05 vs controls).
Fig 6.
Therapeutic potential of 5-NI drug targets.
A MzS and congenic MzR line of G. lamblia WB were tested for susceptibility to the indicated inhibitors of the targets identified by adductomics. The line graphs show concentration-response curves for parasite survival after treatment with the indicated inhibitor concentrations (means ± SE, n = 3–4; horizontal lines represent parasite numbers untreated controls). The bar graphs depict the pEC50 values derived from the concentration-response curves (means + SE, n = 3–4; *p<0.05 vs MzS cells,; ns, not significant).
Fig 7.
In vivo efficacy of inhibitors of 5-NI targets against giardiasis.
A. In vitro activity of the ADI inhibitor, canavanine, and the peroxiredoxin inhibitor, conoidin A, against G. lamblia GS/M, as determined by EC50 assay. Results are mean + SD (n = 3 independent experiments). B. Adult mice infected with G. lamblia GS/M were treated orally with canavanine (2 g/kg) for a total of three doses, or with conoidin A (100 mg/kg) for a total of four doses, or given vehicle alone (Control), over a 3-day period. On day 5, live trophozoites were enumerated in the small intestine. Each data point represents one animal, horizontal bars show geometric means (*p<0.05 vs controls). The dashed horizontal line represents the assay sensitivity.