Fig 1.
Design and validation of a multi-protein expression system for mevalonate pathway genes.
A) Streptococcus pneumoniae mevalonate kinase (mevK), Homo sapiens phosphomevalonate kinase (pmevK), E. coli isopentenyl diphosphate isomerase (idi), and S. pneumoniae mevalonate 5-diphosphate decarboxylase (mvd) were arranged for polycistronic expression in E. coli with a stop codon (red) and ribosomal binding site (black box) between each pair of genes. Digestion with a specific endonuclease (Nuc2) allowed these elements to be removed for expression of a multi-functional proteins linked by flexible three amino acid linkers (3AA). B) The eight possible ways of connecting the four genes (red octagons show retention of the stop codon) were tested by attempting to rescue the growth of E. coli inhibited with the isoprenoid synthesis inhibitor fosmidomycin (Fos). Successful growth in the presence of mevalonate (Mev) showed that all eight constructs rescued fosmidomycin inhibition in a mevalonate-dependent manner, while the negative control only containing the two kinases (KK—) did not.
Fig 2.
The MVA pathway bypasses fosmidomycin toxicity in PfMev parasites.
A) Diagram of the parasite cell showing that the endogenous MEP pathway is localized with the apicoplast organelle of the parasite, with IPP and DMAPP presumably being exported out of the organelle. We engineered the PfMev parasites to express four enzymes forming a MVA pathway in the cytosol capable of producing IPP and DMAPP from exogenous mevalonate supplied in the growth medium. B) PfMev parasites were treated with 25μM fosmidomycin (50x IC50), and supplemented with various concentrations of mevalonate, with growth compared to an untreated control. PfMev parasites grew to wild-type levels in the presence of a typically lethal concentration of fosmidomycin when supplemented with at least 10μM mevalonate, indicating usage of the engineered mevalonate pathway for the generation of essential isoprenoid precursors. PfMev parasites treated with fosmidomycin and not supplemented with mevalonate failed to grow. These data are from quadruplicate independent experiments, each conducted in quadruplicate with error bars representing the standard error of the mean (SEM).
Fig 3.
Fluorescent labeling of the apicoplast organelle in PfMev parasites and confirmation via co-localization.
A) Live fluorescence microscopy of the PfMev line expressing the signal sequence and transit peptide from ACP fused to SFG (green), the mitochondria was stained with MitoTracker (red), and nuclear DNA was stained with DAPI (blue). B) Immunofluorescence co-localization of aSFG with the apicoplast marker ACP, with α-ACP (green), α-GFP (red), with nuclear DNA stained with DAPI (blue) in fixed PfMev parasites. Images depict fields that are 10 microns long by 10 microns wide.
Fig 4.
Generation and characterization of an apicoplast-disrupted PfMev parasite line.
A) Attempted detection of nuclear, mitochondrial, and apicoplast genomes by PCR amplification of the genes ldh (nuclear, N), sufB (apicoplast, A), and cox1 (mitochondrial, M) from PfMev parasites treated with 100nM azithromycin and 50μM mevalonate for one week (red), in addition to an untreated control (blue). Failure to amplify sufB in PfMev parasites treated with azithromycin/mevalonate indicates loss of the apicoplast organelle. B) Live fluorescence microscopy of PfMev parasites after treatment with 100nM azithromycin and 50μM mevalonate for one week, showing a disrupted organelle phenotype, with multiple discrete vesicles instead of a single intact organelle. The apicoplast is labeled with the api-SFG protein (green), the mitochondrion is stained with MitoTracker (red), and nuclear DNA is stained with DAPI (blue). C) Immunofluorescence co-localization of api-SFG with the apicoplast marker ACP, with αGFP (red), αACP (green), with nuclear DNA stained with DAPI (blue) in fixed apicoplast disrupted PfMev parasites. D) Growth curve of apicoplast-disrupted PfMev parasites grown in the presence of different concentrations of mevalonate. The growth of untreated (apicoplast intact) PfMev parasites was also measured in the presence of 50μM or 0μM mevalonate as a control. Supplementation with 10μM mevalonate, or more, restored the growth of apicoplast-disrupted parasites to near wild-type levels. Data from quadruplicate independent experiments, each conducted in quadruplicate are shown with error bars representing the standard error of the mean (SEM). Images depict fields that are 10 microns long by 10 microns wide.
Fig 5.
Metabolic labeling of PfMev parasites with [2-13C]-mevalonate.
PfMev parasite cultures were treated with 25μM fosmidomycin (50x IC50) and supplemented with either 50μM unlabeled mevalonate or [2-13C]-mevalonate for 48 hours. Metabolites were extracted and analyzed via targeted liquid-chromatography mass spectrometry (LC-MS/MS) using selected reaction monitoring (SRM). A) Detection of IPP/DMAPP (black) in parasites treated with 50μM unlabeled mevalonate and 25μM fosmidomycin, with attempted detection of mass-shifted IPP/DMAPP (red). B) Detection of mass-shifted IPP/DMAPP (red) in parasites treated with 50μM [2-13C]-mevalonate and 25μM fosmidomycin, with attempted detection of IPP/DMAPP (black). C) Detection of FPP (black) in parasites treated with 50μM mevalonate and 25μM fosmidomycin, with attempted detection of mass-shifted FPP (red). D) Detection of mass-shifted FPP (red) in parasites treated with 50μM [2-13C]-mevalonate and 25μM fosmidomycin, with attempted detection of FPP (black). The red asterisks on the biomolecules shown denote the positions of the labeled carbons. Metabolite peaks are labeled with the liquid chromatography column retention time in minutes.
Fig 6.
Deletion of the fosmidomycin drug target DXPR.
A) Confirmation of DXPR gene deletion via PCR showing the expected 5’ and 3’ integration of the selection cassette. Δdxpr parasites (red) were also screened for the unmodified 5’ and 3’ loci to detect if there were any residual wild-type parasites and compared to a control (blue). B) PfMev Δdxpr parasites were screened for the presence of the nuclear (ldh), apicoplast (sufB), and mitochondrial (cox1) genomes via PCR (red), and compared to a control (blue). Both lines showed the presence of nuclear, apicoplast and mitochondrial genomes, indicating deletion of DXPR does not result in organelle loss. C) Live epifluorescence microscopy of the PfMev Δdxpr parasites showing the presence of an intact organelle. The apicoplast is labeled with the api-SFG protein (green), the mitochondrion is stained with MitoTracker (red), and nuclear DNA is stained with DAPI (blue). Images depict fields that are 10 microns long by 10 microns wide. D) PfMev Δdxpr parasites grown in the presence of 50μM mevalonate (blue) or no mevalonate (red). Error bars represent the standard error of the mean from three independent experiments, each conducted in quadruplicate. The reliance of PfMev Δdxpr parasites on mevalonate for growth indicates that DXPR is essential for the blood-stage survival of the parasite.
Fig 7.
Transcriptomic analysis of apicoplast-disrupted parasites.
A) Hierarchical clustering of apicoplast genome expression from control and apicoplast-disrupted parasites during the IDC. The apicoplast gene expression data were normalized by the median expression of 5851 genes from corresponding parasites at each time point before performing the clustering. We used Euclidean distance measure, along with Ward’s method, to create the hierarchical clusters in MATLAB. The expression profile of the apicoplast genes from the apicoplast-disrupted parasites is minimal confirming that the organelle is missing. The values are shown on a log2 scale. B) Spearman coefficient values were calculated by comparing the global transcriptional data generated from control (black circles) and apicoplast disrupted (colored circles) parasites at the indicated time points against corresponding data from each time point generated in the high-resolution 3D7 IDC transcriptome from Llinás and coworkers [39]. The y-axis shows Spearman coefficient; the x-axis shows hours post invasion in the IDC data set. The apex of the peak in each graph corresponds to the approximate point in the IDC to which apicoplast-disrupted parasites best correlate at the indicated incubation time. Plots are shown with a Loess fit of the data: 16hrs (red), 32hrs (green), and 48hrs (blue). C) Hierarchical clustering of transcriptomic data, of 96 high-confidence apicoplast genes, from control and apicoplast-disrupted parasites at the indicated time points. The expression profile of each gene was normalized by its expression at time point 0 to visualize variation in their temporal profile, in response to apicoplast disruption. Four distinct clusters are shown (cyan, magenta, blue, and red) and are further described in S3 Fig. The clustering method used here is identical to that described for A).
Fig 8.
Metabolomic analysis of apicoplast-disrupted parasites.
A) Hierarchical clustering of metabolomic data from control and apicoplast-disrupted parasites at the indicated time points. Metabolomic data are shown for 405 metabolites that are present in both data sets. The data represent the abundance of each metabolite after normalization with their median value at the sampled time points. The hierarchical clustering method is the same as described above, except that we also used the k-nearest-neighbor method to impute any missing values in the data. The values are shown on a log2 scale with orange and blue colors indicating amounts higher and lower than the median value, respectively. B) Four clusters of metabolites that have temporal profiles which differ between control and apicoplast-disrupted parasites are colored in red, magenta, green, and cyan.