Apicoplast-Localized Lysophosphatidic Acid Precursor Assembly Is Required for Bulk Phospholipid Synthesis in Toxoplasma gondii and Relies on an Algal/Plant-Like Glycerol 3-Phosphate Acyltransferase

Abstract

Most apicomplexan parasites possess a non-photosynthetic plastid (the apicoplast), which harbors enzymes for a number of metabolic pathways, including a prokaryotic type II fatty acid synthesis (FASII) pathway. In Toxoplasma gondii, the causative agent of toxoplasmosis, the FASII pathway is essential for parasite growth and infectivity. However, little is known about the fate of fatty acids synthesized by FASII. In this study, we have investigated the function of a plant-like glycerol 3-phosphate acyltransferase (TgATS1) that localizes to the T. gondii apicoplast. Knock-down of TgATS1 resulted in significantly reduced incorporation of FASII-synthesized fatty acids into phosphatidic acid and downstream phospholipids and a severe defect in intracellular parasite replication and survival. Lipidomic analysis demonstrated that lipid precursors are made in, and exported from, the apicoplast for de novo biosynthesis of bulk phospholipids. This study reveals that the apicoplast-located FASII and ATS1, which are primarily used to generate plastid galactolipids in plants and algae, instead generate bulk phospholipids for membrane biogenesis in T. gondii.

Fig 1. TgATS1 is a plant-like G3PAT residing in the stroma of the T. gondii apicoplast.

(A) Structure-based protein sequence alignment of G3PATs from T. gondii, TgATS1 (TGGT1_270910), P. falciparum (PF3D7_1318200), P. yoelii (PyapiG3PAT), A. thaliana (AtATS1, [68]) and C. moschata (CmATS1, [47]). Residues strictly conserved between all species are highlighted in green, residues conserved in at least three species in cyan, residues conserved in apicomplexan sequences in yellow, and residues conserved between TgATS1 and plant ATS1 (AtATS1 and CmATS1) in blue. Brown triangles and grey ovals represent residues putatively involved in G3P or FA binding, respectively. Secondary structures [47, 67] are represented above the sequence alignment by blue cylinders for α-helices and orange arrows for β-strands. Domain 1 of the protein (4-helix bundle) is symbolized with dashed lines and Domain 2 (α-β Domain) in solid lines. Residues putatively involved in binding the G3P substrate in CmATS1 (brown triangles) are strictly conserved in TgATS1, while those putatively involved in binding the acyl-ACP substrate are highly conserved (grey ovals). (B) Overlay of the CmATS1 crystal structure ([47]) and the predicted TgATS1 3D structure. The overall structure and surface accessibility of CmATS1 (grey) and TgATS1 (green and magenta) is conserved and highly similar as observed in the ribbon representation. (C) Residues putatively involved in binding substrate (G3P) and those involved in the catalytic motif NHX₄D of CmATS1 form a catalytic pocket with His-167 and Asp-172. (D) Both the motif and topology of the pocket are strictly conserved in TgATS1 (His-574 and Asp-579). (E, F) IFA shows that TgATS1 is a stromal-resident protein of the apicoplast, as confirmed by co-localization with fluorescence of (E) the chimeric apicoplast stromal FNR-RFP reporter protein co-expressed in the TgATS1-iKO parasite line and (F) anti-CPN60, a known marker of the apicoplast stroma [51]. Scale bars represent 2 μm.

Fig 2. TgATS1 is critical for normal intracellular development and division.

(A) Inducible knockdown of TgATS1 in the TgATS1-HA-iKO line. TgATS1 was detected by Western blot analysis using anti-HA antibody as two bands: the pre-mature form (pATS1, ~75KDa) and the mature form (mATS1, ~55KDa). Protein expression was down-regulated to undetectable levels by 3 or 4 days of ATc treatment (1 mg/mL ATc, numbers indicate days of culture). GRA1 (lower panel) served as a loading control. (B) Plaque assays performed in the absence (-) or presence (+) of ATc and fixed after 10 days show an impaired lytic cycle of TgATS1-HA-iKO parasites in the presence of ATc. (C) IFA of TgATS1-HA-iKO parasites using anti-HA antibody and apicoplast stromal markers FNR-RFP (two upper panels) and CPN60 antibody (lower panel) indicates loss of HA and apicoplast signals in the presence of ATc for 4 days (white arrows indicate normal apicoplast while red arrows indicate loss of apicoplast signal), as well as the cytosolic mis-localisation of apicoplast CPN60 (lower panel, zoomed area). (D) IFA of TgATS1-HA-iKO parasites using antibodies against the apicoplast stromal marker CPN60 and the apicoplast outer membrane marker ATRx1 confirms mis-localisation of CPN60 and ATRx1, indicating loss of apicoplast structure (lower panels, zoomed areas). White arrows indicate normal apicoplast and red arrows indicate a normal apicoplast. (E) Quantification of the number of intact apicoplasts relative to parasites and vacuoles in TgATS1-HA-iKO parasites following ATc treatment. A significant loss of apicoplasts was observed in the presence of ATc (upper graph) of up to 60% at days 5 and 6, compared to the wild type (parental) strain that contained 100% apicoplasts in all vacuoles regardless of the number of parasites per vacuole (lower panel). (n = 100 vacuoles). (F) IFA of TgATS1-HA-iKO parasites using anti-IMC1 antibody grown in the presence and absence of ATc indicates IMC structure defect (zoomed areas). Scale bars: 2.5 μm.

Fig 3. TgATS1 disruption affects tachyzoite division and the morphology of intracellular organelles.

Transmission electron micrographs showing a typical vacuole containing 4 TgATS1-iKO intracellular tachyzoites in the absence of ATc, each bearing normal intracellular organelles such as the mitochondrion (mt) and nucleus (n) (A), with an apicoplast surrounded by 4 membranes as indicated by white stars (B). In the presence of ATc, intracellular development of TgATS-iKO parasites was drastically affected, resulting in parasites bearing aberrant organelles shown by red arrows and large electron lucent regions shown by blue arrows (C). Apicoplast biogenesis was also affected in the presence of ATc, with only a few parasites bearing an apicoplast and, of those present, morphological aberrations were observed, including disorganized membranes and atypical stroma (D). Intracellular division also seemed affected upon TgATS1 disruption (E). Parasites often displayed large electron-lucent vesicles containing an unusual ribbon-like material shown by blue arrows (E, F). Scale bars are indicated in each figure.

Fig 4. ¹³C-carbon source labelling strategies to assess the role of FASII and TgATS1 in T. gondii membrane biogenesis.

TgATS1-iKO parasites were grown in the presence of ¹²C-glucose or U-¹³C-glucose in the presence or absence of ATc. Glucose is metabolized via the parasite glycolytic pathway into dihydroxyacetone-phosphate (DHAP) and phosphoenolpyruvate (PEP). These precursors are imported into the apicoplast via the apicoplast phosphate transporter (APT). PEP can then be transformed into acetyl-CoA via the action of the apicoplast pyruvate kinase (PK) and the pyruvate dehydrogenase complex (PDH). Acetyl-CoA serves as a substrate for the FASII pathway to generate growing acyl chains on an acyl carrier protein (ACP) scaffold. Glycerol 3-phosphate (G3P) can be generated from imported DHAP by the apicoplast glycerol 3-phosphate dehydrogenase (Gpda), or potentially by an as yet unidentified cytosolic G3P dehydrogenase (G3PDH). Both labelled and/or unlabelled G3P and FA can be used as a substrate by the apicoplast TgATS1 (glycerol 3-phosphate acyltransferase) to form lysophosphatidic acid (LPA). These LPA species can potentially be exported from the apicoplast towards the ER to be assembled into phospholipids (PL) and/or have its FA elongated via the three ER elongases (ELO). In order to assess ER elongation, TgATS1-iKO parasites were grown in the presence or absence of ATc together with ¹²C-acetate or U-¹³C-acetate, which is metabolized to cytosolic acetyl-CoA by acetyl-CoA synthetase (ACS). Collectively, this strategy enabled the determination of (i) apicoplast FASII-generated FA products, (ii) products assembled via TgATS1, and (iii) PL species generated via the apicoplast. Total lipid extracts from TgATS1-iKO parasites grown in the described conditions were analyzed by two mass spectrometry techniques. GC-MS was used for the measurement of total FA profiles and LC-MS/MS was used to determine the relative amounts of ¹³C incorporation into individual PL molecular species. In the cartoon, unlabelled and labelled moieties are shown in black and red, respectively.

Fig 5. Analysis of FASII fatty acid biosynthesis and elongation in TgATS1-iKO parasites by metabolic labelling using stable isotope precursors.

Tachyzoites of TgATS1-iKO parasites were labelled with U-¹³C-glucose for 2 or 4 days or U-¹³C- acetate for 4 days in the presence or absence of ATc. Lipids were extracted, derivatized to form fatty acid methyl esters (FAMEs), and analysed by GC-MS to determine ¹³C incorporation. (A, B) ¹³C incorporation into fatty acids from U-¹³C-glucose in the absence (dark colour) or presence (light colour) of ATc for 2 (A) and 4 days (B). (C, D) Mass isotopologue distributions (MID) of FA (C14:0) from U-¹³C-glucose in the absence or presence of ATc for 2 (C) and 4 days (D) (colour scheme as above). The x-axis indicates the number of ¹³C atoms in each FAME, where ‘m0’ indicates the monoisotopic mass containing no ¹³C atoms, while ‘mX’ represents that mass with ‘X’ ¹³C atoms incorporated). MIDs for the all detected FAMEs using U-¹³C-glucose are shown in S5 Fig. (E, F) Changes in the overall abundance of FAMEs for TgATS1-iKO parasites grown in the absence or presence of ATc for 2 (E) and 4 days (F). (G) ¹³C label incorporation rate into fatty acids from U-¹³C-acetate in the absence (dark colour) and presence (light colour) of ATc for 4 days. (H, I) MIDs for C14:0 (H) and C18:1 (I) labelled with U-¹³C-acetate in the absence (dark colour) and presence (pale colour) of ATc for 4 days. MIDs for the all detected FAMEs using U-¹³C-acetate are shown in S6 Fig. Error bars indicate standard deviation (n = 4 biological replicates). Stars represent significant (p < 0.05) differences as determined by t-test, corrected by the Holm-Sidak method.

Fig 6. Metabolic labelling and LC-MS/MS analysis reveal that TgATS1 is responsible for the bulk assembly of PC, PI and PE.

(A, B, C) Representative mass spectra of unlabelled PC (A), PI (B), and PE (C) species from TgATS1-iKO parasites grown in the absence of ATc. PC molecular species were obtained by m/z 184 precursor ion scan in positive mode. PI molecular species were obtained by m/z 241 precursor ion scan in negative mode. PE molecular species were obtained by 141 u neutral loss scan in positive mode. (D, E, F) Representative mass spectra of U-¹³C-glucose-labelled PC (D), PI (E), and PE (F) molecular species from TgATS1-iKO parasites grown in the absence (red line) or presence of ATc (green line) for 4 days. Labelled molecular species displayed typical mass shifts of +3, +15 and/or +17 corresponding to fully-labelled G3P, LPA(12:0) and LPA(14:0), respectively. Under ATc treatment, labelling was greatly reduced in the parasite, especially for +15 and +17 mass shifts. The black line represents a parental unlabelled control. (G-L) LC-MS/MS analysis of individual molecular species and relative abundance of PC (G, H), PI (I, J) and PE (K, L). Parental parasite strain, shown in blue bars, and TgATS1-iKO parasites in the presence of ATc, shown in green bars (n = 3 with error bars representing standard deviation).

Fig 7. U-¹³C-glucose incorporation to fatty acids determines the apicoplast generates the FA moieties for production of most PC and PI molecular species.

U-¹³C-glucose was incorporated into ATc-untreated parasites grown in glucose-free medium. LC-MS/MS analysis of labelled molecular species that have incorporated 4 or more ¹³C-atoms allowed quantification of apicoplast-generated species as shown in white, while all other species (mass shift +3) allowed quantification of non-apicoplast-generated species as shown in black. (A, B, C) Relative abundance for total PC (A), PI (B) and PE (C). (D, E, F) Relative abundance of individual molecular species for PC (D), PI (E), and PE (F). The data shows that the apicoplast generates the FA moieties for production of ~70% PC, ~72% PI, and ~42% PE molecular species. Error bars indicate standard deviation, (n = 3 biological replicates).

Fig 8. Analysis of PA and PC biosynthesis in TgATS1-iKO parasites using U-¹³C-glucose labelling.

Tachyzoites of TgATS1-iKO parasites were labelled with U-¹³C-glucose in the presence (dark red) or absence (light red) of ATc for up to four days. Lipids were extracted, derivatized, and the resulting FAMEs were analysed by GC-MS to determine isotope incorporation. (A, B) Mean label incorporation from U-¹³C-glucose into PA in the absence or presence of ATc for 2 (A) and 4 days (B). (C, D) Mean label incorporation from U-¹³C-glucose into PC in the absence or presence of ATc for 2 (C) and 4 days (D). (E-H) Corresponding MIDs of C14:0 from panels A-D. ‘m0’ indicates the monoisotopic mass containing no ¹³C atoms, while ‘mX’ represents that mass with ‘X’ ¹³C atoms incorporated). (I) Quantification of PA after 4 days growth in the presence (dark red) or absence (light red) of ATc. (J) Abundance of each fatty acid species in PA, presented as a fraction (mol. %) of the total PA fatty acid pool (enlarged to show detail of low abundance FAMEs in (K)), colour scheme as above. For all analyses, error bars indicate standard deviation (n = 3 biological replicates). Stars indicate significant differences (p < 0.05) as determined by t-test, corrected by the Holm-Sidak method. (L) TgATS1-HA-iKO parasite plaque assay in the presence of LPA(C14:0) or LPA(C16:0). LPA(C14:0), but not LPA(C16:0), restored growth in the TgATS1-repressed (+ATc) parasites.

Fig 9. Proposed role of the apicoplast lipid assembly pathway.

The glycolytic intermediates, dihydroxyacetone-phosphate (DHAP), and phosphoenolpyruvate (PEP) are imported into the apicoplast by the apicoplast phosphate transporter (APT) and converted to glycerol 3-phosphate (G3P) and acetyl-CoA, respectively. Acetyl-CoA is used by the FASII to generate, predominantly, C12 and C14 FA chains, which are transferred to G3P by TgATS1 to form lysophosphatidic acid (LPA). These LPA species are exported to the endoplasmic reticulum (ER) by an as yet unidentified transport system to generate bulk PC, PI and PE (shown as head groups with C, I and E, respectively). All PC, PI and PE species could then be elongated by elongase enzymes (ELO) before being exported to contribute to general parasite membrane biogenesis. Note: PI may also me assembled by PI synthase outside of the ER, most likely in the Golgi apparatus.