Sphingolipid synthesis and scavenging in the intracellular apicomplexan parasite, Toxoplasma gondii

Abstract

Sphingolipids are essential components of eukaryotic cell membranes, particularly the plasma membrane, and are involved in a diverse array of signal transduction pathways. Mammals produce sphingomyelin (SM) as the primary complex sphingolipid via the well characterised SM synthase. In contrast yeast, plants and some protozoa utilise an evolutionarily related inositol phosphorylceramide (IPC) synthase to synthesise IPC. This activity has no mammalian equivalent and IPC synthase has been proposed as a target for anti-fungals and anti-protozoals. However, detailed knowledge of the sphingolipid biosynthetic pathway of the apicomplexan protozoan parasites was lacking. In this study bioinformatic analyses indicated a single copy orthologue of the putative SM synthase from the apicomplexan Plasmodium falciparum (the causative agent of malaria) was a bona fide sphingolipid synthase in the related model parasite, Toxoplasma gondii (TgSLS). Subsequently, TgSLS was indicated, by complementation of a mutant cell line, to be a functional orthologue of the yeast IPC synthase (AUR1p), demonstrating resistance to the well characterised AUR1p inhibitor aureobasidin A. In vitro, recombinant TgSLS exhibited IPC synthase activity and, for the first time, the presence of IPC was demonstrated in T. gondii lipid extracts by mass spectrometry. Furthermore, host sphingolipid biosynthesis was indicated to influence, but be non-essential for, T. gondii proliferation, suggesting that whilst scavenging does take place de novo sphingolipid synthesis may be important for parasitism.

Graphical abstract

Supplementary Figure 1

Maximum parsimony analyses of Animalae, Fungi, Trypanosomatidae, Plantae and Apicomplexa sphingolipid synthase predicted amino acid sequences. Bootstrap scores >60 indicated. Homo sapiens LPP1 (outgroup) accession number: O14494; T. gondii SLS: TGME49_046490; P. falciparum SMS1&2: PFF1210w and PFF1215w; Arabidopsis thaliana IPCS1-3: At3g54020.1, At2g37940.1, At2g29525.1; T. brucei SLS1-4: Tb09.211.1030, Tb09.211.1020, Tb09.211.1010, Tb09.211.1000; T. cruzi IPCS1&2: Tc00.1047053506885.124, Tc00.1047053510729.290; L. major IPCS: LmjF35.4990; Aspergillus fumigatus AUR1p: AAD22750; Candida albicans AUR1p: AAB67233; Pneumocystis carinii AUR1p: CAH17867; Saccharomyces cerevisiae AUR1p: NP_012922; Schizosaccharomyces pombe AUR1p: Q10142; Caenorhabditis elegans SMS1-3: Q9U3D4, AAA82341, AAK84597; Homo sapiens SMS1&2: AB154421, Q8NHU3; Mus musculus SMS1&2: Q8VCQ6, Q9D4B1.

Supplementary Figure 2

(A) In vitro assay of TgSLS with no donor substrate (−) or PI demonstrated that IPC synthase activity was increased approximately 5-fold in the presence of this donor substrate; (B) In contrast, in an equivalent assay ScAUR1 was non-responsive with the bovine PI utilised. AFU – arbitrary fluorescence units. Mean of 3 independent experiments, standard deviation indicated.

Fig. 1

Identification of a candidate sphingolipid synthase from Toxoplasma gondii (TgSLS). Protein sequence alignment of D1, D2, D3 and D4 from HsSMS1 and 2, PfSMS1 and 2 and TgSLS. The positions highlighted in black are fully conserved; those in dark grey show conservation of strongly similar groups; those in light grey show conservation of weakly similar groups. The 3 residues of the predicted catalytic triad within D3 and D4 are designated by *.

Fig. 2

TgSLS complements a yeast AUR1p auxotrophic mutant. The growth of YPH499 HIS-GAL-AUR1 transformed with pRS426 TgSLS, a positive control (pRS426 ScAUR1) and a negative control (empty pRS426, pRS) was supported on permissive SGR media (galactose as carbon source). In contrast, only pRS426 TgSLS and the positive control could grow on non-permissive SD media (glucose as carbon source).

Fig. 3

TgSLS functions as an inositol phosphorylceramide synthase. HPTLC fractionation of lipids after reaction of CHAPS-washed TgSLS extract with acceptor substrate NBD-C₆-ceramide and either no donor substrate (−) or phosphatidylinositol (PI), phosphatidylethanolamine (PE) or phosphatidylcholine (PC). Only the addition of PI led to a significant increase in the product formation, a species migrating with inositol phosphorylceramide (IPC). A representative image, O was the origin, ceramide (Cer) migrated at the front, markers from extracts of NBD-C₆-ceramide labelled yeast (NBD-IPC) and mammalian Vero cells (NBD-SM).

Fig. 4

Inositol phosphorylceramide identification in Toxoplasma gondii lipid extracts. (A) Selected negative ion m/z 778.5234 (inositol phosphorylceramide, N-hexadecanoyl (N-C16) species, IPC); (B) selected negative ion m/z 747.5652 (sphingomyelin, N-hexadecanoyl (N-C16) species, SM); and (C) selected negative ion m/z 659.5128 (ceramide phosphorylethanolamine, N-hexadecanoyl (N-C16) species, CPE) UPLC–TOF chromatograms of mouse embryonic fibroblast (MEF) host cells and T. gondii lipid extracts. (D) Partial mass spectra (from 590 to 850 amu) corresponding to the 6 and 7.8 min range of a representative chromatogram obtained by UPLC/TOF-ESI(−) analysis of lipid extracts of host cells and T. gondii. Regions amplified (20× or 30×) as indicated. (E) Theoretical mass spectral pattern for the molecular ion region showing an (M–H)-ion for CPE and IPC and (M + HCOO⁻) for SM. CPE and SM were found in both samples, whereas IPC was only identified in T. gondii extracts.

Fig. 5

TgSLS sensitivity to a verified yeast AUR1p inhibitor. (A) Agar diffusion assay of YPH499 HIS-GAL-AUR1 complemented yeast showed that, as expected, ScAUR1 complemented yeast were sensitive to myriocin at 1 mM (Myr) and aureobasidin A at 25 μM (AbA25). In contrast, TgSLS complemented yeast were resistant to AbA at 25 μM and 100 μM (AbA100), but hyper-sensitive to myriocin (1 mM) as evidenced by large zones of exclusion. AGD, the sphingolipid bypass mutant yeast lacking functional SPT and AUR1, acted as the negative control. DMSO was the vehicle control. (B) In vitro assay of the inhibitory effect of aureobasidin A (AbA) on TgSLS demonstrated that the Toxoplasma enzyme is only marginally, but significantly (p < 0.01), sensitive to the drug at a high concentration (100 μM). Fluorescence intensity of IPC was established following fractionation by HPTLC and normalised with respect to an untreated control. Mean of 3 independent experiments, standard deviation indicated.

Fig. 6

Analyses of the role of host serine palmitoyltransferase (SPT) in Toxoplasma gondii proliferation and invasion. Cells were cultured in 10% FCS (A) or in serum-reduced media (B) at the non-permissive temperature (39 °C). In the presence of complete media, T. gondii proliferation was the same in wild type (CHO-K1) and SPT-compromised (SPB-1) host cells (A). However, in serum-reduced media proliferation was significantly (p < 0.001) decreased in SPB-1 cells compared to the control (CHO-K1 and SPB-1 cLCB1) lines (B). All results normalised with respect to proliferation in parental CHO-K1 cells. Analyses of 3 independent experiments performed in triplicate, standard deviation indicated. This effect was not due to significant differential invasion of the host cell lines (p > 0.1; C). Analyses of 3 independent experiments, standard deviation indicated. To analyse the effect of chemical inhibition of SPT, cells were cultured in serum-reduced media at the non-permissive temperature (39 °C) in the presence or absence of the inhibitor, myriocin (D). Myriocin treatment reduced Toxoplasma proliferation (p < 0.05) in wild type (CHO-K1) cells to similar levels to those seen in untreated SPT-compromised (SPB-1) host cells. The compound had no effect on proliferation in SPB-1 cells but exerted a similar effect to wild type in complemented mutant cells (SPB-cLCB1; p < 0.05). All results normalised with respect to proliferation in parental CHO-K1 cells. Analyses of experiments performed in triplicate, standard deviation indicated.