The Trypanosoma brucei dihydroxyacetonephosphate acyltransferase TbDAT is dispensable for normal growth but important for synthesis of ether glycerophospholipids

Glycerophospholipids are the most abundant constituents of biological membranes in Trypanosoma brucei, which causes sleeping sickness in humans and nagana in cattle. They are essential cellular components that fulfill various important functions beyond their structural role in biological membranes such as in signal transduction, regulation of membrane trafficking or control of cell cycle progression. Our previous studies have established that the glycerol-3-phosphate acyltransferase TbGAT is dispensable for growth, viability, and ester lipid biosynthesis suggesting the existence of another initial acyltransferase(s). This work presents the characterization of the alternative, dihydroxyacetonephosphate acyltransferase TbDAT, which acylates primarily dihydroxyacetonephosphate and prefers palmitoyl-CoA as an acyl-CoA donor. TbDAT restores the viability of a yeast double null mutant that lacks glycerol-3-phosphate and dihydroxyacetonephosphate acyltransferase activities. A conditional null mutant of TbDAT in T. brucei procyclic form was created and characterized. TbDAT was important for survival during stationary phase and synthesis of ether lipids. In contrast, TbDAT was dispensable for normal growth. Our results show that in T. brucei procyclic forms i) TbDAT but not TbGAT is the physiologically relevant initial acyltransferase and ii) ether lipid precursors are primarily made by TbDAT.


Introduction
Trypanosoma brucei is a protozoan parasite of the Trypanosomatidae family, which is responsible for important diseases termed sleeping sickness in humans and nagana in domestic animals in Africa. This microorganism alternates between two hosts during its complex digenetic life cycle, the tsetse fly insect vector and the bloodstream of a vertebrate animal.
T. brucei membranes are made predominantly of glycerophospholipids, which represent 80% of total cellular lipids [1,2]. The main function of glycerophospholipids is structural by PLOS  suggesting that T. brucei possesses an additional initial acyltransferase(s) [31]. Furthermore, T. brucei bears a DHAPAT, which may be one of the alternative initial acyltransferase(s), but its molecular identity has not been established [28,31]. Thus, a DHAPAT ortholog gene was searched in T. brucei genome using L. major LmDAT protein sequence (LmjF. 34.1090) as a bait, which lead to the identification of Tb927.4.3160 that was named TbDAT. The latter encodes a protein of 1242 amino acids with a calculated molecular mass of 137 kDa. TbDAT is much larger than the orthologs of higher eukaryotes such as mouse or fly [25]. The N-terminal domain does not show any similarity to any known proteins as the leishmanial ortholog while the C-terminal domain contains the four typical conserved motifs that are likely involved in catalysis and substrate recognition [25].
T. brucei TbDAT suppresses the lethal phenotype of a S. cerevisiae double null mutant lacking endogenous GPAT and DHAPAT activities S. cerevisiae possesses two initial acyltransferases ScGAT1 and ScGAT2 that have the capability to acylate both DHAP and G3P with similar efficiencies [32,33]. Single deletion of either gene is not deleterious for life but inactivation of both genes is lethal. Thus, TbDAT was expressed under the control of the constitutive promoter ADH1 in a S. cerevisiae conditional double null mutant scgat1Δscgat2Δ [GAL1:ScGAT1] that synthesizes its endogenous ScGAT1 in the presence of galactose but not in the presence of glucose [32]. TbDAT restores normal growth, survival in stationary phase and production of normal migrating lipophosphoglycan of the Leishmania null mutant Δlmdat/Δlmdat The function of TbDAT was further investigated by expressing it in a L. major null mutant Δlmdat/Δlmdat. This strain exhibits a slow growth phenotype, poor survival during the stationary phase of growth and synthesizes an altered form of the ether lipid-anchored virulence factor lipophosphoglycan that migrates slower in a SDS-polyacrylamide gel [25]. Expression of TbDAT in the null mutant Δlmdat/Δlmdat restored normal growth and survival during the stationary phase of growth (Fig 2A) TbDAT encodes a DHAPAT enzyme that exhibits preference for palmitoyl-CoA assays were then performed. TbDAT acylated DHAP with very high efficiency but not G3P (Fig 1C), demonstrating that DHAP rather than G3P is the preferred substrate of TbDAT. The fatty acyl-CoA specificity of TbDAT was further investigated. The preferred acyl-CoA donor was palmitoyl-CoA followed by myristoyl-CoA, palmitoleoyl-CoA and oleoyl-CoA ( Fig  1D). Lauryl-CoA, stearoyl-CoA, linoleoyl-CoA and eicosanoyl-CoA were the least effective fatty acyl-CoA donors.
The fatty acyl-CoA specificity of the DHAPAT activity present in T. brucei cell extracts gave a slightly different pattern than that of TbDAT. The best activity was achieved in the presence of oleoyl-CoA and palmitoyl-CoA, followed by myristoyl-CoA, palmitoleoyl-CoA, linoleoyl-CoA, stearoyl-CoA, eicosanoyl-CoA and lastly myristoyl-CoA ( Fig 1E). This result suggests that TbDAT is not the only DHAPAT enzyme in this parasite and that another DHAPAT enzyme(s) with different fatty acyl-CoA preference may exist. TbDAT localizes to the glycosomes To determine the subcellular localization of TbDAT, the enzyme was tagged with a V5 epitope at its N-terminus. A C-terminal tag was not attempted due to the presence of a putative peroxisomal targeting tripeptide SRM, suggesting that TbDAT resides in the glycosomes, which are peroxisomal related organelles in trypanosomes [36]. Immunofluorescence assay was carried out in the presence of anti-V5 antibodies, which gave a punctated stain within the parasite cell, a signal that overlapped nicely with that of glycosomes revealed with anti-glycosomal specific antibodies (Fig 3; [37]). This assay demonstrates that TbDAT resides in the glycosomes and is consistent with cell fractionation data established previously by Opperdoes and colleagues [26,  Immunofluorescence assay in the presence of anti-V5 monoclonal antibodies (ii) and glycosome specific antiserum [37] (iii). i, DIC; iv, overlap of ii and iii. The bar represents 10 μm. 28,29]. In contrast to other eukaryotes, most of the glycolytic enzymes are compartmentalized in the glycosomes in trypanosomatidae. Thus, the substrate DHAP, resulting from the breakdown of fructose 1,6-bisphosphate by a glycosomal aldolase and triose isomerase, is readily available for the TbDAT enzyme and does not need to be imported into the glycosomes [36,38,39]. In contrast, acyl-CoAs need to be transported into the glycosomes by unknown transport proteins as they are produced in the mitochondria by a type II fatty acid synthase complex and in the endoplasmic reticulum by an unusual elongase system [11,[40][41][42].
A conditional null mutant of TbDAT is viable but survives poorly in the stationary phase of growth To assess the role of TbDAT in parasite's biology, a tetracycline-dependent conditional null mutant was created by successive transformation and selection in the presence of appropriate antibiotics as described in the Methods' section. This strategy was chosen as no null mutant could be obtained by traditional successive deletion of the TbDAT alleles and RNA interference was unsuccessful. The conditional null mutant bears replacement of both TbDAT alleles with antibiotic cassettes BSD, that mediates blasticidin resistance, and PAC, that relieves puromycin sensitivity, and expresses an N-terminal V5-tagged TbDAT under the control of a tetracycline TetO regulated promoter ( Fig 4B). The genotype of the conditional null mutant Δtbdat/Δtbdat/TetO:V5:TbDAT was verified by PCR using appropriate primers ( Fig 4C). Three clones were obtained and showed similar phenotypes. Only one clone is described in details in this work. V5-TbDAT expression is maintained in the presence of doxycycline but it is abolished after 5-6 days in the absence of antibiotic as demonstrated by RT-PCR and Western blot analysis (Fig 4D and 4E). The fact that V5-TbDAT expression took 5-6 days to reach undetectable levels seems to suggest that this enzyme or its mRNA may have a very low turnover or a very long half-life.
The importance of TbDAT in cell viability and growth was then investigated. The conditional null mutant Δtbdat/Δtbdat/TetO:V5:TbDAT exhibited similar growth rate in the absence as in the presence of doxycycline (Fig 5A), demonstrating that TbDAT is dispensable for normal growth. This result contrasts with that of L. major null mutant Δlmdat/Δlmdat, which exhibited a twice as long generation time compared to that of the wild type [25]. However, the conditional null mutant reached only half of the wild-type maximal cell density and survived poorly during the stationary phase of growth (Fig 5B and 5C) during which cells become round and necrotic (Fig 5D). The latter phenotype is reminiscent of that of the L. major null mutant Δlmdat/Δlmdat [25].

A conditional null mutant of TbDAT exhibited decreased DHAPAT activity
The DHAPAT activity of T. brucei conditional null mutant of TbDAT was then quantified. This enzymatic activity was present in cells grown in the presence of doxycycline but was diminished in cells grown in its absence by approximately 70% (Fig 6A), indicating that TbDAT is the main DHAPAT and that another enzyme(s) accounts for the remaining activity in procyclic trypanosomes. This result differs from that of L. major, in which LmDAT was the only DHAPAT enzyme in the parasite [18]. In contrast, GPAT activity levels were comparable in cells grown in the presence or absence of doxycycline, consistent with the fact that TbDAT acylates G3P very inefficiently (Fig 6A).
The fatty acyl-CoA specificity of the remaining DHAPAT activity was then assessed. Best activity was achieved with oleoyl-CoA followed by linoleoyl-CoA, palmitoyl-CoA, myristoyl-CoA, palmitoleoyl-CoA, and stearoyl-CoA. Lauryl-CoA and eicosanoyl-CoA mediated the lowest DHAPAT activity (Fig 6B). This profile is reminiscent of that of the GPAT enzyme TbGAT [31]. Thus, the G3P and DHAP substrate specificity of TbGAT was reinvestigated using Leishmania as a heterologous expression system [31]. Expression of TbGAT in the null mutant Δlmgat/Δlmgat lead to a five-fold increase in GPAT activity but also to a modest but significant increase in DHAPAT activity, suggesting that TbGAT can contribute to the remaining DHAPAT activity in the Δtbdat/Δtbdat/TetO:V5:TbDAT cells grown in the absence of doxycycline (Fig 6C). In addition, TbGAT expression in a Δlmdat/Δlmdat mutant, which lacks DHAPAT activity, conferred slight DHAPAT activity as well (Fig 6D). In terms of substrate specificity, TbGAT resembles S. cerevisiae initial acyltransferases ScGAT1 and ScGAT2 that acylate both G3P and DHAP with similar efficiency [32,33] and differs from L. major LmGAT, which acylates primarily G3P and not DHAP [18]. However, it is not excluded that an additional DHAPAT enzyme may exist in this parasite beside TbDAT and TbGAT. The creation of a double null mutant of TbDAT and TbGAT will give a definitive answer.
T. brucei was shown to also scavenge and remodel lipid precursors from the medium as an alternative way to synthesize its cellular glycerophospholipids [2,8,9,44]. Thus, the importance of lipid scavenging was addressed by growing the conditional null mutant in delipidated medium (Fig 5E). Removal of doxycycline did not affect the growth rate, demonstrating that scavenging and remodeling of extracellular lipids are dispensable for the production of parasite's glycerophospholipids in the absence of TbDAT.

TbDAT is essential for ether lipid biosynthesis
The consequence of a lack of TbDAT expression on ester and ether glycerophospholipid cellular composition was then evaluated. Total cellular lipids were isolated and analyzed by comprehensive mass spectrometry. Levels of ether glycerophospholipids PC, PI, PS, and PE were considerably decreased after doxycycline withdrawal (Table 1). In contrast, amounts of ester lipids were similar when TbDAT is expressed or not expressed, except that PI quantities were slightly but significantly increased in cells grown in the absence of doxycycline, maybe to compensate for the lower levels of the corresponding ether glycerophospholipids. Such a mechanism has been observed in mammalian cells defective in ether lipid metabolism and in T. brucei depleted for the ethanolamine-specific phosphotransferase gene TbEPT, which produced increased levels of ester PE to counteract the lower quantities of ether PE [44,45].
Our relative percentages of the individual glycerophospholipid species differ slightly from those reported in [46]. This may be explained by the different lipid extraction protocols applied to isolate total lipids (Folch based versus Bligh and Dyer) or the use of a different type of instrumentation. Lastly, the medium composition (especially FBS content) may account for these differences as well. Organellar proteomics have established that TbDAT is also expressed in the bloodstream form of the parasite [29]. In addition, ether lipids were detected in lipidomics analyses of bloodstream form Trypanosomes, suggesting that TbDAT is active in this parasite form as well [1,46]. Glycolysis is much more active in the bloodstream form of the parasite, thus producing copious amounts of DHAP in the glycosomes to be readily available for TbDAT. It would make sense that the parasite would use TbDAT rather than TbGAT to produce most of its glycerophospholipids, as G3P is a downstream product of DHAP. Thus, TbDAT is likely to be important as well in the bloodstream form of the parasite. The creation of a (conditional) null mutant in the bloodstream form will assess this hypothesis.
In conclusion, this work shows that DHAPAT activity in the procyclic form of T. brucei is primarily mediated by TbDAT, whose role is chiefly in ether lipid biosynthesis (Fig 7). In the absence of TbGAT however, TbDAT can also contribute to ester glycerophospholipid  production via reduction of acyl-DHAP to acyl-G3P [31]. Our studies also suggest that TbGAT is a minor, alternative DHAPAT enzyme, even though it does not contribute to ether but only to ester glycerophospholipid biosynthesis [31].

Strains and growth conditions
The wild-type strain of T. brucei used in this work is the 29-13 cell line [47]. Parasites were grown in a SD79.3 medium. Delipidated medium was prepared using delipidated fetal bovine serum, which was extracted twice with the same volume of a chloroform/methanol mixture (2:1). The organic solvent's contaminants were removed by blowing N 2 (gas) for 3 hr. Leishmania mutants Δlmdat/Δlmdat and Δlmgat/Δlmgat, and complemented lines Δlmdat/Δlmdat [LmDAT NEO] and Δlmgat/Δlmgat [TbGAT NEO] were described elsewhere and propagated in a liquid M199-based medium in the presence of 20 μg/ml G418, as appropriate [18,31,48]. Growth, transformation, and limiting dilution of Trypanosome and Leishmania parasites were accomplished according to [49]. T. brucei transformants were selected and/or maintained in the presence of 15 g/ml G418, 25 μg/ml hygromycin, 5 μg/ml blasticidin, 5 μg/ml phleomycin, 1 μg/ml puromycin, 1 μg/ml doxycycline, as appropriate. For growth curves, cells from a midlog phase culture were diluted to 1x10 6 /ml and cells were counted with a hemocytometer as a function of time. were described previously [25,32].

Plasmids
The plasmid pBEVY-L-TbDAT (Ec312) for expression in S. cerevisiae was constructed by PCR amplifying the TbDAT gene followed by cloning in sense orientation into the BamHI site of pBEVY-L [50]. The same DNA fragment was also inserted in sense orientation into the BamHI site of pXG1a [51] and pLEW100-V5 (Ec877) to give pXG.TbDAT (Ec516) and pLew100-V5:TbDAT (Ec879) for expression in Leishmania and T. brucei, respectively. The episome pLEW100-V5 was created by amplifying the V5 encoding epitope gene by PCR and cloning into the HindIII site of pLEW100 [47].
For inactivation of the TbDAT alleles, antibiotic resistance cassettes BSD and PAC were flanked by approximately 500 bp of TbDAT UTR regions to give the plasmids pUC.TbDAT: BSD (Ec383) and pUC.TbGAT:PAC (Ec874), respectively. The antibiotic cassettes were then integrated into the TbDAT locus by double crossing over. All PCR-amplified DNAs were verified by sequencing.

Creation of the conditional null mutant Δtbdat/Δtbdat/TetO:V5:TbDAT
The conditional null mutant of TbDAT was created by the subsequent transformation of the TbDAT:PAC and the TbDAT:BSD cassettes that were amplified by PCR using pUC.TbDAT: PAC and pUC.TbDAT:BSD as template, respectively, and oligonucleotides 190 (5'-CCGA AAATGCAAACCACACAC-3') and 192 (5'-TCCTGTAGGTGACAGATAATGG-3'). The PCR product was directly used for transformation of wild-type procyclic form parasites followed by selection in the presence of puromycin. The resistant heterozygous TbDAT:PAC/TbDAT clone was verified by PCR and then subjected to a second round of transformation with pLew100-V5:TbDAT, which was previously linearized with NotI, to give phleomycin resistant TbDAT:PAC/TbDAT/TetO: V5:TbDAT cells. Lastly, the latter strain was transformed with the TbDAT:BSD cassette and parasites resistant to blasticidin, phleomycin and hygromycin were selected. Also, a concentration of 1 μg/mg of doxycycline was applied to maintain V5: TbDAT expression as needed. The proper chromosomal integrations were all verified by PCR (Fig 4C).

Enzymatic assays
Leishmania, S. cerevisiae and Trypanosoma whole cell extracts were prepared as described previously [25]. Protein concentration was determined by the bicinchoninic acid assay using bovine serum albumin as a standard.
DHAPAT and GPAT activity assays were performed as described in [18,25,31]. Briefly, GPAT assay was performed in a buffer system containing 20 mM TrisHCl pH7.5, 4 mM NaF, 1 mM DTT, 2 mM MgCl 2 , 1 mg/ml fatty acid free BSA, 100 μM fatty acyl-CoA, 0.4 mM G3P (including 5.7 μM of [U 14 C]L-G3P; specific activity of 25 mCi/mmol; MP Biomedicals, LLC, Santa Ana, CA), and 200 μg of protein extract in a total volume of 200 μl. The reaction was incubated at 30˚C and stopped after10 min with 700 μl of 1% HClO 4 . The products were extracted with 2 ml of a methanol:chloroform (1:1) mixture and the lower, organic phase was washed with 700 μl of 1% HClO 4 . The organic phase was air dried and the radioactivity was quantified with a scintillation counter.
DHAPAT activity was assessed by measuring the acylation rate of DHAP based on a protocol established by Bates and Saggerson [25,31,52]. First, DHAP was produced by catabolism of fructose 1,6-bisphosphate by the enzymes triose isomerase (28 U; Millipore-Sigma, St Louis, MO) and aldolase (0.44 U; Millipore-Sigma, St Louis, MO) in a buffer system containing 50 mM TrisHCl pH7.5, 0.12 M KCl, 1 mM NaF, 4 mM MgCl 2 , 2 mg/ml fatty acid free BSA, 150 μM fatty acyl-CoA, and 0.5 mM fructose 1,6-bisphosphate (including 0.13 µM of [U-14 C] D-fructose 1,6-bisphosphate; specific activity of 295 mCi/mmol; MP Biomedicals, LLC, Santa Ana, CA). The assay was incubated at 30˚C for 15 min after which 0.1% CHAPS and 1.5 mg of cell extracts were added. The reaction was stopped after 10 min by adding 400 μl of 1% HClO 4 . The samples were then extracted as described above for the GPAT assay. Both enzymatic assays were performed at least twice in duplicate. The incorporated radioactivity was quantified by liquid scintillation counting.

Lipid analysis
Parasites were grown in quadruplicate cultures to late log phase and washed three times with cold PBS. Total cellular lipids were purified according to Schneiter et al. [54]. Lipids were profiled by electrospray ionization tandem mass spectrometry (ESI-MS/MS) using the method described by [34], except that internal standards were (with some small variation in amounts in different batches of internal standards): 0.6 nmol di12:0-PC, 0.6 nmol di24:1-PC, 0. In addition to the scans previously described [34], a scan for PG as [M+ NH 4 ] + in the positive mode with NL 189.0 was performed with collision energy of 20 V, declustering potential of 100 V, and exit potential of 14 V. Ether-linked (alk(en)yl, acyl) lipids were quantified in comparison to the diacyl compounds with the same head groups without correction for response factors for these compounds as compared to their diacyl analogs.
For viability assay, T. brucei cells were labeled with the Zombie Green Fixable Viability Kit (BioLegend, San Diego, CA) and washed as stated by the manufacturer's instructions. Cells were then fixed for 10 min in the presence of 4% paraformaldehyde in PBS and mounted on poly-lysine coated cover slips. Green and red cells were visualized with a fluorescence microscope and quantified with a hemocytometer. At least 200 cells were counted. This assay was performed twice in duplicate.

Statistical analyses
Data are represented as mean ± SE. The significance of means' differences was calculated by two-tailed paired t-test using Microsoft Excel data analysis. Differences were considered significant when the p-value was < 0.05.