A novel phospholipase from Trypanosoma brucei

Abstract

Phospholipase A(1) activities have been detected in most cells where they have been sought and yet their characterization lags far behind that of the phospholipases A(2), C and D. The study presented here details the first cloning and characterization of a cytosolic PLA(1) that exhibits preference for phosphatidylcholine (GPCho) substrates. Trypanosoma brucei phospholipase A(1) (TbPLA(1)) is unique from previously identified eukaryotic PLA(1) because it is evolutionarily related to bacterial secreted PLA(1). A T. brucei ancestor most likely acquired the PLA(1) from a horizontal gene transfer of a PLA(1) from Sodalis glossinidius, a bacterial endosymbiont of tsetse flies. Nano-electrospray ionization tandem mass spectrometry analysis of TbPLA(1) mutants established that the enzyme functions in vivo to synthesize lysoGPCho metabolites containing long-chain mostly polyunsaturated and highly unsaturated fatty acids. Analysis of purified mutated recombinant forms of TbPLA(1) revealed that this enzyme is a serine hydrolase whose catalytic mechanism involves a triad consisting of the amino acid residues Ser-131, His-234 and Asp-183. The TbPLA(1) homozygous null mutants generated here constitute the only PLA(1) double knockouts from any organism.

Fig. 1

Multiple sequence alignment of trypanosome and bacterial PLA₁. Amino acid sequences were aligned with MUSCLE (Edgar, 2004) and edited using the BOXSHADE algorithm. Amino acid residues below circles represent those which were mutated to reveal active-site components (closed circles) and non-active-site residues (open circles). The lipase consensus pattern is underscored by a double line and the signal sequence utilized for PhlA secretion (Givskov et al., 1988) is underlined. TcPLA₁, Trypanosoma congolense PLA₁ (product of GeneDB Systematic name congo50h08.q1k_9); TbPLA₁, Trypanosoma brucei PLA₁ (Accession Number CAG29794); TvPLA₁, Trypanosoma vivax PLA₁ (product of GeneDB Systematic name Tviv1355g06.p1k_2); SgPLA₁, Sodalis glossinidius PLA₁ (Accession Number Q2NQT2); YplA, Yersinia enterocolitica PLA₁ (Accession Number O85477); PhlA, Serratia liquefaciens PLA₁ (Accession Number P18952).

Fig. 2

Purification and regiospecificity of recombinant TbPLA₁. A. SDS-PAGE analysis of proteins from TbPLA₁-overexpressing E. coli cells during various purification steps as compared with molecular weight standards (lane 1). The various fractions include insoluble proteins (lane 2), the soluble protein fraction containing His·tagged recombinant TbPLA₁ (lane 3), nickel column flow through (lane 4), nickel affinity eluate (lane 5), purified His·tagged recombinant TbPLA₁ after anion exchange (~3 μg) (lane 6), purified His·tag-cleaved TbPLA₁ after thrombin digestion and anion exchange (~3 μg) (lane 7). B. Phospholipase activity was measured at various pH values using the fluorescent BODIPY®C₁₁-PC assay at a mole fraction of 0.018 with a final concentration of substrate at 0.075 mM. A universal buffer containing 50 mM Tri-sodium citrate, 50 mM Tris and 50 mM NaCl was used to obtain the various pH conditions. Data points represent the activity measured relative to the maximal pH-activity value of pH 7.0. Data represent the average from two experiments performed in triplicate. C. Metabolism of GPCho(16:0/[³H]18:0) after a 5 min reaction is shown as a function of varying amounts of recombinant TbPLA₁. Formation of lysoGPCho(−/[³H]18:0) appears to be dose-dependent, whereas [³H]18:0 acid cleavage from the sn-2 position is undetectable.

Fig. 3

Probing for active-site residues. Mutant TbPLA₁ proteins were expressed and purified from E. coli in parallel as stated in the Experimental procedures and 10 μl of the final eluate was analysed by SDS-PAGE (above graph). The radiolabelled PLA₁ assay was used with 5 ng of purified WT and mutant TbPLA₁ to compare substrate hydrolysis after 10 min; a no-enzyme negative control was tested in parallel. Mutant activities are shown as relative percentages of WT activity (26 μmol min⁻¹ mg⁻¹).

Fig. 4

Validation and characterization of TbPLA₁ mutant cell lines. A. TbPLA₁ null mutants were generated in the procyclic form of T. brucei after sequential TbPLA₁ UTR-targeted homologous recombination with UTR-flanked puromycin and blasticidin resistance genes. Cell lines were analysed by Southern blot after digesting genomic DNA with EcoRV. EcoRV-digested T. brucei WT gDNA reveals one size of DNA fragment detectable by fluorescein-labelled TbPLA₁ ORF probe. In TbPLA₁ null mutant cells (Δpla1) both TbPLA₁ alleles are absent. A tetracycline (Tet)-inducible myc-tagged recombinant ectopic copy of TbPLA₁ (PLA1-myc) cloned into a phleomycin-resistant pLEW100 cassette was introduced into a different locus in WT gDNA to produce transgenic TbPLA₁ overexpression cell lines (ovexPLA1-myc), or introduced in the null mutant gDNA to produce a rescue cell line (Δpla1 rescPLA1-myc). B. Northern blot analysis of TbPLA₁ mutants. Total RNA extracted from WT and TbPLA₁ mutant trypanosomes were hybridized with TbPLA₁ (top panels), loading controls are also presented (bottom panels). C. Western blot analysis of cell lysates of TbPLA₁ mutants. Both native PLA₁ (theoretically 32.4 kDa) and tagged PLA1-myc (theoretically 33.9 kDa) proteins were detected by antibodies raised against the purified recombinant enzyme. * = background bands that show loading in the null mutant lanes.

Fig. 5

TbPLA₁ functions to synthesize long-chain unsaturated lysoGPCho intermediates. A. Positive ion ESI-MS-MS spectrum of m/z 184 lipid precursors in total lipid extracts from T. brucei WT cells. Identities of the major [M+H]⁺ ions are indicated as well as internal standard lysoGPCho [M+H]⁺ peaks. B. Positive ion ESI-MS-MS short-range spectra of the parents of the m/z 184 ion from total lipid extracts from WT and TbPLA₁ mutant cell lines. The major sets of lysoGPCho [M+H]⁺ metabolites detected are boxed and annotated next to the m/z peak from which they are derived. The numbers in place of x:y refer to the total number of sn-2 FA carbon atoms (x) and their degree of unsaturation (y). Panel insets are similar short-range spectra from lipid extracts spiked with lysoGPCho internal standards used for quantitative purposes (Table 2).

Fig. 6

Phospholipase activity from lysates of T. brucei cells. PCF phospholipase activity from 10⁶ cell equivalents was examined after 20 min incubation with GPCho(16:0/[³H]18:0). The TLC autoradiograph shows the metabolism of GPCho(16:0/[³H]18:0) into lysoGPCho(−/[³H]18:0) by PLA₁ in WT and single knockout (Δpla1−/+) lysate. PLA₁ activity is absent in the cell-free control and the Δpla1 double knockout cell line. To stimulate any possible PLA₂ activity in the lysates 10 mM CaCl₂ was added to each assay. Even with the dominant PLA₁ activity being absent from Δpla1 clones, no PLA₂ activity was observed in their lysate. Free [³H]18:0 was used as a standard to monitor PLA₂ activity and its R_f is indicated. O, origin; F, front.

Fig. 7

Cellular distribution of TbPLA₁. A. Subcellular fractions were prepared as described in Experimental procedures and used in immunoblotting with anti-TbPLA₁ antibodies. TbPLA₁ is constitutively expressed and was detected in the cytosolic fraction of both BSF and PCF WT cells (lane 3 and lane 4 respectively) but absent from PCF TbPLA₁ null mutants (lane 5). TbPLA₁ is also absent from the large granular and small granular (microsomal) fractions from BSF WT cells (lane 1 and lane 2 respectively). B. PCF TbPLA₁ null mutant cells transformed with an ectopic copy of an N-terminal EGFP-PLA1 gene fusion (Δpla1 EGFP-PLA1-myc) or a C-terminal PLA1-EGFP gene fusion (Δpla1 PLA1-EGFP) were analysed after tetracycline induction by anti-TbPLA₁ (top panel) or anti-EGFP (lower panel) immunoblotting for the presence of expressed fusion protein (~60 kDa). WT cells and null mutants transformed with an ectopic copy of EGFP only (Δpla1 EGFP) serve as negative controls for fusion protein expression. * = background bands just smaller than native TbPLA₁ that show loading in the null mutant lane. C. Analysis of the subcellular localization by immunofluorescence microscopy of both PLA1/EGFP fusion proteins (right panels) compared with EGFP only confirms a cytoplasmic distribution for soluble TbPLA₁. PCF transgenic mutants are also shown in phase contrast (left panels) and after DAPI staining (middle panels).

Fig. 8

Proposed HGT point for S. glossinidius PLA₁. The arrow indicates the proposed acquisition of S. glossinidius PLA₁ by an ancestor of T. brucei. The maximum likelihood tree was calculated using the Phylip v3.6 Dnamlk algorithm and based on a MUSCLE alignment of glycosomal glyceraldehyde 3-phosphate dehydrogenase (gGAPDH) genes from various kinetoplastids. Values at nodes are maximum likelihood bootstrap values (%) obtained from 100 replicates (Seqboot). The tree is rooted to the corresponding gene of Euglena gracilis. The tree phylogenies agree very well with the more detailed and exact gGAPDH trees previously published (Hamilton et al., 2004; 2005). The gGAPDH sequences were obtained from GenBank and their accession numbers are: Trypanosoma theileri, AJ620282; Trypanosoma lewisi, AJ620272; Trypanosoma cruzi, AJ620269; Trypanosoma rangeli, AF053742; Trypanosoma simiae, AJ620293; Trypanosoma congolense, AJ620291; Trypanosoma godfreyi, AJ620292; Trypanosoma brucei, AJ620284; Trypanosoma vivax, AJ620295; Crithidia fasciculata, AF047493; Leishmania major, AF047497; Euglena gracilis, L21903.