Subcellular localisation of FLAG tagged enzymes of the dynamic protein S-palmitoylation cycle of Trypanosoma cruzi epimastigotes

Dynamic S-palmitoylation of proteins is the addition of palmitic acid by zDHHC palmitoyl transferases (PATs) and depalmitoylation by palmitoyl protein thioesterases (PPTs). A putative PAT (TcPAT1) has been previously identified in Trypanosoma cruzi, the etiological agent of Chagas disease. Here we analyse other 14 putative TcPATs and 2 PPTs in the parasite genome. T. cruzi cell lines expressing TcPATs and TcPPTs plus a FLAG tag at the C terminus were produced for most enzymes, with positive detection by indirect immunofluorescence. Overexpressed TcPATs were mostly found as single spots at the parasite anterior end, while the TcPPTs were dispersed throughout the parasite body.

expressed in different developmental stages of the parasite, with a modified DHYC motif (Batista et al. 2013). Such modified motif is functional in the homologue Akr1p enzyme of Saccharomyces cerevisae (Roth et al. 2002).
It has been recently shown that dynamic protein Spalmitoylation is involved in life cycle progression and virulence in some pathogenic protozoa (Brown et al. 2017). However, no evidence of global PATs or PPTs expression has been yet reported in T. cruzi. Thus, aim of this work was to verify the expression of dynamic protein S-palmitoylation enzymes in T. cruzi Dm28c (Contreras et al. 1988) epimastigote forms. An in silico search for PATs was made in the T. cruzi genomic data base (Tri-trypDB), in parallel with nucleotide BLAST alignment (nBlast-NCBI, Bethesda, MD, USA) of T. cruzi genes with the well characterised S. cerevisiae PAT genes that encode for Erf2 (with DHHC-CRD motif) (Lobo et al. 2002) and Akr1p (with DHYC-CRD motif) (Roth et al. 2002). As a result, 15 PATs genes were found, identical to that formerly identified by Goldston et al. (2014). Size of these genes varied from 768 (TcPAT7) to 2610 (TcPAT1) base pairs and the resulting protein products were between 30 and 95.4 kDa. Sequence identity between the TcPATs was very low, between 14.2% and 26.83%, as assessed using multiple alignment with Clustal Omega (EMBL-EBI, Cambridgeshire, UK). The softwares TM-HMM Server v. 2.0 (Center for Biological Sequence Analysis, CBS, Lyngby, Denmark) and Phyre2 (Kelley et al. 2015) were used to predict transmembrane regions and calculate 3D protein models, respectively. It could be determined that these proteins had three (TcPATs 2 and 6) to seven (TcPAT5) transmembrane domains. By using pFAM software (Sanger Institute, Cambridge, UK) to predict protein domains, it was found that only TcPAT1 had the DHYC motif, while the number of cysteines close to the DHHC/DHYC motif varied from 5 to 9. Only TcPAT4, TcPAT10 and TcPAT14 had both DPG and TTxE structural motifs. On the other hand, TcPATs 5 and 9 had only the DPG motif, while TcPATs 1, 7 and 8 had only the TTxE motif (Fig. 1).
All TcPATs showed similar predicted 3D models, except for TcPAT1 (larger and with ankyrin repeats). TcPPTs 1 and 2 were very different from each other. All 3D models had 100% confidence (Fig. 2).
Aiming to produce transfectant cell lines of T. cruzi epimastigotes expressing TcPATs plus a FLAG tag at the C terminus (FLAGC tagged TcPAT), the genes were amplified using specific primers (Table I) with recombination sites for the Gateway cloning platform (Thermo Fischer Scientific, Waltham, MA, USA) by using the entry plasmid vector pDONR 221 and the destination T. cruzi vector pTcGWFLAGC (Batista et al. 2010, Kugeratski et al. 2015. All genes were cloned, except TcPAT6 and TcPAT1 (already characterised). Three-day-old T. cruzi epimastigotes were transfected with a Gene Pulser XCell BIORAD electroporator (BIORAD Inc., Hercules, CA, USA), selected with 500 µg.mL -1 G418 and maintained with 250 µg.mL -1 of the same antibiotic, as previ- Fig. 1: identification and in silico analysis of Trypanosoma cruzi PATs. Fifteen TcPATs were identified, with predicted molecular mass from 30 to 95.4 kDa. All proteins had transmembrane domains and contained the DHHC motif, with the exception of TcPAT1 (DHYC). S1-SN: membrane crossings; kDa: molecular weight of the predicted protein; TTxE and DPG: structural motifs. Fig. 2: predicted 3D models of Trypanosoma cruzi PATs and PPTs. The software Phyre2 was used. All 3D models had 100% confidence. ously described (Batista et al. 2010). Twelve resistant cell lines could be selected, with the exception of TcPAT4.
For subcellular localisation by indirect immunofluorescence assays (IFA), T. cruzi transfectants were washed twice in PBS, fixed for 10 min with 4% paraformaldehyde, adhered to 0.1% poly-L-lysine coated coverslips, permeabilised with 0.5% Triton/PBS, and incubated for one hour at 37ºC using a mouse anti-flag antibody  Hoechst 33342 (Sigma-Aldrich St. Louis, MO, USA) and the coverslips were mounted with Prolong Gold antifading agent (Thermo Fischer Scientific, Waltham, MA, USA). The slides were observed in a Nikon Eclipse E600 epifluorescence microscope.
As a result, TcPATs 3, 5, 8, 11, 12, 14 and 15 were located as single dots at the anterior region of the parasite, close to the kinetoplast and the flagellar pocket (Fig. 3). Interestingly, most PATs with four transmembrane domains (five out of seven) showed this pattern. The positive reaction was frequently found lateral to the kinetoplast, which suggests Golgi, flagellar pocket or contractile vacuole localisation. TcPAT2 labeling appeared as strong dots distributed throughout the cell body, suggestive of localisation in some cytoplasmic organelle (Fig. 3). TcPAT13 presented a stronger labeling at the perinuclear region (Fig. 3). These patterns were expected, since PATs are usually found at the endoplasmic reticulum, Golgi and plasma membranes (Ohno et al. 2006). No positive reaction was detected for TcPATs 7, 9, 10 (Fig. 3). Transcriptomic data from TritrypDB in-dicate that TcPATs 7 and 9 are expressed in metacyclic trypomastigotes, but not in epimastigotes. Therefore, gene expression of these two enzymes (and possibly also TcPAT10) can be down-regulated in epimastigotes. In summary, these results indicated that at least nine TcPATs could be overexpressed in T. cruzi epimastigotes.
In order to characterise the TcPPTs, a genomic data search was performed as described above, and two genes were identified (Table II). TcPPT1 is an 843 base pairs gene and the product (30.2 kDa) is homologue to H. sapiens acyl-protein thioesterase-1 (APT1) and lysophospholipase genes, which are involved in cytosolic and lysosomal protein depalmitoylation (Long and Cravatt 2011). TcPPT2 is a 951 base pairs gene and the product (35.5 kDa) is homologue to H. sapiens acyl-protein thioesterase-2 (APT2), involved in cytosolic depalmitoylation (Long and Cravatt 2011). Primers were then designed for isolation and amplification of these genes (Table II).
The same steps described above for TcPATs were used to produce T. cruzi cell lines expressing TcPPTs plus a FLAG tag at the C terminus (FLAGC tagged TcPPTs). Resistant cell lines expressing TcPPT1 and TcPPT2 were selected with 500 µg.mL -1 G418. After IFA in the same conditions as described above, both TcPPTs showed strong labeling dispersed through the cell body, suggesting a cytoplasmic localisation (Fig. 3), indicating that T. cruzi epimastigotes overexpressed both TcPPTs, in the expected cytoplasmic localisation.
In conclusion, our data indicate that a dynamic protein S-palmitoylation machinery (nine PATS and two PPTs) could be overexpressed in T. cruzi. Future studies will be crucial to determine the importance of this machinery for the parasite survival. Palmitoylation and depalmitoylation of proteins can play an important role in this parasite, in events as diverse as nutrition, protein traffic, differentiation, host-cell interaction and infection establishment.