Conservation and divergence within the clathrin interactome of Trypanosoma cruzi

Trypanosomatids are parasitic protozoa with a significant burden on human health. African and American trypanosomes are causative agents of Nagana and Chagas disease respectively, and speciated about 300 million years ago. These parasites have highly distinct life cycles, pathologies, transmission strategies and surface proteomes, being dominated by the variant surface glycoprotein (African) or mucins (American) respectively. In African trypanosomes clathrin-mediated trafficking is responsible for endocytosis and post-Golgi transport, with several mechanistic aspects distinct from higher organisms. Using clathrin light chain (TcCLC) and EpsinR (TcEpsinR) as affinity handles, we identified candidate clathrin-associated proteins (CAPs) in Trypanosoma cruzi; the cohort includes orthologs of many proteins known to mediate vesicle trafficking, but significantly not the AP-2 adaptor complex. Several trypanosome-specific proteins common with African trypanosomes, were also identified. Fluorescence microscopy revealed localisations for TcEpsinR, TcCLC and TcCHC at the posterior region of trypomastigote cells, coincident with the flagellar pocket and Golgi apparatus. These data provide the first systematic analysis of clathrin-mediated trafficking in T. cruzi, allowing comparison between protein cohorts and other trypanosomes and also suggest that clathrin trafficking in at least some life stages of T. cruzi may be AP-2-independent.


Results
Isolation of clathrin-interacting proteins from Trypanosoma cruzi. To initiate a systematic and unbiased identification of proteins interacting with the clathrin in T. cruzi we created transgenic epimastigotes harbouring epitope-tagged forms of the clathrin light chain (CLC) and EpsinR, both of which interact with the clathrin heavy chain. Both were tagged at the N-terminus, and expressed in cells as GFP::TcEpsinR or Protein A::TcCLC.
Initially, using TcCLC as affinity handle, coupled with cryomilling, we identified a large cohort of candidate interacting proteins using label-free proteomics. Cryomilling provides a robust method by which one can preserve protein-protein interactions in the cell and has been applied to many organisms and systems (see Obado et al., 2016 for an example in trypanosomes). Analysis of these complexes by 1D SDS-PAGE and visualisation by Silver staining indicated multiple co-isolated proteins (Fig. 1). Significantly, a prominent band was observed at ~200 kDa in the electrophoretogram, and which was subsequently identified as the clathrin heavy chain by Western blotting with monoclonal antibody to TcCHC 21 and subsequently by mass spectroscopy (Fig. 1A, Table 1). Neither TcCLC or TcCHC were detected in control isolates. Following mass spectrometric analysis of these isolations and comparisons with the untagged control, we observed that the affinity-tagged isolations included both conserved and novel clathrin-associated proteins (CAPs) ( Table 1). Similar protein profiles were obtained in two independent immunoprecipitations for TcCLC and three for TcEpsin, indicating that the isolation procedure was reproducible and thus likely robust.
Peptide sequences predicted by MS were used to query the T. cruzi predicted proteome in order to identify proteins that copurified with Protein A::TcCLC. Besides TcCHC (TcCLB.506167.50), over 30 additional proteins were identified (Table 1). Amongst these were TcEpsinR, subunits of the AP-1 and AP-4 complexes and AP180. We applied a cutoff criterion of five-fold greater emPAI score in the test versus the control isolation, together with an exclusion of 0.1 emPAI (see Supplementary data for full MS reporting). The vast majority of proteins was identified in both replicates, with the exception of some low abundance SNARE and Rab proteins and dynamin (TcCLB.508153.20). This latter protein is a frequent contaminant in membrane fractions 32 and whilst it may be involved in endocytic functions, it is unclear from these data.
The second highest ranked protein in the TcCLC isolation was the T. cruzi ortholog of EpsinR. Tagging of this protein with GFP at the N-terminus to produce GFP::TbEpsinR and immunofluorescence using anti-GFP and anti-clathrin heavy chain monoclonal antibody demonstrated significant colocalisation for these two proteins, at the anterior region of the cell and close to the flagellar pocket (Fig. 1B). Whilst the resolution of light microscopy is insufficient to confirm a direct interaction, these data do indicate that TcEpsinR and TcCLC have the potential Scientific RepoRts | 6:31212 | DOI: 10.1038/srep31212 to interact, based on proximity, and provides additional support for this connection. This is also consistent with previous work in T. brucei 18 . Isolation of TcEpsinR-interacting proteins from Trypanosoma cruzi. To strengthen the evidence that the proteins identified by immuno-isolation of TcCLC complexes are genuine clathrin interaction partners, a reciprocal co-immunoprecipitation was performed using GFP::TcEpsinR. Immunoprecipitation of GFP::TcEpsinR using magnetic beads covalently coupled to llama anti-GFP antibody successfully co-precipitated clathrin heavy and light chains from tagged T. cruzi epimastigotes ( Fig. 2A). Again LCMS 2 was used to identify the proteins in these complexes using three replicates, and besides TcCHC (TcCLB.506167.50), over 30 additional proteins were confidently identified ( A cohort of endocytic proteins in T. cruzi. It is significant that a great many proteins identified using GFP::CLC and Protein A::EpsinR were in common (Fig. 3). This orthogonal identification supports the hypothesis that these are indeed bona fide endocytic proteins in T. cruzi. Of these, TcCHC was recovered from all five isolations (two × GFP::CLC and three × Protein A::EpsinR) while TcCLC was also found in all three TbEpsinR isolates. The ortholog of AP180/CALM (TcCLB.503449.30) was recovered from four of five experiments. Together with TcEpsinR these proteins are involved in AP-2-independent clathrin-mediated endocytosis in T. brucei 13 , and the data here suggest a similar configuration in T. cruzi. A clathrin-uncoating protein, the trypanosome auxilin ortholog (TcCLB.510045.30) was also found in four of five independent experiments.
Four candidate clathrin-associated proteins (CAPs) encoded by TcCLB.503595.10 (TcCAP80), TcCLB.507221.70 (TcCAP141), TcCLB.510057.30 (TcCAP37) and TcCLB.507895.170 (TbCAP30) all encode hypothetical proteins (Fig. 3). Apart from a similar structure of predominantly β-sheet at the N-terminus and disordered/α-helical at the C-terminus for TcCAP80 and TcCAP141, these proteins appear quite divergent in secondary structure. All are essentially restricted to trypanosomatids, and even absent from the heterolobosid Naegleria gruberi, a sister lineage (Fig. 3). Orthologs of TcCAP80 and TcCAP141 have also been identified in T. brucei through affinity isolat using the TbCHC as the affinity handle, and mediate endocytosis and morphological features of the flagellar pocket (Manna et al., 2016 submitted), suggesting that this cohort are also likely bona fide players in endocytosis in T. cruzi.

Figure 1. Immunoprecipitation of T. cruzi clathrin-associated proteins (TcCAPs). Panel (A) Protein
complexes isolated by immunoprecipitation from cryolysates of T. cruzi epimastigotes expressing Protein A:TcCLC (+) using Dynabeads M280 coupled to sheep anti rabbit-IgG were resolved by 4-12% gradient SDS-PAG. Wild-type cell lysate (WT) was used as a negative control. Coomassie staining showed the presence of a prominent 192 kDa band (TcCHC), but not in the negative control. Visualization of TcCHC (192 kDa) was by reaction with a monoclonal antibody against TcCHC and the visualization of TcCLC/AC (55k Da) was by reaction with an anti-rabbit secondary antibody, which has affinity for protein A. Panels (B-E) Immunocolocalization of clathrin heavy chain (TcCHC) and TcEpsinR in Trypanosoma cruzi epimastigotes. Nucleus and kinetoplast DNA were stained with Hoechst 33342.Transfected epimastigote expressing EpsinR-GFP incubated with antibody against GFP (TcEpsinR) and TcCHC monoclonal antibody (clathrin). Note co-localization of the GFP and TcCHC signals (D). (E) Differential interference contrast (DIC) image of the parasite body. Scale bar 5 μm. Images are representative of n = 10 cells.  33 . This protein is broadly conserved and present in most kinetoplastids except for the Phytomonas and Leishmania lineages, which significantly also lack the AP-4 complex, evidence that Tepsin is likely also associated with AP-4 in trypanosomatids 34 . In addition, Tepsin represents an additional member of the ANTH/ENTH family of phosphoinositide-binding trafficking proteins, beyond those characterised so far in trypanosomes, i.e. TbEpsinR and TbCALM.

Localisation of TcCAP30. From the TcEpsinR isolation we selected the hypothetical protein
TcCLB.507895.170, on account of its apparent novelty as a candidate clathrin-associated protein in this protozoan, the fact that it has not previously been localied (unlike CAP80 and CAP141, where this has been done in T. brucei (Manna et al., 2016 under revision)) and exclusive presence in trypanosomatids. However, it was more convenient to investigate this protein in T. brucei (Tb927.8.7230: TbCAP30, 30 kDa) bloodstream forms, where clathrin localizes to endomembrane compartments restricted to the region between the kinetoplast and nucleus. As the general organisation of the endosomal system of T. cruzi is similar, we anticipated that bona fide CAP proteins should localize to this region. We determined the location of the gene product TbCAP30 by expression of a C-terminally haemagglutinin (HA)-tagged version of the protein. We verified that the tagged protein had the correct apparent molecular weight (Fig. 2B), and that TbCAP30-HA localized in the region between the nucleus and the kinetoplast, with signal distribution overlapped with TbEpsinR (Fig. 2C). This supports the possibility that TcCAP30 has the potential to interact with clathrin/EpsinR.

Discussion
The surface of infectious organisms forms the interface between the pathogen and host and represents the primary target of immune attack. The trypanosome surface composition 36,37 is highly specialised, and the flagellar pocket constitutes a specific region that facilitates efficient internalization of host macromolecules and restricts access of host immune factors to the exposed, endocytic receptors of the parasite 13,38 . This paradigm is probably common to all pathogenic trypanosomes, but variation in surface molecules indicates fundamental adaptation to the specific demands of the parasite/host interaction. In silico analysis suggests that several major proteins of the endocytic pathway characterised in animals and fungi are absent 16 .
It remains unknown how much diversity is present between the trypanosomatids, but considering the remarkable differences in lifestyles and surface proteins, adaptations are predicted. For example, T. cruzi possesses AP-1 to 4, distinct from Leishmania which lacks AP-4 and T. brucei lacking AP-2. T. cruzi also possesses Rab14, which functions in Golgi to endosome transport 39 and Rab32, which has many roles including phagocytosis 40 ; these are additional to the Rab set shared with T. brucei 41 . Both Rab14 and Rab32 are present in the last common eukaryotic ancestor, suggesting that T. brucei lost these genes, indicating a likely more sophisticated endomembrane system in T. cruzi, and providing evidence for significant divergence. Similar variance has been reported in the Apicomplexa 42 .   We exploited two conserved proteins within the clathrin-mediated transport system of T. cruzi: the light chain of clathrin (TcCLC) and EpsinR (TcEpsin). We identified cohorts of candidate proteins for both TcCLC and TcEpsinR. The clathrin heavy chain (TcCHC) is the most abundant protein 43 and other candidate interacting partners appear to be sub-stoichiometric, similar to CCV isolations from metazoa and trypanosomes, reflecting promiscuity of clathrin interactions 19,44,45 . A range of additional proteins with clear roles in transport also identified.
Surprisingly AP-2 was not present in any of our isolations. While the genes encoding the four subunits of this adaptor complex are absent from the genome of T. brucei 34 , they are present in T. cruzi 20,41 . We predicted AP-2 to be identified, since this complex facilitates clathrin-mediated endocytosis and the pathway is active in T. cruzi epimastigotes 20,21 . Some unicellular organisms, including yeast, can survive without AP-2 46,47 while very rapid neuronal endocytosis is also AP-2 independent 48 . Specific cargo adaptors support clathrin-mediated endocytosis in the absence of AP-2 49 , and therefore, the AP-2 complex is not mandatory. For T. brucei alternate adaptors, such as TbEpsinR and TbCALM, must support clathrin-mediated endocytosis 13 . Since we failed to recover AP-2, but did identify AP-1 and AP-4, this suggests that the result is likely real and unlikely simply failure to maintain clathrin-AP complexes. Therefore the dominant form of endocytosis in T. cruzi may be AP-2 independent, suggesting an unexpected mechanistic similarity to African trypanosomes. This is a surprising finding, potentially unifying AP-2 endocytic mechanisms across a broader range of taxa.
In contrast to AP-2, we recovered all AP-1 subunits with both affinity handles. This complex is mainly associated with transport at the trans-Golgi network and late endosomes in mammalian cells 49,50 and T. brucei 51,52 . It is possible that AP-1 has related functions in T. cruzi, such as targeting lysosomal enzymes like cruzipain and chagasin 53 to reservosomes. It is of interest that cruzipain (TcCLB. 507537.20) was also found and that may represent cargo en route to the lysosome 54 . The precise function of AP-4 is not well defined 55 , but significantly the ε-subunit of AP-4 complex was also identified in a T. cruzi contractile vacuole proteome along with clathrin and AP180 25 , also found here. Significantly, we also recovered Tepsin, a central component of AP-4-containing vesicles 33 . Tepsin and AP-4 have coevolved and organisms lacking AP-4 also lack Tepsin 55 . These data robustly confirm these earlier observations for AP-4. Significantly, we also identified orthologs of TbCAP80 and TbCAP141, recently shown to be involved in endocytosis in African trypanosomes (Manna et al., 2016 under revision), suggesting that these proteins are part of a conserved trypanosome-specific endocytic mechanism.
In conclusion, we report an interactome for clathrin for T. cruzi. The cohort contains many highly conserved members, but also several trypanosome-specific factors. Taken together with recent evidence from African trypanosomes, these data indicate the presence of divergent mechanisms for clathrin function in these pathogenic protozoa.

Cloning and expression of Trypanosoma cruzi TcCLC and TcEpsinR.
To generate transgenic epimastigotes stably expressing TcCLC (Clathrin Light Chain, TcCLB.506211.240) with a protein A and C amino-terminal fusion (TcCLC/AC), TcCLC cDNA was cloned into the pTcGWPTP expression vector. The pTcGWPTP vector encodes proteins A and C and is a modification of the previously described pTcGWGFP vector 58 . The TcCLC gene was used to design forward (5′-ATGGACCCTTTTGAAGGAAGC-3′) and reverse (5′-TTATTGAGCGGTTTCGCCCT-3′) primers flanked by sequences compatible with the Gateway (Invitrogen, USA) cloning platform to enable subsequent subcloning into the target vector. The resulting pTcGWPTP plasmid encoding the TcCLC gene fused to proteins A and C was used to transfect parasites.
T. cruzi transfection. T. cruzi epimastigote cultures were grown at 28 °C in LIT medium supplemented with 10% FBS to a density of approximately 3 × 10 7 cells/ml. Parasites were then harvested by centrifugation at 3,000 g for 5 min at room temperature, washed once in phosphate-buffered saline (PBS, pH 7.2) and resuspended in 0.4 ml of electroporation buffer (140 mM NaCl, 25 mM HEPES,0.74 mM Na 2 HPO 4 , pH 7.5) at a density of 1 × 10 8 cells/ml. Cells were then transferred to a cuvette (0.2 cm gap width) and 10-15 μg DNA was added. The mixture was placed on ice for 10 min and then subjected to two pulses of 450 V/500 μF using the Gene Pulser II (Bio-Rad, Hercules, CA, USA). Following electroporation, cells were cultured in 10 ml LIT medium containing 10% FBS and incubated for 24 h at 28 °C. The antibiotic G418 (500 μg/ml) was then added to the culture medium and stable, resistant cells were obtained approximately 20 days after transfection. Stably transfected cells were maintained in cultures containing 250 μg/ml G418.

One-step PCR-mediated transfection of T. brucei BSF cells for in-situ tagging.
To generate transfected T. brucei BSF cells expressing TbCAP30 with a 3xHA carboxy-terminal fusion (TbCAP30/HA), TbCAP30 cDNA (Tb927.8.7230) was cloned into the pMOTag2H (kindly provided by George Cross, Addgene plasmid #26296) with a puromycin selectable marker and a 3xHA-tag 59 . The TbCAP30 gene sequence was used to design forward ( 5′-GT TG AC GT TG AC CG TG TT TA CG TA CC AG GG AC GG TG GA GG CC GC TA AG GC GC TC GG CA CT TC TG AG AA GC AG GG GT ACAATGCGGTTGTTGGTACCGGGCCCCCCCTCGAG-3′) and reverse (5′-TGCCCATTTCAACCGCTTTCACTGCTTGCCCTTTCCCTTTTCCCCTCTTTCTTTATATAT ATATATATATATCCCCAACCTTCCTCGAAGTGGCGGCCGCTCTAGAACTAGTGGAT-3′) primers. By using one-step PCR the 3′ UTR of the target gene was replaced by a heterologous intergenic region. This replacement directs the correct splicing of the downstream antibiotic resistance marker.
T. brucei transfection: At a cell density of 1.0-1.5 × 10 6 cells/ml, 3.5 × 10 7 BSF were harvested by centrifugation for 10 minutes, 800 g at 4°C. The supernatant was removed and the cells resuspended in 100 μl of Amaxa buffer (Lonza; Basel, Switzerland), mixed with 30 μg of ethanol precipitated linear PCR product and transferred into a sterile cuvette. Electroporation was performed using an Amaxa Nucleofector II (Lonza) as described. Cells were immediately transferred into pre-warmed HMI-9 and cultured at 37 °C to recover. After 6 hours, the antibiotic puromycin (2 μg/ml) was added and the cells were transferred to 24 well plates to enable the isolation of clonal-antibiotic resistant populations after 7-14 days, and were further expanded in continuous presence of antibiotic.

Immunofluorescence in T. cruzi. For colocalisation of endogenous TcCHC (clathrin heavy chain) with
exogenously expressed TcEpsin/GFP in transfected T. cruzi epimastigotes, 3-day-old cells were washed twice with PBS, fixed for 30 min with 4% paraformaldehyde and adhered to poly-L-lysine coated slides and incubated for 1 h at 37 °C with anti-GFP antibody (1:100) and TcCHC monoclonal antibody 21 . After three washes in PBS, the samples were incubated under the same conditions with a secondary Alexa Fluor 488-conjugated goat anti-rabbit antibody (1:600) and an Alexa Fluor 594-conjugated anti-mouse antibody (1:600). A negative control was performed by incubating anti-GFP antibody with wild-type epimastigotes (data not shown). Nuclear and kinetoplast DNA were stained with Hoechst 33342. After extensive washes, the slides were prepared with mounting medium containing N-propyl-gallate as an anti-fade agent. The samples were examined using a Leica SP5 confocal laser-scanning microscope (Leica Microsystems, Mannheim, Germany) at the Microscopy Facility of the Carlos Chagas Institute, Fiocruz-PR. Acquired images were processed for presentation using Adobe Photoshop CS5 (Adobe Systems Incorporated, USA). The slides were dried and mounted with a drop of Vectashield supplemented with 4,6-diamidino-2-phenylindole (DAPI) (Vector Laboratories, USA) to stain DNA. Images were acquired on a Nikon Eclipse E600 epifluorescence microscope with a Hammamatsu ORCA CCD camera and images captured using Metamorph software. Final processing for presentation was done using Adobe Photoshop CS5 (Adobe Systems Inc.).

Cryomilling.
To identify the proteins associated with clathrin in coated vesicles, T. cruzi epimastigotes expressing TcCLC/AC, T. cruzi epimastigotes expressing TcEpsinR/GFP and T. cruzi wild-type epimastigotes were submitted to cryomilling with subsequent immunoprecipitation of associated complexes 32 . This method requires substantial quantities of starting material, but allows retention of protein-protein interactions not otherwise preserved. Briefly, a total of 5 × 10 10 cells were harvested by centrifugation at 3000 g for 10 s and the cells snap frozen in liquid nitrogen and milled using a ball mill in liquid nitrogen (Retsch Planetary Ball Mill PM100, Haan, Germany) to produce a cryogrindate, under essentially native conditions. Reverse co-immunoprecipitation (TcEpsinR/GFP). Immunoprecipitation using a llama polyclonal anti-GFP antibody was performed using T. cruzi epimastigotes expressing TcEpsinR/GFP. For TcEpsinR immunoisolation, 350 μg of cell grindate was ressuspended in CHC Buffer and complexes were subsequently bound to 35 μl of Dynabeads M270 epoxy coupled to llama anti-GFP. After incubation the beads were washed in the same buffer and then processed as described above.

Mass spectrometry. Liquid chromatography tandem mass spectrometry (LC-MS/MS) was performed by
the Proteomic Facility at the University of Dundee. To separate proteins for mass spectrometry analysis, the samples were run 2 cm on a 10% SDS gel (NuPAGE ® Bis-Tris 10% gels, Novex by Life Technologies) in a 1× MOPS SDS running buffer, fixed and stained with Coomassie. The selected 2 cm gel piece was excised and in-gel tryptic digestion (Trypsin, Modified Sequencing Grade, Roche) was carried out for 16 h at 37 °C. Peptides were extracted with 0.1% trifluoroacetic acid in 50% acetonitrile and dried in a SpeedVac. Peptides were then resuspended in 1% formic acid, centrifuged (13,000 rpm, 1 min) and transferred to an HPLC (high performance liquid chromatography) vial. Usually, 5 μl of this suspension was analysed. Samples were analysed using an Ultimate 3000 RSLC nano system coupled on-line to a LTQ OrbiTrap Velos Pro equipped with an Easy-Spray source (Thermo Scientific). Peptides were initially trapped and desalted using an Acclaim ® PepMap100 C18 Nano-trap column (100 μM × 2 cm) with 0.1% formic acid (buffer A). After 3 min, a wash gradient was formed to separate the peptides using a 180 min gradient on an Easy-Spray PepMap RSLC C18 column (75 μM × 50 cm). Samples were transferred to the mass spectrometer via an Easy-Source with the temperature set at 50 °C and a source voltage of 1.9 kV. The mass spectrometer was operated in standard data dependent acquisition mode. Survey full scan MS spectra were acquired with a resolution of 60,000 at m/z 335-1800. The AGC was set to 1 × 10 6 and an ion trap Msn target value of 5000 was used. The top 15 most intense ions were targeted for CID fragmentation (2 Da isolation window), with normalized collision energy of 35% in the linear ion trap. The dynamic exclusion time window was set to 45 sec, with an isowidth of 2 Da. Once part of the mass range has been excluded for the set time it is released again 60 . Lock mass of 445.120024 was enabled for all experiments.
The mass spectra was analyzed using the Mascot search engine tool (Version 2.3.2) (http://www.matrixscience.com/) against the database of protein sequences from T. cruzi UniProt (54,500 sequences) of five different strains of T. cruzi (CL Brener Esmeraldo-like, CL Brener non Esmeraldo-like, Sylvio, Dm28c and Marinkellei). This strategy was used to increase the coverage of identified peptides. The abundance of proteins was deduced from the total number of MS /MS spectra generated from the same related peptides 61 . The approximate relative quantification of these proteins in complex was estimated in label-free mode and through the exponentially modified protein abundance index (emPAI) 62 .
Relative quantitative real time (qRT)-PCR. Total RNA was extracted using the RNeasy mini kit (Qiagen) according to the manufacturer's instructions along with DNase treatment and quantified using a ND-1000 spectrophotometer and Nanodrop software (Nanodrop Technologies). For cDNA synthesis, 2 μg RNA was diluted to 10 μl with diethylpyrocarbonate (DEPC)-treated water and denatured at 70 °C, 5 min. 15 μl of a reaction mix was added (2.5 μl dNTPs (25 mM stock), 5 μl 5× reverse transcription buffer (Invitrogen), 2 μl 100 mM DTT, 0.5 μl RNAseOUT (recombinant ribonuclease inhibitor, 5000 U/μl, Invitrogen), 2 μl oligo dT, (T 30 VN, 10 μM stock) 0.5 μl Superscript II Reverse Transcriptase (200 U/μl Invitrogen), and 2.5 μl DEPC-treated water and incubated at 37 °C for 1 hr, heat-inactivated at 90 °C, 5 min and finally diluted to 200 μl with DEPC-treated water. For qRT-PCR, 5 μl of cDNA was used in a 25 μl reaction including IQ SYBR Green Supermix (BioRad) with 0.4 μM gene-specific forward and reverse primers. qRT-PCR reactions were performed in white thin wall polypropylene multiplate 48-well unskirted PCR plates (BioRad) sealed with microseal 'B' adhesive (BioRad). Reactions were performed in a BioRad MiniOpticon real time PCR detection system and included an initial denaturation at 95 °C for 3 min, 40 cycles of 95 °C 30 seconds, 58°C 30 sec, 72 °C 30 sec (with a signal read at the end of each cycle). In each amplification step, a non-template control was subjected to the reaction to ensure that there was no contamination.