L. tarentolae culture

L. tarentolae P10 strain promastigotes (LEXSY host P10, Jena Biosciences) were cultured at 27 °C in ventilated tissue culture flasks in 10 ml of brain heart infusion (BHI) medium containing 37 mg/mL BHI powder (Lexsy Broth BHI, Jena Biosciences) supplemented with 5 μg/mL hemin as described by the manufacturer (Jena Biosciences). Cultures were diluted in 1:10–1:20 with fresh medium to maintain growth in the mid-log phase at 5 x 10⁷ cells/ml.

Generation of L. tarentolae CRISPR-Cas9 cell lines

An L. tarentolae cell line transiently expressing Cas9 and T7 RNA polymerase was generated using the method reported previously [22,23]. Briefly, L. tarentolae P10 in the mid-log phase were placed in BHI media and mixed with plasmid pT007 (~2–5 µg), which encodes for the humanised Streptococcus pyogenes Cas9 nuclease gene and T7 RNA polymerase, in a pre-chilled 4 mm electroporation cuvette (BioRad) on ice. Electroporation was performed using the BTX electroporator ECM630 at 1500 V, 25 μF with a double pulse (10 s intervals) and the cuvette was kept on ice for 10 min before being transferred to 10 mL LEXSY Broth BHI supplemented with 5 μg/mL hemin and 50 μg/mL penicillin-streptomycin (Pen-Strep, Jena Bioscience) and incubated at 27 °C. After 24h incubation, 20 μg/mL hygromycin B was added to the culture for polyclonal selection.

Generation of chromosomally tagged and knockout L. tarentolae cell lines

Tagged strains were generated according to the CRISPR-Cas9 method developed by Beneke, Madden [22]. Constructs for chromosomal tagging of LtVtc1 and LtVtc4 were amplified by PCR using the pPLOT-mNG-Blast, pPLOT-mCherry-Puro, and pPLOT Puro BirA* plasmids, while gene knockouts were performed using the pTpuro plasmid [22]. The generic primer pairs used for LtVtc4 and LtVtc1 were obtained from the LeishGEdit database (www.leishgedit.net, S1 Table). These primers are designed to include ~30 bp homology flanks (HFs), which match sequences flanking the target gene locus. As a result, the PCR product contains a fluorescent tag (with a Myc epitope), an antibiotic resistance cassette, and the necessary homology arms to guide recombination at the intended genomic site. PCR amplification was performed using a two-step protocol. The sgRNA was expressed in vivo from a double-stranded DNA (dsDNA) template containing a T7 promoter, the target-specific sgRNA sequence, and a scaffold region. For transfection, this dsDNA template was delivered alongside the CRISPR repair cassette. N-terminal tagging required a 5′ sgRNA; C-terminal tagging required a 3′ sgRNA; and gene knockout required both sgRNAs and a matching repair cassette containing a puromycin resistance marker (pTPuro plasmid). The sgRNA templates for this CRISPR system were synthesised by PCR from a forward primer that consisted of a T7 promoter sequence followed by the gene-specific target sequence and an overlapping region with the sgRNA scaffold. A generic reverse primer with a 20 bp overlap to the forward primer was then provided, which encoded for the remainder of the scaffold. These short PCR products (ca. 120 bp) were also not purified for transfection into L. tarentolae cells.

Transfection and selection of L. tarentolae

L. tarentolae cells were transfected as described in the LEXSY protocol (Jena Bioscience). Briefly, the L. tarentolae P10 strain was cultured with hygromycin selection over several passages in LEXSY BHI medium and diluted at 1:10 dilution one day before transfection with the pT007 plasmid. We used the BTX electroporator ECM630 at 1500 V, 25 μF with a double pulse in 10 s intervals to transfect the cells with 1–10 μg tagging or knockout cassette and the corresponding sgRNA. The cells were then incubated for 20 h in fresh media at 27 °C. Next, 32 μg/mL hygromycin B (Invitrogen, cat 10687010) and 50 μg/mL penicillin-streptomycin (Pen-Strep, Jena Bioscience) and blasticidin (10 μg/mL, Gibco cat 46–1120) or puromycin (10 μg/mL, Gibco cat A11130-03) were added and the culture was continued until the resistant cells emerged. Successful transfection was confirmed by PCR (primers listed in S1 Table) and by sequencing (Eurofins, GE).

Acidocalcisome isolation/preparation

L. tarentolae BirA*-LtVtc1 and BirA*-LtVtc4 cell lines, as well as control cells lacking BirA*, were grown in 500 ml of BHI medium supplemented with 50 μM biotin (Sigma-Aldrich, B4501). Biotin was added 24 h after inoculation, and cells were harvested 30 h later.

Cells were collected by centrifugation at 2,000 × g for 3 min, and the pellet (~2 g wet weight) was washed with cold isolation buffer (125 mM sucrose, 50 mM KCl, 4 mM MgCl₂, 0.5 mM EDTA, 5 mM DTT, 20 mM Hepes-KOH, pH 7.2) supplemented with Pierce Protease Inhibitor Mini Tablets, EDTA-free (Thermo Scientific, A32955). Lysis was performed by grinding for 60 s with silicon carbide (stored at –20 °C before use). The mixture was resuspended in isolation buffer and centrifuged sequentially at 100 × g for 5 min, 300 × g for 10 min, and 1,200 × g for 10 min at 4 °C) to remove silicon carbide and cell debris. The clarified lysate was then centrifuged at 15,000 x g for 10 min at 4 °C to pellet the organelles. The pellet was resuspended in isolation buffer and applied to the 34% step of the discontinuous density gradients containing 20, 24, 28, 34, 37, and 40% (w/V) iodixanol in isolation buffer. Following the published protocol in [24] with one modification (S1D Fig), we performed single gradient ultracentrifugation instead of two in order to recover not only the acidocalcisome-enriched fraction but also biotinylated proteins from other cellular compartments. The gradient was centrifuged at 50,000 × g for 1 h at 4 °C in an SW Ti rotor. Fractions were collected from the top of the gradient based on their colour and analysed by SDS–PAGE followed by western blotting with anti-mPPase to identify the acidocalcisome-containing fractions (see detailed method below). The rabbit polyclonal anti-TmPPase antibody was generated against purified Thermotoga maritima TmPPase [25].

Western blot analysis

The cells (washed twice in 1 × PBS) or the acidocalcisome fractions were lysed on ice for 30 min in RIPA buffer (150 mM NaCl, 20 mM Tris/HCl, pH 7.5, 1 mM EDTA, 1% SDS, and 0.1% Triton X-100) containing a protease inhibitor tablet (Thermo Scientific A32955). Protein concentration was determined via Bradford assay (M172-L, VWR Life Sciences). Cell lysates were mixed with 4 × Laemmli sample buffer for separation by SDS-PAGE. After SDS-PAGE, the gels were transferred onto nitrocellulose membranes, blocked with 3% bovine serum albumin (BSA) in 1× TBS buffer with 0.5% Tween for 1 h, RT. The blots were incubated with primary rabbit antibodies against TmPPase (1:10,000) [25] and secondary antibodies against anti-rabbit IgG antibody (1:10,000) (sc-2004 by Santa Cruz Biotechnology, INC) or HRP-conjugated avidin (BioRad cat 170–6628) for 1 h. After washing three times with 1× TBS-T, the immunoblots were visualised using Pierce ECL Western blotting substrate according to the manufacturer’s protocol.

Affinity capture of biotinylated proteins and mass spectrometry

BioID pulldown was done as previously described [20]. The acidocalcisome fraction (BirA*-LtVtc1 or BirA*-LtVtc4) was resuspended in 1 mL of lysis buffer (50 mM Tris pH 7.5, 500 mM NaCl, 0.4% SDS, 1 mM DTT). The samples were solubilised in 2% DDM (n-dodecyl-β-D-maltoside) or 2% FC-12 (Fos-choline-12) on ice for 30 min. The samples were centrifuged at 16,000 × g for 10 min at 4 °C, and 100 μL of StrepTactin Sepharose beads (Iba, 2-1201-025) were added to the supernatant and incubated overnight at 4 °C with gentle rotation. The beads were washed three times for 5 min with 1 mL wash buffer (50 mM Tris-HCl pH 7.4, 8 M Urea). The biotinylated proteins bound to the StrepTactin beads were washed eight times with 8 M urea/0.2 M ammonium bicarbonate. Lys-C peptidase was added and incubated o/n at RT in ~4 M urea/0.1 M ammonium bicarbonate. The peptide digests were centrifuged to pellet the beads, and the supernatant containing the released peptides was collected. Then, trypsin solution was added and the mixture incubated for 4 hrs at 37 °C. The resulting peptide mixtures were desalted using reverse-phase C18 tips (Pierce). Peptides were then vacuum-dried and reconstituted in 1% trifluoroacetic acid (TFA) with brief sonication to ensure they were completely resuspended. Samples were analyzed by liquid chromatography–mass spectrometry (LC-MS). The acquired MS data were searched against TriTrypDB databases (www.tritrypdb.org) [26] for L. tarentolae Parrot Tar II 2019 [27], using Thermo Proteome Discoverer 2.5.

The protein list from the mass spectrometry data was screened for overlap and manually checked for homology to T. brucei proteins using the TriTrypDB-8.1 TREU 927 database (https://tritrypdb.org/tritrypdb/app). Their potential subcellular localisation was analysed using the TrypTag database (http://www.tryptag.org/) [28], a trypanosome genome-wide protein localisation resource. Hits localised to small cytoplasmic organelles [29] were further examined using published proteomes of the acidocalcisome [24], glycosome [30, 31], and RNA-granule/polysomes [32] from Leishmania and Trypanosoma. Additional checks were performed to assess potential localisation in the mitochondria [33, 34], nucleus [35], and for polyP binding capacity [36], based on available proteomic datasets. Raw mass spectrometry data were deposited in the PRIDE repository (ProteomeXchange Consortium) under accession number PXD075501 (https://proteomecentral.proteomexchange.org/ui?pxid=PXD075501).

Pulldown assays

L. tarentolae cultures (10 mL) expressing mNG-tagged LtVtc4 and mCh-tagged binding partner candidates were pelleted by centrifugation (4000 × g, 10 min at 4 °C) and washed with 1 × PBS. The cell pellets were lysed by seven freeze-thaw cycles in 100 µL of lysis buffer (50 mM HEPES pH 7.5, 150 mM NaCl, 1 mM IP₆ (Inositol hexaphosphate, Sigma-Aldrich), 1mM DTT, 1 mM PMSF, 2 µg/mL Pepstatin A (Sigma Aldrich, P5318-5MG), and Pierce Protease Inhibitor Mini Tablets, EDTA-free (Thermo Scientific A32955)). Membrane proteins were solubilised by adding 1% DDM, followed by incubation on a roller (1.5 h, 4 °C). After centrifugation (25,000 × g, 1 h at 4 °C) to remove cell debris, the supernatants were incubated on a roller (o/n, 4 °C) with 25 µL of ChromoTek mNeonGreen-Trap Agarose beads (Proteintech, AB_2827593). Additionally, for mNG-LtVtc4 + mCh-LtVBP3, ChromoTek RFP-Trap Magnetic Particles M-270 (Proteintech, AB_2861253) were used. Beads were collected by centrifugation (1000 × g, 1 min) using a spin column, washed three times with 1 mL of lysis buffer containing 0.05% DDM (centrifugation at 1000 × g, 1 min per wash). For magnetic beads, a magnetic stand was used for collection and washing. Finally, beads were resuspended in 50 µl of lysis buffer containing 0.05% DDM. Samples from each purification step were resolved by SDS-PAGE and analysed by in-gel fluorescence using the Sapphire FL Biomolecular Imager (Azure Biosystems). Fluorescence was detected using 488 nm excitation for mNG and 532 nm for mCh, which also partially excites the mNG emission signal due to spectral overlap. As a result, some mNG signals may appear yellowish in the 532 nm channel (see S2 Fig). A 638 nm channel was used to detect the protein marker (PageRuler Plus Prestained Protein Ladder, 26619, ThermoFisher Scientific).

Membrane preparation and purification of the VTC complex from L. tarentolae

For in vitro assessment of polyP polymerase activity, the VTC complex was purified from L. tarentolae expressing N-terminally c-Myc–mNG-tagged LtVtc4. Frozen cell pellets (corresponding to 1 L culture) were thawed and resuspended in buffer A (50 mM HEPES pH 7.5, 300 mM NaCl, 0.3 M sucrose, 1 mM MnCl₂, 1 mM MgCl₂), supplemented with 1 mM DTT, 1 mM PMSF, 2 µg/mL Pepstatin A (Sigma Aldrich, P5318-5MG) and Pierce Protease Inhibitor Mini Tablets, EDTA-free (Thermo Scientific A32955), to a final volume of 50 mL. Cells were lysed by high-pressure homogenisation using an Emulsiflex system. Cell debris was removed by centrifugation (4000 rpm, 10 min at 4 °C), and the clarified lysate was transferred to ultracentrifuge tubes. Membrane fractions were pelleted by ultracentrifugation using a Beckman Ti45 rotor (42,000 rpm for 1 h at 4 °C). The membrane pellet was resuspended in buffer A using a Dounce homogeniser and subjected to a second ultracentrifugation step under the same conditions to further enrich membrane components. The resulting membrane pellet was resuspended in buffer B (50 mM HEPES pH 7.5, 300 mM NaCl, 10% glycerol), supplemented with 1 mM DTT, 1 mM PMSF, 2 mM IP₆ (Inositol hexaphosphate, Sigma-Aldrich), 2 µg/mL Pepstatin A (Sigma Aldrich, P5318-5MG), and Pierce Protease Inhibitor Mini Tablets, EDTA-free (Thermo Scientific A32955), using a Dounce homogeniser.

Membrane fractions were adjusted to a final total protein concentration of 20 mg/mL using solubilization buffer (50 mM HEPES pH 7.5, 300 mM NaCl, 10% glycerol), supplemented with 2 mM IP₆ (Inositol hexaphosphate, Sigma-Aldrich), 1 mM DTT, 1 mM PMSF, Pepstatin A (Sigma Aldrich, P5318-5MG), Pierce Protease Inhibitor Mini Tablets, EDTA-free (Thermo Scientific A32955) and solubilised under native conditions with 1% DDM, followed by incubation on a roller (1.5 h, 4°C). Insoluble material was removed by ultracentrifugation using a Beckman Ti50.2 rotor (33,000 rpm, 1 h at 4 °C). The clarified lysate was applied to anti-c-Myc agarose resin (Pierce Anti-c-Myc Agarose, 20168) using a gravity-flow column. Following binding, the resin was washed with solubilisation buffer containing reduced detergent concentration (0.05% DDM) to remove non-specifically bound proteins.

The VTC complex was eluted using competing c-Myc peptide (Pierce c-Myc Peptide, 20170), and elution fractions were concentrated using a 100 kDa MWCO Amicon Ultra centrifugal filter (Millipore, UFC8100). All purification steps were performed at 4 °C.

Samples from purification were resolved by SDS-PAGE and analysed by in-gel fluorescence using the Sapphire FL Biomolecular Imager (Azure Biosystems), followed by Coomassie Brilliant Blue staining. Fluorescence was detected using 488 nm excitation for mNG. Gel images were acquired using a Bio-Rad Gel Doc EZ Imager. Purified VTC complex was used directly for subsequent polyP activity assays.

PolyP activity assay and urea-PAGE analysis

PolyP polymerase activity of purified VTC complex was assessed using an enzymatic assay followed by urea-PAGE analysis. Reactions were performed in a buffer containing 50 mM HEPES pH 7.5, 300 mM NaCl, 2 mM MgCl₂, 2 mM MnCl₂, and 0.05% DDM. Purified protein (1mg/mL, 10 µL) was mixed with 30 µL assay buffer, and the reaction initiated by the addition of 10 µL substrate mixture containing 25 mM ATP and 25 mM IP₆ (Inositol hexaphosphate, Sigma-Aldrich). Reactions were incubated at 30 °C for 2 h or overnight. PolyP products were analysed by urea-PAGE. Samples were mixed with DNA loading dye (6X TriTrack DNA loading dye, Thermo Scientific, R1161) and separated on pre-run urea gels in 1× TBE buffer at 150 V. Gels were pre-run for 20 min prior to sample loading and electrophoresed for 40 min. PolyP standard (0.1 µg) (Sodium polyphosphate, EMPLURA, 10361–03) was included for size comparison. Following electrophoresis, gels were stained with DAPI-containing buffer (2 µg/mL DAPI in 20% methanol, 5% glycerol, 50 mM Tris pH 10.5) for 1 h in the dark, followed by destaining in buffer lacking DAPI (20% methanol, 5% glycerol, 50 mM Tris pH 10.5) for 1 h in the dark. PolyP was visualised using Bio-Rad Gel Doc EZ Imager based on DAPI fluorescence shift upon binding.

Immunofluorescence microscopy and statistics

Cells were attached to coverslips by centrifugation (3,000 × g, 5 s) and fixed with 4% (w/v) paraformaldehyde (20 min, RT). After fixation, cells were washed with 1× PBS, permeabilised with 0.25% Triton X-100 in 1× PBS (V/V) for 5 min at RT, washed again with 1× PBS, and blocked with 3% BSA in 1× PBS (w/v) (30 min, RT). The coverslips were incubated with primary antibodies (ChromoTek mouse anti-mNeonGreen (1:250, 32F6, Proteintech), Rabbit Polyclonal anti-mCh (1:400, 26765–1-AP, Proteintech)) diluted in 1× PBS in a humidified chamber (1 h, RT), washed three times with 1× PBS, and then incubated with secondary antibodies (CoraLite 488-Conjugated Goat Anti-Mouse IgG(H + L) (1:750, SA00013–1, Proteintech), Goat anti-Rabbit IgG (H + L) Cross-Adsorbed Secondary Antibody, Alexa Fluor 594 (1:1000, A-11012, ThermoFisher Scientific)) diluted in 1× PBS (1 h, RT). After three washes with 1× PBS, the coverslips were rinsed with MilliQ water and mounted on glass slides using Fluoromount G + DAPI (Southern Biotech).

Cells were imaged using a Leica Stellaris 8 confocal microscope with a DMI8 base, operated using LAS X 3.5.2 acquisition software. The objective used was an HC PL APO CS2 40 × /1.25 GLYC (Glycerol Immersion Medium). Fluorescenffce detection was performed using hybrid detectors (HyD) and solid-state lasers, with the following settings: mNG detected using Alexa 488 (HyD X2; emission range: 510–540 nm), DAPI detected using Alexa 405 (HyD S1; emission range: 420–490 nm), and mCh detected using Alexa 568 (HyD S3; emission range: 590–740 nm). The pinhole size was set to Airy 1, and the scan speed was 400 Hz.

Images were acquired with these settings by capturing Z-stacks ranging from 8 to 18 slices, with slice spacing between 2.16 and 6.13 µm and pixel size ranging from 56.78 to 116.44 nm. Images were processed using FIJI (ImageJ) [37], and deconvolution was performed for 10 cycles using the DeconvolutionLab2 plugin [38]. Pearson’s correlation coefficients were calculated using the Coloc2 plugin [37] by analysing Z-stacks. For each condition, 30–50 randomly selected cells were analysed, depending on the experiment. Statistical comparisons of Pearson’s correlation coefficient values were performed in Python using SciPy, statsmodels, and Pingouin packages [39]. One-way ANOVA and Tukey’s honestly significant difference (HSD) post hoc test were used for comparisons.

Gene ontology analysis

GO analysis of biological processes, cell components, and molecular functions was performed on the 412 overlapping LtVtc1 and LtVtc4 interactors using the TriTrypDB GO analysis tool [40] with the complete L. tarentolae Parrot Tar II 2019 (ATCC 30267) genome as the reference. Fisher’s exact test was applied, with a p-value cutoff of 0.01, and significance was defined as Bonferroni-adjusted p-value < 0.01 and Benjamini-Hochberg FDR < 0.01.

Protein structure prediction and analysis

Protein structure predictions were performed using AlphaFold3 [41] and ColabFold for actifpTM analysis [42,43]. Protein transmembrane topology and orientation were predicted using MEMSAT-SVM via the PSIPRED server (https://bioinf.cs.ucl.ac.uk/psipred/) [44,45]. Molecular analysis and molecular graphics were produced in PyMOL (https://pymol.org). Protein sequence similarity searches were performed using BLAST in the UniProt database [46]. Sequence alignment was performed using Clustal Omega [47] and displayed using ESPript 3.0 [48].

Localisation of L. tarentolae Vtc1 and Vtc4 to acidocalcisomes

To investigate the localisation of LtVtc1 and LtVtc4 in L. tarentolae (orthologous to Tb927.7.3900 and Tb927.11.12220 in Trypanosoma brucei, respectively), we tagged both proteins with mNeonGreen (mNG) using CRISPR-Cas9, following the method of Beneke et al. [22], which has been successfully applied in L. tarentolae [23]. PCR analysis showed clear bands at ~2200 bp, corresponding to integration of the mNG-tagging cassette into the respective VTC subunit loci, with no amplification in the wild-type (WT) control, confirming successful chromosomal tagging (Fig 1B). Expression of mNG-LtVtc1 and mNG-LtVtc4 was confirmed by SDS-PAGE followed by fluorescent imaging (Fig 1C). Immunofluorescence microscopy showed that both fusion proteins colocalised with the acidocalcisomal marker mPPase, confirming their acidocalcisomal localisation (Fig 1D). To verify that the VTC complex remains functionally active, the complex was purified from L. tarentolae and its polyP polymerase activity was assessed (S3A-B Fig). PolyP synthesis was confirmed by urea-PAGE analysis, demonstrating that the natively isolated complex retains enzymatic activity (S3B Fig).

Identification of proteins in close proximity to the VTC complex

After confirming the acidocalcisomal localisation of LtVtc1 and LtVtc4, both genes were N-terminally tagged with BirA* to enable proximity labelling (BioID) of potential VTC interactors from the cytoplasmic side. BioID is an established technique that identifies proteins within a ~ 10 nm radius of the bait protein in live cells, offering insights into potential protein-protein interactions [20,21].

Successful integration of the BirA* cassette into the chromosomes and expression of the fusion proteins were verified by PCR and western blot, respectively (S1A-S1B Fig). In the presence of biotin, L. tarentolae cells expressing BirA*-LtVtc1 and BirA*-Vtc4 showed strong enrichment of biotinylated proteins (S1C Fig). Following cell lysis, iodixanol-gradient ultracentrifugation (panel D in S1 Fig) yielded acidocalcisome-enriched fractions (panel E in S1 Fig), with BirA*-LtVtc1 and BirA*-LtVtc4 predominantly localising to fractions 4 and 5, respectively (S1F-G Fig). The slight difference in fractionation is due to variations in the collected gradient volumes.

Acidocalcisome fractions from BirA*-LtVtc1 and BirA*-LtVtc4 cell lines were solubilised with 2% DDM or 2% FC-12, based on detergent solubilisation screening of mNG-LtVtc4 showing that these detergents efficiently extracted more than 50% of the LtVtc4 (S1H Fig). Affinity purification of the biotinylated proteins followed by liquid chromatography-mass spectrometry (LC-MS) resulted in a total of 3,160 protein identifications across four datasets, corresponding to 1,240 non-redundant (unique) protein hits (Fig 2A and S2-S5 Tables). From the BirA*-LtVtc1 samples, 912 proteins were identified in the DDM-solubilised and 752 in the FC-12-solubilised fractions, while BirA*-LtVtc4 samples yielded 929 and 567 proteins, respectively (Fig 2A). As expected, there was substantial overlap between the conditions, with 414 proteins detected in all datasets (Fig 2A).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Number of proteins observed in the proximity-dependent biotinylation experiment.

(A) Venn diagram showing overlap of proteins observed in the four datasets (BirA*-LtVtc1, FC-12; BirA*-LtVtc4, FC-12; BirA*-LtVtc1, DDM; BirA*-LtVtc4, DDM). (B) Predicted localisation of L. tarentolae BioID hits based on the localisation of their T. brucei orthologs in the TrypTag database (http://www.tryptag.org/).

https://doi.org/10.1371/journal.pntd.0014511.g002

To further characterise the overlapping proteins, we performed gene ontology (GO) enrichment analysis using the L. tarentolae Parrot Tar II 2019 genome as reference in the TriTrypDB GO analysis framework [26,40] to identify enriched biological processes and functions among proteins proximal to the VTC complex. This analysis revealed significant enrichment across multiple GO categories (S4 Fig). Among these, ATP metabolic process (GO 0046034), intracellular organelle (GO 0043229) and proton channel activity (GO 0015252) were notably overrepresented in the biological processes, cell components and molecular functions categories, respectively.

In parallel, we assessed the predicted subcellular localisation of the L. tarentolae BioID hits by mapping them to their T. brucei orthologs annotated in the TrypTag database, a genome-wide protein localisation resource for T. brucei [28], together with published organelle proteomes [24, 30-32]. Many of the identified L. tarentolae proteins were assigned to acidocalcisomes based on this orthology-based analysis, whereas others were associated with glycosomes and RNA granules (Fig 2B and S6 and S7 Tables). Some of these non-acidocalcisomal assignments may reflect contamination during acidocalcisome isolation. However, the high number of non-acidocalcisomal interactors, together with the known presence of polyP in other organelles [36], prompted further analysis.

PolyP-binding proteins were identified by cross-referencing the L. tarentolae BioID datasets with proteins previously reported to interact with polyP in trypanosomatids [36]. This analysis identified 50 putative polyP-binding proteins, which were assigned to acidocalcisomes (11%), glycosomes (12%), nucleoplasm (12%), other compartments (20%), and the cytoplasm (45%), with ribosomal subunits being highly represented (S8 Table). GO enrichment analysis further showed that translation initiation factors and ribosomal proteins were overrepresented among the BioID hits.

In addition to the polyP-binding and translation-related proteins, the BioID datasets also contained proteins assigned to RNA granules based on TrypTag localisation data. Seventeen L. tarentolae BioID hits were mapped to T. brucei orthologs predicted to localise to RNA granules, which are cytosolic, non-membrane-bound structures involved in RNA storage, processing, and degradation (S7 Table). Among these, Alba domain-containing protein 3 (LtALBA3, LtaPh_3424900), Poly(A) binding protein 2 (LtPABP2, LtaPh_2500900) and Zinc finger protein family member (ZC3H41, LtaPh_2713900) were consistently detected across all VTC BioID datasets. LtALBA3 and LtPABP2 also showed high peptide coverage (S7 Table), and their orthologs have been reported in the acidocalcisome proteome of T. brucei [24].

Proximal acidocalcisome proteins and novel interactors of the VTC Complex in Leishmania tarentolae

To identify relevant protein interactions within acidocalcisomes, we manually filtered the 1240 unique L. tarentolae BioID hits by mapping them to their T. brucei orthologs and examining their predicted localisation in small cytoplasmic organelles (acidocalcisomes, glycosomes, lipid droplets, and RNA granules) [29] using the TrypTag database (http://www.tryptag.org/) [28]. This analysis identified 103 L. tarentolae proteins whose T. brucei orthologs localised to these organelles (Fig 2B and S9 Table). Among these, 14 proteins had previously been reported to localise in T. brucei acidocalcisomes [6,49,50], including eight proteins consistently detected across all BioID datasets: mPPase, Ca²⁺-ATPase, ZnT, PAT2, and two V-H⁺-ATPase subunits (-A, -c putative subunit), as well as Vtc1 and Vtc4 (Table 1).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Acidocalcisome proteins near the VTC complex.

https://doi.org/10.1371/journal.pntd.0014511.t001

V-H⁺-ATPase is a multi-subunit proton pump consisting of two sub-complexes, the peripheral V1 complex (composed of eight subunits: A-H), responsible for ATP synthesis/hydrolysis, and the integral-membrane V0 complex (composed of five subunits: a, c, c’, c”, d, and e), responsible for proton translocation [51]. In addition to the two V-H⁺-ATPase subunits present in all datasets, subunits D and a occurred in one and three datasets, respectively (Table 1). These findings suggest that the VTC and V-H⁺-ATPase complex are close to each other, although not all subunits may be accessible to BirA* biotinylation. Supporting this, Vtc1 deletion in yeast decreased the levels of several V-H⁺-ATPase subunits and impaired proton uptake activity, indicating that Vtc1 is important in maintaining V- H⁺-ATPase stability [52].

We also detected a magnesium transporter (LtMgT) and an inward rectifier potassium channel (LtIRK) in two L. tarentolae BioID datasets, whereas phosphate transporter 91 (LtPho91) and acid phosphatase (LtAP) appeared only in one dataset (Table 1). The limited detection of these proteins may reflect spatial separation from the VTC complex within the acidocalcisome membrane; for instance, acid phosphatase is known to localise to the acidocalcisome lumen in T. brucei [6], making it unlikely to be directly biotinylated by BirA*.

The complete protein lists from this analysis, including corresponding L. tarentoale gene IDs (LtaPh_) and T. brucei ortholog accession numbers (Tb927), are provided as searchable supplementary tables (S6-S9 Tables).

Further, we assessed the colocalisation of mNG-tagged LtVtc4 with several of these proteins (mCherry (mCh)-tagged LtV-H⁺-ATPase_A, LtPho91, LtCa²⁺-ATPase, LtZnT and LtVtc1) in the acidocalcisome of L. tarentolae cells using immunofluorescence microscopy; LtVtc1 was the positive control. As expected, LtVtc1 colocalised strongly with LtVtc4, with a Pearson’s correlation coefficient (PCC) of 0.87 (Fig 3). LtCa²⁺-ATPase, LtV-H⁺-ATPase_A, LtZnT, and LtPho91 also colocalised with LtVtc4 with PCC of 0.83, 0.87, 0.84, and 0.87, respectively (Fig 3), suggesting similar kinds of proximities.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Colocalisation analysis of L. tarentolae expressing mNG-LtVtc4 and mCh-tagged by immunofluorescence microscopy.

mNG-LtVtc4 colocalises with mCh-LtVtc1 (Pearson’s r = 0.83) as well as with mCh-LtPho91, mCh-LtCa² ⁺ -ATPase, mCh-LtV-H ⁺ -ATPase_A, and mCh-LtZnT, showing Pearson’s correlation coefficients of 0.87, 0.83, 0.87, and 0.84, respectively. A lower degree of colocalisation is observed with mCh-IP₃R (r = 0.67). Scale bar: 5 µm.

https://doi.org/10.1371/journal.pntd.0014511.g003

To test whether these proteins directly interact with the VTC complex, we performed an affinity pulldown assay from cell lysates using ChromoTek mNeonGreen-Trap Agarose beads to capture mNG-tagged LtVtc4. As a positive control, mCh-LtVtc1 was successfully captured with mNG-LtVtc4, as indicated by two strong bands in the bead fraction after washing (Figs 4A and S5). Inositol 1,4,5-triphosphate receptor (IP₃R), which was not identified in the BioID datasets, served as a negative control. Although it colocalised with LtVtc4 in the acidocalcisome (Fig 3), IP3R was not captured by mNG-Vtc4, confirming the absence of a direct interaction (Figs 4B and S6). Surprisingly, none of the acidocalcisome proteins that colocalised with LtVtc4 were captured in the pulldown assay (Figs 4B and S6), suggesting no direct interaction with LtVtc4. By contrast, LtVtc1 and LtVtc4 appear to form a stable, direct complex interaction (Figs 4A and S5).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Pulldown assay of acidocalcisome proteins with mNG-LtVtc4.

(A) mCh-LtVtc1 is captured by mNG-LtVtc4, indicated by two distinct bands in both input and bound fractions. (B) No capture is detected for negative control mCh-IP₃R or for mCh-LtV-H ⁺ -ATPase_A, mCh-LtPho91, mCh-LtCa² ⁺ -ATPase, mCh-LtZnT and mCh-LtmPPase, as indicated by the absence of mCh signal in the bound fraction. Lane labels: I – input; FT – flow-through; W3 – third wash; B – bound proteins. Green bands correspond to mNG-LtVtc4; some I and B lines appear yellowish due to partial mNG detection in the 532 nm channel (see S2 Fig); red bands correspond to mCh-tagged proteins. The fluorescence signal marked by a white asterisk in the gel represents free mCh tag, visible only in the input due to partial degradation.

https://doi.org/10.1371/journal.pntd.0014511.g004

In addition to the previously identified acidocalcisome proteins, all datasets contained two novel, putative membrane proteins, LtaPh_2928300 (13.6 kDa) and LtaPh_3617400 (60 kDa), which we named VTC Binding Partner 1 (VBP1) and VTC Binding Partner 2 (VBP2). Further analysis revealed a third hypothetical protein (LtaPh_2928300, 34.4 kDa), present in three out of four datasets (Table 2), which we named VTC Binding Partner 3 (VBP3). In T. brucei, they are homologous to Tb927.3.3140 (hypothetical protein), Tb927.10.6180 (Fla1-like protein), and Tb927.4.860 (hypothetical protein), respectively.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Novel VTC binding partners identified by BioID of L. tarentolae.

https://doi.org/10.1371/journal.pntd.0014511.t002

These proteins are unique to kinetoplastids and lack orthologs in other organisms (S7 Fig). While the TrypTag database suggests that each ortholog may localise to the acidocalcisome [28], only the LtVPB2 ortholog was found in the T. brucei acidocalcisome proteome [6], with experimental studies supporting its acidocalcisomal localisation [49]. We used immunofluorescence microscopy to confirm colocalisation of mCh-tagged LtVBP1–3 with mNG-LtVtc4 in acidocalcisomes (Fig 5).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. Colocalisation analysis of L. tarentolae expressing mNG-LtVtc4 and mCh-tagged LtVBP1, LtVBP2, and LtVBP3 by immunofluorescence microscopy.

All three proteins show colocalisation with LtVtc4 in acidocalcisomes. Pearson’s correlation coefficients with LtVtc4 are 0.83 for LtVBP1, 0.85 for LtVBP2, and 0.82 for LtVBP3. Scale bar: 5 µm.

https://doi.org/10.1371/journal.pntd.0014511.g005

To evaluate direct binding between these proteins and the VTC complex, we conducted pulldown assays as described above, using mNG-LtVtc4 as bait. SDS-PAGE fluorescence visualisation detected mNG-LtVtc4 together with each mCh-LtVBP in the bound fraction, suggesting interaction (Figs 6A-6C and S8A-S8C). However, LtVBP3 bound LtVtc4 more weakly than LtVBP1 or LtVBP2, with a fraction detected in the flowthrough (Figs 6C and S8C). To further validate this interaction, we performed a pulldown assay using mCh-specific beads with mCh-LtVBP3 as bait and confirmed that mNG-LtVtc4 was captured by mCh-LtVBP3 (Figs 6D and S8D), although a significant portion of mNG-LtVtc4 was also detected in the flowthrough.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. Pulldown assay of novel VTC binding partners with mNG-LtVtc4 in L. tarentolae.

(A-C) SDS-PAGE with fluorescence detection shows capture of mCh-LtVBP1, mCh-LtVBP2, and mCh-LtVBP3 by mNG-LtVtc4 using mNeonGreen-Trap agarose beads, as indicated by signal in the bound fraction. (D) Pulldown of mNG-LtVtc4 with mCh-LtVBP3 using ChromoTek RFP-Trap beads confirms interaction. Lane labels: I – input; FT – flow-through; W3 – third wash; B – bound fraction. Green bands correspond to mNG-LtVtc4; some I and B lines appear as yellowish due to partial detection of the mNG tag in the 532 nm channel (see S2 Fig). Red bands correspond to mCherry-tagged proteins. The fluorescence signal marked by a white asterisk in the gel represents free mCh tag, visible only in the input due to partial degradation.

https://doi.org/10.1371/journal.pntd.0014511.g006

To investigate whether the VTC binding partners influence the localisation of the VTC complex, we analysed LtVtc4 distribution in cells lacking LtVBP1, LtVBP2, or LtVBP3. Knockout transfectant populations without clonal selection (VBP1-Pop, VBP2-Pop, and VBP3-Pop) were generated in a cell line expressing mNG-LtVtc4 (parental cell line) (S9 Fig) and examined for acidocalcisome colocalisation with the acidocalcisomal marker mPPase by confocal microscopy (Fig 7A). In control parental cells, mNG-LtVtc4 strongly colocalised with mPPase (PCC = 0.74), consistent with its acidocalcisomal localisation (Fig 7A). A similar pattern was observed in VBP2-Pop and VBP3-Pop, resulting in PCC of 0.69 and 0.74, respectively (Fig 7A). In contrast, VBP1-Pop cells showed reduced colocalisation between LtVtc4 and mPPase (PCC = 0.63). Interestingly, we also observed partial redistribution of LtVtc4 towards the DAPI-stained kinetoplast region (PCC = 0.52) (Fig 7A). To quantify these observations, we performed statistical analysis on PCC values obtained from 30 to 50 randomly selected cells using one-way ANOVA. This analysis revealed a strong group effect for acidocalcisomal colocalisation of LtVtc4 (F = 35.23, p = 1.03 × 10^-17). Tukey’s post hoc test showed that VBP1-Pop differed significantly from the parental cell line, VBP2-Pop and VBP3-Pop (p < 0.0001), whereas no differences were detected among the parental cells, VBP2-Pop, and VBP3-Pop (Fig 7B).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 7. Localisation of mNG-LtVtc4 in LtVBP knockout population mutants.

(A) Representative fluorescence microscopy micrographs of L. tarentolae cells expressing mNG-LtVtc4 (parental cell line) and in VBP1-Pop, VBP2-Pop and VBP3-Pop. mNG-LtVtc4 colocalises with the acidocalcisome marker mPPase in the parental cell line, VBP2-Pop and VBP3-Pop cells, whereas reduced colocalisation and partial redistribution towards the kinetoplast are observed in VBP1-Pop. Magnified insets highlight the kinetoplast region. Pearson’s correlation coefficients (PCC) for overlapping channels are indicated. Scale bar: 5 µm. (B) Quantitative colocalisation analysis of LtVtc4 localisation. Violin plots show PCC values calculated from 30 to 50 randomly selected cells for each strain. Blue violin plots represent LtVtc4 colocalisation with kinetoplast DNA stained with DAPI; green violin plots represent LtVtc4 colocalisation with the acidocalcisome marker mPPase. Each dot represents one analysed cell. Statistical significance was assessed using one-way ANOVA followed by Tukey’s post hoc test. ****p < 0.0001.

https://doi.org/10.1371/journal.pntd.0014511.g007

Consistent with the immunofluorescence microscopy observations, analysis of LtVtc4 colocalisation with DAPI-stained kinetoplast DNA also showed a group effect (F = 25.04, p = 1.92 × 10^-13), with VBP1-Pop showing higher correlation values than the parental cell line and the other knockout transfectant populations (Fig 7B). These results indicate that LtVBP1 affects the correct localisation of LtVtc4 to the acidocalcisome, while deletion of LtVBP2 or LtVBP3 does not.

Predicted structures of the VTC complex components and their binding partners support complex assembly

Following confirmation of direct interaction by pulldown assays, we used AlphaFold3 [41] to predict the structures of LtVtc1, LtVtc4, and their binding partners, and assessed the potential for complex formation. The transmembrane topology and orientation of each protein were predicted using MEMSAT-SVM via the PSIPRED server [44,45]. AlphaFold3 predicted well-folded structures for all proteins, with most residues showing high-confidence pLDDT scores (pLDDT > 70), indicating good confidence in the model (Fig 8A). The C-terminal TMHs of both LtVtc1 and LtVtc4 have an average pLDDT above 70. In contrast, the N-terminal region of LtVtc1 (aa1–59) is unstructured, with pLDDT scores below 50, indicating very low confidence in the model. Recently, Bertolini et al. [53] suggested that trypanosomatid Vtc4 lacks an SPX domain in its N-terminal cytoplasmic region. In contrast, our AlphaFold3 analysis shows that LtVtc4 contains an SPX domain (S10 Fig) and catalytic domain (CD), as does the equivalent Trypanosoma cruzi Vtc4. Both domains are well-structured with high pLDDT scores (Fig 8A), except for the interdomain loops. The discrepancy in the occurrence of the SPX domain is likely explained by the low amino acid sequence identity between the trypanosomatid SPX domain and yeast ScVtc4-SPX, which is 20.7% for LtVtc4-SPX and 19.5% for TcVtc4-SPX (panel C in S10 Fig), meaning that conventional sequence homology tools are unlikely to identify the similarity.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 8. Novel VTC complex interactors.

(A) AlphaFold3 model of LtVtc1, LtVtc4, LtVBP1, LtVBP2, and LtVBP3. Transmembrane topology and orientation were predicted using MEMSAT-SVM [45]. The colour is based on the pLDDT colouring scheme. N: N-terminal residue, C: C-terminal residue. Red square: CD domain of Vtc4 and dashed-red square: SPX domain of Vtc4. (B) The predicted structure of the C-terminal domain (CTD) of LtVBP3 (magenta) superimposes on the catalytic domain of LtVtc4 (wheat) with an RMSD of 5.85 Å. (C) The predicted structure of the C-terminus domain of LtVBP3 (magenta) on the structure of inorganic polyphosphatase from Escherichia coli (EcygiF) (PDB: 5a60) (green) with an RMSD of 4.13 Å.

https://doi.org/10.1371/journal.pntd.0014511.g008

Unlike LtVtc1 and LtVtc4, all VBPs have only a single-pass TMH, at the N-terminus for LtVBP1 and LtVBP2 and at the C-terminus for LtVBP3 (Fig 8A). On the lumenal side, LtVBP1 has two β-hairpin motifs; LtVBP2 contains a β-propeller domain composed of six four-stranded β-sheets; and LtVBP3 consists of a C-terminal domain structurally similar to LtVtc4-CD with an RMSD of 5.85 Å (Fig 8B) and to the inorganic polyphosphatase from Escherichia coli (EcygiF) [54] with an RMSD of 4.13 Å (Fig 8C), despite very low sequence identity – 10.6% with LtVtc4-CD and 15.4% with EcygiF-NTD, respectively (S11 Fig).

To analyse complex formation between the VTC components with LtVBPs, we used AlphaFold3 to predict all pairwise interactions among the five proteins and evaluated the inter-chain predicted TM-score (ipTM) [55] for each combination (S12A Fig). The LtVtc1-LtVtc4 complex was used as a positive control, as this interaction has been structurally confirmed by the yeast VTC complex [14]. Although the ipTM score for the LtVtc1-LtVtc4 is only 0.62, which is relatively low, the predicted aligned error (PAE) plot reveals consistent low-error regions between subunits, indicating a good intermolecular structural correlation (S12B Fig). Among the predicted complexes, LtVtc4-LtVBP1, LtVtc4-LtVBP2, and LtVBP1-LtVBP3 have ipTM scores of 0.50, 0.52, and 0.66, respectively (S12A Fig). To further assess these interactions, we calculated the actifpTM (actual interface pTM) scores, which focus specifically at predicted interfacial residues [43]. The results show that the actifpTM scores of LtVtc4 in complexes with LtVtc1, LtVBP1, LtVBP2, and LtVBP1 with LtVBP3 are 0.93, 0.91, 0.92, and 0.96, respectively (S12A Fig), which indicates strong confidence in complex formation. Interestingly, the interacting sites of the complexes align with the transmembrane orientation of each protein predicted by MEMSAT-SVM (Fig 8A).

Based on the predicted structures, and consistent with the yeast VTC complex structure [14], LtVtc4 forms a complex with two LtVtc1 subunits via their transmembrane helices (TMHs) at two distinct interfaces: interface 1 pairs TMH1/TMH3 of the first LtVtc1 with TMH1/TMH2 of LtVtc4, whereas interface 2 pairs TMH1/TMH2 of the second LtVtc1 to TMH1/TMH3 of LtVtc4 (Figs 9A and S12C). Although the yeast complex also forms Vtc1-Vtc1 interactions, the predicted LtVtc1–LtVtc1 homodimer scores were low (ipTM = 0.41; actifpTM = 0.40), suggesting a possible false-negative prediction (S12A Fig). Despite that limitation, the model indicates two additional contact sites on LtVtc4. A cytoplasmic helix-turn-helix of the LtVtc4-CD domain engages a short helix in LtVBP1, while the lumenal end of TMH3 contacts the junction between the β-propeller and single TMH of LtVBP2, respectively (Figs 9B-9C and S12D-S12E). Both interfaces are dominated by hydrophobic contacts with a few hydrogen bonds and salt bridges. No direct interface is predicted for LtVBP3; its low ipTM/actifpTM values agree with a model in which LtVBP3 binds LtVBP1 via its single TMH and adjacent loop, rather than LtVtc4 itself, forming a β-sheet with the first β-hairpin of LtVBP1 (Figs 9D and S12F). This suggests that, unlike LtVBP1 and LtVBP2, which bind directly to LtVtc4, LtVBP3 binds indirectly to the VTC complex via its interaction with LtVBP1.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 9. AlphaFold3 prediction of the complex formed by VTC components and VBPs.

(A) Complex prediction of two LtVtc1 (green) with one LtVtc4 (wheat). The zoomed-in inset shows the arrangement of TMHs of each LtVtc1 and LtVtc4 viewed from the lumenal side. H1, 2, 3 = TMH1, 2, and 3, respectively. Red-dashed lines are interfaces separating TMH of LtVtc1 and LtVtc4. (B) Complex prediction of LtVtc4 (wheat) with LtVBP1 (cyan). Red rectangle: CD domain of LtVtc4. Black rectangle: Interaction region between LtVtc4 and LtVBP1. (C) Complex prediction of LtVtc4 (wheat) with LtVBP2 (light blue). Black rectangle: Interaction region between the TMH3 of LtVtc4 with LtVBP2. (D) Complex prediction of LtVBP1 (cyan) with LtVBP3 (magenta).

https://doi.org/10.1371/journal.pntd.0014511.g009