Mouse fibroblasts cell lines, derived from WT, LAMP-1/2^−/−, or LAMP-2^−/− knockout C57BL/6J mice, were obtained from a collection of cell lines from Dr. Paul Saftig’s laboratory (Biochemisches Institut/Christian-Albrechts-Universität Kiel, Germany), which were previously generated by spontaneous immortalization of primary fibroblasts in culture around passages 10–20 (Eskelinen et al., 2004). The cells were maintained in high-glucose Dulbecco’s modified Eagle’s medium (DMEM) (Thermo Fischer Scientific) supplemented with 10% fetal bovine serum (FBS), 1% penicillin/streptomycin (100 U/ml and 100 μg/ml, respectively) and 1% glutamine (DMEM 10%).

Tissue culture trypomastigotes from the T. cruzi Y strain were obtained and purified from the supernatant of infected LLC-MK2 monolayers, as described previously (Andrews et al., 1987), and used to infect wild-type or LAMP-deficient fibroblasts.

For invasion assays, 5 × 10⁴ cells (WT fibroblasts or L6 murine myoblasts) in high-glucose DMEM 10% were plated per well on a 24-well plate, containing 13-mm round glass coverslips. Cells were plated 24 h before the experiment and incubated at 37°C and 5% CO₂. Cells were then exposed to T. cruzi Y strain trypomastigotes obtained from wild-type (TcY-WT), LAMP-2-deficient (TcY-L2^−/−), or LAMP-1- and LAMP-2-deficient (TcY-L1/2^−/−) fibroblasts for 40 min at 37°C at a multiplicity of infection (MOI) of 50. After parasite exposure, the monolayers were washed four times with phosphate-buffered saline containing Ca²⁺ and Mg²⁺ (PBS^+/+), in order to remove the non-internalized parasites, and fixed in 4% paraformaldehyde overnight. After fixation, cells were processed for immunofluorescence.

After fixation, coverslips with attached cells were washed three times in PBS, incubated for 20 min with PBS containing 2% bovine serum albumin (BSA) and processed for an inside/outside immunofluorescence invasion assay as described previously (Andrews et al., 1987). Briefly, extracellular parasites were immunostained with rabbit anti-T. cruzi polyclonal antibodies in a 1:500 dilution in PBS/BSA for 1 h at room temperature (RT) and washed and labeled with Alexa Fluor-546 conjugated anti-rabbit IgG antibody (Thermo Fischer Scientific) in a proportion of 1:500 in PBS/BSA for 45 min. After that, the DNA of host cells and parasites was stained for 1 min with DAPI (4′,6-diamidino-2-phenylindole, dihydrochloride, Sigma), 0.1 μM in PBS, mounted, and examined on a ZEISS Axio Vert.A1 microscope.

The assays followed the protocol previously described by Schenkman, Diaz, and Nussenzweig (1991) (Schenkman et al., 1991). Briefly, WT fibroblasts or L6 myoblasts were plated in each well of a 24-well plate, containing 13-mm round glass coverslips, 24 h before the experiment and incubated at 37°C and 5% CO₂. Cells were then washed with Hank’s balanced salt solution (HBSS, Sigma), pre-fixed with 2% glutaraldehyde in PBS^+/+ for 5 min at 4°C and incubated for 24 h in 0.16M pH 8.3 ethanolamine solution. Before parasite exposure, cells were washed two times with DMEM containing 0.2% BSA. Cells were exposed to TcY-WT, TcY-L2^−/−, and TcY-L1/2^−/− trypomastigotes at 37°C and 5% CO₂ for 40 min at an MOI of 50. After exposure, cells were washed with PBS, fixed with 4% paraformaldehyde (PFA), and stained using rapid panotic staining. Attached parasites were counted using a microscope.

WT fibroblasts or L6 myoblasts were cultured in a 24-well plate (5 × 10⁴ cells/well) at 37°C in 5% CO₂ in DMEM 10%. After 24 h, cell monolayers were exposed to DMEM 10% containing PI (10 μg/ml) in the presence or absence of TcY-WT, TcY-L2^−/−, or TcY-L1/2^−/− trypomastigotes for 30 min at 37°C at an MOI of 100. Alternatively, cells were exposed to parasites for 40 min, washed three times with PBS^+/+, and then exposed to PI staining for another 40 min. All cells were washed twice with PBS^+/+, trypsinized, and analyzed by flow cytometry (FACS Scan; Becton Dickinson). The data were analyzed using FlowJo v10.1 software (Tree Star, Inc.).

For the calcium signaling assays, 2.5 × 10⁴ cells (WT fibroblasts or L6 myoblasts) in high-glucose DMEM 10% were plated in each well of four-well confocal glass-bottom dishes and, 24 h later, incubated with a fluorescent calcium probe [5 μM Fluo-4/AM (Thermo Fisher), 0.01% pluronic acid and 2.0 mM probenecid (Sigma), for 50 min at 37°C (Luo et al., 2011). Afterwards, cells were exposed to the following conditions: (1) culture medium; (2) ionomicin (5 μM); and (3) TcY-WT, TcY-L2^−/−, or TcY-L1/2^−/− trypomastigotes at an MOI of 100. Cells were analyzed in a Nikon C2 Plus confocal microscope using the time lapsing resource and photographed for 600 s without intervals. The changes observed for the intracellular calcium concentration were expressed as the probe intensity fluorescence fold change accordingly with the equation (F/F₀), F being the maximum fluorescence intensity observed and F₀ the basal fluorescence observed for the point 0 (Pinto et al., 2013; Oliveira et al., 2015).

To evaluate proteolytic activity in TcY-WT, TcY-L1/2^−/−, and TcY-L2^−/− trypomastigotes, cellular extracts were prepared using 5 × 10⁸ trypomastigotes (70% pure regarding amastigote forms contamination) in lysis buffer [5 mM Tris–HCl pH 8.0, 10% glycerol, 1 mM EDTA (Sigma), 1 mM DTT (Promega), 10 mM cOmplete Mini EDTA-free protease inhibitors (Roche)] and protein concentration quantified by Bradford’s method (Bradford, 1976). Gelatin zymography assays were performed as previously described (Leber and Balkwill, 1997) with few adaptations. Briefly, 12.5 μg of total protein were mixed with loading buffer without reducing agents and electrophoresed through a 12% polyacrylamide gel copolymerized with gelatin (Sigma) at a final concentration of 1 mg/ml. Gels were soaked in renaturing buffer [50 mM Tris–HCl, pH 7.5; 2.5% Triton X-100 (Sigma)] for 30 min at (RT), followed by one wash for 30 min at (RT) in development buffer (50 mM Tris–HCl, 200 mM NaCl, 5 mM CaCl₂, pH 7.5) and incubated for 24 h in development buffer at 37°C. Gels were stained with R-250 Coomassie brilliant blue (Sigma). Gelatinolytic activities were detected as pale bands against the dark blue background. Band intensities were quantified using ImageJ software and expressed as the pixels fold change regarding TcY-WT pixels read mean. The molecular weight standard was purchased from BioRad Prestained SDS-PAGE Standards, Broad Range.

To evaluate the cysteine and serine proteolytic activities in TcY-WT, TcY-L1/2^−/−, and TcY-L2^−/− trypomastigotes, cellular extracts were prepared as described above and 5 μg of total protein extract mixed with 50 μM N-acetyl-Tyr-Val-Ala-Asp-7-amido-4-trifluoromethylcoumarin (Ac-YVAD-AFC #A9965, Sigma-Aldrich) caspase-like or 13 μM N-succinyl-Leu-Leu-Val-Tyr 7-Amido-4-Methylcoumarin (Suc-LLVY-AMC, #S6510—Sigma-Aldrich) chymotrypsin-like fluorogenic substrates. The reactions were performed in buffer containing 50 mM Tris–HCl, pH 8.0, 10 mM MgCl₂, and 1 mM DTT, in the presence or absence of 20 μM MG-132 (#M7449, Sigma-Aldrich) inhibitor, and incubated at 37°C from 0 to 120 min with readings every 20 min. The readings were performed in the Synergy 2 Multi-detection Microplate (BioTek Instruments, Inc.) with excitation wavelength of 380/20 and emission wavelengths of 508/20 or 440/40 for Ac-YVAD-AFC or Suc-LLVY-AMC substrates, respectively.

To obtain the membrane proteins fraction, about 1 × 10⁸ trypomastigotes forms released from wild-type or LAMP-deficient fibroblasts, presenting about 70%–80% purity regarding amastigote forms contamination, were centrifuged during 10 min at 2,200g at 4°C and washed two times with PBS (pH 7.2). Samples were resuspended in 0.1M sodium carbonate lysis buffer (pH 11.5) supplemented with protease inhibitor cocktail (Protease Inhibitor Mix, GE Healthcare, USA), submitted to two ultrasonic pulses during 20 s on ice and incubated for 1 h on ice before ultracentrifugation at 150,000g for 90 min at 4°C (Sorvall Ultra Pro 80 Ultracentrifuge). Pellets were washed one time with 500 mM triethylammonium bicarbonate (TEAB) and one time with 50 mM TEAB.

Samples were suspended in 6M urea, 2M thiourea, and 50 mM TEAB buffer and reduced with 10 mM dithiothreitol (DTT) at (RT) for 1 h. After reduction, free thiols were alkylated with 20 mM iodoacetamide (IAA) in the dark for 1 h (RT). Samples were digested for 3 h (RT) in 1:100 ratio with Lys-C. Afterwards, samples were diluted eightfold with 50 mM TEAB (pH 8.0) and proteins digested with 1:100 ratio trypsin (Promega, Madison, WI) for overnight at 37°C. The reaction was quenched by adding formic acid (FA; 1% final concentration, Sigma-Aldrich, St. Louis, MO) followed by centrifugation at 14,000g for 10 min.

The dried and digested peptides were resuspended in 100 mM HEPES (pH 8.5) with about 10 μg of peptide, as determined using a Qubit Protein Assay Kit (Thermo Fisher), from each sample used for TMT labeling. Six tags (tags 126, 127N, 128N, 129N, 130N, and 131N) from a TMT 11-plex kit (Thermo Fisher) were used to label the samples from each biological replicate. Aliquots (0.8 mg) of each reagent in 41 μl of anhydrous acetonitrile were incubated with peptide samples for 2 h at RT. The reaction was quenched by the addition of 8 μl of 5% hydroxylamine followed by incubation for 15 min at R. For each biological replicate, equal amounts of the six TMT-labeled samples were mixed, and the TMT-labeled peptides were purified on an Oasis HLB 10 mg cartridge (Waters Corporation), which had been wetted with 100% acetonitrile and equilibrated twice with 0.1% aqueous trifluoroacetic acid (TFA). Adsorbed peptides were washed with 2 ml 0.1% TFA twice, eluted in 300 μl of 30% acetonitrile and 300 μl of 60% acetonitrile, and dried using a SpeedVac evaporator. Afterwards, the dried TMT-labeled peptides were resuspended in 2% acetonitrile in 10 mM ammonium formate (pH 9.3) for high-pH reverse phase chromatography.

TMT-labeled peptides were injected onto an C18 Acquity UPLC MClass CSH 1.7 µm, 300 µm × 100 mm (Waters), using a Dionex Ultimate 3000 HPLC system. Buffer A was composed of 2% acetonitrile in 10 mM ammonium formate (pH 9.3) and buffer B of 80% acetonitrile in 10 mM ammonium formate (pH 9.3). Columns were run at 5 µl/min at 30°C, starting at 2% buffer B, and rising to 40% B over the course of a 27-min linear gradient. Buffer B was increased to 50% for 4 min; to 70% for 4 min; to 95% for 10 min, followed by a drop back to 2% B for 1 min, which was maintained until the end of the run at 71 min. Fractions were collected from the 4th to the 44th min with 180 s per fraction, producing 30 fractions. Fractions were concatenated into 10 samples, for example, the 1st, 11th, and 21st fractions were pooled in the same well of a 96-well plate. The 11 samples (10 samples plus the flow through—peptides eluted in the beginning of the starting conditions of high pH separation) were dried using a SpeedVac evaporator and solubilized in 0.1% FA.

Samples were analyzed by an EASY-nano LC system (Proxeon Biosystems, Odense, Denmark) coupled online to an Q Exactive HF—Orbitrap mass spectrometer (QE-HF) (Thermo Scientific, Waltham, USA). Peptides from each fraction (2 µg each) were loaded onto a 18-cm fused silica column with integrated emitter (75 µm inner diameter) packed in-house with reverse phase resin ReproSil-Pur C18-AQ 3 µm (Dr. Maisch GmbH, Germany) and eluted using a gradient from 97% phase A (0.1% FA) to 28% phase B (0.1% FA, 95% acetonitrile) for 64 min, for all 11 fractions, 28–45% phase B for 10 min, 45–100% phase B for 3 min, and 100% phase B for 8 min in a total of 86 min at 250 nl/min. After each run, the column was washed with phase B and re-equilibrated with phase A. Mass spectra were acquired in positive ion mode applying data-dependent acquisition. Each MS scan in the Orbitrap (mass range of m/z of 350–1,500 and resolution 120,000 at m/z 200) was followed by MS/MS of the 20 most intense ions (Top 20). Fragmentation was performed by higher energy collisional dissociation (HCD), and selected ions were dynamically excluded for 20 s. Raw data were viewed in Xcalibur v.4.2 (Thermo Scientific, Waltham, USA). Data processing was performed using Proteome Discoverer v.2.4 (Thermo Scientific, Waltham, USA), by searching raw files with SequestHT algorithm against a T. cruzi Y strain database (release 47, downloaded in May 2020), containing the proteins of the parasite found in the newly sequenced T. cruzi Y strain (Callejas-Hernandez et al., 2018). Common contaminant proteins were also added to the database, and all contaminant proteins identified were filtered and removed from the result lists. The searches were performed with the following parameters: MS accuracy, 10 ppm; MS/MS accuracy, 0.02 Da; trypsin digestion with up to two missed cleavage allowed, carbamidomethyl modification of cysteine and peptide N-terminus TMT-6plex modification as fixed modifications and oxidized methionine and lysine modification by TMT-6plex as variable modifications. Number of proteins, protein groups, and number of peptides were filtered for false discovery rate (FDR) <1%, peptides with rank 1 and proteins with at least 2 peptides using Proteome Discoverer. ProteinCenter software (Thermo Scientific, Waltham, USA) was used to generate FASTA formatted files of groups of proteins of interest. Proteins were considered as regulated when their log₂ fold change was higher than the mean plus two times the standard deviation or lower than the mean minus two times the standard deviation and their coefficient of variation lower than 40%. Transmembrane domain predictions were performed using TMHMM v.2.0 (https://services.healthtech.dtu.dk/service.php?TMHMM-2.0) (Sonnhammer et al., 1998; Krogh et al., 2001), and glycosylphosphatidylinositol (GPI) prediction was performed using the web tools GPI-SOM (http://gpi.unibe.ch/) and big-PI Predictor (https://mendel.imp.ac.at/gpi/gpi_server.html) (Eisenhaber et al., 1998; Sunyaev et al., 1999; Eisenhaber et al., 1999; Eisenhaber et al., 2000). DeepLoc (http://www.cbs.dtu.dk/services/DeepLoc/index.php) (Almagro Armenteros et al., 2017) was used to predict the proteins subcellular localization. SignalP v.5.0 (https://services.healthtech.dtu.dk/service.php?SignalP-5.0) (Almagro Armenteros et al., 2019) was used to predict proteins secreted by classical ER/Golgi pathway, and SecretomeP v.2.0 (http://www.cbs.dtu.dk/services/SecretomeP/) (Bendtsen et al., 2005) was used to predict non-classical protein secretion. Overrepresented categories were investigated through Gene Ontology (GO) enrichment analysis using TriTrypDB (http://tritrypdb.org/tritrypdb/), and redundant GO terms generated were summarized in clusters by REVIGO tool (http://revigo.irb.hr/) (Supek et al., 2011).

The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium (Deutsch et al., 2020) via the PRIDE (Perez-Riverol et al., 2019) partner repository with the dataset identifier PXD028897.

Quantitative data were expressed as mean ± standard deviation (SD) of three independent experiments unless otherwise indicated. Statistical analysis was carried out using GraphPad Prism 7.0 software via one-way ANOVA followed by Fisher’s least significant difference (LSD) test or by controlling the false discovery rate (FDR) using the original procedure of Benjamini and Hochberg for multiple comparisons. Means were considered significantly different when p <0.05.

T. cruzi multiplication is increased in a LAMP-deficient environment (LAMP1/2^−/− or LAMP2^−/− fibroblasts) (Albertti et al., 2010), which suggests that differences in the intracellular environment can modify parasite biological behavior inside the host cell. In order to evaluate whether this LAMP-deficient environment may also influence the biological behavior of parasites released from these cells, T. cruzi Y strain trypomastigotes collected from a single infection cycle in WT (TcY-WT), LAMP1/2^−/− (TcY-L1/2^−/−) or LAMP2^−/− (TcY-L2^−/−) fibroblasts were evaluated concerning their ability to invade L6 myoblasts and WT fibroblasts. For this, these cells were exposed to TcY-WT, TcY-L1/2^−/−, and TcY-L2^−/− infective trypomastigote forms at an MOI of 50 for 40 min.

For L6 myoblasts, TcY-L2^−/− parasites presented the highest invasion rates, as evidenced by the number of intracellular parasites per 100 cells, when compared to TcY-WT and TcY-L1/2^−/− parasites, which did not show statistically significant differences between them ( Figure 1A ). On the other hand, for WT fibroblasts, TcY-L2^−/− parasites presented the lowest invasion rates, when compared to the other parasites, TcY-L1/2^−/− and TcY-WT ( Figure 1B ). Additionally, for these fibroblasts, TcY-WT invasion rate was higher than that observed for TcY-L2^−/− parasites, which were the less infective ones. As expected, for both host cells, the number of parasites per infected cell was about one, for all tested groups.

T. cruzi adhesion to the host cell has been considered as the initial step for the invasion process (Andrews and Colli, 1982; Piras et al., 1982; Piras et al., 1983), since this proximity could aid the interaction between the parasite and host cell surface proteins, which is responsible for triggering the necessary signaling pathways that lead to parasite internalization (Fernandes and Andrews, 2012). To investigate whether the differences found in the parasite invasion rates were related to their ability in adhering to these cells, we exposed pre-fixed L6 myoblasts or WT fibroblasts to TcY-WT, TcY-L1/2^−/−, and TcY-L2^−/− trypomastigotes at an MOI of 50, for 40 min, and analyzed parasites that remained attached to the cell surface after several washes.

The adhesion assay in L6 myoblasts, revealed that TcY-L1/2^−/− trypomastigotes presented the highest adhesion rates when compared to TcY-WT and TcY-L2^−/− parasites ( Figure 1C ). TcY-WT parasites showed an intermediate adhesion rate compared to TcY-L2^−/− parasites, which in turn showed the lowest adhesion rates ( Figure 1C ). For WT fibroblasts, TcY-L1/2^−/− trypomastigotes presented the lowest adhesion rates when compared to TcY-WT and TcY-L2^−/− parasites ( Figure 1D ). There was no statistically significant difference between TcY-L2^−/− and TcY-WT parasites. Despite the differences found in the adhesion rates among the studied parasites in the two studied cell types, these results revealed that adhesion and invasion rates did not necessarily correlate with each other.

As mentioned before, T. cruzi invasion into host cells is dependent on the parasite’s ability to trigger host cell intracellular calcium signaling and subverting the host cell plasma membrane repair mechanism. T. cruzi is able to initiate this calcium signaling by both causing microlesions in the host cell plasma membrane or inducing calcium signaling events through engagement of host cell surface receptors (Fernandes and Andrews, 2012). The increase in intracellular calcium concentration leads to lysosomal exocytosis and compensatory endocytosis events, culminating in parasite internalization (Tardieux et al., 1994). Since calcium ions play a crucial role in T. cruzi invasion, we decided to investigate whether the changes observed for the invasion rates could correlate to parasite ability in triggering calcium signals. For that, L6 myoblasts or WT fibroblasts were loaded with the fluorescent calcium-binding probe Fluo4-AM and exposed to TcY-WT, TcY-L1/2^−/−, and TcY-L2^−/− infective trypomastigote forms at an MOI of 100, while recording the fluorescence increase in real time by confocal microscopy. For L6 myoblasts, calcium signaling induced by TcY-L2^−/− was significantly higher than the signaling induced by TcY-WT or TcY-L1/2^−/− parasites, while those two did not present statistically significant differences between them ( Figure 2A , Supplementary Figure S1 ). On the other hand, for WT fibroblasts, calcium signaling induced by TcY-L2^−/− parasites was significantly lower than the signaling induced by the other two parasites, TcY-WT and TcY-L1/2^−/−, being the highest calcium signals induced by TcY-L1/2^−/− parasites ( Figure 2B , Supplementary Figure S2 ). For both host cell types, myoblasts and fibroblasts, the calcium signals induced by the different parasites directly correlated with their invasion rates, meaning the higher the calcium signaling triggered by the parasite, the higher the cell invasion efficiency. These results reinforce the important role played by calcium signaling events during the T. cruzi invasion process.

To investigate further the differences in the calcium signaling triggered by the distinct parasites in the different cell types, we evaluated whether the calcium-signaling events were generated by micro-injuries at the host cell plasma membrane or triggered by activation of host cell surface receptors. To investigate the calcium signaling induced by plasma membrane injuries, we exposed L6 myoblasts or WT fibroblasts to TcY-WT, TcY-L1/2^−/−, and TcY-L2^−/− trypomastigotes at an MOI of 100 in the presence of propidium iodide (PI) for 30 min. PI is a red-fluorescent nuclear and chromosome counterstain, not permeant to live cells. Therefore, PI staining indicates membrane permeabilization by injury. For L6 myoblasts, TcY-L2^−/− parasites induced injury to a higher number of cells when compared to TcY-WT and TcY-L1/2^−/− parasites, corroborating their invasion profile and suggesting that for these parasites, calcium signals are triggered through injuries at the host cell membrane ( Figure 2C ). No statistically significant differences were observed between the number of PI positive cells for TcY-WT and TcY-L1/2^−/− parasites, indicating that these parasites likely induce similar levels of injury to L6 myoblasts ( Figure 2C , Supplementary Figure S3 ). Again, these results corroborate with the invasion profile observed for these parasites. Alternatively, cells were incubated with parasites in the absence of PI, washed, and then incubated with PI. As expected, no statistically significant differences were observed among PI staining of cells incubated with PI post T. cruzi infection, for all parasites, and the negative control (cells incubated with PI only, without being previously exposed to the parasite) ( Supplementary Figure S4 ). For WT fibroblasts, TcY-L1/2^−/− parasites induced injury to a higher number of cells when compared to TcY-WT and TcY-L2^−/− parasites, corroborating their invasion profile in these cells ( Figure 2D and Supplementary Figure S5 ). TcY-WT parasites induced injury to the lowest number of cells, when compared to the others, even though calcium signaling triggered by these parasites in WT fibroblasts was higher than that observed for TcY-L2^−/− parasites ( Figures 2B, D ).

Once parasites released from the distinct cell types presented different biological behaviors, we decided to investigate whether changes in these parasites’ surface membrane protein profiles could account for the biological phenotypes obtained. For this, we performed quantitative mass spectrometry-based proteomic analysis to identify and quantify the membrane-enriched protein preparation from TcY-WT, TcY-L1/2^−/−/−, and TcY-L2^−/− trypomastigotes. T. cruzi Y strain trypomastigotes were cultured in the wild-type or LAMP-deficient fibroblasts, collected, lysed, and then ultra-centrifuged to separate the membrane-enriched protein fraction. Extracted proteins were digested, reduced, and alkylated, and the tryptic peptides were labeled with TMT 6-plex (Thermo Fisher), analyzed by mass spectrometry-based proteomics, and submitted to bioinformatics tools to perform functional analyses. For protein identification from the peptide sequences, we used the Y strain annotated proteins from TriTrypDB website (released on April 23, 2020 and downloaded on May 2020) (Callejas-Hernandez et al., 2018).

We identified 3,806 proteins with high confidence level and containing at least one unique peptide in all three samples ( Supplementary Table S1 ). A principal component analysis was performed to evaluate the biological replicates similarity, using the normalized quantitative abundance values ( Supplementary Figure S6 ). This analysis showed an acceptable level of similarity between replicates within the same group. A total of 1,365 proteins were annotated as uncharacterized or hypothetical proteins, corresponding to about 36% of the identified proteins in our data set. The high percentage of uncharacterized proteins is not surprising, since for the T. cruzi Y strain genome, 56.94% of the proteins correspond to hypothetical proteins of unknown function (Callejas-Hernandez et al., 2018). Because of this, we performed a GO analysis with Fisher exact test filtering for FDR <0.05, using the GO enrichment analysis tool on the TriTrypDB in order to verify if our samples were in fact enriched in membrane terms. Regarding the GO cellular component level, the analysis showed that our data set presented an enrichment in many terms related to membrane, such as intrinsic component of membrane, integral component of membrane, and endomembrane system ( Supplementary Figure S7 ). Even though the most enriched term is cytoplasm, the QuickGO definition of this term comprises the contents of a cell excluding the plasma membrane and nucleus, but including other subcellular structures (https://www.ebi.ac.uk/QuickGO/GTerm?id=GO:0005737). Therefore, many organelle membrane proteins present in our data set fell into this GO cytoplasm term.

We then compared the proteomes of TcY-L1/2^−/− and TcY-L2^−/− trypomastigotes’ membrane-enriched proteins preparation to that from TcY-WT trypomastigotes. We considered as regulated proteins, all members that presented a log2 fold change higher than the mean plus 2 times the standard deviation or lower than the mean minus 2 times the standard deviation and coefficient of variation lower than 40%—referred from here as the regulated subproteome. We found 75 and 90 proteins in the TcY-L1/2^−/− and TcY-L2^−/− regulated subproteomes, respectively ( Supplementary Table S2 ) that matched these criteria. From these proteins, either a considerable percentage, in both regulated subproteomes, had predicted transmembrane domains or to be classically or non-classically secreted ( Table 1 ). Additionally, important T. cruzi virulence factors, such as members of trans-sialidase superfamily, and other surface proteins families, such as MASP and surface protease gp63 (Goncalves et al., 1991; Trocoli Torrecilhas et al., 2009; Bayer-Santos et al., 2013), were found to be differentially regulated in these parasites ( Table 2 ).

In the TcY-L1/2^−/− regulated subproteome, 51 proteins were upregulated and 24 were downregulated, and in the TcY-L2^−/− regulated subproteome, 55 proteins were upregulated and 35 were down regulated, when compared to TcY-WT proteome. There were only four upregulated proteins, three of them being hypothetical proteins, and nine downregulated proteins shared between TcY-L1/2^−/− and TcY-L2^−/− regulated subproteomes ( Figure 3 ).

For further understanding the possible role of these differentially regulated proteins in the biological behavior observed for TcY-L1/2^−/− and TcY-L2^−/−, we classified and grouped them in distinct biological processes, based on their gene ontology, using the enrichment test available on TriTrypDB with Fisher exact test filtering for FDR <0.05. Regarding the TcYL1/2−/− regulated subproteome regulated subproteome, there was no enriched biological processes (BP) from the up regulated proteins and we found 15 BP enriched from the down regulated. We removed the redundant terms from them using the REVIGO tool (http://revigo.irb.hr/), and after summarizing the terms, we found eight BP downregulated (protein folding, cellular ketone metabolic process, quinone biosynthetic process, pentose-phosphate shunt, glyceraldehyde-3-phosphate metabolic process, microtubule-based movement, cellular process, movement of cell or subcellular component) in comparison to TcY-WT proteome ( Figure 4A and Supplementary Table S3 ). Six proteins were relevant for two or more downregulated BP in theTcY-L1/2^−/− regulated subproteome: 6-phosphogluconate dehydrogenase, decarboxylating, putative (TcYC6_0069460), ubiquinone biosynthesis protein COQ7-like, putative (TcYC6_0110760), kinesin-13 3, putative (TcYC6_0081520), kinesin, putative (TcYC6_0110780), T-complex protein 1, delta subunit, putative (TcYC6_0023210), and chaperonin/T-complex protein 1 gamma subunit, putative (TcYC6_0030620). We next performed a GO biological process enrichment analysis using the membrane-associated proteins that presented transmembrane domains and/or were assigned as membrane by the DeepLoc analysis (referred from here as membrane subproteome). From this analysis, we did not find any enriched BP either from the up- or the downregulated proteins list. However, for the upregulated membrane subproteome, we found the sphingolipid metabolism (ec00600:PK:KEGG) pathway enriched, using the Metabolic Pathway Enrichment analysis tool on TriTrypDB with KEGG as pathway source and filtering FDR <0.05.

For TcY-L2^−/− regulated subproteome, there was no enriched BP from the downregulated proteins, and we found 25 BP enriched from the upregulated. After summarizing the terms with the REVIGO tool, we found five BP upregulated (proteolysis involved in cellular protein catabolic process, catabolic process, proteolysis, cellular protein-containing complex assembly, mitochondrion organization) in comparison to TcY-WT proteome ( Figure 4B and Supplementary Table S3 ). Six proteins were present in two or more upregulated BP: proteasome alpha 1 subunit, putative (TcYC6_0071370), proteasome alpha 2 subunit, putative (TcYC6_0023180), proteasome beta 3 subunit, putative (TcYC6_0044600), proteasome beta 6 subunit, putative (TcYC6_0041700), proteasome beta 7 subunit, putative (TcYC6_0091650), and putative mitochondrial chaperone BCS1 (TcYC6_0123290). Using the Metabolic Pathway Enrichment analysis tool on TriTrypDB with KEGG as pathway source and filtering FDR <0.05, we found two upregulated pathways [arginine and proline metabolism (ec00330:PK:KEGG), and Diterpenoid biosynthesis (ec00904:PK:KEGG)] and one downregulated pathway [ether lipid metabolism (ec00565:PK:KEGG)]. The GO biological process enrichment analysis of TcY-L2^−/− membrane subproteome revealed one upregulated BP, the vacuolar proton-transporting V-type ATPase complex assembly. We did not find any enriched pathways for this membrane subproteome.

In order to establish a better understanding of the LAMP-deficient cells-derived parasites, we compared the TcY-L1/2^−/− and TcY-L2^−/− proteomes. These parasites were the most different regarding their invasion characteristics. In fact, the proteomic analysis revealed a higher number of differentially regulated proteins between them, 136 proteins—55 downregulated and 81 upregulated (using the same previously described criteria) ( Supplementary Table S2 ). From these proteins, a considerable part is predicted to have transmembrane domains or to be classically or non-classically secreted ( Table 3 ). Interestingly, we also found a higher number of membrane surface proteins regulated between TcY-L1/2^−/− and TcY-L2^−/− trypomastigotes ( Table 4 ).

The GO BP enrichment analysis from these differentially regulated proteins between TcY-L1/2^−/− and TcY-L2^−/− trypomastigotes showed 13 BPs from the down regulated proteins list and no BP from the upregulated protein list. After summarizing the terms using the REVIGO tool, three downregulated BP remained (proteolysis involved in cellular protein catabolic process, catabolic process, and proteolysis) ( Figure 4C and Supplementary Table S3 ). Five proteasome proteins were present in all three BPs: proteasome alpha 1 subunit, putative (TcYC6_0071370), proteasome alpha 2 subunit, putative (TcYC6_0023180), proteasome alpha 7 subunit, putative (TcYC6_0044220), proteasome beta 3 subunit, putative (TcYC6_0044600), and proteasome beta 7 subunit, putative (TcYC6_0091650). The metallo-peptidase, Clan MF, Family M17, putative (TcYC6_0033380), cytoskeleton-associated protein CAP5.5, putative (TcYC6_0107360), and surface protease GP63, putative (TcYC6_0169090) proteins also contributed to the proteolysis enrichment.

Since many proteolysis-related terms were enriched in the TcY-L2^−/− regulated subproteome compared to TcY-WT and TcY-L1/2^−/− proteomes, we decided to further investigate the parasites proteolytic activity. For that, we evaluated the parasites’ proteolytic activity through gelatin zymography assays. The results showed a higher gelatinolytic activity for TcY-L2^−/− trypomastigotes compared to TcY-WT and TcY-L1/2^−/− ( Figure 5 ), thus corroborating the proteomic analysis.

In order to investigate deeper the higher proteolytic profile observed for TcY-L2^−/− trypomastigotes, we decided to test the protein extracts for serine proteases and cysteine proteases activity. Enzyme activity assays were performed using Ac-YVAD-AFC caspase-like and Suc-LLVY-AMC chymotrypsin-like fluorogenic substrates, for cysteine and serine proteases, respectively. For caspase-like activity, TcY-L1/2^−/− and TcY-L2^−/− protein extracts showed slightly higher activities than TcY-WT protein extract ( Figures 6A, C ). For all protein extracts, the inhibitor MG-132 decreased caspase-like activity in 2.23 ± 0.055, 1.98 ± 0.25, and 1.94 ± 0.34 times for TcY-WT, TcY-L1/2^−/−, and TcY-L2^−/−, respectively ( Figures 6A, C ). For chymotrypsin-like activity, TcY-L2^−/− protein extract showed the highest activity, while TcY-L1/2^−/− presented the lowest ( Figures 6B, D ). Here again, for all protein extracts, the inhibitor MG-132 considerably decreased their activities in 12.12 ± 1.53, 8.91 ± 0.48, and 15.73 ± 0.53 times for TcY-WT, TcY-L1/2^−/−, and TcY-L2^−/−, respectively. For Suc-LLVY-AMC substrate, the kinetic activity also showed a similar profile to the one revealed by the zymography assays.