Ethics statement

For working with human PBMCs, we obtained buffy coats of anonymized blood donors from the blood donation center (DRK-Blutspendedienst Hessen GmbH), Frankfurt am Main, Germany, which did not require approval by an Ethics committee.

L. major parasite culture

Leishmania major (MHOM/IL/81/FEBNI) promastigotes were cultured on a semi-solid rabbit blood-agar system as described before [26]. Cultures were seeded on day 0 with 50,000 viable L. major in100 µl per well of a 96-well plate. Logarithmic-phase (logPh) promastigotes, 3 days post passaging, or stationary-phase (statPh) promastigotes from 6-8 days post passaging were used for experiments. Stationary-phase promastigotes typically consist of approximately 2–10% metacyclic parasites and 50–60% dead parasites exhibiting apoptosis-like features, e.g., phosphatidylserine exposure (Fig A in S1 Appendix).

Axenic amastigotes were generated as described in [27] and washed in tissue culture plates prior to experiments. Parasites either expressed DsRed as fluorescent marker or a Cas9/T7 RNA polymerase [28]/ dimerizable Cre (diCre) construct [29]. Parasites were cultured with respective selection markers: 30 µg/ml Hygromycin B Gold, 10 µg/ml Blasticidin, 40 µg/ml G418, 40 µg/ml Puromycin (all InvivoGen), 100 µg/ml Bleomycin (Jena Bioscience).

L. major proteomics

Nucleotidase assay

Nucleotidase activity of 2x10⁶ intact L. major was determined by the release of free phosphate (P_i) from 3’-AMP or 5’-AMP (both SigmaAldrich). After 3 hours of parasite incubation in 100 µl reaction buffer (116 mM NaCl, 5.4 mM KCl, 5.5 mM D-Glucose, 50 mM HEPES pH 7-7.6) [37] at 27 °C, parasites were separated into pellet and supernatant by centrifugation and 100 µM substrate added for 1 hour at 27 °C. Intact pelleted parasites represented membrane-bound ecto-nucleotidase activity, the supernatant fraction secreted ecto-nucleotidase activity. Released P_i was quantified using a malachite green based phosphate assay kit (SigmaAldrich) according to instructions. A_620nm was measured with a ClarioStarPlus plate reader (BMG Labtech). A sample containing only substrate was used as blank, a control containing only parasites was always checked for P_i carryover. All reagents were tested for unspecific P_i content before use. Nucleotidase activity was expressed as µM released P_i/2 × 10⁶ L. major/h.

Endonuclease assay

Nuclease activity of 2x10⁶ intact L. major was determined by visualizing enzymatic degradation of a DNA substrate on a 0.8% TAE-agarose (BioRed) gel stained with GelRed (biotium). After 3 hours of parasite incubation in 100 µl reaction buffer (30 mM sodium acetate pH 5.3, 100 mM NaCl, 2 mM ZnCl₂) [7] at 27 °C, parasites were separated into pellet and supernatant by centrifugation and 500 µg substrate added for 1 hour at 27 °C. Intact pelleted parasites represented membrane-bound ecto-nuclease activity, the supernatant fraction secreted ecto-nuclease activity. For endonuclease activity, single-stranded or double-stranded circular M13 phage genome (New England Biolabs) was used as substrate. For nuclease activity on NETs, respective NET-enriched supernatant was used. Reaction times were set as specified. Remaining DNA substrate was semi-quantified on gel images relative to a 500 µg undigested control using ImageJ 1.53c. A sample containing substrate and DNaseI (Invitrogen) was used for positive digestion control on every gel. Relative nuclease activity was calculated as difference to the undigested substrate.

His₆ enrichment and Western blot

12x10⁶ L. major logPh promastigotes were lysed in 20 µl Laemmli buffer (500 mM Tris/HCl pH 6,8, 38% glycerin, 10% SDS, 0.6 M DTT, 0,01% bromophenol blue) [38]. Corresponding culture supernatant was incubated with 15 µl HIS-Select nickel magnetic agarose beads (SigmaAldrich) for 30–60 min at room temperature in an overhead shaker and washed twice in PBS. Proteins were eluted in 20 µl Laemmli buffer at 65 °C for 10 min.

Protein samples were denatured for 5 min at 95 °C and separated on a 12% SDS-PAGE. After blotting, PVDF membranes were incubated with 1 µg/ml anti-αTubulin (GeneTex GTX628802) and 776 ng/ml anti-His₆ (Cell Signaling Technology 2366) antibodies overnight at 4 °C, followed by 1 hour at room temperature with HRP-conjugated anti-mouse secondary antibody (Santa Cruz sc516102). The membrane was incubated with Immobilon Forte Western HRP substrate (Millipore) for 1 min at room temperature, signal was visualized in an ECL Chemostar (Intas).

Human monocyte-derived macrophage (hMDM) generation

Monocytes were isolated from peripheral mononuclear cells (PBMCs) of anonymized healthy human donors as described before [39]. Buffy coats were provided by the DRK-Blutspendedienst Hessen GmbH (Frankfurt am Main). CD14⁺ monocytes were isolated by magnetic cell sorting (Miltenyi Biotec) and differentiated with 50 ng/ml M-CSF (R&D Systems) or 30 ng/ml GM-CSF (Sanofi) for 5–7 days at 37 °C, 5% CO₂ in a humidified atmosphere. Further activation was done with 50 ng/ml IFNγ (Sigma) for 24 hours.

Autologous peripheral blood lymphocyte (PBL) isolation

CD14^- PBMCs were stored at -80 °C for the time of macrophage differentiation and thawed upon usage [39].

Isolation of polymorphonuclear cells (PMNs)

PMNs of healthy individuals were isolated from fresh blood as described in [40] or from buffy coats as previously described [26]. 12x10⁶ cells were seeded on poly-D-lysine (Gibco) coated 12-well plates. Neutrophil extracellular trap (NET)-enriched supernatants were induced and harvested as reported before [41]. DNA content was determined using a Quant-it RiboGreen assay (Invitrogen).

Determination of hMDM infection

Differentiated macrophages were infected with L. major^DsRed promastigotes from the statPh at a multiplicity of infection (MOI) of 10 for 3 hours at 37 °C before remaining extracellular parasites were washed off. After the specified time post infection (pi), the infection rate of viable single cells was quantified by detecting the parasite’s fluorescence marker by flow cytometry on either LSR Fortessa or Symphony A3 Cell Analyzer (both BD Bioscience). Parasite burden was determined as DsRed mean fluorescence intensity (MFI) of single, viable, infected macrophages. Cell viability was monitored using staining with 1:1000 ZombieAqua (BioLegend).

Alternatively, cells were fixed in 4% PFA (Sigma) and permeabilized using 0.5% saponin (AppliChem PanReact), then intracellular parasites were stained with 1:10,000 anti-L. major rabbit serum and 1:1000 goat anti-rabbit AlexaFluor 568 secondary antibody (Invitrogen A-11036).

If applicable, adenosine (SigmaAldrich) or 3’-AMP (Santa Cruz Biotechnology) were added during infection at indicated concentrations.

Analysis of PBL proliferation in a co-culture with hMDM

Macrophages and PBLs were stained in 1 µM CellTrace Far Red (Invitrogen) for 20 min at 37 °C. 0.2 × 10⁶ stained macrophages were infected with L. major promastigotes from the statPh at MOI = 20 for 3 hours at 37 °C before 1x10⁶ stained PBLs were added. Infection was carried out in U-shaped 96-well plates with 2x10⁴ hMDMs per well and 10 technical replicates per condition which were pooled before analysis. PBL proliferation was analyzed by flow cytometry 4 days pi as CellTrace_low fraction of viable, single PBLs. Infection rate and parasite burden of hMDM were determined as stated before. Cell viability was monitored using staining with 1:1000 ZombieAqua (BioLegend).

If applicable, adenosine or 3’-AMP were added together during infection at indicated concentrations.

Neutrophil killing assay

Isolated granulocytes were tested for activation status by CD66b and CD62L staining beforehand. 2x10⁶ PMNs in a 24-well plate were treated with 10 µg/ml Cytochalasin D (SigmaAldrich) for 20 min at 37 °C. 130 U DNaseI were added at the same time, if applicable. L. major statPh promastigotes were added at MOI = 1 and incubated for 2 hours at 37 °C and then transferred to 27 °C for two days. Corresponding controls containing reagents and parasites but no PMN were treated the same. Motile, viable parasites were counted in a Neubauer chamber and numbers normalized to the respective control gave the percentage of viable parasites.

ELISA

Cell free supernatants from infected samples were analyzed for soluble cytokine content using enzyme-linked immunosorbent assay (ELISA) (DuoSet, R&D Systems) following instructions. TNFα content was measured as 1:12.5 dilutions. Signals were developed using horseradish peroxidase substrate solution (BioLegend) and stopped by addition of 2 N H₂SO₄ (Carl Roth). Absorbance was measured at 450 nm using 570 nm as reference wavelength at a ClarioStarPlus. Total cytokine concentration was determined by 4-parameter regression of the standard row in Mars v4.00. Samples were measured in duplicates.

Staining of surface markers

Human primary cells were stained for surface markers in PBS pH 7.2 plus 0.5 g/L BSA and incubated for 30 min at 4 °C. Antibodies were added accordingly: anti-A_2a-AlexaFluor488, anti-A_2b-AlexaFluor647 (both R&D Systems, 1 µg/1 × 10⁶ cells). Cells were also stained with 1:1000 ZombieAqua (BioLegend) or 1µM SYTOXBlue (Invitrogen) as viability marker. Expression was analyzed by flow cytometry as percentage positive cells.

Generation of L. major inducible knockout strains

L. major expressing Cas9/T7 RNA polymerase and a dimerizable Cre (diCre) construct were used to integrate a 3’ and a 5’ flox cassette in two consecutive steps by CRISPR/Cas9-facilitated homologous recombination, adapting the method of [29]. Each cassette was amplified using 0.5 µM forward and reverse primer (sequences see Table 1), 15 ng/µl plasmid template (pGL2314 for 3’ flox [29], pTNeo-eGFP-loxP (modified based on [28]) for 5’ flox), 4 µl GC Enhancer, 10 µl Platinum Mastermix (both Invitrogen) and 3 µl H₂O in a S1000 Thermal Cycler (BioRad) set to 35 cycles of 10 s at 98 °C and 3 or 4 min (3’ flox/ 5’ flox) at 72 °. Single guide RNA (sgRNA) templates were amplified using 2 µM primer, 2 µM scaffold, 5 µl PWO Mastermix (Roche) and 1 µl H₂O for 35 cycles of 10 s at 98 °C, 30 s at 60 °C and 15 s at 72 °C. Products were checked on a 1% agarose gel and sterilized for 5 min at 94 °C.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Oligonucleotide sequences.

https://doi.org/10.1371/journal.ppat.1014197.t001

60x10⁶ promastigotes from the logPh were pulsed once with the PCR products in Human T-cell Nucleofector buffer (Lonza) using a Nucleofector 2b (Lonza) with program X-001. Electroporated cells were incubated overnight in 9 ml L. major medium at 27 °C. After addition of the respective antibiotics and 10% fetal calf serum (SigmaAldrich), parasites were distributed in 100 µl on a 96-well rabbit blood agar plate. After 14–21 days, grown clones were checked for mCherry (3’ flox) and eGFP (5’ flox) fluorescence using flow cytometry and single cell clones were generated by serial dilution.

Recombination in full-floxed strains was induced by addition of 100 nM rapamycin (SigmaAldrich) and fluorescence monitored over 9 days. Single cell clones were generated as above and checked for p1/s1 and eGFP status by PCR (2 µl genomic DNA, 1.5 µl forward and reverse 10 µM primer each, 5 µl PWO Mastermix) by 3 min at 95 °C and 25 cycles of 30 s at 95 °C, 30 s at 59 °C, 45 s at 72 °C plus final extension for 10 min at 72 °C. Amplified products were checked on a 1% agarose gel.

Generation of L. major addback strains

p1/s1 (LmjF.30.1500) was inserted into the multiple cloning site of pLexLm plasmid, a plasmid based on pLEXSY-ble2.1 (Jena Bioscience, Cat. No. EGE-271) in which the rRNA promotor region of L. donovani has been exchanged by an L. major rRNA promotor region. Identity of the insert as LmjF.30.1500 was confirmed by Sanger sequencing.

For addback strain generation, 60x10⁶ L. major^Δp1/s1 clone E10 logPh promastigotes were electro-porated with SwaI (New England Biolabs) linearized and sterilized plasmid. Electroporated cells were incubated overnight in 9 ml L. major medium at 27 °C. After addition of bleomycin and 10% FCS, parasites were spread in 100 µl on a 96-well rabbit blood agar plate. After 14–21 days, grown clones were checked for p1/s1 status using PCR and single cell clones were generated.

Generation of L. major strains overexpressing p1/s1-His₆

p1/s1 (LmjF.30.1500) was inserted into the multiple cloning site of pLexLm plasmid including an in-frame fusion with the encoded His_6-tag. Identity of the insert as LmjF.30.1500 was confirmed by Sanger sequencing.

For generation of strain overexpressing p1/s1-His_s, 60x10⁶ L. major^Cas9/T7/eGFP logPh promastigotes were electroporated with SwaI (New England Biolabs) linearized and sterilized plasmid. To generate strains overexpressing p1/s1, a L. major^Cas9/T7/eGFP strain was used. Electroporated cells were selected and cultured as described above for L. major addback strains.

Cell death assays

L. major logPh promastigotes were treated with increasing concentrations of miltefosine (Calbiochem) for 48 h, and parasites were used in different assays to quantify different hallmarks of cell death:

Metabolic rate of 5x10⁶ promastigotes was determined by colorimetric turnover of 1.1 mM Thiazolyl blue tetrazolium bromide (MTT, abcam) in phenolred-free RPMI (SigmaAldrich) for 4 hours at 27 °C. Formazan crystals were dissolved in DMSO (AppliChem PanReact) and absorbance was measured at 540 nm, using 700 nm as reference wavelength, at a ClarioStarPlus. Values were blanked to a sample containing parasites but no MTT. Relative cell metabolism representing viability was calculated by normalizing to the DMSO control.

2x10⁶ promastigotes were stained with 1:1000 AnnexinV-FITC (Miltenyi Biotec) in Ringer solution (B. Braun) for 20 min at 4 °C. Labelling was analyzed by flow cytometry, gating was set using a control stained in calcium-free PBS.

2x10⁶ promastigotes were permeabilized in 70% (v/v) EtOH for 30 min at -20 °C and stoichiometrically stained for DNA content using 5 µM SYTOXGreen (Invitrogen) in PBS containing 50 mM EDTA pH 7.4 and RNase (Thermo Fisher) for 30 min at 37 °C. Parasites were analyzed by flow cytometry on LSR II SORP (BD Bioscience), gating was determined on an unstained control. The share of cell cycle phases was determined based on DNA content.

Effective concentration (EC)₅₀ values were calculated based on 4-parameter regression in GraphPad Prism with constraints for bottom and EC₅₀ ≥ 0 and stated as ± 95% confidence interval (CI).

Statistics

All graphs display mean ± standard deviation and were generated in GraphPad Prism v9.5.0. Significance was determined using a two-sided Welch’s t-test with * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001, **** p ≤ 0.0001.

A set of 9 proteins, including p1/s1, shows increased abundance in L. major upon various cell death pressure

Identification of stress-related proteins in Leishmania is of great interest as they can feature potential drug targets [44– 46] or be involved in mediating apoptotic-like cell death, a mechanism used by Leishmania to promote disease development and circumvent effective adaptive immune responses [39, 47]. In order to designate such proteins in L. major, we applied stress to the parasites by treatment with low doses of miltefosine, an antileishmanial drug, or staurosporine, an unspecific kinase inhibitor, that would lead to cell death after 24 hours. Following treatment with miltefosine for 3 hours or staurosporine for 6 hours, with early cell death processes being induced (Fig A in S1 Appendix) but not being lethal at these time points, we performed a quantitative proteome analysis to compare protein expression. Volcano plots of proteins that show significantly increased abundance in the treated or untreated conditions are displayed in Fig 1A and 1B. Here, proteins with increased abundance in one condition that is displayed by a segregated log2 fold change, had not be detected in the other condition. Additionally, stationary-phase (statPh) promastigotes, harboring parasites that undergo intrinsic cell death, were analyzed and compared to untreated logPh parasites (Fig 1C). By choosing these conditions, proteins of immediate stress responses prior to cell death or survival promotion can be detected in the intersection of more abundant hits (Fig 1D). Proteins involved in drug-specific responses or metacyclogenesis of the statPh are solely detected in single conditions and are thus excluded.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Quantitative proteomic analysis of cell-death induced L. major.

A- Volcano plot of the quantified proteome of L. major treated with 25µM miltefosine for 3 h, B- 25µM staurosporine for 6 h and C- statPh promastigotes, always compared to untreated, logPh promastigotes. A-C- Volcano plot visualized with VolcaNoseR [42], proteins with commonly increased abundance are labelled (Table 2). Cutoffs were set at log₂FC < -1.5, > 1.5 and p-value < 0.01. Visible gaps in the fold change separate protein hits that revealed increased abundance in one condition but had not been detected in the other condition. In order to plot these proteins, we set an arbitrary value for each undetected protein. D- Venn diagram displaying the number of proteins with significantly increased abundance in each labeled condition versus the untreated logPh control. Intersections represent proteins detected as more abundant in several conditions. Template created using BioVenn [43]. Analysis based on 7 biological replicates and 3 technical replicates each.

https://doi.org/10.1371/journal.ppat.1014197.g001

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Common proteins with increased abundance in L. major undergoing cell death; annotations extracted from UniProt [49]. All peptides detected for p1/s1 could be assigned to the conserved C-terminal region of proteins in the cluster (Fig 2A).

https://doi.org/10.1371/journal.ppat.1014197.t002

Across the three comparisons, 4116 proteins were detected in total, corresponding to 51.73% of the reference Leishmania proteome. Taking the intersection of hits with increased abundance, only five proteins were commonly more abundant in all three comparisons, four were exclusively detected in parasites upon cell death induction or statPh promastigotes, but not in the viable control samples (Fig 1D and Table 2). Among these hits are several proteins involved in protein biogenesis and turnover but also proteins of diverse predicted functions. One of the exclusively detected hits, p1/s1, is the only protein with predicted enzymatic activity, namely a class I nuclease, and with proposed involvement in host-parasite interactions. In Leishmania, such enzymes are involved in essential purine salvage by eliciting 3’-nucleotidase and endonuclease activity and are presumably involved in cellular functions as well as in host-parasite interactions (reviewed in [48]). Given p1/s1 expression in stress-induced and infectious statPh L. major, this hit stands out as central link between parasite metabolism and disease progression. Studying its multiple functions and its potential role in parasite cell death could establish p1/s1 as potent drug target.

Ecto-nucleotidase and -endonuclease activity of L. major increase from the logarithmic to stationary growth phase

Indeed, detected peptides assigned to p1/s1 can derive from six open reading frames arranged in a tandem array (LmjF.30.1460 to LmjF.30.1510). Sequence details were retrieved from EnsemblProtists release 61. All open reading frames share a common C-terminal domain. Whereas the four genes 1460–90 contain a putative transmembrane domain (TMD), 1500 and 1510 code for an N-terminal signal sequence for secretion (Fig 2A), as predicted by DeepLoc 2.1 [50]. We verified secretion by overexpressing a C-terminally His₆-tagged LmjF.30.1500 construct in L. major^Cas9/T7/eGFP. Parasite lysate was directly used for SDS-PAGE, whereas the corresponding culture supernatant was subjected to His₆ enrichment over magnetic Ni-NTA beads. This eluate was also separated by SDS-PAGE. p1/s1-His₆ and the housekeeping protein α-Tubulin were detected by Western blot. Whereas no His₆-tagged proteins were detected in the lysate, a His₆-tagged protein approximately at the predicted size of p1/s1 (35.1 kDa including signal sequence) was visible in the culture supernatant for both overexpressing clones H9 and F3 (Fig 2B), confirming p1/s1 secretion.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Ecto-nucleotidase and -endonuclease activities of L. major increase in the stationary-phase.

A- Depiction of the genomic p1/s1 cluster of six genes (LmjF.30.1460 – 1510) sharing a conserved domain but encoding either a signal sequence or a transmembrane domain (TMD). Indicated 5’sgRNA and 3’sgRNA were used for CRISPR Cas9-mediated inducible knockout of the cluster. Created in BioRender. Schmelzle, S. (2026) https://BioRender.com/mlghi1j. B- Western blot to determine p1/s1-His₆ localization in 12x10⁶ L. major promastigotes overexpressing a C-terminally His₆-tagged LmjF.30.1500 construct (clones H9 and F3) and a control parental strain (par.). Secreted, tagged proteins were enriched with magnetic Ni-NTA beads from the corresponding culture supernatant. a-Tubulin was used as loading control for pelleted parasites. Representative blot of n = 3. C- Superimposition of the AlphaFold-predicted structure of Q4Q7F3 (orange, encoded by LmjF.30.1510) onto the experimentally determined structure of a homologous enzyme in Arabidopsis thaliana (green; PDB ID: 3W52). The active site residues and the bound adenine base (cyan) in 3W52 are shown as sticks. Coordinated zinc ions are represented as gray spheres, and the sulfate ion trapped among the zinc ions is shown in ball-and-stick representation. D- Multiple sequence alignment of Q4Q7F8-F3, as encoded in the LmjF.30.1460-1510 cluster. The five regions corresponding to homologous active sites in homologous enzymes are highlighted in yellow boxes. E- Ecto-3’-nucleotidase activity of 2x10⁶ intact logPh or statPh L. major and secreted enzymes in the corresponding supernatant (SN) was measured in a colorimetric assay as µM P_i released from 3’-AMP in one hour. n = 4. F- Ecto-5’-nucleotidase activity of 2x10⁶ intact logPh or statPh L. major and secreted enzymes in the corresponding supernatant (SN) was measured in a colorimetric assay as µM P_i released from 5’-AMP in one hour. n = 3. G- Relative endonuclease activity measured by degradation of double-stranded M13 phage genome on 2x10⁶ intact logPh or statPh L. major promastigotes. Pellet and supernatant were separated and circular, double-stranded M13 DNA (dsM13) added for 60 min. Remaining intact substrate was assessed on a 0.8% TAE-agarose gel, semi-quantified using ImageJ and normalized to intact substrate not incubated with parasites. n = 3. H- Representative gel of G, box denotes area of semi-quantification. dsM13: undigested product; Lm: control with parasites but no substrate; DNase: Substrate incubated with DNase, first and last lane: DNA ladder.

https://doi.org/10.1371/journal.ppat.1014197.g002

To identify the putative homologous active site of p1/s1, AlphaFold-predicted structures of proteins coded in the LmjF.30.1460 -1510 cluster were superimposed on homologous template structures, neglecting the putative signal sequence (Fig 2C). Based on homology to Endonuclease 2 from Arabidopsis thaliana (3W52), five regions were identified to participate in the p1/s1 active site (Fig 2D), coinciding with previous reports on the L. donovani 3’-nucleotidase/nuclease homologue LdNuc^S [8]. In contrast to the putative membrane-bound proteins Q4Q7F8-F5, encoded by LmjF.30.1460-90, only the secreted enzymes Q4Q7F4-3, encoded by LmjF.30.1500-10, contain this active site. Beyond p1/s1, further enzymes (Q66VY6, Q4QGQ3, Q4Q630) possess a homologous active site, indicating a putative functional redundancy.

Further, to assess overall 3’-nucleotidase activity arising from ecto-enzymes of L. major, and to compare p1/s1 expression and 3’-nucleotidase activity patterns, we quantified the release of free phosphate (P_i) from 3’-AMP. Intact parasites, harboring all membrane-bound ecto-enzymes, were separated from the supernatant, which contains secreted enzymes. Besides the secreted p1/s1, the L. major genome codes for additional 3’-nucleotidases/nucleases predicted to be membrane-localized. Comparing logPh and statPh promastigotes, a significant increase in 3’-nucleotidase activity in both the pellet and the supernatant fraction was observed, corresponding well to the elevated p1/s1 protein levels in statPh (Fig 1C). Overall, activity associated with pelleted parasites was higher (Fig 2E), likely related to non-secreted 3’-nucleotidases/nucleases. In contrast, degradation of 5’-AMP through 5’-nucleotidases was generally much lower than 3’-nucleotidase activity (Fig 2F).

Accordingly, ecto-endonuclease activity was determined using double-stranded M13 phage genome (dsM13) as substrate. Its degradation was visualized on an agarose gel for semi-quantification. Similar to 3’-nucleotidase activity, endonuclease activity is increased in statPh and a higher activity was observed on pelleted parasites than in the supernatant (Fig 2G and 2H). These results align with the pattern seen for 3’-nucleotidase activity and p1/s1 protein expression, suggesting both functions arise from p1/s1.

Presence of 3’-AMP during L. major hMDM in vitro infection mimics anti-inflammatory effects of extracellular adenosine

As shown in Fig 2E, L. major statPh promastigotes can hydrolyze extracellular 3’-AMP to adenosine and P_i. Extracellular adenosine is a well-characterized anti-inflammatory stimulant of immune cells. GM-CSF-differentiated macrophages stimulated with IFNγ exhibit the highest share of cells among tested macrophage phenotypes positive for its cell surface receptors A_2a and A_2b (Fig BA and BB in S1 Appendix). Activation of these receptors can result in anti-inflammatory immune modulation. Thus, we examined whether macrophage exposure to adenosine or 3’-AMP during L. major infection can reduce the pro-inflammatory cytokine response. Indeed, TNFα release from macrophages 3 hours post Leishmania infection decreases dose-dependently with increasing concentrations of adenosine. At 100 µM adenosine, TNFα levels align with those of uninfected controls (Fig 3A and Fig BC in S1 Appendix). Addition of 3’-AMP to the infection can mimic the effect observed with adenosine to some extent.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Presence of adenosine or 3‘-AMP during L. major^DsRed infection of hMDM decreases TNFα secretion and PBL proliferation.

hMDM (GM-CSF+IFNγ) were infected with statPh L. major^DsRed promastigotes.3 hours post infection (pi), extracellular parasites were washed off and autologous peripheral blood lymphocytes (PBLs) stained with CellTrace Far Red were added in a 1 macrophage: 5 PBLs ratio. During infection and coculturing, adenosine (Ado) or 3’-AMP were maintained in the indicated concentrations. A- Soluble TNFα secreted from hMDM (GM-CSF + IFNγ) was determined in supernatant collected 3 hours post infection with MOI = 10 using ELISA and normalized to the infected sample without additives. n = 5-7 donors in 4 independent experiments. Absolute values available in Fig BC in S1 Appendix. B- Relative parasite burden measured in flow cytometry as DsRed mean fluorescence intensity of infected hMDM (GM-CSF + IFNγ) in a hMDM:PBL coculture and normalized to the infected control without additives, determined 4 days pi with a MOI = 20 of L. major^DsRed. Absolute values available in Fig BD in S1 Appendix. C- Percentage of proliferated PBLs in a hMDM:PBL coculture was determined as viable CellTrace_low lymphocytes 4 days pi with a MOI = 20 of L. major^DsRed and normalized to the infected control without additives. Absolute values available in Fig BE in S1 Appendix. D- Pearson correlation analysis of relative PBL proliferation against relative TNFα secretion and E- relative PBL proliferation and relative parasite burden. r: Pearson correlation coefficient. B-E: n = 9-10 donors in 4 independent experiments.

https://doi.org/10.1371/journal.ppat.1014197.g003

TNFα is a known driver of the antileishmanial host response [51]. Neutralization of TNFα using specific monoclonal antibodies was previously shown to reduce Leishmania-induced proliferation of autologous peripheral blood lymphocytes (PBLs) in a hMDM:PBL coculture [52]. T cells are the predominant cell type within PBLs. Thus, we tested whether extracellular adenosine exposure impacts CD14^- PBL proliferation 4 days post infection. Although parasite burden of macrophages was not significantly affected, an underlying trend towards an increased parasite burden upon addition of adenosine or 3’-AMP was observed (Fig 3B and Fig BD in S1 Appendix). Relative to the control, reduction in proliferation of PBLs and TNFα secretion indeed was apparent for samples treated with adenosine, 3’-AMP again mimicked this (Fig 3C). Addition of adenosine or 3’-AMP to uninfected cocultures did not affect the share of proliferating PBLs. The proportion of proliferating cells increased greatly upon infection (Fig BE in S1 Appendix). Additionally, a mild to moderate correlation between relative PBL proliferation and TNFα secretion or parasite burden in infected samples was identified (Fig 3D and 3E), suggesting accessory factors induced by adenosine signaling also impact the observed effects. Thus, by generating adenosine through their 3’-nucleotidase activity, L. major can reduce certain elements of an early immune response.

p1/s1 is dispensable for L. major but knockout mutants show reduced 3’-nucleotidase/nuclease activity

To confirm that the observed ecto-3’-nucleotidase/nuclease activity of L. major arises from p1/s1, the generation of knockout parasites lacking the p1/s1 cluster was attempted. However, CRISPR-Cas9-based double homologous replacement approaches [28] failed to yield viable null mutants when sgRNAs binding downstream of LmjF.30.1460 and upstream of LmjF.30.1510 were used (Fig 2A). Targeting single genes in the p1/s1 cluster was not possible due to very high homology between coding and intergenic regions within the cluster. Neither facilitated approaches, using stable genetic integration of a LmjF.30.1500 construct into the SSU locus, or nutritional complementation by supplementing 100 µM adenosine to the medium, yielded viable clones.

Thus, an inducible diCre approach [29] was applied. Excision of LmjF.30.1460-1510 was induced by addition of rapamycin, that catalyzes dimerization of the Cre recombinase in two clones (E10 and F4) carrying a floxed p1/s1 cluster (Fig 4A). The eGFP gene of the 5’ flox casette is excised together with the gene cluster during the recombination event. Therefore, loss of eGFP signal over time was used to track successful recombination. Up to 7 days post rapamycin induction, eGFP positivity was lost continuously in both clones, while a 3’ integrated mCherry signal was retained (Fig 4B and 4C). Following serial dilution to yield single-cell clones, genomic loss of LmjF.1500/10 and eGFP was confirmed by diagnostic PCR (Fig 4D and 4E). One confirmed L. major^Δp1/s1 single-cell clone per floxed strain was chosen for further experiments as induced p1/s1 knockouts (iKOs). Based on L. major^Δp1/s1 iKO E10, an addback strain was established by genomic integration of LmjF.1500 into the SSU locus. In comparable culture conditions, both p1/s1 iKO strains show greatly reduced parasite density in the logPh and statPh compared to the parental strain, whereas the addback strain showed intermediate growth behavior (Fig 4F).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. DiCre based inducible knockout approach yields L. major p1/s1 null mutants.

A- Schematic visualization of the inducible, dimerizable Cre (diCre) system used to generate L. major^Δp1/s1. Full floxed parasites with flox cassettes integrated up- and downstream of the p1/s1 locus (top) expressing a diCre construct were induced with rapamycin to initiate Cre-mediated excision of the p1/s1 cluster alongside the eGFP gene. Successful recombination (bottom) can be traced by the loss of eGFP signal. Created in BioRender. Schmelzle, S. (2026) https://BioRender.com/62mz7co. B, C- eGFP and mCherry signal (circles) upon diCre induction with rapamycin (+rapa) measured in flow cytometry until 9 days post induction in two p1/s1 full floxed clones E10 and F4. As controls, respective strains not induced with rapamycin (-rapa, squares) were included. D- Diagnostic PCR to detect eGFP and p1/s1 loci in genomic DNA of single cell clones derived from floxed clones E10 and E- F4 post rapamycin induction. Confirmed L. major^Δp1/s1 clones used for further experiments are marked in gray and further named as E10 and F4 induced knockouts (iKO). L. major^{Cas9/T7/diCre} (DiCre) and induced, floxed parasites before limiting dilution (pre.dil.) and were used as controls. F- Cell numbers of viable and dead L. major^{Cas9/T7/diCre/3‘flox}, L. major^Δp1/s1 clones E10 & F4 and L. major^{p1/s1 addback} were counted on day 3 (logPh) and 7 (statPh) post passaging for viable and dead parasites. All strains were grown under selection pressure of the same antibiotics and were seeded as 0.5x10⁶ viable parasites on day 0. n = 3 of consecutive passages.

https://doi.org/10.1371/journal.ppat.1014197.g004

When using L. major^Δp1/s1 for 3’-nucleotidase activity assays, both null mutants show considerably diminished activity in the supernatant fraction, whereas the addback partly restored activity. 3’-nucleotidase activities on the pellet fraction are inconclusive but suggest no difference between parental and knockout strains (Fig 5A). Similarly, L. major^Δp1/s1 exhibit declined endonuclease activity on double-stranded substrate (Fig 5B and 5C) and delayed degradation of single-stranded substrate (Fig 6D-F), both in the pellet and the supernatant fractions. L. major^Δp1/s1 addback promastigotes here behaved like the parental strain. Thus, behavior of the parental strain coincided with 3’-nucleotidase/nuclease activity observed in Fig 2, whereas null mutants exhibit a distinct phenotype with delayed or reduced degradation of substrates. The L. major genome codes for further 3’-nucleotidases/nucleases besides p1/s1. These redundant enzymes Q66VY6, Q4Q630 and Q4QGQ3 all possess an active site homologous to p1/s1, so residual enzymatic activity in mull mutants was expected.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. L. major p1/s1 null mutants show reduced 3’-nucleotidase/nuclease activity.

A- 3‘-nucleotidase activity was measured as µM released P_i from 3‘-AMP on 2x10⁶ total, intact L. major statPh promastigotes (L. major^{Cas9/T7/diCre/3‘flox} L. major^Δp1/s1 clones E10 & F4 and L. major^{p1/s1 addback}) and the corresponding supernatant (SN). Parasite pellet and supernatant were separated and 3‘-AMP as substrate added for 1 h, P_i content was determined with a malachite green based assay. n = 4. B- Endonuclease activity measured by degradation of double-stranded M13 phage genome on 2x10⁶ total L. major statPh promastigotes (L. major^{Cas9/T7/diCre/3‘flox} L. major^Δp1/s1 clones E10 & F4 and L. major^pSaddback). Pellet and supernatant (SN) were separated and circular, double-stranded M13 DNA added for 90 min. Remaining intact substrate was assessed on a 0.8% TAE-agarose gel, semi-quantified using ImageJ and normalized to intact substrate not incubated with parasites (dsM13). Representative gel shown. Also included are controls containing only parasites (Lm) and dsM13 incubated with DNase. n = 3. Uncropped gel available in Fig C in S1 Appendix. C, D- Endonuclease activity on L. major pellet and supernatant (SN) fraction was determined as in B, but on single-stranded M13 phage genome (ssM13) for the stated reaction time. n = 3. E- Representative gels of C & D.

https://doi.org/10.1371/journal.ppat.1014197.g005

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. L. major p1/s1 null mutants bypass loss of p1/s1 via redundant purine salvage.

A- Ecto-5’-nucleotidase activity of 2x10⁶ intact statPh L. major promastigotes (L. major^{Cas9/T7/diCre}, L. major^Δp1/s1 clones E10 & F4) and secreted enzymes in the corresponding supernatant (SN) was measured in a colorimetric assay as µM P_i released from 5’-AMP in one hour. n = 3. B- Principal component analysis of the quantitative proteome data generated on L. major^Δp1/s1 clones E10 & F4 and a parental strain (L. major^{Cas9/T7/diCre/3‘flox}) by label-free mass spectrometry. C- Venn diagram showing the number of proteins with increased abundance in L. major^Δp1/s1 clones compared to their parental strain. Template created using [43]. D, E- Volcano plot of quantified protein abundance of L. major^Δp1/s1 clones E10 & F4, respectively, in comparison to their parental strain. Cutoffs were set at log₂FC < -0.5, > 0.5 and p-value < 0.01. Denoted are highly abundant proteins of the purine salvage pathway listed in Table 3 Volcano plots visualized with VolcaNoseR [42]. B-E Data analysis based on n = 7 biological replicates in 3 technical repeats each.

https://doi.org/10.1371/journal.ppat.1014197.g006

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Purine salvage proteins with increased abundance in both L. majorΔp1/s1 clones compared to a parental strain, ranked by p-value in E10 comparison.

https://doi.org/10.1371/journal.ppat.1014197.t003

L. major can compensate for p1/s1 loss by enrichment of purine salvage proteins

Essentiality of the p1/s1 cluster remains elusive, as conventional and facilitated CRISPR-Cas9 methods were lethal to the parasites (n = 6) but inducible gene deletion using the same sgRNA templates yielded viable mutants. This observation raised the hypothesis that the gradual depletion of p1/s1 during diCre induction allowed adaptation of the parasites to compensate for this deletion. To test this, we measured the parasites’ 5’-nucleotidase activity. This activity, serving as an alternative source of extracellular adenosine for purine salvage, can aid the parasites in supplying metabolite needs. Indeed, we found 5’-nucleotidase activity to be elevated on L. major^Δp1/s1 (Fig 6A).

To quantitatively assess global proteomic dynamics in p1/s1 null mutants and expression of further purine salvage-linked proteins, a label-free mass spectrometry analysis was applied. In total, 5890 proteins could be detected, corresponding to 73.28% of the reference proteome.

A principal component analysis revealed similarity between replicates of both p1/s1 iKO strains, whereas the parental strain clustered separately (Fig 6B). Successful depletion of p1/s1 was confirmed at the protein level. We found 604 proteins to be differentially expressed between parental strain and iKO E10, 788 between parental strain and iKO F4, with 467 regulated proteins being shared by both comparisons (Fig DA in S1 Appendix). Focusing only on more abundant proteins in p1/s1 null mutants, 175 are shared by both deletion strains (Fig 6C). Strikingly, among the proteins with strongly increased abundance in L. major^Δp1/s1 strains are enzymes with redundant 3’-nucleotidase-nuclease activity, membrane-bound phosphatases and purine transporters (Fig 6D and 6E; Table 3). Gene ontology terms for purine nucleotide binding, adenyl nucleotide binding, and nucleotide binding were significantly enriched for p1/s1 iKO E10 but not for p1/s1 iKO F4 (Fig DB and DC in S1 Appendix). All of these hits can be involved in purine salvage by nucleic acid and nucleotide catabolism or purine precursor transport, providing the parasites with purines despite their reduced 3’-nucleotidase/nuclease activity. Notably, a xanthine phosphoribosyltransferase (Q4QCC2) as key enzyme in intracellular Leishmania purine metabolism [53], was more abundant in p1/s1 iKO E10. Since no copy number variations were detected in the genomic loci of these compared to a parental strain (NCBI SRA data: PRJNA1330523; https://www.ncbi.nlm.nih.gov/sra/PRJNA1330523, Fig E in S1 Appendix), upregulation by genomic adaptation is unlikely. This rather points to a posttranscriptional regulation, as typical for Leishmania.

Loss of p1/s1 does not impact cell death hallmarks upon miltefosine treatment

A 3’-nucleotidase/nuclease in L. infantum was shown to play a role in parasite susceptibility to miltefosine, an anti-leishmanial drug [54]. As we have identified p1/s1 as highly abundant in pre-cell death stressed promastigotes, we speculated whether it has a role in parasite cell death, potentially as executer by fragmenting DNA. However, upon treatment with increasing doses of miltefosine, null mutants did not show altered sensitivity compared to the parental or addback strains, neither in viability, phosphatidylserine exposure nor DNA fragmentation (Fig 7A-C). Half-maximal effective concentrations (EC₅₀) of miltefosine (Fig 7D) neither varied for the null mutants nor indicated opposed effect for the addback strain. Thus, p1/s1 does not seem to be involved in intracellular cell death pathways like DNA degradation or miltefosine sensitivity of L. major. Alternatively, its involvement could be masked by compensation through redundant enzymes.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 7. L. major p1/s1 null mutants do not show altered cell death hallmarks upon miltefosine treatment.

L. major logPh promastigotes (L. major^{Cas9/T7/diCre/3‘flox}, L. major^Δp1/s1 clones E10 & F4 and L. major^{p1/s1 addback}) were treated with increasing concentrations of miltefosine (0.1-60µM) for 48 h and cell death hallmarks analyzed by different assays. A- Cellular metabolism of 5x10⁶ parasites was quantified using a colorimetric MTT assay, absorbance at 540 nm was normalized to the sample containing 1.5% DMSO as carrier control; n = 3. B- Phosphatidylserine exposure on the cell surface of 2x10⁶ parasites was stained using FITC labelled AnnexinV. Percentage of phosphatidylserine positive parasites (AnnV+) was quantified by flow cytometry; n = 3–5. C- 2x10⁶ parasites were analyzed for fragmented DNA, as seen in the SubG₁ phase, by stochiometric DNA staining using SYTOXGreen analyzed by flow cytometry; n = 3–5. D- Summarized EC₅₀ values obtained from non-linear 4-parameter regression of all cell death assays for L. major^{Cas9/T7/diCre/3‘flox} (par.), L. major^Δp1/s1 clones E10 & F4 and L. major^{p1/s1 addback} (addb.); ± 95% confidence interval.

https://doi.org/10.1371/journal.ppat.1014197.g007

p1/s1 null mutants cannot decrease PBL proliferation in hMDM:PBL coculture infection and are more susceptible for neutrophil killing

To determine whether p1/s1 serves as virulence factor during macrophage infection, L. major^Δp1/s1 were used in an in vitro infection model. In the promastigote stage, p1/s1 null mutants showed growth deficiency (Fig 4F) and reduced utilization of extracellular 3’-AMP and nucleic acids (Fig 5). Presuming that both effects are transferable to intracellular amastigotes, they could lead to reduced parasite persistence in host macrophages. However, in an in vitro hMDM:PBL coculture system, infectivity of L. major^Δp1/s1 compared to the parental strain neither decreased in terms of infection rate nor parasite burden (Fig 8A and 8B). This also confirmed that the slightly decreased proportion of metacyclic promastigotes in the p1/s1 iKO (Fig FA in S1 Appendix) does not to affect macrophage infection. As shown above, we observed a reduction in PBL proliferation when adenosine or 3’-AMP was added to a hMDM:PBL coculture during infection (Fig 3C). p1/s1 deletion alone did not influence absolute PBL proliferation (Fig 8C). A suppressive effect comparable to Fig 3C was apparent for the parental strain in the presence of adenosine or 3’-AMP. However, L. major^Δp1/s1 lost the ability to suppress PBL proliferation when 3’-AMP was added (Fig 8D), likely due to reduced degradation to adenosine. Addition of adenosine as a positive control still reproduced this reduction.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 8. L. major p1/s1 null mutants cannot decrease PBL proliferation upon presence of 3’-AMP and show decreased degradation of neutrophil extracellular traps.

hMDM (GM-CSF + IFNγ) were infected with statPh L. major^{Cas9/T7/diCre/3‘flox}, L. major^Δp1/s1 clones E10 or F4 promastigotes at an MOI=20. 3 hours post infection (pi), extracellular parasites were washed off and autologous peripheral blood lymphocytes (PBLs) stained with CellTrace Far Red were added in a 1 macrophage: 5 PBLs ratio for 4 days. During infection and coculturing, adenosine (Ado) or 3’-AMP were maintained at 100 µM, if applicable. Intracellular L. major were stained with α-Lm mouse serum and labelled with α-mouse-AlexaFluor647 antibody. A- L. major infection rate of macrophages in a hMDM:PBL coculture 4 days pi. The percentage of AlexaFluor647 + viable macrophages was determined in flow cytometry to give the share of infected macrophages. B- Parasite burden 4 days pi in a hMDM:PBL coculture. Parasite burden was measured as AlexaFluor647 mean fluorescence intensity of viable, infected macrophages and normalized to the control infected with parental strain. C- Percentage of proliferated PBLs in a hMDM:PBL coculture was determined as viable CellTrace_low lymphocytes 4 days pi. D- Normalized share of proliferated PBLs in a hMDM:PBL coculture 4 days pi, relative to the respective strain’s infected control without adenosine or 3‘-AMP. Absolute percentage of PBL proliferation is provided in Fig FB in S1 Appendix. A-D: n = 6 donors in 3 independent experiments, Welsh’s t-test. Human primary neutrophils were used to study p1/s1 dependent degradation of neutrophil extracellular trap (NET) degradation. E- Nuclease activity measured by degradation of NET enriched supernatants on 2x10⁶ intact L. major statPh promastigotes (P, L. major^{Cas9/T7/diCre/3‘flox} L. major^Δp1/s1 clones E10 & F4) and the respective supernatant (SN). Pellet and supernatant were separated and 500µg NET DNA added for 2 hours. Remaining intact NET DNA scaffold was assessed on a 0.8% TAE-agarose gel, semi-quantified using ImageJ and normalized to substrate not incubated with parasites. n = 3 donors in three independent experiments. F- Representative gel of E, box indicates area used for semi-quantification with ImageJ. G- L. major statPh promastigotes (L. major^{Cas9/T7/diCre/3‘flox} L. major^Δp1/s1 clones E10 & F4) were incubated with human neutrophils at MOI = 1, optionally in the presence of DNase, for 3 hours at 37 °C. CytochalasinD (CytD) inhibits phagocytosis of parasites into neutrophils. After further incubation at 27 °C for 2 days, motile, viable promastigotes were counted for each sample and normalized to the respective control incubated without neutrophils. n = 4–5 donors in 2–3 independent experiments.

https://doi.org/10.1371/journal.ppat.1014197.g008

Besides the p1/s1 3’-nucleotidase activity that potentially counteracts inflammatory signaling by generating extracellular adenosine, the endonuclease activity of p1/s1 can also bring advantages for parasite survival in mammalian hosts. Ecto-nuclease activity of p1/s1 can be beneficial for Leishmania promastigotes when encountering neutrophils in the early stages of infection. In response to pathogen detection, neutrophils can extrude neutrophil extracellular traps (NETs) that have antimicrobial capacity. Several pathogens possess extracellular nucleases to evade NET-mediated killing [55– 58]. L. major was shown to induce NET formation upon neutrophil infection [59, 60]. Whether p1/s1 allows L. major to degrade and survive NETs was examined on human primary neutrophils (PMNs) and cell-free NET-enriched supernatants. Control promastigotes of the parental strain were able to efficiently degrade NET DNA, whereas L. major^Δp1/s1 were less able to do so (Fig 8E and 8F). This effect was prominent for both parasite pellet and supernatant, comparable to nuclease activity shown in Fig 5B. To test for antileishmanial effects of NETs on parental and p1/s1 deficient strains, parasites were incubated with PMNs whose phagocytosis was inhibited by Cytochalasin D. The majority of parental strain promastigotes survived this co-incubation but both L. major^Δp1/s1 strains were significantly reduced in number of surviving parasites. When we added DNaseI to degrade NETs, these leishmanicidal effects were lost and survival of the iKO strains resembled that of the parental strain (Fig 8G).

Taken together, both of the dual p1/s1 activities are able to support evasion of different branches of the innate immune response. Thus, this ecto-enzyme is key in host-Leishmania interactions during early infection.