The SET7 protein of Leishmania donovani moderates the parasite’s response to a hostile oxidative environment

Abstract

SET domain proteins methylate specific lysines on proteins, triggering stimulation or repression of downstream processes. Twenty-nine SET domain proteins have been identified in Leishmania donovani through sequence annotations. This study initiates the first investigation into these proteins. We find LdSET7 is predominantly cytosolic. Although not essential, set7 deletion slows down promastigote growth and hypersensitizes the parasite to hydroxyurea-induced G1/S arrest. Intriguingly, set7-nulls survive more proficiently than set7^+/+ parasites within host macrophages, suggesting that LdSET7 moderates parasite response to the inhospitable intracellular environment. set7-null in vitro promastigote cultures are highly tolerant to hydrogen peroxide (H₂O₂)-induced stress, reflected in their growth pattern, and no detectable DNA damage at H₂O₂ concentrations tested. This is linked to reactive oxygen species levels remaining virtually unperturbed in set7-nulls in response to H₂O₂ exposure, contrasting to increased reactive oxygen species in set7^+/+ cells under similar conditions. In analyzing the cell’s ability to scavenge hydroperoxides, we find peroxidase activity is not upregulated in response to H₂O₂ exposure in set7-nulls. Rather, constitutive basal levels of peroxidase activity are significantly higher in these cells, implicating this to be a factor contributing to the parasite’s high tolerance to H₂O₂. Higher levels of peroxidase activity in set7-nulls are coupled to upregulation of tryparedoxin peroxidase transcripts. Rescue experiments using an LdSET7 mutant suggest that LdSET7 methylation activity is critical to the modulation of the cell’s response to oxidative environment. Thus, LdSET7 tunes the parasite’s behavior within host cells, enabling the establishment and persistence of infection without eradicating the host cell population it needs for survival.

The SET domain proteins get their name from the three Drosophila proteins the domain was first identified in: suppressor of variegation [Su(var)3-9], enhancer of zeste [E(z)], trithorax (reviewed in (1)). Initially identified as proteins that methylate histones at specific lysine residues, SET proteins were subsequently found to target non-histone substrates as well, and can mediate monomethylation, dimethylation, and/or trimethylation of their target residues. Common histone targets include H3K4, H3K9, H3K27, H3K36, and H4K20 (reviewed in (1, 2, 3)). The impacts of these histone methylation events vary widely, and thus while H3K4 methylation is associated with transcriptional upregulation, H3K9 and H3K27 methylation marks mediate transcriptional repression. Non-histone substrates include tumor suppressors (such as p53 and Rb), transcription factors (like GATA4 and TAF₁₀), and signaling proteins (such as STAT3) (reviewed in (1, 2, 3)). SET domain proteins modulate a wide range of cellular processes, such as DNA replication, DNA repair, transcriptional activation as well as silencing, mRNA splicing, heterochromatin formation, and X-chromosome inactivation. Abrogation of SET domain protein functions have been linked to various types of cancer (1, 3).

While exhaustively studied in yeast and mammalian cells, limited information is available about protozoan SET proteins. The most widely investigated are the Plasmodium falciparum SET proteins, where six of the ten proteins identified appear to be essential for survival of blood stage parasites. The histone target residues of six PfSET proteins have been identified and their functional roles characterized to varying degrees (4, 5, 6). The Tetrahymena EZL3 SET protein has been found to modulate development and progeny viability (7). The non-nuclear Toxoplasma SET domain protein modulates host cell invasion as well as the parasite’s exit from the host cell (8). In vitro biochemical assays have uncovered the target residues of three SET domain proteins identified in Entamoeba histolytica, two of which appear to be involved in phagocytosis (9).

Leishmania species are trypanosomatids that are endemic to 88 countries and according to WHO around 12 million people are at risk globally (https://www.who.int/news-room/fact-sheets/detail/leishmaniasis). Leishmaniases are manifested in three forms: cutaneous, subcutaneous, and visceral and different Leishmania species cause the different forms of the disease. In South Asia the prevalent form of the disease is visceral leishmaniasis or kala-azar, caused by Leishmania donovani. Visceral leishmaniasis is treatable, but treatment regimens are lengthy, expensive, and have toxic side effects. Adding to that the problems of growing drug resistance, risks due to HIV-Leishmania coinfection, and emerging threat of post kala-azar dermal leishmaniasis in the Indian subcontinent, this parasite’s cellular biology remains an area of interest to several research groups around the world. While no one has directly investigated the role of SET domain proteins in Leishmania species, the Leishmania tarentolae SET domain protein LtaP35.2400 has been found to be a part of a multiprotein complex that also carries the JBP3 protein, which has been previously shown to play a role in regulating transcription termination events (10). Twenty nine SET domain proteins have been identified in Trypanosoma brucei by whole-genome sequence annotations (11), and through orthology searches in the Tritryp DB (www.tritrypdb.org) we have been able to identify orthologs of all 29 in L. donovani. The subcellular localization of the T. brucei SET proteins has been examined in blood form as well as procyclic form (PF), by epitope-tagging coupled to immunofluorescence (11, 12, 13). Eight of the twenty-nine proteins localize to the nucleus in the blood form parasites, though also found in the cytoplasm. Four of these eight proteins localize to the nucleus in the PF parasites as well. Six SET proteins localize to the nucleus (though also found in the cytoplasm) in the PF parasites only. Two proteins, TbSET26 and TbSET27, have been found to be enriched at the transcriptional start regions (TSRs) (11). Furthermore, TbSET27 has been found to exist as part of a multiprotein complex (SPARC) that modulates the accuracy of transcription initiation at the TSRs (14).

This report presents the data from the first study across Leishmania species that directly investigates a SET protein, examining the functional role of the protein encoded by the L. donovani 1S ortholog of the gene LdBPK_360230.1, which we name LdSET7 as it is the Leishmania ortholog of the TbSET7 protein. Our study reveals that the gene encoding LdSET7 (set7) is not essential for cell survival. The LdSET7 protein is found to play a role in mediating the parasite’s response to an oxidative environment in both promastigotes (the extracellular form in insect host) and amastigotes (the intracellular form in mammalian host).

Results

Leishmania donovani SET7 is constitutively expressed and is predominantly cytosolic

In the absence of genome sequence information of L. donovani 1S (Ld1S), the set7 gene (∼1.42 kb) of Ld1S was cloned by amplification using genomic DNA as template with primers designed against its ortholog in L. donovani BPK282A1 (LdBPK_360230.1), whose sequence was obtained from TriTrypDB ((15), www.tritrypdb.org). The cloned amplicon was sequenced to confirm its veracity (GenBank Accession no: OR479702). It was found to carry only one SNP in comparison with the LdBPK_360230.1 gene, which resulted in a change from leucine to valine. A comparison of the derived amino acid sequence of LdSET7 with orthologs of the same protein in other trypanosomatid species using Clustal Omega analysis (16) and blastp (https://blast.ncbi.nlm.nih.gov.in) revealed that LdSET7 showed 35 to 40% identity over a coverage of 91% with SET7 of Trypanosoma species and 90 to 99% identity with SET7 of other Leishmania species over 100% coverage (Fig. S1). Analysis of the SET7 amino acid sequence revealed that the ∼51.5 kDa protein carries a SET domain between amino acids 304 to 422, and a post-SET domain at the C-terminal end between residues 430 to 446 (Fig. S2A). SET domains are often flanked by pre-SET and post-SET domains. While the pre-SET domain is believed to overall help stabilize the SET protein structure via its interactions with the residues of the SET domain, the post-SET domain forms part of the active site hydrophobic channel (17).

The digenetic Leishmania parasites exist extracellularly as promastigotes in the insect host, initially as free-living flagellate parasites in the midgut and subsequently in the salivary glands of the insect, whereupon they are introduced into the mammalian host when the insect feeds on the host’s blood. In the mammalian host the parasite is taken up by macrophages where they take up residence and propagate as intracellular aflagellate amastigotes. The expression of LdSET7 in logarithmically growing as well as stationary phase L. donovani promastigotes was analyzed in whole-cell extracts using anti-LdSET7 antibodies already available in the lab and found to be comparable at both stages (Fig. S2B). Examination of whole-cell lysates isolated from procyclics (early stage promastigotes normally existing in insect midgut) and metacyclics (late stage promastigotes found usually in insect salivary glands; the infective stage) revealed that expression was more or less equivalent in both types of promastigotes (Fig. S2C). Upon examining LdSET7 expression at different stages of the cell cycle using promastigotes that had been synchronized at the G1/S boundary with a hydroxyurea (HU) block and then released into S phase (described in Experimental procedures), it was observed that LdSET7 expression was maximum in S phase (Fig. S2D). This was in contrast to expression of another L. donovani SET protein, encoded by LdBPK_212120.1 (ortholog of TbSET29 and thus named LdSET29), which was expressed more or less equivalently at all stages of cell cycle (Fig. S2D). As the antibodies did not work in immunocytochemistry experiments, the set7 gene was expressed in fusion with a C-terminal FLAG tag in L. donovani promastigotes (Fig. S3A), and the subcellular localization of LdSET7-FLAG analyzed by indirect immunofluorescence as described in Experimental procedures, using kinetoplast morphology and segregation pattern as cell cycle stage marker (18). LdSET7-FLAG was found to be primarily cytosolic at all stages of the cell cycle (Fig. S3B). Collectively, these data demonstrate that LdSET7 is robustly expressed in promastigotes (in both procyclics and metacyclics) and is constitutively expressed, predominantly in the cytoplasm, with maximal expression detectable in S phase.

Parasites partially depleted of LdSET7 display greater competency in survival within host macrophages than WT parasites

To investigate the role of LdSET7 in the Leishmania cell, we replaced the set7 genomic alleles sequentially by homologous recombination (as detailed in Experimental procedures). The first allele was replaced with a hygromycin resistance cassette and the authenticity of recombination at both ends verified by PCRs across the deletion junctions (Fig. 1A). Western blot analysis revealed a corresponding decrease in the levels of the LdSET7 protein (Fig. 1B). Analysis of the growth pattern of these set7-heterozygous knockout (set7^−/+) promastigotes revealed that the observed reduction in LdSET7 expression upon knockout of one set7 allele did not have any impact on growth and survival under normal conditions (Fig. 1C). The cell cycle progression pattern of set7^−/+ cells, synchronized at the G1/S boundary with HU, and then released into S phase (as described in Experimental procedures), was similar to that of set7^+/+ cells (Fig. S4). To examine the effect of loss of one set7 allele on amastigote growth and propagation, we infected set7^−/+ parasites into murine macrophage cells J774A.1 (as described in Experimental procedures) and followed their survival over 48 h after infection. As seen in Figure 1D, the intracellular load of set7^−/+ parasites was comparable to that of set7^+/+ parasites at the end of the 5 h parasite-host cell incubation period (5 h time point), signifying that the ability to infect host cells per se was not altered in set7^−/+ cells. However, unexpectedly, set7^−/+ parasites survived more competently than set7^+/+ parasites within the host cells, as determined by the higher intracellular parasite load 24 h and 48 h after infection (Fig. 1D). Thus, the levels of LdSET7 protein in set7^−/+ parasites were sufficient for the promastigotes to remain unperturbed under normal in vitro growth conditions but insufficient to maintain the normal parasite-host cell balance in case of the amastigotes.

Figure 1.

Effect of partial depletion of LdSET7 on parasite growth and survival.A, creation of set7^−/+. Recombination at both ends was checked by PCRs across the deletion junctions. Positions of primers used are indicated with arrows. F1, F2, F3, F4: forward primers. R1, R2, R3: reverse primers. PCNA served as input template DNA control. Lanes 1: Ld1S genomic DNA template. Lanes 2: set7^−/+ genomic DNA template. M: DNA ladder. B, Western blot analysis of whole-cell extracts isolated from 1 × 10⁸ logarithmically growing promastigotes using anti-SET7 antibodies (1:2500 dilution). Loading control: tubulin. Quantitation was carried out using ImageJ (https://imagej.net/ij/) analysis and plotted values represent average of two experiments. Error bars represent SD, and statistical significance was determined using two-tailed unpaired student’s t test. ∗∗ represents p value <0.005. C, analysis of growth of parasites. Values plotted represent average of three experiments. Each experiment was performed with two technical replicates. Error bars represent SD. D, analysis of parasite survival in host macrophages. Z-stack imaging using confocal microscopy was employed to count the number of intracellular parasites (4′,6-diamidino-2-phenylindole-stained nuclei served as marker). The experiment was performed thrice and average values are plotted in the bar graph. Error bars depict SD, and two-tailed unpaired student’s t test was used to assess statistical significance. ∗∗∗p value <0.0005. ns: statistically not significant. set7^+/+ cells: Ld1S::hyg cells. set7^−/+ cells: set7-heterozygous KO cells. Ld1S, Leishmania donovani 1S.

Promastigotes partially depleted of LdSET7 are more tolerant to an oxidative environment than WT promastigotes

Considering the oxidative intracellular environment of macrophage host cells, we examined the growth of set7^−/+ promastigotes in conditions where oxidative stress has been induced. Thus, set7^−/+ parasite cultures were initiated from stationary phase cultures, and hydrogen peroxide (H₂O₂) varying over 25 to 200 μM was added on day 3 (as described in Experimental procedures). The parasites were incubated in this H₂O₂-containing medium (M199) for 5 h before replacing it with fresh H₂O₂-free medium and continuing incubation. Growth was monitored over the next few days, and on comparing the behavior of set7^−/+ promastigotes with that of set7^+/+ promastigotes, it was observed that while neither cell type displayed any obvious response to lower concentrations of H₂O₂ (25–50 μM), they behaved quite differently in response to higher H₂O₂ concentrations. set7^+/+ promastigotes grew slower in response to 100 μM H₂O₂ treatment and displayed severely compromised growth in response to 200 μM H₂O₂ treatment; in contrast, set7^−/+ parasites displayed a very moderate response to 100 μM H₂O₂ treatment, and were able to recover and continue to grow well in response to 200 μM H₂O₂ treatment as well (Fig. 2A). The altered behavior of LdSET7-depleted parasites upon exposure to H₂O₂ was more apparent when parasites were incubated in 100 μM H₂O₂ over several days (with replenishment of H₂O₂ every 10 h), wherein growth of set7^+/+ cells was distinctly compromised while set7^−/+ cells continued to thrive (Fig. 2B). These data indicate that LdSET7 regulates the parasite’s response to an oxidative milieu under in vitro conditions as well.

Figure 2.

Effect of LdSET7 depletion on promastigote growth and survival in response to H₂O₂exposure.A, cultures were initiated from stationary phase cultures, and 0 to 200 μM H₂O₂ was added to the cultures 48 h after initiation. At the time of addition of H₂O₂, the cultures were split into two, with half the cells receiving H₂O₂ treatment and the other half being carried forward as the untreated culture. After a 5-h exposure to H₂O_2, the parasites were refed with fresh H₂O₂-free medium, and cells were counted every 24 h. Graphs show growth patterns of the cells. The experiment was done thrice, with two technical replicates in each experiment. Values plotted are average of three experiments, and error bars depict SD. The experimental data has been split into four panels for easier viewing and thus the (-H₂O₂) data is identical in all four panels. B, cultures were initiated from stationary phase cultures, and 100 μM H₂O₂ was added to the cultures 48 h after initiation. At the time of addition of H₂O₂, the cultures were split into two, with half the cells receiving H₂O₂ treatment and the other half being carried forward as the untreated culture. H₂O₂-treated cultures were maintained in H₂O₂-containing medium, with H₂O₂ being replenished every 10 h, and cells were counted every 24 h. Left panel: growth analysis of the cultures. Right panel: Survival percent of H₂O₂-treated cultures with reference to untreated cultures was determined by dividing the number of cells in treated cultures by the number of cells in untreated cultures and multiplying by 100. The experiment was done thrice, with two technical replicates in each experiment. Values plotted are average of three experiments, and error bars depict SD. Student’s t test (two-tailed unpaired) was used to assess statistical significance. ∗∗∗p value <0.0005. set7^+/+ cells: Ld1S::hyg cells. set7^−/+ cells: set7-heterozygous KO cells. H₂O₂, hydrogen peroxide; Ld1S, Leishmania donovani 1S.

The set7 gene is not essential for cell survival

The second set7 allele was replaced with a neo^r cassette and genuine recombination at the 3′end checked by PCRs across the deletion junction, while the authenticity of recombination at the 5′end was checked by inverse PCRs across the deletion junction (Fig. S5). Western blot analysis confirmed the absence of LdSET7 expression in set7-nulls (set7^−/−, Fig. 3A), and analysis of growth revealed that promastigotes grew slower when completely devoid of LdSET7, entering log phase 3 days later than set7^+/+ cells and never reaching the same cell density as set7^+/+ cells before entering stationary phase (Fig. 3B). The generation time of set7^−/− cells was found to be considerably longer than that of set7^+/+ cells (∼23 h as compared to the usual ∼9.7 h; Fig. 3C). Examination of cell cycle progression patterns of HU-synchronized promastigotes by flow cytometry analysis, revealed that set7^−/− cells displayed a heightened sensitivity to HU-induced G1/S arrest, with a large fraction of the cells failing to be released into S phase upon removal of HU (Fig. 3D). While the fraction of cells that got released from the HU-induced block appeared to traverse S phase and G2/M in a manner comparable to control cells, no definitive conclusion regarding the cause for increased generation time could be drawn from this data. It is possible that the span of G1 phase is longer in set7^−/− cells; this needs further investigation. From the data in Figures 1 and 3, we concluded that under normal in vitro growth conditions while a partial depletion of LdSET7 (to ∼50% WT levels) did not have any impact on growth and cell cycle progression, complete elimination of LdSET7 expression slowed down growth and increased the generation time more than 2-fold. The survival of the parasite in the absence of the set7 gene indicates that set7 is not essential to the parasite.

Figure 3.

Effect of elimination of set7 gene on promastigote growth and cell cycle progression.A, Western blot analysis of whole-cell extracts isolated from 1 × 10⁸ logarithmically growing promastigotes using anti-SET7 antibodies (1:2500 dilution). Loading control: tubulin. Quantitation was carried out using ImageJ analysis and plotted values represent average of three experiments. Error bars represent SD, and statistical significance was determined using student’s t test. ∗∗∗ represents p value <0.0005. B, analysis of growth of parasites. Cultures were initiated at 1 × 10⁶ cells/ml from stationary phase cultures. Values plotted represent average of three experiments. Each experiment was performed with two technical replicates. Error bars represent SD. C, generation time of cells was determined by initiating cultures from logarithmically growing cultures, at 1 × 10⁶ cells/ml, and diluting the cultures to 1 × 10⁶ cells/ml every 24 h after counting them. The experiment was done thrice and average values are plotted, with error bars depicting SD. D, flow cytometry analysis of HU-synchronized promastigotes. Time points at which cells were sampled are indicated on the left of each row of histograms. Thirty thousand events were analyzed at every time point. M1, M2, and M3 represent gating for cells in G1, S, and G2/M, respectively. Percent cells in each cell cycle phase are indicated in upper right-hand corner of each histogram. The experiment was done thrice, with comparable data, and data of one experiment are shown. set7^+/+ cells: Ld1S::neo-hyg cells. set7^−/+ cells: set7-heterozygous KO cells. set7^−/− cells: set7-nulls. HU, hydroxyurea; Ld1S, Leishmania donovani 1S.

set7-nulls survive more proficiently than WT cells within host macrophages

Even more interestingly, these slower growing set7^−/− parasites (Fig. 3, B and C) survived more proficiently than set7^+/+ parasites within host macrophages, though not to the same extent as set7^−/+ parasites, reflecting the longer generation time of set7^−/− cells as compared to set7^−/+ cells (Fig. 4A). Thus, LdSET7 appears to be moderating the parasite’s response to the inhospitable intracellular environment of host cells. The differential response of set7^−/− promastigotes to an oxidative growth environment in vitro was even more pronounced. While set7^−/− cells in general grew slower than set7^+/+ promastigotes, the parasites did not demonstrate any apparent response to a 5-h H₂O₂ treatment over concentrations of 25 to 200 μM (Fig. 4B) and nor was any response evident when the parasites were incubated in 100 μM H₂O₂ over several days (Fig. 4C). When treated with H₂O₂ over concentrations of 500 to 1000 μM for 5 h, however, set7-nulls did not survive (Fig. 4B, inset in lower right panel). Elimination of LdSET7 from the cell thus appears to make the parasites extremely tolerant to the effects of an H₂O₂-induced oxidative environment.

Figure 4.

Effect of oxidizing environment on set7-null parasites.A, analysis of parasite survival in host macrophages. Z-stack imaging using confocal microscopy was employed to count the number of intracellular parasites (4′,6-diamidino-2-phenylindole-stained nuclei served as marker). The experiment was performed thrice, and average values are plotted in the bar graph. Error bars depict SD, and student’s t test (two-tailed, unpaired) was used to assess statistical significance. ∗∗∗p value <0.0005, ∗∗p value <0.005, and ns: statistically not significant. B, effect of H₂O₂ exposure on promastigote growth and survival. Cultures were initiated from stationary phase cultures, and 0 to 200 μM H₂O₂ was added to the cultures at 7 to 9 × 10⁶ cells/ml. At the time of addition of H₂O₂ the cultures were split into two, with half the cells receiving H₂O₂ treatment and the other half being carried forward as the untreated culture. Parasites were re-fed with fresh H₂O₂ -free medium after a 5 h-exposure to H₂O₂, and cells counted every 24 h. The experiment was done thrice, with two technical replicates in each experiment. Values plotted are average of three experiments and error bars depict SD. The experimental data has been split into four panels for easier viewing and thus the (-H₂O₂) data is identical in all four panels. Inset in lower right panel shows effect of higher H₂O₂ concentrations on set7-null cultures. Cultures were initiated at 1 × 10⁶ cells/ml, and hydrogen peroxide (500 or 1000 μM) added on reaching a cell density of ∼7 to 9 × 10⁶ cells/ml (Day 6). Cells were counted every 24 h thereafter. C, effect of prolonged H₂O₂ exposure on promastigote growth and survival. Cultures were initiated from stationary phase cultures, 100 μM H₂O₂ added to the cultures at cell density 7 to 9 × 10⁶ cells/ml (at the time of addition of H₂O₂ the cultures were split into two, with half the cells receiving H₂O₂ treatment and the other half being carried forward as the untreated culture), and cultures maintained in H₂O₂-containing medium (with H₂O₂ being replenished every 10 h), with cells being counted every 24 h. Upper panel: growth analysis of the cultures. Lower panel: survival percent of H₂O₂-treated cultures with reference to untreated cultures, determined by dividing the number of cells in treated cultures by the number of cells in untreated cultures and multiplying by 100. The experiment was done thrice, with two technical replicates in each experiment. Values plotted are average of three experiments, and error bars depict SD. Student’s t test (two-tailed, unpaired) was used to assess statistical significance. ∗∗∗p value <0.0005. D, Western blot analysis of whole-cell extracts isolated from 8 × 10⁷ cells that were exposed to 100 μM H₂O₂ for 5 h, using anti-SET7 antibodies (1:2500 dilution). Loading control: tubulin. Time points indicate hours after start of the H₂O₂ exposure. set7^+/+ cells: Ld1S::neo-hyg cells. set7^−/+ cells: set7-heterozygous KO cells. set7^−/− cells: set7-nulls. H₂O₂, hydrogen peroxide; Ld1S, Leishmania donovani 1S.

The effect of H₂O₂ treatment on LdSET7 expression in L. donovani promastigotes was analyzed by incubating cells with H₂O₂ (100 μM) for 5 h, followed by continued growth in H₂O_2-free medium and isolation of whole-cell lysates at various time intervals thereafter. The lysates were analyzed using Western blotting with anti-SET7 antibodies, and as seen in Figure 4D, LdSET7 expression levels did not change in response to H₂O₂ treatment. To see if H₂O₂ treatment affected the subcellular localization of LdSET7, L. donovani promastigotes expressing LdSET7-FLAG were treated with H₂O₂ (100 μM) for 5 h, collected by centrifugation, and analyzed by indirect immunofluorescence. No change in subcellular localization was apparent (Fig. S6). The results presented in Figure 4 reinforce the fact that LdSET7 regulates the cell’s response to an oxidative milieu.

LdSET7-depleted promastigotes do not display detectable DNA damage in response to an oxidative environment

Cells in an oxidative environment experience a variety of ill effects. To determine if the discernible differential growth patterns of LdSET7-depleted cells in response to H₂O₂ were due to faster recovery of these cells following exposure to the oxidative agent or due to complete lack of response to H₂O₂ at concentrations up to 200 μM, we examined one of the consequences typically suffered by cells under these conditions: double strand DNA breaks (DSBs). H₂O₂-induced cellular oxidative stress leads to ssDNA and dsDNA breaks due to the production of hydroxyl free radicals (^•OH) within the cell, which attack the bases and sugar groups in the double helix. In trypanosomatids, time-course kinetics of recruitment of various repair proteins to DSBs induced by ionizing radiation has revealed that DSBs are primarily repaired by the homologous recombination pathway involving Exo1, RPA, and RAD51 (19). RAD51-ssDNA filaments play a major role in homology recognition and strand invasion, thus promoting eventual strand exchange in the repair process, and DSBs trigger the activation of RAD51 expression and formation of distinct RAD51 foci at the DSBs 20, 21, 22. In examining the effect of H₂O₂ on the induction of DNA damage, we initially adopted the route of using RAD51 as a marker for DSB repair. For this, cultures of set7^−/− and set7^+/+ cells were initiated from stationary phase cultures and H₂O₂ (100 μM) added when cells reached a density of ∼7 to 9 × 10⁶ cells/ml (set7^−/− cultures being initiated 3 days earlier to enable simultaneous H₂O₂ treatment of both lines), incubation carried out for 5 h, the medium replaced with fresh H₂O₂-free medium, and cells sampled at various time intervals thereafter for isolation of whole-cell lysates. The lysates were probed for RAD51 activation using anti-RAD51 antibodies already available in the lab. As seen in Figure 5A, unlike in set7^+/+ parasites where RAD51 was activated over time in response to the oxidative stress induced by H₂O₂, in keeping with previous results from experiments with Trypanosoma cruzi (20), RAD51 levels did not increase at any of the sampled times in set7^−/− cells. Interestingly, we observed the expression of RAD51 in untreated cells also to be significantly lower in set7^−/− than set7^+/+ cells. The reason for this is not understood at present and needs further exploration.

Figure 5.

Effect of set7 elimination on H₂O₂-induced DNA damage.A, analysis of DNA damage response in H₂O₂-treated cells. Western blot analysis of whole-cell lysates, isolated from 4 × 10⁷ cells that were exposed to 100 μM H₂O₂ for 5 h, using anti-RAD51 antibodies (available in the lab, 1:1000 dilution). Loading control: tubulin. Time points indicate hours after start of the exposure. Quantitation was carried out using ImageJ analysis and plotted values represent average of three experiments. Error bars represent SD, and statistical significance was determined using student’s t test (two-tailed, unpaired). ∗∗∗p value <0.0005, ∗∗p value <0.005, ∗p value <0.05, and ns: statistically not significant. B, analysis of DNA damage in H₂O₂-treated cells. Microscopic analysis of TUNEL assay reactions that were performed on cells that were exposed to 200 μM H₂O₂ for 5 h. DAPI stains nucleus and kinetoplast in each cell. Fluorescein labels free 3′OH ends of DNA, as generated by breaks. The magnification bar represents 10 μm. Images were captured by Z-stack analysis using confocal microscopy. set7^+/+ cells: Ld1S::neo-hyg cells. set7^−/− cells: set7-nulls. DAPI, 4′,6-diamidino-2-phenylindole; H₂O₂, hydrogen peroxide; Ld1S, Leishmania donovani 1S.

To rule out the possibility of absence of RAD51 activation in set7-nulls reflecting repair of DNA strand breaks occurring through a RAD51-independent pathway in these cells, we directly assessed DNA damage using the TUNEL assay, which detects DNA strand breaks by the TdT-mediated uptake of fluorescein-tagged dUMP at the free 3′OH groups generated by the breaks. Accordingly, L. donovani promastigotes (set7^+/+ and set7^−/−) were treated with H₂O₂ following the same regimen, for 5 h, followed by analyses for DNA breaks using the TUNEL reaction. As seen in Figure 5B, untreated parasites of both types (set7^+/+ and set7^−/−) showed labeling of kinetoplast DNA in some cells, signifying dUMP incorporation in replicating kinetoplasts (23). While hardly any nuclei were labeled in these cells, it was observed that almost twice as many set7^−/− cells exhibited nuclear labeling relative to set7^+/+ cells (Table S1), suggesting that a higher fraction of these cells might carry DNA breaks, and perhaps reflecting the lower basal levels of RAD51 in set7^−/− cells. This aspect needs further study. Contrasting to untreated cells, set7^+/+ cells displayed compelling evidence of breaks in nuclear DNA after treatment with H₂O₂, with more than 95% of the nuclei being strongly labelled with dUMP upon exposure to 200 μM H₂O₂. However, almost no evidence of damage was detectable in nuclear DNA in response to similar H₂O₂ treatment in set7^−/− cells (Figs. 5B and S7; Table S1). Taken together, the data in Figure 5 underscore the fact that set7^−/− cells do not suffer any discernible damage to nuclear DNA in response to H₂O₂-induced oxidative environment.

Ectopic expression of LdSET7 in set7-nulls rescues the aberrant phenotypes associated with LdSET7 depletion

To verify that the observed phenotypes were due to LdSET7 depletion, LdSET7-FLAG was ectopically expressed in set7^−/− promastigotes as described in Experimental procedures (Fig. 6A) and growth of set7^−/−::SET7⁺ promastigotes monitored as earlier. Ectopic expression of LdSET7 largely rescued the growth defects of set7-nulls (Fig. 6B). At ∼12 h, the generation time of set7^−/−::SET7⁺ promastigotes was found to be near to that of set7^+/+ cells (Fig. 6C). Flow cytometry analysis of HU-synchronized promastigotes revealed an alleviation of the defects observed in set7^−/− cells, upon ectopic expression of LdSET7 in them (Fig. 6D). The effect of H₂O₂-induced oxidative stress on set7^−/−::SET7⁺ cells was examined by incubating cells with H₂O₂ (100 μM) and monitoring growth. The data in Figure 6E demonstrate that LdSET7-FLAG expression in set7-nulls allowed the parasite to largely overcome the mutant phenotype, with set7^−/−::SET7⁺ promastigotes being vulnerable to H₂O₂ exposure almost as much as set7^+/+ cells. This was also reflected in the RAD51 activation profile of set7^−/−::SET7⁺ cells in response to H₂O₂ exposure (Fig. 6F). The observation of only a partial rescue of set7^−/− phenotypes could perhaps be due to differential expression of LdSET7-FLAG in the many cells of the population.

Figure 6.

Effect of ectopic expression of LdSET7 in set7-nulls, on the phenotypes associated with set7 deletion.A, Western blot analysis of whole-cell extracts isolated from 8 × 10⁷ logarithmically growing promastigotes using anti-SET7 and anti-FLAG antibodies (1:1000 dilution). Loading control: tubulin. B, analysis of growth of parasites. Cultures were initiated at 1 × 10⁶ cells/ml, from stationary phase cultures. Values plotted represent average of three experiments, with each experiment being performed with two technical replicates. Error bars represent SD. C, generation time were determined by initiating cultures from logarithmically growing cultures at 1 × 10⁶ cells/ml and diluting the cultures to 1 × 10⁶ cells/ml every 24 h after counting them. Values plotted are average of three experiments, with error bars depicting SD. D, flow cytometry analysis of HU-synchronized promastigotes. Sampling time points are indicated on the left of each row of histograms. Thirty thousand events were analyzed at every time point. M1, M2, and M3 represent gating for cells in G1, S, and G2/M, respectively. Percent cells in each cell cycle phase are indicated in upper right-hand corner of the histograms. The experiment was carried out twice, with comparable results, and one dataset is shown here. E, effect of prolonged H₂O₂ exposure on promastigote growth. Cultures initiated from stationary phase cultures were treated with 100 μM H₂O₂, for 3 days, with cells being counted every 24 h. F, analysis of DNA damage response. Western blot analysis of whole-cell lysates was isolated from 4 × 10⁷ cells that were exposed to 100 μM H₂O₂ for 5 h, using anti-RAD51 antibodies (1:1000 dilution). Loading control: tubulin. Lysates were isolated 24 h after H₂O₂ exposure. set7^+/+ cells: Ld1S::neo-hyg cells carrying empty vector. set7^−/− cells: set7-nulls. set7^−/−::SET7⁺ cells: Transfectant set7-nulls expressing SET7-FLAG ectopically. H₂O₂, hydrogen peroxide; HU, hydroxyurea; Ld1S, Leishmania donovani 1S.

set7-null parasites do not exhibit activation of reactive oxygen species in response to exposure to H₂O₂

The data in Figure 2, Figure 3, Figure 4, Figure 5, Figure 6 strongly indicated that LdSET7 depletion modulates the cell’s response to an oxidizing environment. Particularly intriguing in set7-nulls was the apparent absence of detectable DNA damage, expected to be induced by the production of reactive oxygen species (ROS) upon exposure to H₂O₂ (Fig. 5). The production of ROS in set7^−/− cells was compared with that in set7^+/+ cells using the dichlorodihydrofluorescein diacetate (DCFDA) assay (24). This assay is based on the principle that DCFDA taken up by cells is deacetylated by cellular esterases to dichlorodihydrofluorescein. Intracellular ROS oxidize the dichlorodihydrofluorescein to dichlorofluorescein, whose fluorescence is measured spectrofluorimetrically (488 ex/529 em). The assay was carried out by incubating set7^+/+ and set7^−/− promastigotes in medium carrying H₂O₂ (100 μM), before incubation with DCFDA and analysis by measurement of fluorescence emission at 529 nm (as detailed in Experimental procedures). It was observed that while ROS levels increased in set7^+/+ cells over 45 min to 24 h after exposure to H₂O₂ before gradually dropping, no appreciable change in ROS (in comparison with untreated cells) was detectable in set7^−/− cells over the same period of time except a slight elevation detected just after exposure to H₂O₂ (Fig. 7A). Ectopic expression of LdSET7 in set7^−/− parasites partially rescued the mutant phenotype (Fig. 7B). The apparent absence of induction of ROS production in response to H₂O₂ treatment in set7^−/− cells may be the likely reason for these cells not suffering DNA damage under these conditions and could also explain why the parasites are apparently resistant to fairly high concentrations of H₂O_2. This observation also points toward the possibility of a significantly higher efficiency of ROS scavenging in set7^−/− parasites.

Figure 7.

Effect of set7 deletion on ROS production.A, analysis of ROS production in set7^+/+ and set7^−/− cells treated with 100 μM H₂O₂, relative to the respective cells which had not been exposed to H₂O_2. Sampling time is marked with reference to time of start of H₂O₂ treatment. The three graphs represent three separate experiments that were each performed thrice (as detailed in Experimental procedures), and values plotted are average of three experiments, with error bars depicting SD. Statistical significance was determined using two-tailed unpaired student’s t test. ∗∗∗p value <0.0005, ∗∗p value <0.005, ns: statistically not significant. B, analysis of ROS production in set7^+/+, set7^−/−, and set7^−/−::SET7⁺ cells treated with 100 μM H₂O₂, relative to respective cells which had not been exposed to H₂O_2. Values plotted are average of three experiments, with error bars depicting SD. Statistical significance was determined using two-tailed unpaired student’s t test. ∗∗∗p value <0.0005, ∗∗p value <0.005, ∗p value <0.05, ns: statistically not significant C, schematic representation of trypanothione peroxidase scavenging system in trypanosomatids. D, analysis of peroxidase activity (as detailed in Experimental procedures). Left panel: activity in untreated set7^−/− cells relative to untreated set7^+/+ cells. Right upper panel: activity in set7^+/+ and set7^−/− cells treated with 100 μM H₂O₂, relative to respective cells which had not been exposed to H₂O_2. Sampling times are with reference to start of H₂O₂ treatment. Right lower panels: Coomassie-stained gels of cell inputs used in the reactions (input loading controls). The experiment was performed thrice (as detailed in Experimental procedures), and values plotted are average of three experiments, with error bars depicting SD. Statistical significance was determined using two-tailed unpaired student’s t test. ∗∗∗p value <0.0005, ∗p value <0.05, ns: statistically not significant. E, real-time PCR analysis of transcripts of the trypanothione peroxidase system. Fold difference in expression was determined using 2^−▵▵Ct method. Tubulin served as internal control for each RNA sample. Left panel: expression in untreated set7^−/− cells with reference to untreated set7^+/+ cells. For each cell type, the Ct values of the genes and of tubulin were determined, from these the ▵Ct values (Ct of gene − Ct of tubulin) obtained, and thereafter the ▵▵Ct values (▵Ct of set7^−/− − ▵Ct of set7^+/+) were calculated. The bar graphs represent the 2^−▵▵Ct values (fold change). Right panel: expression in H₂O₂ -treated set7^+/+ and set7^−/− cells with reference to the respective untreated cells. For each cell type and each condition (untreated and H₂O₂-treated), the Ct values of the genes and of tubulin were determined, and from these the ▵Ct values (Ct of gene − Ct of tubulin) obtained. Thereafter, for each cell type, the ▵▵Ct values (▵Ct of treated cells − ▵Ct of untreated cells) were obtained, and the 2^−▵▵Ct values calculated. The bar graphs represent the 2^−▵▵Ct values (fold change in expression). Data plotted are average of three experiments, each experiment being performed with technical duplicates. Bar graphs represent 2^−▵▵Ct values. Error bars signify SD. Statistical significance was assessed using two-tailed unpaired student’s t test. ∗∗∗p value <0.0005, ∗∗p value <0.005, ∗p value <0.05, and ns: statistically not significant. H₂O₂, hydrogen peroxide; ROS, reactive oxygen species; TR, trypanothione reductase; TXN1, tryparedoxin 1; TXN2, tryparedoxin 2; TXN3, tryparedoxin 3; TxnPx: tryparedoxin peroxidase.

Whereas the mammalian defense against oxidative stress is largely glutathione-dependent, with intracellular superoxide radicals being converted to molecular oxygen and H₂O₂ by superoxide dismutase, and H₂O₂ in turn being converted to water and oxygen by enzymes like glutathione peroxidase and catalase, trypanosomatids employ a somewhat different modus operandi. Lacking catalase and selenium-dependent peroxidases, instead, they possess a unique trypanothione-dependent system with a set of enzymes that act concertedly for the detoxification of peroxides. The three main components of this system are trypanothione reductase (TR), tryparedoxin (TXN), and tryparedoxin peroxidase (TxnPx) (Fig. 7C). Of the three TxnPxs harbored by the parasite, only one is cytosolic and the cytosolic TxnPx (cTxnPx) is the primary mediator of hydroperoxide scavenging in the macrophage environment. While TR reduces the thiol trypanothione (TS₂) to T(SH)₂ using NADPH, TXN (a thiol disulphide oxidoreductase) serves as the conduit through which the reducing equivalents flow from T(SH)₂ to cTxnPx (a 2-Cys peroxiredoxin), and the reduced cTxnPx reacts with and catalyzes the breakdown/reduction of hydroperoxides. The absence of detectable ROS induction in response to H₂O₂ treatment led us to consider if this trypanothione-dependent system was overactive in set7^−/− cells. Thus, we examined peroxidase activity in set7^−/− parasites, as described in Experimental procedures. Results of activity assays revealed that untreated set7^−/− parasites possessed significantly higher peroxidase activity relative to untreated set7^+/+ parasites (Fig. 7D, left panel). Interestingly, while peroxidase activity was enhanced in set7^+/+ parasites over 1 h to 72 h after exposure to H₂O₂, set7^−/− parasites did not demonstrate any visible change in peroxidase activity in response to H₂O₂ over the same period, suggesting that elimination of LdSET7 was causing loss of regulation of peroxidase activity, and the higher levels of basal peroxidase activity in set7^−/− cells were sufficient to rapidly scavenge ROS (Fig. 7D).

In contemplating the reasons for enhanced basal peroxidase activity in set7-nulls, the possibility of higher expression of TXN peroxidase was considered. We were unable to check protein expression levels directly in the absence of available antibodies. Thus, we analyzed transcript levels of the enzyme, as well as of the other components of the trypanothione-dependent peroxide detoxification system, in set7^−/− cells in comparison with set7^+/+ cells. This was done by real-time PCR analysis of RNA isolated from untreated and H₂O₂-treated (8 h after start of the 5 h-treatment with 100 μM H₂O₂) cells. On comparing expression in set7^−/− and set7^+/+ untreated cells, it was observed that cTxnPx was expressed almost 2-fold higher in these cells (Fig. 7E, left panel). In response to H₂O₂ treatment, none of the transcripts were significantly elevated in either type of parasite (Fig. 7E, right panel). The enhanced peroxidase activity in set7^+/+ cells in response to H₂O₂ exposure (Fig. 7D) reflects previous results from Iyer et al., who reported activated expression of the TXN peroxidase protein in L. donovani in response to H₂O₂ treatment (25); our data indicates that this is not due to elevated transcript levels.

The data in Figure 7, D and E suggest that LdSET7 regulates peroxidase activity by controlling cTxnPx expression: deletion of set7 is coupled to significantly higher levels of txnPx transcripts in the cells regardless of exposure to an oxidative environment, in turn resulting in higher peroxidase activity which probably leads to rapid scavenging of the ROS produced in the cell in response to H₂O₂ treatment, such that ROS activation escapes detection and the effects of ROS activation (such as DNA damage) are not experienced.

Ectopic expression of LdSET7-Y421A in set7-nulls does not rescue the deviant phenotypes associated with SET7 depletion

The structure of the SET domain is largely loops and turns, with several small β sheets carrying a few short strands (26). The C-terminal end of the SET domain carries a short conserved stretch of residues harboring an invariant tyrosine residue. This stretch is preceded by a pseudoknot structure. The pseudoknot and invariant tyrosine residue are essential for AdoMet binding and catalysis, with the conserved tyrosine playing a critical role in catalysis. Mutation of this tyrosine residue has been linked to loss of the protein’s methylation activity (27). The data in Figure 1, Figure 2, Figure 3, Figure 4, Figure 5, Figure 6, Figure 7 suggest that LdSET7 plays a role in moderating the cell’s response to oxidative stress. To determine if LdSET7 methylation activity is critical to this function, we resorted to mutating this conserved tyrosine residue of LdSET7 (Fig. S1; mutagenesis detailed in Experimental procedures). The SET7-Y421A-FLAG protein was expressed ectopically in set7^−/− promastigotes (Fig. 8A) and when the growth pattern of set7^−/−:: SET7-Y421A⁺ promastigotes was monitored it was found to resemble that of set7^−/− cells (Fig. 8B). set7^−/−::SET7-Y421A⁺ promastigotes displayed a response similar to set7-nulls when exposed to H₂O₂ as well (Fig. 8C). Analyses of ROS production (Fig. 8D), peroxidase activity (Fig. 8E), and txnPx transcript levels (Fig. 8F) revealed that ectopic expression of the SET7-Y421A protein in set7-nulls could not rescue the phenotypes associated with set7 deletion, underscoring the significance of this residue in LdSET7 function, and implicating a role for LdSET7 methylation activity in moderating the cell’s response to oxidative stress.

Figure 8.

Effect of ectopic expression of LdSET7-Y421A in set7-nulls, on the phenotypes associated with set7 deletion.A, Western blot analysis of whole-cell extracts isolated from 8 × 10⁷ logarithmically growing promastigotes using anti-SET7 and anti-FLAG antibodies (1:1000 dilution). Loading control: tubulin. B, analysis of growth of parasites. Cultures were initiated at 1 × 10⁶ cells/ml from stationary phase cultures. Values plotted represent average of three experiments, with each experiment being performed with two technical replicates. Error bars represent SD. C, effect of prolonged H₂O₂ exposure on promastigote growth. Cultures initiated from stationary phase cultures were treated with 100 μM H₂O₂, for 3 days, with cells being counted every 24 h. D, analysis of ROS production in set7^+/+, set7^−/−, set7^−/−::SET7⁺, and set7^−/−::SET7-Y421A⁺ cells treated with 100 μM H₂O₂, relative to respective cells which had not been exposed to H₂O_2. Values plotted are average of three experiments, with error bars depicting SD. Statistical significance was determined using two-tailed unpaired student’s t test. ∗∗∗p value <0.0005, ∗∗p value <0.005, ∗p value <0.05, and ns: statistically not significant. E, analysis of peroxidase activity. Left panel: activity in untreated set7^−/−, set7^−/−::SET7⁺, and set7^−/−::SET7-Y421A⁺ cells, relative to untreated set7^+/+ cells. Right upper panel: activity in set7^+/+, set7^−/−, set7^−/−::SET7⁺, and set7^−/−::SET7-Y421A⁺ cells treated with 100 μM H₂O₂, relative to the respective cells which had not been exposed to H₂O_2. Sampling times are with reference to start of H₂O₂ treatment. Right lower panels: Coomassie-stained gels of cell inputs used in the reactions (input loading controls). The experiment was performed thrice, and values plotted are average of three experiments, with error bars depicting SD. Statistical significance was determined using two-tailed unpaired student’s t test. ∗∗∗p value <0.0005, ∗∗p value <0.005, ∗p value <0.05, ns: statistically not significant. F, real-time PCR analysis of txnPx transcripts: expression in untreated set7^−/−/WT complement/Y421A complement cells with reference to untreated set7^+/+ cells. Fold difference in expression was determined using 2^−▵▵Ct method. Tubulin served as internal control for each RNA sample. For each cell type, the Ct values of the txnPx gene and of tubulin were determined, from these the ▵Ct values (Ct of txnPx − Ct of tubulin) obtained, and thereafter the ▵▵Ct values (▵Ct of set7^−/−/WT complement/Y421A complement − ▵Ct of set7^+/+) were calculated. The bar graphs represent the 2^−▵▵Ct values (fold change). set7^+/+ cells: Ld1S::neo-hyg cells carrying empty vector. set7^−/− cells: set7-nulls. set7^−/−::SET7⁺ cells: transfectant set7-nulls expressing SET7-FLAG ectopically. set7^−/−::SET7-Y421A⁺ cells: transfectant set7-nulls expressing SET7-Y421A-FLAG ectopically. H₂O₂, hydrogen peroxide; ROS, reactive oxygen species.

Discussion

As the cause of Visceral Leishmaniasis, the mechanisms by which L. donovani’s physiological processes are regulated remain an area of intensive research. Gene regulation is somewhat unusual in L. donovani and other trypanosomatids, with the genes being organized in long unidirectional clusters of functionally unrelated genes that are by-and-large coordinately and constitutively transcribed polycistronically, from divergent strand switch regions, with the trans-splicing of short leader sequences at the 5′ ends of the individual gene units and polyadenylation at the 3′ ends of these units, leading to the production of individual mRNAs ready for translation. No consensus sequences or defined promoter elements have been recognized at these TSRs, nor have the wide repertoire of canonical transcriptional activators found in more conventional eukaryotes, been identified as yet in these organisms. Epigenetic mechanisms appear to control gene expression in trypanosomatids to a fair extent, and the roles of specific histone acetylations in modulating transcription and other DNA-related processes in Leishmania as well as Trypanosoma species have been documented (28, 29, 30, 31, 32, 33, 34). The functions of histone methylations and the proteins mediating them, however, remain largely unexplored in these organisms, with studies thus far revealing the enrichment of H3K4 methylation marks at TSRs (34, 35), and the role of the Trypanosoma Dot1 proteins in modulation of DNA replication (via H3K76 methylation) being uncovered (36). Histone methylations at lysine residues are mediated by SET-domain proteins (Dot1 proteins being the only other histone lysine methyltransferases), and while 29 SET-domain proteins have been identified in T. brucei, all of which are conserved in L. donovani, their target substrates have not been identified yet, and the cellular role of only one of them (TbSET27) has been elucidated thus far (11, 14). SET proteins that have been extensively studied across eukaryotes, particularly yeast and mammalian cells, have been found to target non-histone substrates as well (1, 37), highlighting the fact that they regulate cellular processes through a wide range of downstream substrates. In several instances one or more functions of a SET protein have been unearthed without the target substrate(s) being identified. These proteins often display stringent catalytic activity as they may mediate monomethylation, dimethylation, or trimethylation of the target lysine residue, translating into tighter regulation of the pertinent cellular process as the extent of methylation of a particular lysine residue may govern protein–protein interactions. SET protein–mediated methylations also impact protein activity and protein stability.

In investigating the LdSET7 protein, we found the primarily cytosolic LdSET7 was expressed in promastigotes throughout the cell cycle (Figs. S2 and S3). LdSET7 was not essential to the parasite, but set7^−/− promastigotes exhibited slower growth and a heightened sensitivity to HU-induced G1/S arrest in comparison to set7^+/+ cells (Fig. 3). Contrastingly, set7^−/− amastigotes survived more proficiently than set7^+/+ amastigotes within host macrophages (Fig. 4A), indicating that LdSET7 plays a role in moderating the parasite’s response to the hostile intracellular oxidative environment of host cells. Interestingly, this effect was tightly controlled by LdSET7 expression levels: partial depletion of LdSET7 to ∼50% its usual expression, while not having any effect on growth or cell cycle progression of promastigotes (Figs. 1C and S4), led to much higher proficiency of survival of amastigotes in host macrophages (Fig. 1D), even higher than that of set7^−/− amastigotes (Fig. 4A). The impact of LdSET7 depletion on amastigote growth and survival was mirrored in the response of promastigotes to an oxidative growth environment in vitro, wherein partial depletion of LdSET7 allowed the parasites to tolerate H₂O₂-induced oxidative stress better than usual (Fig. 2), while complete elimination of LdSET7 made the parasites almost completely resistant to stress induced by H₂O₂ at concentrations up to 200 μM (Fig. 4, B and C). The resistance of set7^−/− promastigotes to H₂O₂-induced stress was apparent not only in the growth kinetics of the parasites but also in the almost complete absence of DNA damage in response to H₂O₂ exposure (Figs. 5 and S7; Table S1). ROS-induced stress in mammalian cells is known to alter the subcellular localization of the p65/RelA subunit of the NF-kB transcription protein complex, relocating the cytoplasmic protein to the nucleus where it exercises its ability to activate transcription of several genes such as Mn-SOD and catalase (37, 38). No obvious alteration in the subcellular localization of LdSET7 was detectable in response to H₂O₂ exposure (Fig. S6), and LdSET7 expression was not upregulated upon H₂O₂ exposure either (Fig. 4D). The finding that LdSET7 elimination makes the parasite resistant to H₂O₂-induced oxidative stress and allows amastigotes to survive more proficiently suggests that the parasite (WT) harmonizes its response with the host environment, to withstand the onslaught of the host defense system and proliferate in an orderly fashion, thus establishing itself firmly in the host. The higher survival and proliferation rates associated with set7 deletion would lead to a depletion of host cells with time, thence being detrimental to the persistence of infection.

Amastigotes express a gamut of proteins to tolerate and overcome the stress induced by the oxidative environment of the macrophages. While the macrophages assault the parasites early in infection by releasing a burst of ROS, the parasite employs a barrage of proteins to fight the ROS wave. These include antioxidant proteins and enzymes to scavenge the ROS, as well as proteins to repair DNA lesions induced by the oxidative environment. Hydroperoxides in Leishmania are scavenged by a trypanothione-dependent system. Three TXN genes have been annotated in the L. donovani genome and of the three TxnPxs that have been identified in Leishmania, one is cytosolic (cTxnPx: LdBPK_151140.1) and essential for cell survival, the second is mitochondrial (mTxnPx: LdBPK_230050.1), and the other is glycosomal (gTxnPx:LdCL_150016500) (25, 39). The cytosolic TxnPx is the primary mediator of hydroperoxide scavenging within host macrophages and has been reported to be secreted in Leishmania infantum and Leishmania major (40). The data in Figure 7 suggests that LdSET7 controls expression and activity of TxnPx through transcript levels (Fig. 7, D and E), with set7 deletion leading to rapid scavenging of any ROS generated (Fig. 7A). Moreover, the observations of experiments carried out with the LdSET7-Y421A protein (Fig. 8) strongly suggest that SET7-mediated methylation activity is critical to the protein’s impact on host–parasite interactions.

While the mechanism by which LdSET7 controls txnPx transcript levels remains unknown to us, two general possibilities are contemplatable. In the first scenario, LdSET7 may regulate txnPx transcripts either through an epigenetic mark on a histone residue, which would have an impact on global gene expression or through the methylation of a transcription factor which downregulates txnPx transcription. Considering that LdSET7 is predominantly cytosolic, it is unlikely to mediate an activating/repressive histone methylation mark, which would usually be added after histone deposition. While txnPx transcripts are upregulated in set7-nulls (Fig. 7E), other genes lying in the same polycistronic cluster remain unaffected by set7 deletion in untreated as well as H₂O₂-exposed cells (data not shown), indicating that LdSET7 does not modulate global gene activation/repression. Contemplating the alternate possibility of LdSET7 controlling the expression of txnPx transcripts through the methylation of a transcriptional activator: previous studies have reported SET protein–mediated methylation to impact activity of such proteins. Intracellular ROS accumulation in mammalian cells shuttles the p65/RelA subunit of the NF-kB transcription factor into the nucleus, ultimately triggering activation of genes whose products fight the oxidative burst. While Set7/9-mediated methylation of p65/RelA at a specific lysine residue is believed to be critical to RelA’s ability to activate transcription, methylation of RelA at other lysine residues by the same SET protein has been reported to repress its ability to activate transcription (reviewed in (37)). The abundantly cytosolic human SET protein SMYD2 methylates p53, which shuttles between the nucleus and cytoplasm, and this methylation (at K370) inhibits p53 transactivation activity (41). Thus, LdSET7-mediated methylation of a transcriptional activator may inhibit its activity, tightly controlling txnPx transcript levels, and deletion of set7 may alleviate this negative regulatory effect. While no evidence in support of gene regulation through transcriptional activators has been uncovered in trypanosomatids thus far, with transcription being primarily polycistronic and constitutive, a small subset of genes have been identified to be transcriptionally activated in a cell cycle–dependent manner. These genes are scattered throughout the genome and are turned on by promoters immediately upstream of them (28). It is possible that txnPx is activated by a promoter of this kind; this needs to be further investigated.

In the second scenario, in light of the fact that posttranscriptional processes are crucial to regulating gene expression in these organisms, the roles of RNA-binding proteins (RNA-BPs or RBPs) and/or RNA methylation may be a vital factor. 5′ and 3′ UTRs of stable processed transcripts often carry cis-acting elements that serve as regulatory motifs to which RBPs bind. The RBP–RNA interactions modulate mRNA stability, decay, transport, and can also modulate translation efficiency. The Trypanosoma cruzi UBP1 protein has been reported to bind to a specific sequence UBP1m in the 3′UTR of the gene transcripts whose expression it regulates (mostly those encoding surface glycoproteins), thus stabilizing the mRNAs (42). Two yeast SET proteins have themselves been found to be RBPs, with evidence of RNA binding activity in vitro as well as in vivo (43, 44). These proteins carry RNA recognition motifs (RRMs) and are enriched on only a small number of transcripts, with the SET–mRNA interactions being hypothesized to contribute toward holding the SET proteins in proximity with their target histone residues. Although it is possible that LdSET7 may itself be an RBP, no RRM has been identifiable. In Trypanosoma brucei, while almost 350 transcripts have been found to be enriched in N⁶-methyladenosine, with RNA methylation in the poly(A) tails of VSG transcripts being found to stabilize VSG mRNAs, txnPx transcripts are not among those identified with N⁶-methyladenosine marks thus far (45, 46). Considering also that to date there is no evidence of SET proteins being RNA methylases and LdSET7-mediated methylation activity appears to be critical to modulating txnPx transcript levels (Fig. 8F), it seems unlikely that LdSET7 is directly controlling txnPx transcript levels. It is possible that LdSET7 mediates the methylation of an RNA methylase that targets txnPx transcripts, with SET7-mediated methylation having an inhibitory effect on activity; these possibilities need further in-depth studies.

While PTMs on proteins have been widely studied and are generally known to regulate protein function through their impact on localization, protein–protein interactions, stability, and activity, not much information is currently available on the impact of PTMs in case of RNA-BPs. In Saccharomyces cerevisiae, the Set3 protein methylates Nab3 of the Nrd1–Nab3–Sen1 complex, and mutation of the target lysine residue which lies in its RRM leads to reduced binding to RNA in vitro and transcription termination defects in vivo (47). An in-depth analysis of the PTM landscape on RBPs in human cells has been carried out through collation of existing curated datasets and published research (48). PTMs have been identified in almost 2400 RBPs, with the vast majority of these carrying multiple PTMs. The commonest classes of PTMs identified were phosphorylation, acetylation, ubiquitination, and methylation. Methylations on RBPs have been found in over 1100 proteins and while typically occurring at arginine residues, methylation of RBPs at lysine residues have also been identified in several human cancer cell lines as well as in clinical samples drawn from different body organs (48). Thus, it is possible that LdSET7 mediates its effect through the methylation of one or more RBPs, the negative impact of which would be reflected in tightly controlled levels of target transcripts, in this case, txnPx. Depletion of LdSET7 and corresponding abrogation of the RBP methylation event would be coupled to loss of this regulation, resulting in an overall increase in txnPx transcript expression levels. The mammalian SET7/9-mediated E2F1 methylation marks it for degradation via the ubiquitination–proteasomal degradation pathway (49); a similar mechanism could gradually deplete the parasite of an RNA-BP that is vital to stabilizing txnPx transcripts.

This study reports the first data directly investigating the functional role of a SET protein in Leishmania species, with the results of our experiments demonstrating that LdSET7 modulates the persistence of Leishmania infection by working toward maintaining the delicate balance between parasite and host cell. The distinct phenotypes obtained upon set7 deletion suggest that although 29 SET proteins have been identified in these organisms, functional redundancy may be limited. While the global landscape of histone modifications has not been uncovered yet in Leishmania species, a large number of methylation events have been identified in T. brucei and T. cruzi histones, and several of these SET proteins must target histone substrates. The identification of target histone substrates of specific SET proteins will allow us to gain new insights into how these proteins function.

Experimental procedures

Leishmania cultures and manipulations

Ld1S cultures were grown in M199 medium supplemented with fetal bovine serum (Invitrogen), adenine, and hemin (Sigma Aldrich), at 26 °C as described earlier (50). Growth and survival patterns were analyzed as described earlier (30). Generation time in the logarithmic phase of growth was determined as described earlier (28). Synchronization of parasites using HU (5 mM for 8 h) and flow cytometry analyses were carried out as described previously (18). Transfections were done and clonal lines generated as described (29, 30). Whole-cell extracts were isolated using the M-PER kit (Thermo Fisher Scientific) as per manufacturer’s instructions. Procylics and metacyclics were isolated as described earlier (51). For treatment with H₂O₂, cultures were initiated from stationary phase cultures, at 1 × 10⁶ cells/ml, and H₂O₂ (25–1000 μM) added on reaching a cell density of ∼7 to 9 × 10⁶ cells/ml (day 3 for set7^+/+ and set7^−/+ cells, day 6 for set7^−/− cells). The set7^−/− cultures were initiated 3 days earlier to enable addition of H₂O₂ at the same time as the set7^+/+ and set7 ^−/+ cells. At the time of addition of H₂O₂, the cultures were split into two, with half the cells receiving H₂O₂ treatment and the other half being carried forward as the untreated culture. The cultures were incubated further at 26 °C for varying time intervals and sampled for further experimentation.

Cloning of set7 gene for expression in Leishmania

For expression in Leishmania parasites, the pLEXSY-FLAG vector was created from the pLEXSY-CYC9-FLAG plasmid (28) by digesting it with BglII (which would drop out the CYC9 gene but not the FLAG sequence it is tagged with) and ligating the vector’s BglII ends. For expressing the set7 gene in L. donovani promastigotes, the gene was amplified off Ld1S genomic DNA using primers SET7-FLAG-F (5′-TAGGATCCATGCCCATCAGCCAG-3′) and SET7-FLAG-R (5′- TAGGATCCAGGAAGAAGAGGCTTCA-3′). The amplicon was cloned into the BglII site of the pLEXSY-FLAG vector using the BamHI sites in the primers, creating plasmid pLEXSY/SET7-FLAG. The SET7-Y421A mutation was created using overlap PCR. For this, the N-terminal fragment of the gene was amplified using primers SET7-FLAG-F and SET7-Y421A-R (5′-AAAGGTGGCGGCGTCCATGCTCAAGT-3′), while the C-terminal fragment was amplified using primers SET7-Y421A-F (5′- ACTTGAGCATGGACGCCGCCACCTTT-3′) and SET7-FLAG-R. The full-length amplicon was obtained using a mix of the two amplicons as template, with the help of primers SET7-FLAG-F and SET7-FLAG-R. The amplicon carrying the mutation was cloned into the BglII site of the pLEXSY-FLAG vector, creating plasmid pLEXSY/SET7-Y421A-FLAG.

Creation of knockout and rescue lines

The first set7 genomic allele was knocked out by replacing it with a hyg^r cassette, while the second allele was replaced with a neo^r cassette. The donor plasmid for replacing the second allele was constructed using the pLEXSY_I-neo3 vector as backbone (Jena Bioscience). To replace the first allele, first the neo^r cassette in pLEXSY_I-neo3 was replaced with a hyg^r cassette using the BamHI-SpeI sites flanking the neo^r cassette, generating the vector pLEXSY/hyg, which served as the backbone for constructing the donor plasmid. The ∼800 bp region immediately upstream of the set7 gene was amplified using Ld1S genomic DNA with primers SET7-5′FL-F (5′- TAGCGGCCGCATTTAAATGGTTTTTTCTGCCTTCTCTTG-3′) and SET7-5′FL-R (5′- TAGCGGCCGCGGTCTGCGACCGATATACCTCGG-3′) and the amplicon cloned into the NotI site of the two vector backbones. The ∼ 800 bp region immediately downstream of the set7 gene was amplified using primers SET7-3′FL-F (5′- TAACTAGTATACATGAGTGAGACGCTCCGCGG-3′) and SET7-3′FL-R (5′- TAACTAGTATTTAAATTGCTTCTCAAAGCCCTTGTCA-3′) and the amplicon cloned into the SpeI site of the two clones carrying the 5′flank sequence, thus creating the donor plasmids SET7-KO/hyg and SET7-KO/neo.

For making the set7 heterozygous knockout (set7^−/+) the donor cassette was released from plasmid SET7-KO/hyg using SwaI digestion, transfected into L. donovani promastigotes by electroporation, and clonal lines selected for and expanded in the presence of hygromycin (16 μg/ml), as described earlier (29). For making the set7-null (set7^−/−) the donor cassette was released from plasmid SET7-KO/neo using SwaI digestion, transfected into set7^−/+ promastigotes by electroporation, and clonal lines selected for and expanded in the presence of G418 (50 μg/ml) and hygromycin (16 μg/ml). Clonal lines were maintained in liquid culture under selection pressure (G418 at 100 μg/ml and hygromycin at 32 μg/ml) except for flow cytometry experiments where the drugs were withdrawn a week before setting up the experiment.

For making the rescue line, the set7 gene was cloned into the BamHI-EcoRV sites of pXG-FLAG (bleo) vector (30), using primers SET7-pXG-F (5′-GAGGATCCGCCACCATGCCCATCAGCCAG-3′) and SET7-pXG-R (5′-AGGATATCTCCAGGAAGAAGAGGCTT-3′) to amplify the gene for cloning. The plasmid pXG-SET7-FLAG (bleo) was transfected into set7^−/− promastigotes and clonals selected for using G418, hygromycin, and phleomycin (50 μg/ml, 16 μg/ml and 2.5 μg/ml, respectively).

Isolation of RNA and real-time PCRs

Total RNA was isolated from 5 × 10⁷ promastigotes with the help of the PureLink RNA mini kit (Invitrogen). The RNA was treated with DNaseI (1 U DNase I per 2 μg RNA at 37 °C for 30 min) prior to cDNA synthesis, to eliminate any genomic DNA contamination. Total cDNA was synthesized as per the manufacturer’s instructions using the iScript cDNA synthesis kit (Bio-Rad). For expression analysis using real-time PCR, one-twentieth of the cDNA synthesis reaction was used as template per reaction. Tubulin expression was analyzed using primers Tubulin RT-F1 (5′-CTTCAAGTGCGGCATCAACTA-3′) and Tubulin RT-R2 (5′-TTAGTACTCCTCGACGTCCTC-3′). TR expression was analyzed using primers designed against LdBPK_050350.1 (TR-RT-F: 5′- CACAACATCAGCGGCAGCAAG-3′ and TR-RT-R: 5′-TCGGCGCTCGTCGGGTGGA-3′). Expression of TXN1 was analyzed using primers designed against LdBPK_291250.1 (TXN1-RT-F: 5′- GAGTTCTACGAGAAGCATCACA-3′ and TXN1-RT-R: 5′-TCAGCGTCGGAATCGATTCCA -3′). Expression of TXN2 was analyzed using primers designed against LdBPK_291240.1 (TXN2-RT-F: 5′- CAACAAACACGCGAAGTCGAAG-3′ and TXN2-RT-R: 5′-CGACGCCGATCAGCGTCGGA-3′). Expression of TXN3 was analyzed using primers designed against LdBPK_312000.1 (TXN3-RT-F: 5′- GACTACTACTGCCTGCCGTAC-3′ and TXN3-RT-R: 5′- GGCTGCTGCGGCTCTGCATC-3′). Expression of TxnPx was analyzed using primers designed against LdBPK_151140.1 (TxnPx-RT-F: 5′- GCCTACCGCGGTCTCTTCATC-3′ and TxnPx-RT-R: 5′-TTCCAGTTCGCGGGGCACAC-3′). Expression of genes in set7^−/− cells relative to in set7^+/+ cells were determined using the 2^−△△Ct method (52), with tubulin gene expression serving as the internal control in each sample type. Expression of genes in H₂O₂-treated versus untreated cells were likewise determined using the 2^−△△Ct method using the tubulin gene as internal control. Real-time PCR experiments were done three times, with technical duplicates in each experiment. Values plotted are the average of three experiments and error bars show SD. Student’s t test was applied for analyzing statistical significance.

Immunofluorescence analysis

Indirect immunofluorescence was carried out as described earlier (30). Briefly, exponentially growing promastigotes expressing LdSET7-FLAG were fixed with 2% paraformaldehyde, cell spreads prepared on poly-lysine coated coverslips, cells permeabilized with 0.1% Triton X-100, blocked with 5% chicken serum, incubated with anti-FLAG antibody for 2 h (Sigma Aldrich, 1:100 dilution), antibody washed off, incubated with Texas Red-labeled secondary antibody for an hour (Jackson Immunoresearch Laboratories, 1:100 dilution), and finally mounted in antifade solution containing 4′,6-diamidino-2-phenylindole (Vectashield, Vector Laboratories). Cells were viewed using the Leica TCS SP8X confocal microscope (with 100× (in oil) objective), and images captured and analyzed using the Leica Application Suite X (https://www.leica-microsystems.com/products/microscope-software/p/leica-las-x-ls/) software.

TUNEL assay

To assess DNA strand breaks in L. donovani promastigotes (set7^+/+ and set7^−/−), cultures initiated at 1 × 10⁶ cells/ml from stationary phase cultures were grown to a cell density of 7 to 9 × 10⁶ cells/ml (day 3 in case of set7^+/+ and day 6 in case of set7^−/− cells; set7^−/− cultures were initiated 3 days earlier to enable addition of H₂O₂ to all cultures at the same time), and H₂O₂ (100 μM or 200 μM) added. Cultures were incubated further for 5 h, and aliquots of 1 × 10⁷ cells removed immediately after to perform the TUNEL assay. For this, the cells were collected by centrifugation at 1448g, washed with 1× PBS, fixed with 2% paraformaldehyde, and cell spreads prepared on poly-lysine coated coverslips (2–4 × 10⁶ cells per coverslip). After cell permeabilization with 1× PBS-0.2% Triton X-100 for 10 min, the TUNEL assay was performed using the DeadEnd Fluorometric TUNEL System (Promega) as per the manufacturer’s instructions. Briefly, the adhered cells were incubated with the dUTP tailing reaction mix at 37 °C for an hour before stopping the reaction, washing the coverslips with 1× PBS to remove unincorporated dUTP, and mounting in antifade solution carrying 4′,6-diamidino-2-phenylindole. Cells were viewed using the Leica TCS SP8X confocal microscope (with 100× (in oil) objective), and images captured and analyzed using the Leica Application Suite X software.

Measurement of ROS

ROS in L. donovani promastigotes were assessed using the DCFDA assay as described (24). Toward this, L. donovani promastigote cultures (set7^+/+ and set7^−/−) were initiated at 1 × 10⁶ cells/ml from stationary phase cultures, and H₂O₂ (100 μM) added when cells reached a density of 7 to 9 × 10⁶ cells/ml (day 3 in case of set7^+/+ and day 6 in case of set7^−/−; set7^−/− cultures were initiated 3 days earlier to enable addition of H₂O₂ to all cultures at the same time). Cultures were incubated further at 26 °C for 45 min, the medium was replaced with fresh H₂O₂-free medium and incubation at 26 °C continued. Aliquots of 1 × 10⁷ cells were removed at various time intervals thereafter to perform the DCFDA assay.

For this, the cells were collected by centrifugation at 1448g, medium was completely aspirated, and collected cells were washed twice in Hepes/NaCl buffer (21 mM Hepes (pH 7), 137 mM NaCl, 5 mM KCl, 6 mM glucose, 0.7 mM NaH₂PO₄) before suspending the cells in 1 ml of the same buffer. The DCFDA reagent (5 μM; Sigma Aldrich) was added to the cell suspension, reaction mixed by inverting the tube, and incubated in the dark at 26 °C for 45 min. This was followed by collecting the cells by centrifugation, washing them in the Hepes/NaCl buffer, and making a cell suspension in 1 ml of the buffer. Two hundred microliters aliquots of the cell suspensions were excited at 488 nm and fluorescence emission detected at 529 nm using a Tecan plate reader (Infinite 200 PRO).

Control reactions carried out without addition of DCFDA yielded no fluorescence. Control reactions carried out without cells gave “background” fluorescence readings, and the values of these control reactions (which were set up at every time point) were subtracted from the values obtained in the reactions with the two cell types (set7^+/+ and set7^−/−). At every time point, reactions with the two cell types were analyzed in case of both untreated and treated cells. The ratio between values obtained (after subtracting control reaction value) under treated versus untreated conditions was used as a measure of ROS activation in response to H₂O₂, for each cell type. The experiments were done thrice and mean values are presented in the bar charts; error bars indicate SD. Student’s t test was used to analyze statistical significance.

Measurement of peroxidase activity

To analyze peroxidase activity in L. donovani set7^+/+ and set7^−/− promastigotes, cultures were initiated at 1 × 10⁶ cells/ml from stationary phase cultures, and incubated to a cell density of 7 to 9 × 10⁶ cells/ml (set7^−/− cultures were initiated 3 days earlier so that it reached the cell density at the same time as set7^+/+ cells), before adding H₂O₂ (100 μM) and allowing incubation at 26 °C for 5 h, then replacing the medium with fresh H₂O₂-free M199 and continuing incubation at 26 °C. Aliquots of 1×10⁷ cells were removed at various time intervals thereafter to perform the Amplex Red assay (53, 54). For this, the cell aliquots were washed with 1× PBS and resuspended in 500 μl assay mix (1× PBS carrying 64 μM digitonin, 10 μM Amplex Red (Invitrogen), 1 mM H₂O_2, 1 mM protease inhibitors mix). The reaction mixes were incubated in the dark at 26 °C for 30 min, cell remnants removed by centrifugation, and fluorescence of the supernatant (200 μl aliquot) measured (excitation at 535 nm, emission at 590 nm).

Reactions carried out without addition of Amplex Red yielded no fluorescence. Reactions carried out without cells gave “background” fluorescence readings, and the values of these control reactions (set up with every time point) were deducted from the values obtained in the reactions with the two cell types (set7^+/+ and set7^−/−). At every time point, reactions with the two cell types were analyzed in case of both untreated and treated cells. The ratio between values obtained (after subtracting control reaction value) under treated versus untreated conditions was used as a measure of activation of peroxidase activity in response to H₂O_2, for each cell type. The ratio between values obtained in set7-nulls versus set7^+/+ in cells that had not been treated with H₂O₂ was used as a measure of basal peroxidase activity in set7-nulls relative to set7^+/+ cells. The experiments were done thrice and mean values are presented in the bar charts. Error bars indicate SD, and student’s t test was used to analyze statistical significance.

Macrophage infection experiment

Metacyclic parasites were incubated with macrophages (J774A.1) for infection as described previously (28, 29). Each experiment was carried out with three biological replicates and the data presented in the bar charts show the mean of three experiments, with error bars representing SD. The two-tailed student t test was applied to analyze statistical significance of the obtained data, and p values are mentioned in Figure legends.

Data availability

The authors declare that all data supporting the findings of this study are available within the paper and its Supporting information files. Ld1S set7 sequence has been deposited in GenBank. Accession no: OR479702. All raw data are available from the corresponding author upon request.

Supporting information

This article contains supporting information.

Conflict of interest

The authors declare that they have no conflicts of interest with the contents of this article.

Reviewed by members of the JBC Editorial Board. Edited by Brian D. Strahl