- Open Access
Genetic validation of Leishmania genes essential for amastigote survival in vivo using N-myristoyltransferase as a model
Parasites & Vectors volume 13, Article number: 132 (2020)
Proving that specific genes are essential for the intracellular viability of Leishmania parasites within macrophages remains a challenge for the identification of suitable targets for drug development. This is especially evident in the absence of a robust inducible expression system or functioning RNAi machinery that works in all Leishmania species. Currently, if a target gene of interest in extracellular parasites can only be deleted from its genomic locus in the presence of ectopic expression from a wild type copy, it is assumed that this gene will also be essential for viability in disease-promoting intracellular parasites. However, functional essentiality must be proven independently in both life-cycle stages for robust validation of the gene of interest as a putative target for chemical intervention.
Here, we have used plasmid shuffle methods in vivo to provide supportive genetic evidence that N-myristoyltransferase (NMT) is essential for Leishmania viability throughout the parasite life-cycle. Following confirmation of NMT essentiality in vector-transmitted promastigotes, a range of mutant parasites were used to infect mice prior to negative selection pressure to test the hypothesis that NMT is also essential for parasite viability in an established infection.
Ectopically-expressed NMT was only dispensable under negative selection in the presence of another copy. Total parasite burdens in animals subjected to negative selection were comparable to control groups only if an additional NMT copy, not affected by the negative selection, was expressed.
NMT is an essential gene in all parasite life-cycle stages, confirming its role as a genetically-validated target for drug development.
The kinetoplast parasites, Leishmania spp., alternate between two distinct life-cycle stages: the flagellated and motile extracellular promastigotes, and the immotile intracellular amastigotes bearing their rudimentary flagellum [1, 2]. Extracellular promastigotes develop within the female blood-feeding sand fly vector prior to transmission to the mammalian host during a blood meal. Following uptake by professional phagocytes (e.g. macrophages, dendritic cells), promastigotes then differentiate into intracellular amastigotes. These infections result in a spectrum of diseases termed the leishmaniases, the most severe forms of which are fatal in man [3, 4]. Currently, there are only a handful of licensed drugs available to treat these infections, with most having severe side effects while being difficult to administer and often requiring patient hospitalization . In addition, resistance has developed in the field against some current drugs and all show varying degrees of efficacy against the differing species of infecting Leishmania parasite .
Despite some advances in drug re-purposing [6, 7], novel drug development for the leishmaniases has not been a priority for the pharmaceutical industry, even though there is an urgent need for new approaches to the treatment of these deadly infections. Current efforts are focused on the identification of compounds that target and kill intracellular amastigotes [8, 9]. Two approaches are generally available for such screening programmes: phenotypic screening or screening against a known drug target [10, 11]. Phenotypic screening has the advantage of identifying selective “cidal” compounds from high throughput screens of intracellular amastigotes, potentially identifying previously unknown/unexplored therapeutic pathways. A disadvantage of this approach, however, is that modes of action of specific compounds may be difficult to determine, although with the advent of metabolomic and chemical proteomic approaches, target deconvolution should be possible. In addition, compound optimisation is difficult when the parasite target is unknown. These challenges are obviated in target-based screening, where both the target identity and its mode of action can be studied in detail, leading to compound optimisation guided by structural constraints and definition of a structure activity relationship [12, 13].
Given the potential advantages of using target-based screening approaches and the necessity for testing against intracellular amastigotes it is crucial that the target in question is essential for parasite viability within the intracellular environment. The target has to be a known gene coding for an essential protein required for parasite viability in the host. At present, any target gene of interest is assumed to be essential for viability if it can only be deleted from its genomic locus when its product is expressed ectopically in the extracellular promastigote stage of the parasite life-cycle [14, 15]. In practice, these gene replacements can be technically complex, often requiring use of more than two selectable markers, dependent on the location of the gene of interest, to account for the variable chromosomal ploidy of Leishmania species . The data generated from such genetically manipulated promastigote strains are then often correlated with gene function in intracellular amastigotes, despite the lack of phenotypic information derived from genuine gene knockouts generated in amastigotes. Generating such information is not only important for the robust identification of new drug targets in intracellular parasites, but will also inform our understanding of parasite biology, ensuring, for example, that there are no appropriate biochemical “escape” pathways to facilitate amastigote survival rendering drug treatment futile and to predict adaptation of the parasites to inhibition.
To date, and to our knowledge, no gene essential for promastigote viability has also been definitively shown by genetic manipulation to be essential for viability in the Leishmania amastigote. This is principally due to the lack of a robust inducible expression system  and a functioning RNAi machinery  that works in all Leishmania species.
N-myristoyltransferase (NMT) catalyses the covalent attachment of a myristate moiety to the N-terminal glycine of selected eukaryotic target proteins, with 60 proteins predicted to be myristoylated (reviewed in [19, 20]). Experimentally 30 proteins were confirmed with high and 18 with lower confidence to be a substrate of NMT . Gene knockout by homologous recombination has been used to demonstrate that NMT is an essential enzyme in extracellular promastigotes of Leishmania [22, 23] as well as in both insect and mammalian extracellular stages of Trypanosoma brucei . As a consequence, NMT is being exploited as a potential target for chemotherapeutic intervention across the leishmaniases, with high-throughput screening leading to the identification of first generation compounds with specificity and sensitivity against Leishmania promastigotes [21, 24,25,26].
Plasmid shuffle is based on the concept that a plasmid carrying an essential gene is only dispensable in the presence of another functional copy of the gene or if necessary metabolic products are present compensating for the lack thereof . It has already been applied to investigate gene function in Leishmania, but only in promastigote stages [28,29,30,31]. Methodologically, the gene of interest is expressed ectopically from a plasmid before it is deleted from its genomic locus. To direct the plasmid shuffle in Leishmania, this plasmid also encodes for thymidine kinase (TK) a negative selective marker, to force the death of the parasite or the loss of this plasmid upon negative selection with ganciclovir (GCV), a nucleoside analogue. TK phosphorylation occurs and the resulting nucleotide can be used by the parasites, interfering with strand elongation during DNA replication. If parasites die or the plasmid is only dispensable in the presence of another functional gene copy upon negative selection, then the gene can be regarded to be essential. Here, we have applied plasmid shuffle methodology to provide supportive evidence that NMT is essential in both life-cycle stages of Leishmania parasites, thereby confirming its potential as a fully validated drug target in these species.
Leishmania in vitro culture
Promastigotes of L. donovani MHOM/ET/67/L28 (LV9 strain) were grown at 26 °C in RPMI 1640 medium (Gibco, Paisley, UK) supplemented with 20% heat-inactivated FBS, 100 µM adenine, 20 mM 2-[N-morpholino] ethanesulphonic acid (pH 5.5), 5 µM hemin, 3 µM biopterin, 1 µM biotin (all Sigma-Aldrich, Irvine, UK), 100 U/ml penicillin and 100 μg/ml streptomycin (GIBCO, UK).
In vivo infections
BALB/c mice were obtained from Charles River (UK) and infected with 3 × 107L. donovani stationary phase promastigotes, in 200 μl of RPMI 1640 (GIBCO), intravenously (i.v.) via the tail vein. Ganciclovir (GCV, Invivogen, Toulouse, France) was dissolved in H2O pH 12 at 10 mg/ml and then diluted to 1.25 mg/ml in 0.1 M HEPES pH 7.4. Solvent control was prepared in an identical manner but without the GCV. Mice were treated with 3 mg/kg b.i.d (morning and evening (10–12 h apart) by intraperitoneal injection. Control mice were injected with the respective volume of solvent solution.
Generation of NMT-expression plasmids
NMT was amplified from genomic DNA (forward primer: 5′-ACG AGA TCT ATG TCT CGC AAT CCA TCG AAC TC-3′; reverse primer: 5′-GCT AGA TCT CTA CAA CAT CAC CAA GGC AAC C-3′) and cloned into the BglII restriction site (restriction sites are underlined throughout) of pXNG4-SAT (also encoding TK and kindly provided by S. Beverley, Washington University, St Louis, USA and described elsewhere ) to obtain pXNG4-SAT-NMT. The plasmids pX-NEO-tdtom and pX-NEO-NMT-tdtom (tdtom denotes tandem tomato fluorescent protein - tdTomato) were generated by amplifying LamDH intergenic region (IR) (forward primer: 5′-ACC TCT AGA ACA TCG ATT GTG GAA GCA CAA AGC GCA C-3′; reverse primer: 5′-AAT TCT AGA CAT ATG CAA GCT GAT CCA GAG GAC GTG-3′) and Lam NAGT IR (forward primer: 5′-TCA GGA TCC GAT CCA GTA GTG CCA ATA GAG-3′; reverse primer: 5′-ATT GCG GCC GCT CAT GTT TGA CAG CTT ATC ATC-3′) from pXNG4-SAT and cloning into the XbaI and BamHI/NotI restriction sites, respectively, of the pX expression vector . Tdtom was generated by amplification from pSSU-tomato ; forward primer: 5′-GAG CAT ATG GTG AGC AAG GGC GAG GAG-3′; reverse primer: 5′-ATA GCG GCC GCA CGC GCC GGG CAT CGC TGC-3′) and cloned into the previously introduced NdeI site (reverse primer Lam NAGT IR) and NotI site; NMT was amplified from genomic DNA (forward primer: 5′-GAA CCC GGG ATG TCT CGC AAT CCA TCG AAC TC-3′; reverse primer: 5′-CCA TCT AGA CTA CAA CAT CAC CAA GGC AAC C-3′) and cloned into the XmaI and XbaI sites to obtain pX-NEO-NMT-tdtom.
Transfected Leishmania lines
For generation of conditional double NMT allele replacement mutants, L. donovani NMTHYG/+ promastigotes  were transfected with pXNG4-SAT-NMT. Replacement of the second NMT allele was achieved by targeted gene replacement as described . One confirmed genomic homozygous NMT knockout carrying the pXNG4-SAT-NMT plasmid line was then used for transfection with pX-NEO-tdtom or pX-NEO-NMT-tdtom.
Transfections were performed sequentially with 5 μg purified DNA by nucleofection (using the Human T Cell Nucleofector Kit, Lonza, Basel, Switzerland) of 2 × 107 mid-log phase promastigotes. Mutants were selected on RPMI medium/1% agar plates with respective antibiotics (32 μg/ml hygromycin B; 40 μg/ml G418; 80 μg/ml puromycin; 100 μg/ml nourseothricin). PCR screening to confirm the correct targeting of the 2nd allele was performed with the following forward primers: PACint (5′-ACC TGG TGC ATG ACC CGC AAG-3′), HYG (5′-CCT GAA CTC ACC GCG ACG TC-3′) and LdNMT (5′-CTA TGC CCA CCG AGC TAC ATC C-3′) together with the reverse primer: Syntaxin Rev (5′-GCC AGC TGC GTC AAA CGC AT-3′).
Intracellular in vitro analysis
Bone marrow was extracted from 6–8 week-old female BALB/c mice and differentiated to macrophages with L929 conditioned medium. Alternatively, THP-1 cells were transformed to macrophages with phorbol 12-myristate 13-acetate (PMA) or retinoic acid as described in  or , respectively. Macrophages were infected with a MOI of 10 with stationary promastigotes. Upon transformation to amastigotes, in our hands it took 72 h, the cells were either treated with 50 μg/ml GCV or solvent. At the appropriate time points, PMA stimulated cell were lifted of the well bottom upon two washes with PBS with a cell scraper and then collected as described in . Retinoic acid-stimulated cells were harvested as described in . Cells were then subjected to flow cytometric or microscopic analysis.
GFP and tdtom expression was assessed in live promastigotes suspended in PBS by flow cytometry. Cells were analysed on a DAKO CyAn ADP analyser (DakoCytomation, Ely, UK) and data were analysed with FlowJo v10.0.6 (TreeStar Inc., Ashland, OR, USA).
Splenic parasite burdens were determined by limiting dilution assay as described . Spleens were dounce-homogenized and then plated out as described, in the presence/absence of the following concentrations of selection antibiotics as appropriate: hygromycin (32 μg/ml); nourseothricin (100 μg/ml); puromycin (80 μg/ml); G418 (40 μg/ml).
Analysis of statistically significant differences between groups were performed by Mann–Whitney test using GraphPad Prism v8 (GraphPad Software Inc., San Diego, CA, USA)
Generation and characterisation of conditional genomic NMT allele deletions
NMT had previously been shown to be essential for the viability of L. donovani promastigotes . In the present study, in order for us to replace both genomic NMT alleles in these experiments, it was necessary to express an NMT gene copy at an ectopic locus, in this case, from one of two or two distinct plasmids expressing different reporters GFP or tdTom. The strategy for generation and use of the resulting parasite lines in these new plasmid shuffle experiments is shown in Fig. 1. Flow cytometry was used to monitor the fate of either plasmid by detecting GFP or tdtom expression. This allowed us to determine whether NMT is an essential or dispensable gene for parasite viability. The open reading frame of L. donovani NMT (ORF) was cloned into the plasmid pXNG4Sat (hereafter called NMT-TK; coding for GFP, TK and the selection marker, S-streptothricin acetyltransferase (SAT; ). This plasmid was transfected into L. donovani heterozygous for NMT, NMTHYG/+ . The ectopic expression of NMT from the plasmid made it possible to replace the second genomic NMT allele (Fig. 2a-d). The resulting parasite line, LV9 ΔNMTHYG/PAC [NMT-TK], was then transfected with a second plasmid, either encoding tdTomato and neomycin phosphotransferase (NEO) alone (hereafter called tdtom) or tdTomato and NEO plus LdNMT (hereafter called NMT-tdtom). The second plasmid was engineered so that the tdTomato was under the same regulatory control as GFP in the NMT-TK plasmid. The histograms depicting the fluorescent characteristics of the parasite lines generated are also shown on the right in Fig. 1. All of these parasite lines moderately overexpressed NMT compared to wild type (wt) levels (Additional file 1: Figure S1) but below the lethal ~5-fold limit as reported for L. major . Table 1 summarises the three plasmids and their characteristics used here.
Qualitative and quantitative in vitro analysis
First, it was necessary to establish that the plasmid shuffle approach works in the extracellular life-cycle stage, in which NMT has already been shown to be essential for viability  and to monitor the effect of GCV on these promastigotes. The hypothesis was that upon negative selection, the NMT-TK plasmid could only be lost in the presence of another expressed NMT copy present on a second plasmid. In the absence of this second NMT copy, we expected the parasites to die or to maintain the NMT-TK plasmid.
No effects were observed following GCV treatment of L. donovani NMT single knockout (KO) promastigotes transfected with the tdtom plasmid (Fig. 3a). However, the absence of the antibiotics hygromycin and G418 in the GCV-treated samples led to the loss of this plasmid in 14% of the parasite population (Fig. 3a). Double NMT KO promastigotes transfected with NMT-TK did not show altered fluorescent characteristics even upon GCV treatment for 5 passages (~50 generations) and in the absence of nourseothricin to select for this plasmid. The main population remained GFP-positive as in the untreated parasites. However, the lower GFP fluorescent-positive (dim) population increased from 10.6% to 30.2% (Additional file 2: Figure S2) gate spanning ~1 log scale across the GFP dim population shown in Fig. 1, this gate is slightly shifted along the X-axis to include the majority of the GFP-dim parasites), similar to that described earlier for essential genes [28, 31] and attributed to either a low copy number of the pXNG4 plasmid  or reduced GFP expression, possibly due to cellular stress .
Leishmania donovani double NMT KO promastigotes, only expressing NMT from the NMT-TK plasmid, did not lose this plasmid, even after ~50 generations in the presence of 50 μg/ml GCV (Fig. 3b). The majority of these parasites were still GFP/tdtom-positive, ~45%, or purely GFP-positive, ~30%, with again a dim GFP-positive population (based on gates shown in Additional file 2: Figure S2), ~30%, of which ~14% were also tdtom-positive (Fig. 3b). A similar distribution was recorded after 2 passages (~15 generations) with GCV. There was no apparent change between the control parasites maintained for 2 or 5 passages in the appropriate antibiotics to select for the plasmids (Fig. 3b) 89.1% and 92.2% double positive, respectively.
Leishmania donovani promastigotes with double NMT KOs expressing NMT from the second plasmid, NMT-tdtom, did lose the NMT-TK plasmid over time when cultured in the presence of 50 μg/ml GCV, as expected. This plasmid was completely lost after ~50 generations with the parasites remaining solely tdtom-positive (Fig. 3c). In comparison, the control parasites maintained in the appropriate selecting antibiotics did not differ from the starting population (Fig. 3c). Table 2 shows the median fluorescence intensity (MFI) values of the GFP fluorescence for the GFP dim populations. Results indicate a clear and consistent skew towards higher GFP fluorescence in the groups where the NMT-TK plasmid is present in the GCV-untreated groups and in the double NMT KOs where NMT-TK had to be maintained. These data confirm that a copy of the NMT gene is essential for promastigote viability in a direct in vitro assay. The targeted loss of NMT-TK upon GCV treatment in parasites carrying an additional copy on NMT confirms that the GCV selection is functional.
Intracellular in vitro analysis
Next, we wanted to apply the plasmid shuffle method to test whether NMT is also essential in the intracellular life-cycle stage of Leishmania. At first an in vitro approach was pursued, with two sources of phagocytes used for infection with the promastigote mutant lines described above: murine bone marrow-derived macrophages and PMA  and retinoic acid  stimulated human THP-1 cells. Upon transformation of promastigotes to amastigotes inside these cells, macrophages were treated with GCV or solvent. However, during this transition, it was noted that the GFP fluorescent intensity had dropped considerably when compared to the starting promastigotes, such that it was not possible to discriminate between single GFP-positive cells, dim GFP-positive cells or dim GFP/tdtom-positive cells (data not shown). It was concluded from this observation that the differentiated amastigotes did not proliferate at a sufficient rate for robust detection in either of these cellular systems and the doublings needed for an efficient loss of the plasmid in a single round of in vitro macrophage infection are not achievable. Hence, in vivo whole animal experiments were initiated as an alternative approach.
Qualitative in vivo analysis
To investigate to what extent and under what circumstances the NMT-TK plasmid could be lost in vivo, groups of BALB/c mice were infected with stationary phase L. donovani promastigotes of the different transgenic lines as shown in Fig. 4. Mice were then treated twice daily with GCV starting at one week post-infection in order to make sure promastigotes have transformed into amastigotes and the infection had time to establish itself; Additional file 3: Figure S3 shows flow cytometry plots of the parasites used in these experiments. After a further 4 weeks, mice were sacrificed, and the spleens removed to determine the level of parasitaemia by limiting dilution assay (parasites are not cleared from the spleen over this time period; ). The extent to which parasites were resistant to the different combinations of antibiotics, and its correlation with retention of specific plasmids was also measured. Table 3 summarises the mutant lines of L. donovani used for the qualitative and quantitative plasmid retention experiments and the applied selection conditions. To further investigate the dynamics of plasmid retention, the same samples were also re-differentiated into promastigotes, grown with the respective antibiotics for single or double NMT KO s, and subsequently analysed by flow cytometry (Fig. 5). Each column represents a different mutant line and each row represents for which plasmid(s) was selected for by respective antibiotics, no plasmid i.e. total population, isolated from infected mice that were GCV-untreated (-) or GCV-treated (+) (Fig. 5, rows 1 and 2). In order to gain more insight how the population was composed, the parasite samples were subjected to the respective antibiotics to select for either plasmid or both together (Fig. 5, rows 3–8).
The predominant population of splenic single NMT KOs had lost the NMT-TK plasmid (Fig. 5, column C, row 1 (C1 and C2) and Additional file 4: Figure S4), even in the untreated group. However, we were able to recover a small subset of parasites were this plasmid was still present, both treated or untreated as identified under appropriate selection with hygromycin and nourseothricin (Fig. 5, C3, C4). The NMT-TK plasmid, however, was retained in double NMT KOs under all selection conditions (Fig. 5, D1–D4; Additional file 5: Figure S5).
Double NMT KOs also possessing ectopic NMT alleles encoded by the NMT-TK or the NMT-tdtom plasmid were isolated from spleens of BALB/c mice. Both the GCV-treated and GCV-untreated parasites were exclusively tdtom+ (Fig. 5, A1, A2; Additional file 6: Figure S6), consistent with possession of only a single NMT allele. It was still possible to identify parasites as a GFP+ or as double positive (GFP+/tdtom+) population when selected for the NMT-TK plasmid (Fig. 5, A3, A4). No difference was observed in the qualitative analysis when these mutants were selected for the NMT-tdtom plasmid (Fig. 5, A5, A6). Both GCV-treated and GCV-untreated parasites resulted in a tdtom+ population, reflecting the result obtained in the majority of the population (Fig. 5, A1, A2). In GCV-untreated parasites, selection for NMT-TK and NMT-tdtom plasmids (Fig. 5, A7) resulted in tdtom+ and GFP+ populations, as in the original parasites injected. However, when we selected for parasites still carrying both plasmids, there were almost no double positive parasites left in the population extracted from mice treated with GCV (Fig. 5, A8). This result is in agreement with previous results (Fig. 5, C1–C4), confirming that the NMT-TK plasmid was lost while the NMT-tdtom plasmid was almost exclusively retained (i.e. Fig. 5, A2 and A6).
Double NMT KO parasites only possessing an ectopic NMT allele, encoded by the NMT-TK plasmid, when isolated from the spleens of GCV-treated and GCV-untreated BALB/c mice, were GFP+ or GFP+/tdtom+ in both groups (Fig. 5, B1 and B2; Additional file 7: Figure S7). Selecting for the NMT-TK plasmid (Fig. 5, B3, B4) in parasites isolated from untreated mice resulted in an almost exclusive GFP+ population. Similarly, selection for the NMT-TK plasmid in the GCV-treated group resulted in GFP+ (46%) and GFP+/tdtom+ (37%).
Selecting for the tdtom plasmid not encoding NMT in the GCV-untreated group (Fig. 5 B5) resulted in a mainly double positive (90% GFP+/tdtom+) population as in the original injected population. However, almost no parasites could be detected in the GCV-treated group and the few that were detected were also double positive (Fig. 5 B6). The double negative population was comprised of splenic cells as confirmed by microscopy.
In GCV-treated and GCV-untreated parasites, selection for the NMT-TK and tdtom plasmid (Fig. 5 B7, B8) resulted in GFP+/tdtom+ populations identical to the profile of the originally injected parasites.
In summary, the NMT-TK plasmid does not need to be maintained in the presence of another endogenous or ectopic gene copy of NMT even in the absence of negative drug pressure (Fig. 5, A1 and C1). In contrast, the NMT-TK plasmid is retained in the absence of another NMT gene (Fig. 5, B1 and D1).
The average median fluorescence intensity (MFI) values of the GFP fluorescence in the tdtom gate for the 4–5 animals are shown in Table 4. For the wt population the median MFI for GFP was determined by drawing a gate around the population and it was 2 and 2.1 in the GCV-untreated and GCV-treated population, respectively. The GFP MFI in the tdtom+ gate for double NMT KO [NMT-TK][tdtom] was 5.6 and double NMT KO [NMT-TK][NMT-tdtom] was 4.8 on the day of the injection. In the double NMT KO parasites possessing ectopic NMT alleles encoded by the NMT-TK or the NMT-tdtom plasmid in the GCV-treated groups the GFP MFI in the tdtom+ gate was very low (2.6 and 2.8, respectively) when only selected for the double KO and the red plasmid, respectively. Similarly the MFI values in the tdtom+ gate in the GCV-untreated groups, possessing only ectopic NMT alleles encoded by the NMT-TK or the NMT-tdtom plasmid and selection was only performed for the double KO or the red plasmid, were also very low (3.2 and 3.0, respectively). This confirms that these populations mainly or only possess the tdtom plasmid. In all other cases, the GFP MFI in the tdtom+ gate was between 4.0–4.4 as there was also a contribution of the NMT-TK plasmid with its GFP dim population.
In the double NMT KO parasites only possessing an ectopic NMT allele, encoded by the NMT-TK plasmid there was a clear skew towards higher MFI values (the lowest was 3.6 and the highest was 6.2). The lower values were only observed when only very few events were recorded in the tdtom+ gate. This confirms that there is always a contribution of the GFP dim population in the tdtom+ gate, highlighting the fact that the parasites need to maintain the NMT-TK plasmid.
Quantitative in vivo analysis
The total burden of parasites in the spleen after 35 days of infection and 28 days of treatment was determined by limiting dilution assay (Fig. 6). There was no difference observed in the total splenic numbers of L. donovani wt parasites in solvent control or GCV-treated mice. No significant differences were observed in mice infected with single NMT KOs and expressing NMT from an ectopic locus nor if NMT was expressed solely ectopically from both plasmids LV9 NMT+/HYG [NMT-TK] and LV9 ΔNMT [NMT-TK] [NMT-tdtom], although the parasite burdens were slightly less in the GCV-treated animals. However, parasite burdens were significantly reduced upon treatment with GCV in mice infected with NMT double KO parasites and with NMT only encoded by the NMT-TK plasmid (LV9 ΔNMT [NMT-TK] and LV9 ΔNMT [NMT-TK][tdtom]; Mann–Whitney U-test: U = 0, P = 0.0079).
All mutant lines resulted in a lower splenic parasite burden compared to the wt parasites. It was further noted that the parasite burdens were markedly reduced (Mann–Whitney U-test: U = 0, P = 0.0079) when NMT was only expressed from the NMT-TK plasmid as compared to the mutants in which NMT was also expressed from one genomic allele or from the NMT-tdtom plasmid, indicating that the correct regulation of NMT expression is important in this context. The burdens upon GCV treatment were comparable when NMT was only expressed from the NMT-TK plasmid (Fig. 6), although the burdens were lower in the solvent-treated mice when the parasites also carried the tdtom plasmid (Fig. 6).
The splenic parasite burden in the GCV untreated groups varied (Fig. 6). This suggested the over-expression of NMT and the regulation of its expression had an effect on the infectivity and/or the doubling-time of the parasites. Of note, this could be an effect due to the genetic manipulation in general and/or the general loss of infectivity observed when parasites are cultured axenically, a process particularly evident in L. donovani [38, 39]. Due to technical reasons, mutant lines were not passaged through mice as it was assumed that without selection for the plasmids by the respective antibiotics in vivo, dispensable plasmids would be lost rapidly; this assumption was later confirmed (see Fig. 5, A1, B1, C1 and D1).
In the surviving total splenic parasite population, it was further assessed as to what extent the plasmids were retained. This was also done by a limiting dilution assay but in the presence of the respective antibiotics (Fig. 7). Significantly lower burdens were observed upon GCV treatment compared to solvent treatment (Mann–Whitney U-test: U = 0, P = 0.0159 and Mann–Whitney U-test: U = 0, P = 0.0079), when subsequently selected for the NMT-TK plasmid. There were almost no parasites present in the spleen still carrying this NMT-TK plasmid (Fig. 7a). This was the case for all mutant lines in which NMT was present either as a genomic allele or expressed from an ectopic locus. On the other hand, a significant difference (Mann–Whitney U-test: U = 0, P = 0.0079) was observed in the splenic parasite burdens in mice upon treatment with GCV compared to the solvent-treated and subsequently selected for the tdtom plasmid (Fig. 7b). As expected, if NMT is essential and no genomic NMT allele is present, no significant difference in the parasite burden was observed when selected for the NMT-tdtom plasmid in the GCV or solvent-treated group (Fig. 7b), indicating the necessity for the presence of NMT for parasite viability.
In agreement with the hypothesis that NMT is an essential gene in amastigotes, a significant difference (Mann–Whitney U-test: U = 0, P = 0.0159) in the parasite burden was observed upon treatment with GCV and the subsequent selection for the NMT-TK and the NMT-tdtom plasmids as compared to the solvent group (Fig. 7c). This was due to loss of the NMT-TK plasmid. However, no difference was observed in the parasite burdens in mice treated with GCV or the solvent and subsequent selection for the NMT-TK and the tdtom plasmid (Fig. 7c), indicating that there is no selection in mice for the plasmid encoding only tdtom. This is in agreement with the qualitative data (Fig. 5, B1 and B2) where it was shown that the predominant population in the solvent-treated group had mostly lost the plasmid coding only for tdtom, while in the GCV-treated group the predominant population was only carrying the NMT-TK plasmid.
Determining the essentiality of a gene in Leishmania species, especially in intracellular amastigotes, is challenging. At present, an absolute requirement for parasite viability is assumed when it is only possible to fully delete a target gene of interest from its genomic locus when an ectopically expressed additional gene copy is present. This type of analysis has been exclusively performed in extracellular promastigote stages of the parasite life-cycle to date, with the information generated then correlated with similar function in intracellular amastigotes. Plasmid shuffle  provides an alternative approach for the positive discrimination of essential genes as it involves not only the deletion of the gene of interest from its genomic locus in the presence of an ectopic copy, but testing of whether this ectopic copy is dispensable upon negative selection [28,29,30]. Morales et al.  employed this approach to demonstrate an absolute requirement for the co-chaperone STI1 in parasite viability and its essential residues necessary for phosphorylation, while Dacher et al.  used similar methods to show that activity of the protein kinase LmaMPK4 is also required. However, both studies were carried out using extracellular promastigotes.
TK-expressing amastigotes of both L. donovani and L. major have already been described elsewhere. Muyombwe et al.  infected murine and human macrophages with L. major promastigotes expressing a TK plasmid and, following differentiation into amastigotes and subsequent treatment with GCV, showed decreased intracellular infection levels over time. Similarly, Ghedin et al.  expressed TK under the control of the A2 gene in L. donovani promastigotes, prior to infection of murine primary macrophages. Subsequent GCV treatment decreased infection levels substantially.
Here, we wanted to take the plasmid shuffle technique a step further and use it to prove the essentiality of a potential drug target gene in intracellular amastigotes. Using this approach, we first confirmed our previous observations that NMT is an essential gene in promastigotes in two Leishmania species [22, 23]. These experiments also demonstrated that the plasmid shuffle system was reproducible in our hands. We observed that during negative selection with GCV, a replicating population of parasites was necessary. Our data indicate that the TK-encoding plasmid is not actively lost but rather decreases in abundance through unequal division/propagation during mitosis and cytokinesis. Parasites without this plasmid then have a growth advantage and will outgrow the population which still contains the plasmid; as a result, 3–5 passages (i.e. ~25–50 generations) are needed to lose the plasmid in vitro. This also explains why we were unable to perform these experiments with different macrophage populations using the transgenic L. donovani parasites, which did not replicate fast enough intracellularly for detection in vitro. Other parasite species such as Leishmania mexicana or L. amazonesis could be more amenable to use of this in vitro strategy, as these strains replicate more rapidly in macrophage cell lines.
However, promastigotes expressing NMT only from the NMT-TK plasmid propagated even under GCV selection. This might be due to the inactivation of TK under negative selection as has been described in T. brucei, due to a single point mutation or a frame shift . Although we cannot exclude that the GCV negative selection lead to TK inactivation here, it has been reported in previous Leishmania promastigote plasmid shuffle experiments that negative selection did not cause a similar TK mutation. Instead GCV was tolerated by the parasites . Taken together, this indicates that Leishmania promastigotes can deal with GCV stress much better than T. brucei.
Given our focus on L. donovani, we therefore decided to work in vivo and chose a susceptible mouse strain that develops persistent splenic infection with a constant increase in splenic parasite burden over at least 112 days . Mice were infected intravenously with promastigotes, with GCV treatment or control injections commencing 7 days later, after promastigote transformation and establishment of an amastigote infection. After 4 weeks of treatment, spleen parasite burdens and compositions were determined, setting up limiting dilution assays with the respective antibiotics to select for parasites still carrying both or either plasmid. The quantitative analysis revealed that NMT is essential for viability in intracellular amastigotes of L. donovani, hence confirming the validity of this enzyme as a target for drug development. The data also showed that a plasmid coding for an essential gene and TK can only be lost in vitro when a functional copy of the essential gene is present, even in the presence of GCV. In vivo, on the other hand, it appears that these parasites die either by necrosis or immune clearance. However, clearance of all parasites from spleens was not complete as promastigotes for flow cytometric analysis grew out in 8 out of 9 animals if NMT was only encoded by an episome. This highlights the fact that some parasites might undergo quiescence  upon negative selection pressure and replication of parasites resumes once negative selection pressure is removed. This circumstance could lead to recrudescence in a chemical intervention. It is also possible that the treatment has just not been long enough as no promastigotes grew out of the spleen for one animal indicating complete clearance is possible. The qualitative analysis appears to indicate that the GCV treatment in vivo is not necessary, as there is no apparent difference between GCV-treated and untreated. However, these discrepancies come from the fact that the flow cytometric analysis was done some time after the parasites have been freed from the spleens, thus what was present at the end of the GCV treatment was then again amplified. There was a 10-fold difference in the quantitative analysis in the amount of parasites present in the spleen between the treated and untreated group if NMT is only encoded on a single chromosome or plasmid. Quantitative PCR on the day of parasite isolation could be more precise in determining the plasmid levels present in the parasite population. However, the NMT-TK plasmid appears to have an intrinsic toxicity as it is not maintained over the period of the experiments in the GCV-untreated single NMT KOs group or in the double NMT KOs if the NMT-tdtom plasmid is present. In contrast, in the NMT KOs the NMT-TK plasmid is maintained if it is the sole source for NMT expression. Possibly the use of almost identical plasmids only differing by the fluorescent marker and encoding for an inactivated TK will reduce the inherent negative selection pressure due to varying expression levels of the gene of interest, the size of the plasmids, and replication efficiency of plasmids. The data also show that analysis of the plasmid retention in the total population confirms the essentiality of NMT sufficiently and the subsequent dissection of the respective retention is not necessary.
Taken together, the results support that NMT is essential in vivo. As demonstrated here and in the absence of a robust inducible expression system, plasmid shuffle is a useful and powerful tool to validate the essentiality of a gene of interest in an intracellular pathogen in vivo and hence robustly confirm its suitability as a therapeutic target. This approach could also be used to characterize a gene product both temporally and functionally in order to, e.g. determine whether an active phosphorylation site is required for establishment of infection or for longer-term intracellular survival, or to what extent protein domains are utilized in infective amastigotes as compared to vector-transmitted promastigotes of Leishmania. Overall, these in vitro and in vivo data are consistent with the hypothesis that NMT is an essential gene for parasite viability and show, to our knowledge for the first time, that this target gene is not only essential in extracellular promastigotes but also in the intracellular amastigotes of L. donovani, the causative agent of human visceral leishmaniasis.
Availability of data and materials
Data supporting the conclusions of this article are included within the article and its additional files. Raw data are available from the corresponding author upon request.
bis in die (twice a day)
green fluorescent protein
median fluorescence intensity
open reading frame
phorbol 12-myristate 13-acetate
tandem tomato fluorescent protein (tdTomato)
Alexander J, Satoskar AR, Russell DG. Leishmania species: models of intracellular parasitism. J Cell Sci. 1999;112:2993–3002.
Gluenz E, Hoog JL, Smith AE, Dawe HR, Shaw MK, Gull K. Beyond 9+0: noncanonical axoneme structures characterize sensory cilia from protists to humans. FASEB J. 2010;24:3117–21.
Murray HW, Berman JD, Davies CR, Saravia NG. Advances in leishmaniasis. Lancet. 2005;366:1561–77.
Reithinger R, Dujardin JC, Louzir H, Pirmez C, Alexander B, Brooker S. Cutaneous leishmaniasis. Lancet Infect Dis. 2007;7:581–96.
Romero GAS, Costa DL, Costa CHN, de Almeida RP, de Melo EV, de Carvalho SFG, et al. Efficacy and safety of available treatments for visceral leishmaniasis in Brazil: a multicenter, randomized, open label trial. PLoS Negl Trop Dis. 2017;11:e0005706.
Barrett MP, Croft SL. Management of trypanosomiasis and leishmaniasis. Br Med Bull. 2012;104:175–96.
Andrews KT, Fisher G, Skinner-Adams TS. Drug repurposing and human parasitic protozoan diseases. Int J Parasitol Drugs Drug Resist. 2014;4:95–111.
Nwaka S, Besson D, Ramirez B, Maes L, Matheeussen A, Bickle Q, et al. Integrated dataset of screening hits against multiple neglected disease pathogens. PLoS Negl Trop Dis. 2011;5:e1412.
Siqueira-Neto JL, Moon S, Jang J, Yang G, Lee C, Moon HK, et al. An image-based high-content screening assay for compounds targeting intracellular Leishmania donovani amastigotes in human macrophages. PLoS Negl Trop Dis. 2012;6:e1671.
Freitas-Junior LH, Chatelain E, Kim HA, Siqueira-Neto JL. Visceral leishmaniasis treatment: what do we have, what do we need and how to deliver it? Int J Parasitol Drugs Drug Resist. 2012;2:11–9.
Swinney DC, Anthony J. How were new medicines discovered? Nat Rev Drug Discov. 2011;10:507–19.
Flannery EL, Chatterjee AK, Winzeler EA. Antimalarial drug discovery - approaches and progress towards new medicines. Nat Rev Microbiol. 2013;11:849–62.
Hughes JP, Rees S, Kalindjian SB, Philpott KL. Principles of early drug discovery. Br J Pharmacol. 2011;162:1239–49.
Barrett MP, Mottram JC, Coombs GH. Recent advances in identifying and validating drug targets in trypanosomes and leishmanias. Trends Microbiol. 1999;7:82–8.
Cruz A, Coburn CM, Beverley SM. Double targeted gene replacement for creating null mutants. Proc Natl Acad Sci USA. 1991;88:7170–4.
Rogers MB, Hilley JD, Dickens NJ, Wilkes J, Bates PA, Depledge DP, et al. Chromosome and gene copy number variation allow major structural change between species and strains of Leishmania. Genome Res. 2011;21:2129–42.
Kushnir S, Gase K, Breitling R, Alexandrov K. Development of an inducible protein expression system based on the protozoan host Leishmania tarentolae. Protein Expr Purif. 2005;42:37–46.
Lye LF, Owens K, Shi H, Murta SM, Vieira AC, Turco SJ, et al. Retention and loss of RNA interference pathways in trypanosomatid protozoans. PLoS Pathog. 2010;6:e1001161.
Gordon JI, Duronio RJ, Rudnick DA, Adams SP, Gokel GW. Protein N-myristoylation. J Biol Chem. 1991;266:8647–50.
Wright MH, Heal WP, Mann DJ, Tate EW. Protein myristoylation in health and disease. J Chem Biol. 2010;3:19–35.
Wright MH, Paape D, Storck EM, Serwa RA, Smith DF, Tate EW. Global analysis of protein N-myristoylation and exploration of N-myristoyltransferase as a drug target in the neglected human pathogen Leishmania donovani. Chem Biol. 2015;22:342–54.
Brannigan JA, Smith BA, Yu Z, Brzozowski AM, Hodgkinson MR, Maroof A, et al. N-myristoyltransferase from Leishmania donovani: structural and functional characterisation of a potential drug target for visceral leishmaniasis. J Mol Biol. 2010;396:985–99.
Price HP, Menon MR, Panethymitaki C, Goulding D, McKean PG, Smith DF. Myristoyl-CoA: protein N-myristoyltransferase, an essential enzyme and potential drug target in kinetoplastid parasites. J Biol Chem. 2003;278:7206–14.
Bell AS, Mills JE, Williams GP, Brannigan JA, Wilkinson AJ, Parkinson T, et al. Selective inhibitors of protozoan protein N-myristoyltransferases as starting points for tropical disease medicinal chemistry programs. PLoS Negl Trop Dis. 2012;6:e1625.
Hutton JA, Goncalves V, Brannigan JA, Paape D, Wright MH, Waugh TM, et al. Structure-based design of potent and selective Leishmania N-myristoyltransferase inhibitors. J Med Chem. 2014;57:8664–70.
Paape D, Bell AS, Heal WP, Hutton JA, Leatherbarrow RJ, Tate EW, et al. Using a non-image-based medium-throughput assay for screening compounds targeting N-myristoylation in intracellular Leishmania amastigotes. PLoS Negl Trop Dis. 2014;8:e3363.
Sikorski RS, Boeke JD. In vitro mutagenesis and plasmid shuffling: from cloned gene to mutant yeast. Methods Enzymol. 1991;194:302–18.
Dacher M, Morales MA, Pescher P, Leclercq O, Rachidi N, Prina E, et al. Probing druggability and biological function of essential proteins in Leishmania combining facilitated null mutant and plasmid shuffle analyses. Mol Microbiol. 2014;93:146–66.
McCall LI, El Aroussi A, Choi JY, Vieira DF, De Muylder G, Johnston JB, et al. Targeting ergosterol biosynthesis in Leishmania donovani: essentiality of sterol 14 alpha-demethylase. PLoS Negl Trop Dis. 2015;9:e0003588.
Morales MA, Watanabe R, Dacher M, Chafey P, Osorio y Fortea J, Scott DA, et al. Phosphoproteome dynamics reveal heat-shock protein complexes specific to the Leishmania donovani infectious stage. Proc Natl Acad Sci USA. 2010;107:8381–6.
Murta SM, Vickers TJ, Scott DA, Beverley SM. Methylene tetrahydrofolate dehydrogenase/cyclohydrolase and the synthesis of 10-CHO-THF are essential in Leishmania major. Mol Microbiol. 2009;71:1386–401.
LeBowitz JH, Coburn CM, McMahon-Pratt D, Beverley SM. Development of a stable Leishmania expression vector and application to the study of parasite surface antigen genes. Proc Natl Acad Sci USA. 1990;87:9736–40.
Beattie L, Peltan A, Maroof A, Kirby A, Brown N, Coles M, et al. Dynamic imaging of experimental Leishmania donovani-induced hepatic granulomas detects Kupffer cell-restricted antigen presentation to antigen-specific CD8 T cells. PLoS Pathog. 2010;6:e1000805.
Seifert K, Escobar P, Croft SL. In vitro activity of anti-leishmanial drugs against Leishmania donovani is host cell dependent. J Antimicrob Chemother. 2010;65:508–11.
Ogunkolade BW, Colomb-Valet I, Monjour L, Rhodes-Feuillette A, Abita JP, Frommel D. Interactions between the human monocytic leukaemia THP-1 cell line and Old and New World species of Leishmania. Acta Trop. 1990;47:171–6.
Kropf P, Kadolsky UD, Rogers M, Cloke TE, Müller I. 13 - The leishmaniasis model. In: Dieter K, Stefan HEK, editors. Methods in microbiology, vol. 37. London: Academic Press; 2010. p. 307–28.
Cotterell SE, Engwerda CR, Kaye PM. Enhanced hematopoietic activity accompanies parasite expansion in the spleen and bone marrow of mice infected with Leishmania donovani. Infect Immun. 2000;68:1840–8.
De T, Roy S. Infectivity and attenuation of Leishmania donovani promastigotes: association of galactosyl transferase with loss of parasite virulence. J Parasitol. 1999;85:54–9.
Giannini MS. Effects of promastigote growth phase, frequency of subculture, and host age on promastigote-initiated infections with Leishmania donovani in the golden hamster. J Protozool. 1974;21:521–7.
Muyombwe A, Olivier M, Ouellette M, Papadopoulou B. Selective killing of Leishmania amastigotes expressing a thymidine kinase suicide gene. Exp Parasitol. 1997;85:35–42.
Ghedin E, Charest H, Zhang WW, Debrabant A, Dwyer D, Matlashewski G. Inducible expression of suicide genes in Leishmania donovani amastigotes. J Biol Chem. 1998;273:22997–3003.
Valdes J, Taylor MC, Cross MA, Ligtenberg MJ, Rudenko G, Borst P. The viral thymidine kinase gene as a tool for the study of mutagenesis in Trypanosoma brucei. Nucleic Acids Res. 1996;24:1809–15.
Mandell MA, Beverley SM. Continual renewal and replication of persistent Leishmania major parasites in concomitantly immune hosts. Proc Natl Acad Sci USA. 2017;114:E801–10.
Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 1970;227:680–5.
Price HP, Hodgkinson MR, Wright MH, Tate EW, Smith BA, Carrington M, et al. A role for the vesicle-associated tubulin binding protein ARL6 (BBS3) in flagellum extension in Trypanosoma brucei. Biochim Biophys Acta. 2012;1823:1178–91.
We thank Stephen Beverley (Washington University, USA) for generous provision of the pXNG4-SAT plasmid; Richard McCulloch and Jeremy Mottram (University of Glasgow, UK) for proof reading and helpful discussions.
This study was funded by the Wellcome Trust (Grant No. 077503) and the Department of Biology, University of York.
Ethics approval and consent to participate
All animal care and experimental procedures were regulated under the UK Animals (Scientific Procedures) Act 1986 (revised under European Directive 2010/63/EU) and were performed under UK Home Office License (‘Immunity and Immunopathology of Leishmaniasis’) Ref # PPL 60/3708, with approval from the Animal Procedures and Ethics Committee of the Department of Biology, University of York.
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Paape, D., Prendergast, C.T., Price, H.P. et al. Genetic validation of Leishmania genes essential for amastigote survival in vivo using N-myristoyltransferase as a model. Parasites Vectors 13, 132 (2020). https://doi.org/10.1186/s13071-020-3999-1
- Plasmid shuffle
- Mouse infection
- Therapeutic target validation