Abstract
Leishmania parasites include important pathogens and model organisms and are even used for the production of recombinant proteins. However, functional genomics and the characterization of essential genes are often limited in Leishmania because of low-throughput technologies for gene disruption or tagging and the absence of components for RNA interference. Here, we tested the T7 RNA polymerase-dependent CRISPR-Cas9 system by Beneke et al. and the glmS ribozyme-based knock-down system in the model parasite Leishmania tarentolae. We successfully deleted two reference genes encoding the flagellar motility factor Pf16 and the salvage-pathway enzyme adenine phosphoribosyltransferase, resulting in immotile and drug-resistant parasites, respectively. In contrast, we were unable to disrupt the gene encoding the mitochondrial flavoprotein Erv. Cultivation of L. tarentolae in standard BHI medium resulted in a constitutive down-regulation of an episomal mCherry-glmS reporter by 40 to 60%. For inducible knock-downs, we evaluated the growth of L. tarentolae in alternative media and identified supplemented MEM, IMDM and McCoy’s 5A medium as candidates. Cultivation in supplemented MEM allowed an inducible, glucosamine concentration-dependent down-regulation of the episomal mCherry-glmS reporter by more than 70%. However, chromosomal glmS-tagging of the genes encoding Pf16, adenine phosphoribosyltransferase or Erv did not reveal a knock-down phenotype. Our data demonstrate the suitability of the CRISPR-Cas9 system for the disruption and tagging of genes in L. tarentolae as well as the limitations of the glmS system, which was restricted to moderate efficiencies for episomal knock-downs and caused no detectable phenotype for chromosomal knock-downs.
Introduction
Kinetoplastid parasites cause important diseases, such as sleeping sickness, Chagas disease as well as cutaneous, mucocutaneous or visceral leishmaniasis1,2. Furthermore, kinetoplastid parasites are valuable unicellular model organisms and have led to seminal discoveries in molecular cell biology, including RNA editing3 or the structure of the glycosylphosphatidylinositol anchor4. The gecko parasite Leishmania tarentolae, for example, is nonpathogenic for humans, grows fast, serves as a valuable model system for RNA editing5 and mitochondrial protein import6 and is also used for the constitutive or tetracycline-induced production of recombinant proteins7,8. However, gene disruption or tagging in diploid L. tarentolae so far relied on rather laborious traditional genetics. The recent introduction of the CRISPR-Cas9 technology in kinetoplastid parasite research9–11 now provides the opportunity to overcome this limitation, especially taking into account that a high-throughput method for gene disruption or tagging was successfully established in L. major, L. mexicana and L. donovani12–14.
Molecular tools to alter intracellular RNA or protein concentrations are crucial for functional genomics, in particular, to address essential genes that cannot be deleted. One common approach is to down-regulate a protein of interest with the help of a genetically encoded protein tag that is degraded in the absence of a stabilizing ligand15. However, protein tagging may affect the localization and function of the protein of interest and is sometimes even used to interfere with protein trafficking16. Furthermore, stabilization of the protein tag usually requires the long-term addition of an expensive ligand to the cell culture medium. An alternative approach is to down-regulate the protein-encoding RNA of interest by RNA interference (RNAi)17,18. RNAi is a widely used method for post-transcriptional gene repression in eukaryotes but requires a specific RNAi machinery for RNA degradation19. This machinery has been lost in many eukaryotes, e.g., in several important apicomplexan and kinetoplastid parasites20–22. A transcriptional knock-down approach is to generate transgenic cells with a T7 RNA polymerase and a tetracycline-controlled trans-activator, which is a fusion protein between a tetracycline repressor and a transactivation domain23. This system also requires an artificial operon that comprises a T7 promoter, the gene of interest and a tetracycline operator. Depending on the type of transactivation domain, the trans-activator binds to the operator and activates the T7 RNA polymerase either in the absence or presence of tetracycline24. One disadvantage of this versatile transcriptional system is that tetracycline or related antibiotics may also block translation in mitochondria or plastids in a time- and concentration-dependent manner25. A fourth knock-down approach utilizes genetically encoded fusion constructs between an RNA of interest and the glmS ribozyme, which is derived from a part of the 5’ untranslated region of the Bacillus subtilis glucosamine-6-phosphate synthase gene26. The glmS ribozyme is activated by glucosamine-6-phosphate or, to a lesser degree, by the precursor glucosamine, and is inhibited by glucose-6-phosphate26,27. Hence, fusion constructs between the RNA of interest and the glmS ribozyme are destabilized by the addition of glucosamine. The amino group of bound glucosamine-6-phosphate is part of the active site and participates in the self-cleavage of an AG-motif at the 5’-end of the ribozyme26,28,29. Replacement of the AG-motif with a CC-motif in the so-called glmS M9 mutant prevents self-cleavage and can be used as a negative control26. The advantages of the glmS ribozyme system are its simplicity and the low costs for glucosamine. However, depending on the investigated organism or cell system, high concentrations of glucosamine might be toxic30, and the activity of the glmS ribozyme might depend on the medium composition and the ratio between ribozyme activators and inhibitors31.
Here, we successfully adopted the LeishGEdit CRISPR-Cas9 method by Beneke et al.12,32 to generate knock-out and knock-in strains and to test the glmS ribozyme-based knock-down system in L. tarentolae. Deletion of the flagellar motility factor PF16 resulted in immotile parasites, deletion of adenine phosphoribosyltransferase (APRT) rendered parasites resistant to 4-aminopyrazolopyrimidine, and deletion of the mitochondrial flavoprotein ERV was lethal. Using a non-invasive plate reader assay for monitoring fluorescent reporter proteins, we found that the medium composition influences the glmS ribozyme-mediated knock-down effect. Standard brain-heart-infusion (BHI) medium resulted in a constitutive down-regulation of an episomal mCherry-glmS reporter, whereas supplemented minimal essential medium (MEM) allowed a glucosamine concentration-dependent down-regulation. However, chromosomal glmS-tagging of the genes encoding Pf16, APRT or Erv did not result in a detectable knock-down phenotype.
Results
Generation of L. tarentolae knock-out strains using the CRISPR-Cas9 system
L. tarentolae promastigotes were transfected with plasmid pTB007 encoding T7 RNA polymerase and FLAG-tagged Cas9 as described previously for L. major and L. mexicana12,13 (Fig. 1a). Western blot analysis confirmed the presence of Cas9 (Fig. 1b) and the functionality of the T7 RNA polymerase was shown by fluorescence microscopy using an mCherry reporter under control of a T7 promoter (Fig. 1c). Furthermore, growth analyses revealed no detrimental effect for plasmid pTB007 in L. tarentolae (Fig. 1d).
As proof-of-principle experiments and because of the expected characteristic phenotypes that could also serve as references for subsequent knock-down studies, we then deleted either the gene for the flagellar motility factor PF16 or the adenine phosphoribosyltransferase (APRT) (Fig. 2). These modifications should render L. tarentolae promastigotes either immotile, due to the loss of PF16, or resistant to the pro-drug 4-aminopyrazolopyrimidine, due to the loss of APRT, as reported previously for L. mexicana12 and L. donovani33,34, respectively. Briefly, L. tarentolae parasites with T7 RNA polymerase and Cas9 were used to drive the in vivo transcription of two different gene-specific single guide RNAs (sgRNA) for the excision of the gene of interest12,32. The two sgRNA-encoding DNAs were generated by PCR using an universal sgRNA antisense primer and a sense primer that encodes the 24 nucleotide T7 promoter, the site-specific 20 nucleotide guide sequence and a 20 nucleotide sequence that is complementary to the universal sgRNA antisense primer. Plasmid pTPuro was used to amplify a puromycin resistance cassette by PCR using primers with 30 nucleotide homology regions upstream and downstream of the first and second sgRNA-induced double-strand break, respectively12,32. Parasites with plasmid pTB007 were co-transfected with the three PCR products (encoding the two sgRNAs for the excision and the puromycin cassette for the repair of the double-strand breaks) and selected on BHI agar plates with 20 μg/mL puromycin. Colonies appeared after 7-14 days and were transferred to liquid culture containing 20 μg/mL puromycin. PCR and motility analyses confirmed the loss of PF16 after a single round of selection, resulting in dividing but completely immotile promastigotes (Fig. 2b,c). PCR analysis and dose-response curves for 4-aminopyrazolopyrimidine also confirmed the loss of APRT after a single round of selection, resulting in drug-resistant parasites (Fig. 2e,f).
We previously showed that the gene encoding the mitochondrial flavoenzyme Erv cannot be deleted in the related human pathogen L. infantum using standard genetics35. This enzyme is thought to play an important role in mitochondrial protein import in kinetoplastid parasites6,36–38. We therefore wanted to test whether the CRISPR-Cas9 method might allow us to select knock-out parasites that can be missed in standard genetic experiments (for example, slow-growing parasites that are usually overgrown by plasmid-containing wild-type parasites without negative selection). Using the same strategy as for PF16 and APRT, we were unable to delete ERV in L. tarentolae. No colonies appeared following transfection with the correct PCR products and selection with puromycin in two independent experiments. Hence, ERV is most likely essential in Leishmania parasites. In summary, we showed that the PCR- and T7 RNA polymerase-based CRISPR-Cas9 system by Beneke et al. can be also used in L. tarentolae. Furthermore, in accordance with previous findings in other Leishmania species, we show that (i) PF16 is crucial for promastigote motility, (ii) deletion of APRT confers resistance towards the pro-drug 4-aminopyrazolopyrimidine and (iii) ERV is refractory to disruption.
Limitations of initial experiments with GFP-glmS and BHI medium
In order to test the glmS ribozyme system in L. tarentolae, we initially chose to quantify the potential down-regulation of plasmid-encoded green fluorescent protein (GFP) in intact cells. A GFP-glmS fusion construct was cloned into vector pX, and transfected clonal L. tarentolae cell lines were subsequently cultured in standard BHI medium with or without 10 mM glucosamine (Fig. S1a). A rather weak fluorescence was observed for control cultures without glucosamine (Fig. S1b). Although parasite cultures with glucosamine appeared to have an even 50% lower fluorescence, we could not obtain statistically meaningful quantitative data from three biological replicates because of the variable fluorescence of the control culture. The rather weak GFP fluorescence in the absence of external glucosamine might have resulted from an unfavorable, medium-dependent ratio between glmS activators and inhibitors. The residual fluorescence in the presence of glucosamine might have been caused by an inefficient down-regulation or by the mitochondrial autofluorescence (which was previously shown to overlap with the fluorescence of GFP and to give a false positive signal, particularly at a low signal-to-noise ratio39). In order to overcome these potential limitations, we tested alternative media and used mCherry as a fluorescent reporter in subsequent experiments.
Identification of alternative cell culture media
To address a putative effect of the medium composition on the glmS ribozyme system, as reported in yeast31, we first had to establish an alternative cell culture protocol. Therefore, we analyzed the growth of L. tarentolae in six different (defined) liquid media that are also available as SILAC media for potential future knock-down studies in combination with quantitative mass spectrometry. The media were supplemented with different concentrations of folic acid, hemin and fetal bovine serum (FBS) resulting in 216 different media as described in the Materials and Methods section. DMEM, DMEM/F12 and RPMI-1640 media as well as media without FBS were altogether excluded from further analysis because parasites died within one to three days regardless of the supplement concentrations. In contrast, parasites from selected FBS-containing cultures with supplemented MEM, IMDM and McCoy’s 5A medium survived. These media were subsequently analyzed in more detail and compared to standard BHI medium as a control. In general, parasite growth was better with 10% FBS compared to 5% FBS. Highest maximum cell densities were observed in McCoy’s 5A and BHI medium followed by IMDM and MEM (Fig. 3a). The morphology and motility of parasites in FBS-containing MEM, IMDM and McCoy’s 5A medium were similar to the control in BHI medium (Fig. 3b). Attempts to replace FBS with 0.5% AlbuMax II resulted in cell aggregation and replacement with dialyzed FBS lead to cell death. Hence, we could not identify a suitable medium with dialyzed FBS for potential SILAC experiments. Nevertheless, we decided to test the glmS ribozyme system in supplemented MEM containing 5% FBS, because MEM contains 50% less glucose (the precursor of the glmS activator glucosamine and the inhibitor glucose-6-phosphate) than BHI medium and has the lowest glucose concentration among the three defined media (1.0, 3.0 and 4.5 g/L in MEM, McCoy’s 5A medium and IMDM, respectively). The low toxicity of glucosamine in MEM was another positive factor for subsequent glmS studies. Single passages of parasite cultures to MEM containing up to 15 mM glucosamine had no toxic effect and concentrations above 15 mM glucosamine reduced the final cell density by only one-fourth (Fig. 3c). In summary, we identified supplemented MEM, IMDM and McCoy’s 5A medium with 5 or 10% FBS as suitable liquid media for the growth of L. tarentolae and selected MEM supplemented with 10 μg/mL hemin chloride, 25 mM HEPES and 5% FBS for subsequent glmS studies because of its low glucose content and glucosamine toxicity profile.
An inducible episomal glmS ribozyme knock-down system in L. tarentolae
To overcome the potential limitations described above, we cloned an mCherry-glmS fusion construct in vector pX and analyzed the glmS ribozyme system for transfected clonal cell lines of L. tarentolae that were cultured in supplemented MEM (Fig. 4a). As a control, we also fused mCherry to the inactive glmS M9 mutant to study the potential effect of endogenous glmS activators and inhibitors. The mCherry fluorescence was quantified in a microplate reader after three passages in supplemented MEM with or without 5 or 10 mM glucosamine (Fig. 4b). This protocol was chosen to dilute the intracellular signal from stable mCherry that was already synthesized before the addition of glucosamine. In contrast to single glucosamine treatments in Fig. 3c, the repeated addition of 10 mM glucosamine resulted in a time-dependent growth retardation (Fig. 4c), which was compensated in the fluorescence measurements by seeding the same number of parasites per well. The wild-type glmS construct was reproducibly down-regulated in the presence of glucosamine in a concentration-dependent manner (Fig. 4d). Furthermore, even in the absence of external glucosamine, the mCherry signal of the culture with the wild-type glmS construct was about 35% lower than the fluorescence of the glmS M9 control, suggesting that endogenous glucosamine-6-phosphate causes an intrinsic glmS ribozyme activation and down-regulation of mCherry. The addition of 10 mM glucosamine to the medium significantly increased the difference between the fluorescence of cultures with wild-type glmS and the glmS M9 control and resulted in a down-regulation of mCherry by more than 70%. In summary, we established an inducible glmS knock-down system for a plasmid-encoded reporter in L. tarentolae using supplemented MEM as an alternative cell culture medium.
A constitutive episomal glmS ribozyme knock-down system in L. tarentolae
To address whether the glmS ribozyme system can be also used in BHI medium, we repeated the experiments for plasmid-encoded wild-type and mutant mCherry-glmS according to the protocol in Fig. 4b using standard BHI medium instead of supplemented MEM. The mCherry fluorescence of cultures with the wild-type glmS construct was lowered by up to two-thirds compared to the glmS M9 control (Fig. 5). The results were highly reproducible for two independent clonal cell lines and among biological replicates. External glucosamine seemed to have no effect on the mCherry fluorescence, suggesting that internal glmS activator(s) and inhibitor(s) overlay the impact of external glucosamine in BHI medium. A constitutive intrinsic down-regulation of wild-type glmS constructs in BHI medium without external glucosamine also explains (i) a passage-independent constant mCherry fluorescence at days 1, 2 and 3 (data not shown) and (ii) the rather low GFP fluorescence in our initial experiments that were performed without a glmS M9 control (Fig. S1). In summary, the glmS ribozyme system can be used in BHI medium for constitutive knock-down experiments using plasmid-encoding genes. Control experiments with glmS M9 constructs enable a quantification of the degree of constitutive down-regulation.
Chromosomal glmS-tagging of PF16, APRT and ERV
In the final set of experiments, we used the CRISPR-Cas9 system for gene tagging in order to study a glmS ribozyme-dependent knock-down of chromosomal genes in L. tarentolae. As a proof-of-principle, the methodology was first tested for mCherry-tagged Pf16 and Erv using plasmid pPLOT-mCherry-puro12. C-terminally mCherry-tagged Pf16 was shown to localize to the flagellum (Fig. S2), thus confirming the suitability of the CRISPR-Cas9 method for gene tagging in L. tarentolae. In contrast, neither N-nor C-terminal mCherry-tagging of Erv resulted in viable parasites, suggesting that the tag interferes with an essential protein function. Next, we fused PF16 (Fig. 6), APRT (Fig. 7) and ERV (Fig. 8) with the wild-type glmS ribozyme as well as the glmS M9 control. We therefore generated a template plasmid for PCR-based 3’-tagging of a gene of interest that is compatible with the LeishGEdit system12,32. Parasites with plasmid pTB007 were co-transfected with PCR products encoding (i) a sgRNA for the introduction of a double-strand break close to the stop codon of the gene of interest and (ii) an insert for a triple HA-tag, the glmS ribozyme and a resistance cassette for the repair of the double-strand break (Fig. S3). Following transfection and selection on BHI agar plates, colonies appeared after 7-14 days.
Tagging of both PF16 alleles required simultaneous selection with 20 μg/mL puromycin and 10 μg/mL blasticidin (Fig. 6a). PCR analyses confirmed the knock-in and replacement of untagged PF16 (Fig. 6b). Sequencing of PCR products from genomic DNA revealed the correct sequences for wild-type glmS ribozyme and the glmS M9 control (Fig. 6c). However, all strains remained fully motile in the presence or absence of 10 mM glucosamine in supplemented MEM. Hence, the glmS knock-down system did not result in a comparable phenotype to the knock-out control in Fig. 2c.
Tagging of both APRT alleles with glmS was achieved after a single round of selection with 20 μg/mL puromycin (Fig. 7a). PCR analyses confirmed the expected modification of the APRT locus (Fig. 7b) and sequencing of PCR products from genomic DNA revealed the correct sequences for wild-type glmS ribozyme and the glmS M9 control (Fig. 7c). All strains remained sensitive towards 4-aminopyrazolopyrimidine in the presence or absence of 10 mM glucosamine in supplemented MEM, regardless whether wild-type glmS or the M9 control was fused to APRT (Fig. 7d). The susceptibilities of the APRT fusion constructs with wild-type glmS or the M9 control remained also highly similar when the cells were pretreated with or without glucosamine before the addition of 4-aminopyrazolopyrimidine (Fig. 7e). Thus, the knock-out phenotype for APRT in Fig. 2f could not be mimicked using the glmS knock-down system.
Tagging of both ERV alleles was achieved after simultaneous selection with 20 μg/mL puromycin and 10 μg/mL blasticidin (Fig. 8a). PCR analyses confirmed the knock-in and replacement of untagged ERV (Fig. 8b). However, tagging of ERV with the wild-type glmS ribozyme only slightly decreased cell growth in supplemented MEM as compared to the M9 control. The addition of 15 mM glucosamine appeared to have only a minor effect on both cell lines (Fig. 8c). Sequencing of PCR products from genomic DNA confirmed the expected sequences for wild-type glmS ribozyme and the glmS M9 control (Fig. 8d) and a more detailed growth analysis also revealed no significant growth defects for wild-type glmS-tagged ERV in the presence of up to 15 mM glucosamine (Fig. 8e). In summary, we showed that the LeishGEdit CRISPR-Cas9 system can be also applied to gene tagging and knock-in studies in L. tarentolae whereas the glmS ribozyme system appears to lack efficiency for general chromosomal knock-down studies in this organism.
Discussion
The CRISPR-Cas9 system has been recently established for a variety of Leishmania parasites10–13,32,40. Here we demonstrated that a robust and simple PCR- and T7 RNA polymerase-based CRISPR-Cas9 system can be also used for the rapid and reliable generation of knock-out and knock-in cell lines in L. tarentolae. We successfully deleted the genes encoding the flagellar motility factor Pf16 and the salvage-pathway enzyme adenine phosphoribosyltransferase. The results confirmed their expected relevance for parasite motility and drug susceptibility, respectively12,13,33. Furthermore, using the CRISPR-Cas9 system we were unable to delete the L. tarentolae gene that encodes the mitochondrial flavoenzyme Erv in accordance with previous results that were obtained for L. infantum using traditional genetics35. Hence, two unrelated genetic studies in Leishmania, as well as RNAi experiments in T. brucei38,41, show that Erv is crucial for parasite survival. The results are encouraging for potential intervention strategies taking into account that parasite Erv homologues and human Erv/ALR have very different structures and also employ a deviating enzyme mechanism37,41.
Whilst the CRISPR-Cas9 system will facilitate the identification of essential genes in important human pathogens, the characterization of these genes and the suitability of Leishmania parasites as model organisms in the post-genomic era will probably depend on the development of alternative knock-down methods. To date, functional genomics in kinetoplastid parasites and the down-regulation of a gene of interest are restricted to only a few species because of a limited method repertoire42. For example, CRISPR interference43 would cause a simultaneous down-regulation of numerous genes in kinetoplastid parasites, which appear to lack gene-specific promoters and have head-to-tail oriented genes that are co-transcribed as polycistronic units44,45. Furthermore, while RNAi is commonly used in Trypanosoma brucei18,22 and was also shown to be functional in L. braziliensis21,46, attempts to implement RNAi in T. cruzi, L. major or L. donovani failed47,48. These and many other kinetoplastid parasites including L. tarentolae have lost key components of the RNAi machinery21,22. With the exception of the functional RNAi system in the Leishmania subgenus Viannia21, knock-down experiments in Leishmania rely so far on protein-destabilization strategies49–52 or tetracycline repressor systems53,54. Applications of these knock-down methods in human pathogens, as reported for L. major, L. mexicana, L. braziliensis and L. donovani, might have drawbacks, for example, because of protein-tagging or antibiotic toxicity. Interference with protein function could, for example, explain why we were unable to add an mCherry tag to L. tarentolae Erv in contrast to Pf16.
The glmS ribozyme system has become a valuable alternative knock-down method in eukaryotes since its discovery in 2004 (ref. 26). For example, the system is now commonly used in the apicomplexan malaria parasite Plasmodium falciparum for the analysis of drug targets as well as essential genes and processes30,55–58. A glmS ribozyme system has also been established for yeast and, very recently, for T. brucei and T. cruzi31,59,60. The glmS ribozyme system now offers the opportunity for medium-dependent constitutive or inducible knock-down experiments of episomal constructs in L. tarentolae. Although the knock-down efficiency for our episomal mCherry-glmS reporter was below 80%, the glmS M9 controls revealed that the inducible down-regulation with 10 mM glucosamine in supplemented MEM was slightly stronger than the constitutive down-regulation in standard BHI medium. Since GFP and mCherry are rather stable proteins, the down-regulation of other proteins might be more efficient. Higher glucosamine concentrations could be used for short-term knock-downs as long as potential glucosamine-dependent off-target effects are addressed with glmS M9 controls. Glucosamine toxicity is not a concern for the constitutive down-regulation in standard BHI medium, which could be explained by relatively high endogenous glucosamine-6-phosphate concentrations as suggested recently for the glmS ribozyme system in T. cruzi60. Under these conditions, episomal glmS M9 constructs only serve as controls in order to quantify the degree of down-regulation. Whether a glmS ribozyme-dependent down-regulation of episomal constructs is sufficient for phenotypic and functional analyses in L. tarentolae might depend on the protein of interest. Nevertheless, the simplicity of the constitutive knock-down system in standard BHI medium might facilitate the generation and phenotypic screening of episomal knock-down libraries. Regarding chromosomal glmS knock-down studies, tagging of three different genes did not result in the expected phenotypes (as shown for our characterized knock-out strains). Thus, the glmS system has intrinsic limitations that impede its general application for chromosomal knock-down studies in L. tarentolae. In our opinion, the glmS ribozyme system might give similar results in other Leishmania species despite the presence of N-acetyl glucosamine 6-phosphate deacetylase, which was recently suggested to be one key factor for the functionality of the system in kinetoplastid parasites60. Whether the suitability of the glmS ribozyme system in Leishmania depends on (invariable) biological factors in addition to the medium composition remains to be addressed in future studies.
In conclusion, the CRISPR-Cas9 system facilitates the genetic manipulation of L. tarentolae and provides a powerful technology for future analyses of this nonpathogenic model organism. In contrast, the glmS system failed to produce a detectable phenotype for chromosomal knock-downs but might be used for medium-dependent constitutive or inducible episomal knock-downs in L. tarentolae. The development of an unbiased and highly efficient knock-down system for important pathogens and model systems remains one of the biggest challenges in Leishmania molecular biology.
Materials and Methods
Generation of L. tarentolae CRISPR-Cas9 knock-out cell lines
L. tarentolae promastigote mutants were obtained by reproducing the CRISPR-Cas9 system reported by Beneke et al. using plasmids pTPuro, pTBlast and pTB00712,13. Donor DNA for the repair of double-strand breaks (containing a selectable marker and 30 nt homology arms) as well sgRNA templates for the excision of PF16, APRT or ERV were generated according to the LeishGEdit method12,32. Primer sequences for the generation of sgRNA templates and the amplification of targeting cassettes for strain L. tarentolae ParrotTarII were obtained online (www.leishgedit.net) and are listed in Supplementary Tables 1–4. Puromycin or blasticidin resistance cassettes with 5’- and 3’-homology regions for the repair of excised PF16, APRT or ERV were amplified by PCR using plasmid pTPuro or pTBlast. Transfections were performed as outlined below in a parental cell line that transiently expressed Cas9 nuclease and T7 RNA polymerase from plasmid pTB007.
Generation of GFP- and mCherry-glmS reporter constructs
Plasmids encoding GFP- or mCherry-reporters were cloned using the primers that are listed in Supplementary Table 5. To clone reporter construct pX-GFP-glmS, the fusion construct GFP-glmS was PCR-amplified using primers 114 and 115 with vector pARL-GFP-glmS as a template. The PCR product was subsequently cloned into the BamHI and HindIII restriction sites of vector pX (ref.61). To generate pX-mCherry-glmS, mCherry was first PCR-amplified from vector pLEXY_IE-blecherry4 (JenaBioscience) using primers 127 and 128 and cloned into pCR 2.1 TOPO (Thermo Fisher Scientific) to generate pCR 2.1 TOPO-mCherry. The M9 mutation reported by Winkler et al.26 (glmSM9) was introduced by PCR using primer 129 with the desired mismatch, antisense primer 130 and vector pARL-GFP-glmS as a template. The wild-type glmS ribozyme-encoding sequence (glmSwt) was PCR-amplified from vector pARL-GFP-glmS using primers 131 and 130. The glmSwt and glmSM9 PCR products were cloned into the BsrGI and HindIII restriction sites of pCR 2.1 TOPO-mCherry to generate pCR 2.1 TOPO-mCherry-glmSwt and pCR 2.1 TOPO-mCherry-glmSM9, respectively. The inserts mCherry-glmSwt and mCherry-glmSM9 were subsequently excised from the according pCR 2.1 TOPO constructs and subcloned into the BamHI and HindIII restriction sites of pX. All constructs were confirmed by Sanger sequencing (SEQ-IT GmbH & Co. KG). The sequences for GFP-glmS, mCherry-glmSwt and mCherry-glmSM9 are listed in the Supplementary material.
Generation of chromosomally tagged L. tarentolae cell lines
Constructs for chromosomal mCherry-tagging of PF16 or ERV were PCR-amplified using plasmid pPLOT-mCherry-puro as a template12,32. For chromosomal glmS-tagging of PF16, APRT or ERV, we generated plasmids that are compatible with the LeishGEdit system using the primers that are listed in Supplementary Table 6. We therefore modified plasmids pMOTag-glmS-4H wt and M9 from ref.59 (addgene plasmids #106378 and #106379, respectively) so that they contain the puromycin or blasticidin cassette with primer binding sites 4 and 5 from the LeishGEdit system. To generate these plasmids, (i) we amplified the 3xHA-glmS wt and M9 sequences from pMOTag-glmS-4H wt and M9 by PCR with primer 173, which contains the ApaI restriction site and the primer binding site 5, and primer 174, which contains an NcoI restriction site. (ii) Resistance marker cassettes against puromycin or blasticidin suitable for 3’-tagging were amplified by PCR from plasmids pTPuro or pTBlast with primers 175 and 176 containing NcoI and BamHI restriction sites, respectively. (iii) Fragments generated from steps (i) and (ii) were ligated using the NcoI restriction site and finally subcloned into the pMOTag-glmS-4H backbone using the ApaI and BamHI restriction sites. The resulting plasmids served as templates to generate the donor DNA for the repair of double-strand breaks in combination with the sgRNA templates for PF16, APRT or ERV as described above. The sequences of the the 3xHA-glmS-Puro/Blast cassettes are listed in the Supplementary material.
Standard L. tarentolae culture and screen of alternative culture media
BHI powder (#237200) was purchased from BD Bioscience, AlbuMax II lipid-rich BSA, heat inactivated FBS as well as DMEM (#42430), DMEM/F-12 (#11330), RPMI-1640 (#52400), IMDM (#12440), MEM (#31095) and McCoy’s 5A medium (#26600) were from Thermo Fisher Scientific. A 2 mg/mL stock solution of hemin chloride from Calbiochem was prepared in 0.05 M NaOH, sterile filtered (0.2 μm) and stored at −20°C. L. tarentolae promastigotes were cultured in 25 cm2 culture flasks under standard conditions at 27°C in BHI medium containing 37 mg/mL BHI powder and 10 μg/mL hemin as described previously62. MEM and McCoy’s 5A medium were supplemented with 25 mM HEPES and sterile filtered (0.2 μm) prior to use. Except for BHI medium, all media were supplemented with either 0.5% (w/v) AlbuMax II or 0%, 5% or 10% (v/v) FBS. The media were also supplemented with folic acid (0 μg/mL, 10 μg/mL or 50 μg/mL) and hemin chloride solution (5 μg/mL, 10 μg/mL or 20 μg/mL) resulting in 216 different test conditions that were compared with standard BHI medium as a control. L. tarentolae parasites from standard cultures were centrifuged in the mid-logarithmic phase at 1500 × g for 10 min at room temperature and washed in the different media without any supplements. After centrifugation, cells were resuspended in the according supplemented medium and diluted to an initial concentration of 5 × 106 cells/mL. Technical triplicates of the test cultures were transferred to 48-well plates (500 μL per well) and incubated at 27°C for three days. Parasite growth and morphology were qualitatively evaluated every 24 h by standard and differential interference contrast (DIC) light microscopy. Based on the screen in 48-well plates, growth curves were subsequently determined in 25 cm2 T-flasks for the following media containing 10 mg/mL hemin chloride: IMDM supplemented with 5% FBS, 10% FBS or 0.5% AlbuMax™, MEM supplemented with 25 mM HEPES and 5% or 10% FBS, McCoy’s 5A Medium supplemented with 25 mM HEPES and 5% or 10% FBS, and standard BHI medium as a control. L. tarentolae parasites from standard cultures were centrifuged in the mid-logarithmic phase at 1500 × g for 10 min at room temperature and washed once with one of the supplemented media. After centrifugation, cells were resuspended in the according supplemented medium and diluted in a total volume of 10 mL to an initial concentration of 5 × 106 cells/mL. The parasites were cultured at 27°C on a Rotamax 120 shaker at 50 rpm. Aliquots (100 μL) were removed every 24 h and fixated with paraformaldehyde before the cell density was determined with a Neubauer chamber. In order to avoid a putative nutrient depletion, the cultures were subsequently centrifuged at 1500 × g for 10 min at room temperature and the cells were resuspended in 10 mL fresh medium. The parasite morphology was documented for cultures in the mid-logarithmic phase on a microscope slide that was sealed with paraffin using a Zeiss LSM780 microscope and the software ZEN2010.
Transfection and selection of L. tarentolae
For transfection, 107 L. tarentolae promastigotes in mid-logarithmic phase were centrifuged at 1500 × g for 3 min at room temperature and washed with 1 mL transfection buffer (21 mM HEPES, 137 mM NaCl, 5 mM KCl, 0.7 mM NaH2PO4, 6 mM glucose, pH 7.4). Parasites were subsequently resuspended in 100 μL nucleofector solution of the Basic Parasite Nucleofector Kit 2 (Lonza), combined with 50 μL plasmid solution containing 5-10 μg DNA, and pulsed in a Lonza Nucleofactor IIb using program U-033. Transfected parasites were allowed to recover in 10 mL standard BHI medium for 24 h before parasites were centrifuged and plated on BHI medium that was supplemented with 0.8% (w/v) agar, 0.08% (w/v) folic acid, 10% (v/v) FBS, 20 μg/mL hemin chloride and 40 μg/mL G418 disulfate (Calbiochem). Single colonies appeared 7-10 days after transfection and clonal parasites were subsequently transferred to liquid media. Targeting cassettes and sgRNA templates from 40 μL unpurified PCR product were transfected into 107 parasites as described previously12,63 using the Lonza Nucleofactor IIb program X-001. Electroporated cells were allowed to recover in 2mL BHI medium without antibiotics for 16 h before being plated on BHI agar to obtain isogenic lines as described above. Single colonies appeared 7-14 days after transfection. The genome modifications were confirmed by analytical PCR using the primers in Supplementary Table 7 and by Sanger sequencing (SEQ-IT GmbH & Co. KG) of PCR products with genomic DNA as a template.
Fluorescence measurements
Stock solutions of 500 mM D-(+)-glucosamine (#G4875 Sigma) were prepared in BHI medium or MEM. The pH was adjusted to 7.5 with 5 M NaOH before the stock solutions were sterile filtered (0.22 μm) and stored at −20 °C. In order to evaluate the glmS ribozyme-dependent down-regulation of the GFP reporter, clonal L. tarentolae cell lines with vector pX-GFP-glmS were grown at 27°C in standard BHI medium that was supplemented with 0.1 mg/mL G418 disulfate. Cultures (10 mL) in 25 cm2 T-flasks were diluted to an initial density of 5 × 106 cells/mL, supplemented with 0 or 10 mM glucosamine and incubated for 24 h. Around 3 × 107 cells were harvested by centrifugation at 1500 × g for 5 min and used for further fluorescence measurements as outlined below. To assess the glmS ribozyme-dependent down-regulation of the mCherry reporter, clonal L. tarentolae cell lines with vector pX-mCherry-glmSwt or pX-mCherry-glmSM9 were grown at 27°C either in MEM that was supplemented with 10 μg/mL hemin, 5% (v/v) FBS, 25 mM HEPES, pH 7.4 and 0.1 mg/mL G418 disulfate or in standard BHI medium with 0.1 mg/mL G418 disulfate. Cultures (10 mL) in 25 cm2 T-flasks were diluted to an initial density of 5 × 106 cells/mL or an OD600nm of 0.2, supplemented with 0, 5 or 10 mM glucosamine and incubated for 24 h. Every 24 h, parasites were diluted to an OD600nm of 0.2 in fresh supplemented MEM media with 0, 5 or 10 mM glucosamine. Three days after inoculation, cultures were harvested at 1500 × g for 5 min. The fluorescence from reporter cell lines was determined in a CLARIOstar fluorescence plate reader (BMG Labtech). Parasites were resuspended in 400 μL 100 mM MES/Tris buffer pH 6.0 to a final OD600nm of 1.5. Technical duplicates of the suspensions (190 μL each) were transferred to a flat-bottom 96-well microplate (#353219, BD Biosciences). The microplate was centrifuged for 5 min at 30 × g and the fluorescence was subsequently measured with the following setting: GFP: λExc = 485-15, dichroic 497.8, λEmis = 530-20, N° cycles 10, flashes/well 40, cycle time 6 s; mCherry: λExc = 583-15, dichroic 602, λEmis = 623,5-20, N° cycles 6, flashes/well 40, cycle time 5 s. The data was averaged from three or four biological replicates as indicated. Statistical analyses were performed in SigmaPlot 13 using the one-way ANOVA method.
Author contributions
G.L.T., L.L. and M.D. designed the project and experiments. G.L.T. performed the knock-out and knock-down studies as well as CLARIOstar measurements. L.S. performed the ERV knock-down studies. L.L. performed the growth analyses in alternative media. G.L.T., L.S. and L.L. analyzed the data. M.D. supervised the study and wrote the manuscript. All authors edited and approved the manuscript.
Conflict of interest
The authors declare that they have no conflict of interest.
Supplementary Material and Methods
Sequence of GFP-glmS construct
Sequence of mCherry-glmSwt construct
Sequence of mCherry-glmSM9 construct
Sequence of the glmS-tagging cassette in pMOTag-4-glmS-puro-5
Sequence of the glmS-tagging cassette in pMOTag-4-M9-puro-5
Sequence of the glmS-tagging cassette in pMOTag-4-wt-blast-5
Sequence of the glmS-tagging cassette in pMOTag-4-M9-blast-5
Acknowledgements
G.L.T. was funded by the German Academic Exchange Service (DAAD). This work was in part funded by the DFG (grants DE 1431/9-1, DE 1431/10-1 and DE 1431/10-2 to M.D.). We thank Eva Gluenz for plasmids pTPuro, pTBlast, pPLOT-mCherry-puro and pTB007, Roberto Docampo for plasmids pMOTag-glmS-4H wt and M9, Michael Lanzer for plasmid pARL-GFP-glmS, Simone Eggert, Jessica Kehrer, Katharina Quadt, Carolina Andrade, Stefan Kins and Freddy Frischknecht for help with microscopy, and Prince Saforo Amponsah and Bruce Morgan for help with the CLARIOstar measurements.