Targeted Deletion of Centrin in Leishmania braziliensis Using CRISPR-Cas9-Based Editing

Abstract

Leishmania braziliensis is the main causative agent of Tegumentary Leishmaniasis in the Americas. However, difficulties related to genome manipulation, experimental infection, and parasite growth have so far limited studies with this species. CRISPR-Cas9-based technology has made genome editing more accessible, and here we have successfully employed the LeishGEdit approach to attenuate L. braziliensis. We generated a transgenic cell line expressing Cas9 and T7 RNA polymerase, which was employed for the targeted deletion of centrin, a calcium-binding cytoskeletal protein involved in the centrosome duplication in eukaryotes. Centrin-deficient Leishmania exhibit growth arrest at the amastigote stage. Whole-genome sequencing of centrin-deficient L. braziliensis (LbCen^-/- ) did not indicate the presence of off-target mutations. In vitro, the growth rates of LbCen^-/- and wild-type promastigotes were similar, but axenic and intracellular LbCen^-/- amastigotes showed a multinucleated phenotype with impaired survival following macrophage infection. Upon inoculation into BALB/c mice, LbCen^-/- were detected at an early time point but failed to induce lesion formation, contrary to control animals, infected with wild-type L. braziliensis. A significantly lower parasite burden was also observed in mice inoculated with LbCen^-/- , differently from control mice. Given that centrin-deficient Leishmania sp. have become candidates for vaccine development, we propose that LbCen^-/- can be further explored for the purposes of immunoprophylaxis against American Tegumentary Leishmaniasis.

Keywords: LeishGEedit; attenuation; genetic manipulation; leishmaniasis; vaccine development.

Figure 1

Homology modeling of the derived protein sequence from putative L. braziliensis centrin (LbrM.22.1290). (A) Amino-acid sequence alignment of L. braziliensis (Lbr.M.22.190) and Trypanosoma brucei (Tb927.7.3410) centrin used as the template for the structure prediction with 99.9% confidence and 100% sequence coverage. (B) Three-dimensional structure of the predicted protein and (C) Ramachandran analysis of the predicted L. braziliensis centrin: Red: most favored regions, Yellow: additional allowed, generously allowed regions and Light Yellow- disallowed regions.

Figure 2

Generation of centrin^−/− L. braziliensis. (A) Western blot of whole cell lysates probed with anti-Cas9 antibody and anti-b-Actin. LbWT, parental cell line, LbCas9T7, L. braziliensis expressing Cas9 and T7. (B) In silico representation of the CRISPR-Cas9 based deletion of the L. braziliensis putative centrin gene (i) centrin genomic locus indicating sgRNA (guide RNA) binding sites at both 5` and 3` UTRs, 3`& 5` HF (homology flank) or flanking regions (30 bp) and primers for the correct integration (flag). PAC and NEO forward primers (PAC F and NEO F), Cen ORF forward and reverse (CEN ORF F and CEN ORF R) diagnostic primers for detection of centrin gene (amplification of 344 bp fragment). (ii) & (iii) Donor cassettes containing PAC and NEO antibiotics markers, indicating diagnostic reverse primers (PAC R and NEO R and size of expected amplicons for the confirmation of correct integration of the cassettes (788 and 886 bp, respectively). (C) PCR analysis of generated cell lines: test for the presence of the Centrin in Lb WT parental line and in LbCas9T7; test for the integration of the PAC and NEO-resistance genes in LbCen^−/−. PCR products were analyzed on a 1% agarose gel.

Figure 3

Deletion of the LbrM.22.1290 centrin gene as confirmed by whole genome sequencing. (A) MHOM/BR/75/M2904 L. braziliensis chromosome 22 encompassing the 527,855–547,147 region covered by LbCas9T7 and LbCen^−/− genomic read libraries. Blue boxes represent L. braziliensis genes, drawn in scale; LbrM.22.1290 centrin gene region is highlighted by red box. Gray histograms above the genes represent the read depth in each genomic position for each genomic library, where colored markings denote SNPs in the reads when compared to the reference genome. (B) Read mapping in the genomic region encompassing the LbrM.22.1290 gene. Mapping of individual reads is represented by gray boxes. An expected number of reads mapped into the LbrM.22.1290 centrin gene region in the LbCas9T7 isolate, whereas no read from LbCen^−/− library mapped into this gene.

Figure 4

Read depth coverage (RDC) alterations in LbCas9T7 and LbCen^−/−. (A) Absolute values of the RDC difference from LbCas9T7 and LbCen^−/− genomic libraries in each L. braziliensis gene normalized by genome coverage. Each line of the X axis corresponds to a gene, ordered by its genome position from the smallest to the largest chromosome. The Y axis represents the difference of RDC absolute values for each gene from LbCas9T7 and LbCen^−/− lines. The LbrM.22.1290 centrin gene is highlighted in red. (B) Density plot of the RDC differences in LbCas9T7 and LbCen^−/− lines. The X axis corresponds to the difference of RDC absolute values of each gene from LbCas9T7 and LbCen^−/− lines, the Y axis represents the distribution of occurrence of these values. Most of the differences are below 0.3.

Figure 5

Kinetics of LbCas9T7 and LbCen^−/− promastigote growth in vitro. (A) Promastigote cultures were started at 5 × 10⁵ parasites/ml and were maintained at 26 °C for 7 days in supplemented Schneider media. Parasite numbers were determined daily by counting using a hemocytometer. Data is plotted as mean ± SEM and is from a representative experiment, performed in triplicate. (B) DIC and fluorescent (DAPI) representative micrographs of LbWT, LbCas9T7, and LbCen^−/− promastigotes after 96 h of culture. (C) Axenic amastigote cultures were started at 1 × 10⁶ parasites/ml and were maintained at 34°C for 5 days in supplemented Schneider media, pH 5.5. Parasite numbers were determined daily by counting using a hemocytometer. Data is plotted as mean ± SEM, and is from a representative experiment, performed in quadruplicate, *p < 0.05. (D) DIC and fluorescent (DAPI) representative micrographs of LbWT, LbCas9T7, and LbCen^−/− amastigotes after 96 h of culture.

Figure 6

Ultrastructural analysis of LbCen^−/− axenic amastigotes. LbWT, LbCas9T7, and LbCen^−/− axenic amastigotes were harvested, fixed and prepared for scanning (A) or transmission electron microscopy (B). Transmission electron micrographs of axenic amastigotes showing the presence of a single nucleus (N) in LbWT and in LbCas9T7 and the presence of multi nuclei in LbCen^−/−. Scale bars, 0.5 µm (LbWT and LbCas9T7); 1 µm (LbCen^−/−).

Figure 7

Macrophage infection with LbCen^−/−. BMDM were infected with LbWT or LbCen^−/− (10:1, parasite/macrophage ratio) promastigotes for 24 h. Cells were extensively washed and further cultured for 48, 72, 96, 120 or 144 h. Cells were stained with H&E and evaluated for the percentage of infection (A) and the number of amastigotes per 200 macrophages (B) by optical microscopy. (C) Photomicrographs showing infected macrophages at 96 h. Data (mean ± SEM) are pooled from four independent experiments, each performed in quadruplicate. *p < 0.05.

Figure 8

Parasite load in mice inoculated with LbCen^−/−. BALB/c mice (10 per group) were infected with 2 × 10⁵ LbWT or LbCen^−/− promastigotes, in the ear dermis and parasite load was determined, four days later, at the inoculation site (ear) (A) and in draining lymph nodes (B) by Limiting Dilution Analysis. Data (mean ± SEM) are from one representative experiment. BALB/c mice (10 per group) were infected as described and lesion development was measured weekly (C). Six (D) and twelve (E) weeks post infection, parasite load was evaluated by Limiting Dilution Analysis. Data (mean ± SEM) are from one representative experiment.