Tubulin detyrosination shapes Leishmania cytoskeletal architecture and virulence

Abstract

Tubulin detyrosination has been implicated in various human disorders and is important for regulating microtubule dynamics. While in most organisms this modification is restricted to α-tubulin, in trypanosomatid parasites, it occurs on both α- and β-tubulin. Here, we show that in Leishmania, a single vasohibin (LmVASH) enzyme is responsible for differential kinetics of α- and β-tubulin detyrosination. LmVASH knockout parasites, which are completely devoid of detyrosination, show decreased levels of glutamylation and exhibit a strongly diminished pathogenicity in mice, correlating with decreased proliferation in macrophages. Reduced virulence is associated with altered morphogenesis and flagellum remodeling in detyrosination-deficient amastigotes. Flagellum shortening in the absence of detyrosination is caused by hyperactivity of a microtubule-depolymerizing Kinesin-13 homolog, demonstrating its function as a key reader of the trypanosomatid-tubulin code. Taken together, our work establishes the importance of tubulin detyrosination in remodeling the microtubule-based cytoskeleton required for efficient proliferation in the mammalian host. This highlights tubulin detyrosination as a potential target for therapeutic action against leishmaniasis.

Keywords: dynamics; kinesin-13; microtubules; trypanosomatids; vasohibin.

Fig. 1.

Distinct distribution of α- and β-tubulin detyrosination in Leishmania. (A) Sequence alignment of tubulin tails. T. brucei α-tubulin isotype located on chromosome 1, L. mexicana α-tubulin isotype located on chromosome 13, T. brucei β-tubulin isotype located on chromosome 1, L. mexicana β₁-tubulin isotype located on chromosome 8 and 32 and L. mexicana β₂-tubulin isotype located on chromosome 21. Blue and green colors correspond to conserved and equivalent residues, respectively. (B–E) UExM analysis of tyrosinated and detyrosinated α and β-tubulin in L. mexicana parental promastigotes. Maximum intensity projections (B, C, and E) and a single section (D) of z-stack confocal images are presented. The αΔ1 antibody recognizes detyrosinated α-tubulin. The YL1/2 antibody recognizes tyrosinated α-tubulin. The β₁Δ1 antibody recognizes detyrosinated β-tubulin. The βtyr antibody recognizes tyrosinated β-tubulin. (C and D) White arrows indicate the mitotic spindle. p: posterior;.a: anterior; of: old flagellum; nf: new flagellum; MtQ: Microtubule quartet; mBB: mature basal body: pBB: pro-basal body. Nb: only one MtQ, mBB, and pBB is highlighted in each dividing cells. (Scale bar: 10 µm.)

Fig. 2.

Deletion of LmVASH reveals differential kinetics of detyrosination and a cross-talk with polyglutamylation. (A) Immunofluorescence and (B) immunoblot analysis of tyrosinated and detyrosinated α and β-tubulin in cytoskeletons from L. mexicana parental and LmVASH knockout promastigotes. Note the absence of α and β-tubulin detyrosination and increase levels of α and β-tubulin tyrosination signal in LmVASH knockout cells. LACK is used as a loading control. (Scale bar: 5 µm.) (C) Immunoblot analysis of tubulin PTMs in cytoskeletons from L. mexicana parental and LmVASH knockout promastigotes. In the absence of detyrosination, LmVASH knockout cells showed decreased levels of polyglutamylation. The GT335 antibody recognizes glutamate side chains of any length. The PolyE antibody recognizes polyglutamylation with at least four glutamate residues. α-tubulin is used as a loading control. (D and E) In vitro time course analysis of the recombinant wild-type LmVASH-mediated α- and β- tubulin detyrosination. Samples were analyzed by immunoblotting (SI Appendix, Fig. S3F) and relative optical density of three independent assays was measured (n = 3; error bars represent SEM). (D) Comparison of detyrosination activity on polymerized (MT) and free tubulin purified from LmVASH knockout promastigotes was performed based on relative optical densities generated by the αΔ1 and the β₁Δ1 antibodies. Note that LmVASH enzyme displays detyrosinase activity toward soluble and polymerized Leishmania tubulin. (E) Comparison of detyrosination activity on polymerized tubulin (MT) from LmVASH knockout promastigotes was performed based on relative optical densities generated by the YL1/2 and the βtyr antibodies. Note that LmVASH enzyme displays slower kinetics of β-tubulin detyrosination toward polymerized Leishmania tubulin. The αΔ1 antibody recognizes detyrosinated α-tubulin. The YL1/2 antibody recognizes tyrosinated α-tubulin. The β₁Δ1 antibody recognizes detyrosinated β-tubulin. The βtyr antibody recognizes tyrosinated β-tubulin.

Fig. 3.

Removal of LmVASH carboxypeptidase activity reduces replication within macrophages and pathogenicity in mice. (A) Representative growth curve (log scale) of L. mexicana promastigotes cultured over 4 d from parental, LmVASH knockout and wild-type LmVASH add-back cell lines. Cell density was determined by counting at 24 h intervals (n = 3; error bars represent SD). (B) Growth curve (log scale) of L. mexicana axenic amastigotes obtained after 3 d of differentiation and cultured over a 4-d course from parental, LmVASH knockout, wild-type and enzymatically dead LmVASH complemented cell lines. Cell density was determined by counting at 24 h intervals (n = 3; error bars represent SD). The P value was calculated using two-tailed unpaired Student’s t test comparing each cell line at day 4. **P < 0.01, *P < 0.05. (C) Schematic representation of THP-1 macrophage infection. Axenic amastigotes of parental, LmVASH knockout and LmVASH add-back cells were used to infect differentiated THP-1 monocytes. Note that deletion of LmVASH leads to reduction of the parasitic index (PI) in infected THP-1 cells. The PI was defined as the percentage of infected macrophages x number of intracellular parasites/macrophage. ***P < 0.001, **P < 0.01. (D) Schematic representation of the experimental mouse model used to analyze Leishmania virulence in vivo. Mice footpads were infected either with the parental, null mutant or add-back stationary phase promastigotes and the infection progress was evaluated over a 6-wk time course. (E) Measurement of mean footpad lesion size during a 6-wk infection time course showing that infection with LmVASH knockout produces smaller lesions compared to parental and LmVASH add-back stationary-phase promastigotes. Error bars represent SDs. The P value was calculated using two-tailed unpaired Student’s t test comparing each cell line at week 6. **P < 0.01, *P < 0.05. (F) Measurement of parasite burden at the end of the 6-wk infection time course in the footpad lesion showing that infection with LmVASH knockout leads to a significant drop of parasite burden compared to the parental and LmVASH add-back cells. The parasite number from each infection is plotted, with the mean and the 95% SEM interval indicated. **P < 0.01 (Student’s t test).

Fig. 4.

Lack of detyrosination impacts morphogenesis and flagellar pocket shape. (A) Fluorescence micrographs of axenic amastigotes cultured over 7 d after differentiation from parental and LmVASH knockout cells expressing SMP1-mCherry. Note the longer flagellum and cell body in LmVASH knockout cells. (Scale bar: 5 µm.) (B–D) Measurements of cell body, flagellum, and kinetoplast to cell body end distances in parental and LmVASH knockout axenic amastigotes (n = 240,170 and 230 cells, respectively, from three independent experiments for each cell line). (E) Immunofluorescence micrographs of THP1 cells infected with parental and LmVASH knockout amastigotes probed with βtyr antibody. (Scale bar: 5 µm.) Note the longer cell body in LmVASH knockout cells. Red and green channels represent extracellular (absent in this picture) and intracellular parasites, respectively. (F) Cell body length measurements of intracellular amastigotes from parental and LmVASH knockout cells after 24H and 48H postinfection (PI) inside THP1 macrophages (n = 300 at 24 h from three independent experiments and n = 150 at 48 h from two independent experiments for each cell line). (G) Cell body length measurements of axenic amastigotes from parental and LmVASH knockout cells cultured over 6 d after differentiation and treated during 24 h with 200 nM paclitaxel (n = 200). Cells were fixed with PFA at a density of 4 × 10⁷ cells/mL. n indicate cells from three independent experiments for each cell line. Dotted lines indicate the median, upper, and lower quartiles. ****P < 0.0001 (ANOVA-one way). (H) Cartoon summarizing the effect of dynamic instability in the cell morphology of amastigotes from LmVASH knockout cells. Paclitaxel-induced microtubule stability restores cell body length in the LmVASH knockout with neutral effects on the parental cell line. Not to scale. (I) Growth curve (log scale) of axenic amastigotes from parental and LmVASH knockout cells cultured after 3 d of differentiation and treated during 4 d with 200 nM paclitaxel. Cell density was determined by counting at 24 h intervals (n = 3; error bars represent SD). *P < 0.05, ***P < 0.001 (Unpaired Student t test). (J) Representative electron micrograph of longitudinal section of the flagellar pocket in parental and LmVASH knockout axenic amastigote cells. (i), (ii), and (iii) represent the flagellar pocket length, flagellar pocket neck lengths and flagellar width at the constriction point, respectively. (Scale bar: 1 µm.) (K) Flagellar pocket length measurements of axenic amastigotes from parental and LmVASH knockout cells (n = 75). (L) Fluorescence micrographs of axenic amastigotes from parental and LmVASH knockout cells expressing SMP1-mNGreen and the flagellar pocket neck (FP Neck) protein (LmxM.28.1990) fused with mCherry. Note the longer flagellar pocket neck in the LmVASH knockout cell. (Scale bar: 2 µm.) (M) Flagellar pocket neck length measurements of axenic amastigotes from parental and LmVASH knockout cells (n = 150). n represents cells from three independent experiments. (B–D, F, K, and M) Dotted lines indicate the median, upper, and lower quartiles. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (Mann–Whitney test).

Fig. 5.

Tubulin detyrosination promotes the remodeling of the amastigote flagellum. (A) Schematic overview of the axoneme remodeling from promastigote to amastigote form. Not to scale. (B) Western blot of detergent-extracted cytoskeletons showing delayed loss of PFR2 in LmVASH knockout axenic amastigotes compared to parental cells cultured over 1 d (1d) and 3 d (3d) after differentiation. 5 × 10⁷ cytoskeletons were loaded on a 10% SDS-PAGE, transferred and immuno-probed with anti-PFR2 and anti-LACK (loading control) antibodies. (C) Immunofluorescence assay of whole cells from parental and LmVASH knockout axenic amastigotes after 24 h postdifferentiation expressing the flagellar membrane marker SMP1-mCh and probed with anti-mCherry and anti-PFR2 antibodies. (Scale bar: 10 μm.) (D) Western blot of detergent-extracted cytoskeletons showing increased levels of IFT52-3xHA in LmVASH knockout axenic amastigotes compared to parental cells cultured over 7 d after differentiation. 5 × 10⁷ cytoskeletons were loaded on a 10% SDS-PAGE, transferred, and immuno-probed with anti-HA and anti-LACK (loading control) antibodies. (E) Immunofluorescence assay of whole cells from parental and LmVASH knockout axenic amastigotes expressing the flagellar membrane marker SMP1-mCh and IFT52-3xHA. (F) Western blot of detergent-extracted cytoskeletons showing increased levels of PF16-3xHA in LmVASH knockout axenic amastigotes compared to parental cells cultured over 7 d after differentiation. 5 × 10⁷ cytoskeletons were loaded on a 10% SDS-PAGE, transferred, and immuno-probed with anti-HA and anti-LACK (loading control) antibodies. (G) Immunofluorescence assay of whole cells from parental and LmVASH knockout axenic amastigotes expressing the flagellar membrane marker SMP1-mCh and PF16-3xHA. Cells were probed with anti-mCherry and anti-HA antibodies. (Scale bar: 2 μm.) (H) Quantification of 9+0, 9+1, and 9+2 axoneme configurations and central material (CM) in transversal section of transmission electron micrographs from parental (n = 82) and LmVASH knockout (n = 110) axenic amastigotes cultured over 7 d after differentiation. (I) Transmission electron micrographs of axenic amastigotes from parental and LmVASH knockout cells cultured over 7 d after differentiation. Transversal sections of the axonemal structure in the parental cell line with a 9+0 configuration and the LmVASH null mutant with nine doublets surrounding a central electron-dense core. (Scale bar: 100 nm.)

Fig. 6.

Microtubule depolymerizing activity of the flagellar LmKIN13.2 is increased in the absence of detyrosination. (A) Scanning electron microscopy micrographs of axenic amastigotes showing a shorter external flagellum in LmVASH knockout cells compared to parental, LmKIN13.2 knockout and LmVASH-LmKIN13.2 double knockout cell lines cultured over 3 d after differentiation. White arrows show external flagella. (Scale bar: 5 μm.) (B) External flagellum length measurements of axenic amastigotes showing that LmKIN13.2 is responsible for shortening of the external flagellum in LmVASH-KO parasites. Measurements were performed on parental, LmVASH knockout, LmKIN13.2 knockout and LmVASH-LmKIN13.2 double knockout cell lines expressing SMP1-mCherry. The external flagellum length corresponds to the distance from the distal end of the SMP1 signal to the cell tip (the flagellum exit point from the cell body). Amastigotes were cultured over 3 d after differentiation and were fixed with PFA at a density of 3 × 107 cells/mL. Fifty cells were counted per sample (n = 3). Dotted lines indicate the median, upper, and lower quartiles. ****P < 0.0001 (ANOVA-one way). (C) Immunofluorescence showing localization of LmKIN13.2-3xHA to the flagellum tip in axenic amastigote. Cells coexpressing HA-tagged LmKIN13-2 and SMP1-mCherry were probed with anti-mCherry and anti-HA antibodies. (Scale bar: 5 μm.) (D) Microtubule-depolymerizing activity of the LmKIN13.2. Immunofluorescent analysis of tubulin in RPE1 cells expressing LmKIN13-2-HA. Blue and red squares show microtubules and depolymerized tubulin, respectively. (Scale bar: 10 μm.) (E) External flagellum length measurements of stationary phase promastigotes showing that LmKIN13.2 is responsible for shortening of the external flagellum in LmVASH-KO parasites. Measurements were performed on parental, LmVASH knockout, LmKIN13.2 knockout and LmVASH-LmKIN13.2 double knockout cell lines expressing SMP1-mCherry. The external flagellum length corresponds to the distance from the distal end of the SMP1 signal to the cell tip (the flagellum exit point from the cell body). Promastigotes were fixed with PFA at a density of 1.7 × 107 cells/mL. Fifty cells were counted per sample (n = 3). Dotted lines indicate the median, upper, and lower quartiles. ****P < 0.0001 (ANOVA-one way). (C) Immunofluorescence showing localization of LmKIN13.2-3xHA to the flagellum tip in axenic amastigote. Cells coexpressing HA-tagged LmKIN13-2 and SMP1-mCherry were probed with anti-mCherry and anti-HA antibodies. (Scale bar: 5 μm.) (D) Microtubule-depolymerizing activity of the LmKIN13.2. Immunofluorescent analysis of tubulin in RPE1 cells expressing LmKIN13-2-HA. Blue and red squares show microtubules and depolymerized tubulin, respectively. (Scale bar: 10 μm.) (F) Schematic representation of the flagellum shortening in the promastigote and amastigote form due to increased activity of the microtubule depolymerizing activity of the flagellar LmKIN13.2 in the absence of detyrosination. Not to scale.