CSAP localizes to polyglutamylated microtubules and promotes proper cilia function and zebrafish development

Abstract

The diverse populations of microtubule polymers in cells are functionally distinguished by different posttranslational modifications, including polyglutamylation. Polyglutamylation is enriched on subsets of microtubules including those found in the centrioles, mitotic spindle, and cilia. However, whether this modification alters intrinsic microtubule dynamics or affects extrinsic associations with specific interacting partners remains to be determined. Here we identify the microtubule-binding protein centriole and spindle-associated protein (CSAP), which colocalizes with polyglutamylated tubulin to centrioles, spindle microtubules, and cilia in human tissue culture cells. Reducing tubulin polyglutamylation prevents CSAP localization to both spindle and cilia microtubules. In zebrafish, CSAP is required for normal brain development and proper left-right asymmetry, defects that are qualitatively similar to those reported previously for depletion of polyglutamylation-conjugating enzymes. We also find that CSAP is required for proper cilia beating. Our work supports a model in which polyglutamylation can target selected microtubule-associated proteins, such as CSAP, to microtubule subpopulations, providing specific functional capabilities to these populations.

FIGURE 1:

CSAP localizes to centrioles, mitotic spindle microtubules, and cilia. (A) Images of HeLa cells stably expressing GFP-CSAP. CSAP localizes to centrioles throughout the cell cycle and to mitotic spindle microtubules. Scale bar, 10 μm. (B) Immunofluorescence images showing the localization of CSAP (anti-CSAP antibodies) to the mitotic spindle (DM1α antibodies). Scale bar, 10 μm. (C) CSAP localizes to basal bodies and cilia. Images show serum-starved hTERT-RPE1 cells stably expressing GFP-CSAP. Scale bar, 10 μm. (D) Immunofluorescence images showing the colocalization of CSAP (anti-CSAP antibodies) with acetylated tubulin, a marker for cilia. Scale bar, 10 μm. (E) Analysis of CSAP centriole and cilia localization by immuno-EM. Totals of 203 and 131 gold particles were quantified for unciliated and ciliated centrioles, respectively. A nearly identical localization pattern was observed with CSAP gold label overlapping with the microtubule cylinder walls for 80% of the gold and 20% in the centriole lumen. A representative example is provided in three serial sections (left). Arrow denotes cilia. The compiled localization pattern is described by the relative distribution of 25 red spots (right). Scale bar, 100 nm.

FIGURE 2:

CSAP requires polyglutamylation to target to specific microtubule subpopulations. (A, B) CSAP precisely colocalizes with polyglutamylated tubulin but not centrin. Immunofluorescence images showing CSAP (anti-CSAP antibodies), centrin (a marker for centrioles), and (A) GT335 or (B) PolyE (markers for polyglutamylated microtubules). (C) In hTERT-RPE1 cells, CSAP (anti-CSAP antibodies) colocalizes with acetylated tubulin and PolyE along the length of the cilia. (D) CSAP colocalizes with B3 along neuronal tracks. Immunofluorescence images showing CSAP (anti-CSAP antibodies) and B3 (a marker for neuronal microtubules) in differentiated human iPS cells. (E) Western blot showing the cosedimentation of GST-CSAP with bovine brain microtubules. (F) CSAP localization to the mitotic spindle requires TTLL5. Immunofluorescence images of HeLa cells treated with control siRNAs or siRNAs against TTLL5. Percentages indicate the relative level of CSAP or PolyE compared with microtubule staining in the TTLL5 depletion vs. control. Antibodies to CSAP, PolyE, and DM1α were used to assess localization. n = 10 cells/condition. (G) Immunofluorescence images as in D, showing CSAP and PolyE localization to the cilia in serum-starved RPE-1 cells after control and TTLL5 RNAi. N = 10 cells/condition. Scale bars, 10 μm.

FIGURE 3:

CSAP is expressed throughout the developing zebrafish embryo and localizes to basal bodies. (A) In situ hybridization showing the expression of CSAP in the developing zebrafish embryo. CSAP is expressed throughout the developing embryo but is enriched in brain tissue and somites. The sense probe serves as a negative control. (B) Colocalization of mCherry-CSAP with the ciliary marker Arl13b-GFP in live transgenic zebrafish embryos. CSAP localizes to the basal bodies at the base of cilia along the apical side of the neuroepithelium, shown here on the posterior side of the midbrain–hindbrain boundary. MHBC, Midbrain–hindbrain boundary constriction; H, hindbrain. Embryos are oriented with the anterior to the left. The enlargements are taken from the designated region of the MHBC. Scale bar, 20 μm.

FIGURE 4:

CSAP is required for proper zebrafish development. (A) Live brightfield imaging of 24-hpf zebrafish embryos injected with 10 ng of control MO or 10 ng of CSAP splice-site MO in conjunction with p53 MO. Top, a dorsal view of the brain. H, Hindbrain; asterisk, the ear. A tracing of the neuroepithelium and brain ventricles is highlighted below. Bottom, a lateral view of whole embryos. Morphant embryos have abnormal brain development, somite defects, and cell death. Asterisk designates the ear. (B) Lateral view of live brightfield imaging of 48- and 72-hpf zebrafish injected with 10 ng of control MO or 10 ng of CSAP splice-site MO in conjunction with p53 MO. Morphant embryos display a prominent heart edema (arrow) at this time point. These older embryos continue to display brain defects but lack hydrocephaly and kidney cysts and have normal otoliths. (C) Percentages of embryos with the CSAP phenotype (brain defects, heart positioning, tail defects). All phenotypes can be rescued by coinjecting CSAP mRNA with the CSAP MO. The rescue of the touch response is highlighted at the bottom.

FIGURE 5:

CSAP is required for neuronal development. (A) Immunofluorescence images of 24-hpf embryos, measuring apoptotic index (indicated by the percentage). Embryos are stained for TUNEL (green) and propidium iodide (red). Cells were counted by placing a 400-nm box around a region from the ear to the midline. N = 15 embryos/condition. Embryos are positioned anterior to posterior. H, Hindbrain; asterisk, the ear. Scale bar, 50 μm. (B) Immunofluorescence images of 24-hpf embryos measuring mitotic index (indicated by the percentage). Embryos are stained for PH3 (green) and propidium iodide (red). Cells were counted as in A. n = 15 embryos/condition. Scale bar, 50 μm. (C) Immunofluorescence images of axon tracts in the brain stained with acetylated tubulin. Top, embryos are oriented laterally with the anterior to the left. Major forebrain/midbrain tracts are labeled as follows: ac (anterior commissure), sot (supraoptic tract), tpc (tract of the posterior commissure), and tpoc (tact of postoptic commissure). CSAP MO-injected embryos have defects in axon tract formation, specifically the tpoc and tpc, as designated by asterisks. The eyes have been removed to easily visualize axon tracts. Bottom, dorsal view of embryos positioned with the anterior at the top. Axon tracts in the hindbrain are labeled with acetylated tubulin. In CSAP morphants, the density of axon tracts is reduced. R3 and R5, rhombomeres 3 and 5, respectively. The ear is noted by an asterisk. Scale bar, 100 μm. (D) The 48-hpf embryos fixed and labeled with RMO44 (a marker for reticulospinal neurons). Dorsal view of the embryos with the anterior at the top. In CSAP morphants, the density of Mauthner interneurons is reduced. The ear is noted by an asterisk. Scale bar, 100 μm.

FIGURE 6:

CSAP is required for heart laterality and cilia beating. (A) The 24-hpf embryos fixed and labeled with an antisense probe against cardiac myosin light chain 2 (cmlc2). Control MO-injected embryos display normal heart development on the left side, whereas CSAP MO-injected embryos display heart development on the right side and in the middle of the embryo at a rate higher than controls. This defect is rescued by coinjection of CSAP mRNA. n > 75 embryos/condition. Scale bar, 100 μm. (B) Table of laterality defects with rescue experiments. (C) Still images taken from time-lapse sequences visualizing the beating of cilia along the neuroepithelial lining along the apical surface of the neuroepithelium on the posterior side of the midbrain–hindbrain boundary. Images were acquired at 5 or 6 frames/s. Ciliary movement was visualized by the kymographs shown at the bottom. Kymographs were generated by drawing a line across the tip of cilia (red hashed line) to track movement as a function of time. Also see Supplemental Movies S1–S4.