Bug22p, a conserved centrosomal/ciliary protein also present in higher plants, is required for an effective ciliary stroke in Paramecium

Abstract

Centrioles, cilia, and flagella are ancestral conserved organelles of eukaryotic cells. Among the proteins identified in the proteomics of ciliary proteins in Paramecium, we focus here on a protein, Bug22p, previously detected by cilia and basal-body high-throughput studies but never analyzed per se. Remarkably, this protein is also present in plants, which lack centrioles and cilia. Bug22p sequence alignments revealed consensus positions that distinguish species with centrioles/cilia from plants. In Paramecium, antibody and green fluorescent protein (GFP) fusion labeling localized Bug22p in basal bodies and cilia, and electron microscopy immunolabeling refined the localization to the terminal plate of the basal bodies, the transition zone, and spots along the axoneme, preferentially between the membrane and the microtubules. RNA interference (RNAi) depletion of Bug22p provoked a strong decrease in swimming speed, followed by cell death after a few days. High-speed video microscopy and morphological analysis of Bug22p-depleted cells showed that the protein plays an important role in the efficiency of ciliary movement by participating in the stroke shape and rigidity of cilia. The defects in cell swimming and growth provoked by RNAi can be complemented by expression of human Bug22p. This is the first reported case of complementation by a human gene in a ciliate.

Fig. 1.

Sequence conservation of Bug22p in eukaryotes and positions that distinguish ciliary versus nonciliary species. The ClustalW alignment of 58 Bug22p sequences, representative of all the clades of eukaryotic species in which it can be found, was visualized using Boxshade and illustrates the strong conservation of the Bug22p protein through evolution. Careful examination of the divergences at seven specific positions (amino acids 30, 38, 45, 55, 66, 67, and 111 according to the numbering of the Homo sapiens Bug22p sequence) revealed that species having cilia or centrioles in at least some stage of their life cycle have a consensus at these positions (denoted in purple) that dramatically differs from the consensus found in plants (denoted in green), as displayed at the bottom of the figure. Note that both the moss Physcomitrella patens and the fern Selaginella moellendorfii (in the middle of the alignment), which both have several Bug22 genes in their genomes and are plants with occurrence of flagellated cells in their life cycles, display a patchwork of the plant consensus and the ciliary consensus at these positions. In addition, the occurrence of a BUG22 gene with ciliary consensus in the genome of Chlorella vulgaris, harboring a centriole that never develops a cilium (39), suggests that Bug22p has a centriolar function. The species studied and the accession numbers of the proteins are as follows: Aa, Aedes Aegypti XP_001663334.1; At, Arabidopsis thaliana AAG51069.1; Am, Apis mellifera XP_395824.1; Bd, Batrachochytrium dendrobaditis jgi ∣Batde5∣ 32931; Bm, Bombyx mori NP_001040392.1; Brm, Brugia malayi XP_001894987.1; Bt, Bos taurus NP_032213.2; Cca, Capitella capitata jgi ∣Capca1∣77038; Ccl, Citrus clementina EST:FC926856.1, GI:218791863; Ci, Ciona-intestinalis XP_002131975.1; Ce, Caenorhabditis elegans NP_871706.1; Cv, Chlorella vulgaris jgi ∣Chlvu1∣29644∣e; Cr, Chlamydomonas reinhardtii XP_001692489; Cp, Carica papaya EST: EX271746.1, GI:186814042; Cq, Culex quinquefasciatus XP_001869024.1; Dm, Drosophila melanogaster NP_609402.1; Dr, Danio rerio NP_957105.1; Ee, Euphorbia esula EST: DV114861.1; Eg, Elaeis guineensis ACF06559; Gg, Gallus gallus NP_001005833.1; Gh, Gossypium hirsutum EST: EX168390.1 GI:164311590; Gl, Giardia lamblia XP_001708694.1; Gm, Glycine max gb:ACU16157.1; Hm, Hydra magnipapillata XP_002163101.1; Hs, Homo sapiens NP_037374.1; Lb, Leishmania_braziliensis XP_001564754.1; Lj, Lotus japonicus EST: GO028159.1, GI:223439164; Ls, Lactuca serriola EST:BQ986870.1, GI:22404395; Me, Manihot esculenta EST:DV448997; Mg, Mimulus guttatus EST:GR148812.1, GI:238398755; Mm, Mus musculus NP_032213.2; Mp, Micromonas_pusilla jgi ∣MicpuC2∣70275∣AZW; Mt, Medicago trunculata ACJ85502.1; Ng, Naeglaeria gruberi JGI Protein ID: 81674; Ot_1p, Oxytricha trifallax OXADaaa05e09; Ot_2p, Oxytricha trifallax OXAEaaf21g04; Ot_3p, Oxytricha trifallax OXAEaaa06d08; Pf, Plasmodium falciparum XP_001351642.1; Pp_1, Physcomitrella patens XP_001775608.1; Pp_2, Physcomitrella patens XP_001760380.1; Pp_3, Physcomitrella patens XP_001784194.1; Ps, Picea sitchensis ABK26040.1; Pt, Paramecium teraurelia (this article); Ptr, Populus trichocarpa XP_002304809.1; Sb, Sorghum bicolor XP_002460501.1; Sm_1, Selaginella moellendorfii jgi ∣Selmo1∣172529; Sm_2, Selaginella moellendorfii jgi ∣Selmo1∣96514; Sm_3, Selaginella moellendorfii jgi ∣Selmo1∣130480; Sm_4, Selaginella moellendorfii jgi ∣Selmo1∣117290; Ta, Trichophlax adherens jgi ∣Triad1∣ 23937; Tb, Trypanosoma brucei XP_822473.1; Tc, Tribolium castaneum XP_975505.1; Tg, Toxoplasma gondii XP_002364709.1; Tt, Tetrahymena thermophila XP_001007313.2; Tp, Thalassiosira pseudonana XP_002286290.1; Vv_1p, Vitis vinifera XP_002268386.1; Vv_2p, Vitis vinifera XP_002265223.1; Xp, Xenopus tropicalis NP_001004835.1; Zm, Zea mays NP_001148848.1.

Fig. 2.

Ciliary and basal-body localization of Paramecium and human Bug22p GFP fusions in Paramecium. (A and B) Wild-type paramecia immunofluorescently labeled with the ID5 monoclonal antibody to visualize basal bodies (A) and with the anti-ciliary tubulin to show cilia (B). (C) Paramecium expressing GFP-PtBug22p showing clear labeling of basal bodies and cilia. (D) Same as panel C for a Paramecium expressing GFP-HsBug22p. Bar = 10 μm.

Fig. 3.

Western blot evidence that Bug22p is the antigen of a commercial anti-GTL3 antibody. Lanes: A1, molecular mass markers for lane A2; A2, anti-GTL3 labeling of a 10,000 × g supernatant of whole homogenized paramecia showing a positive band between 20 and 25 kDa; B1 and C1, total extracts of E. coli expressing GFP in the pPXV plasmid normally used for expression in Paramecium but which allows low expression in the bacteria; B2 and C2, E. coli expressing the GFP-Bug22p fusion; C3, molecular mass markers for lanes B1 to C2. The B series lanes were revealed with the anti-GTL3 antibody and the C series with the anti-GFP antibody. The asterisk indicates the level of the fusion protein that is recognized by both GTL3 and GFP antibodies.

Fig. 4.

Localization of Bug22p in Paramecium by postembedding immunoelectron microscopy. (A to F) Labeling of wild-type cells with the GTL3 antibody using 5-nm gold particles. (G to N) Labeling of GFP-Bug22p-expressing cells with an anti-GFP antibody using 10-nm gold particles. The arrowheads point to gold particles. (A and H) Labeling of the terminal plates of basal bodies. (B, D, and I) Labeling of the transition zone to the cilium. (F and J to N) Labeling in the vicinity of the outer doublets of the axoneme. (C, E, and J) Labeling between the axoneme and the membrane. (G) Labeling within the basal body close to its proximal part. (J and K) Examples of regular disposition of the gold particles along the axoneme. ci, cilium; bb, basal body; tp, terminal plate; tz, transition zone. Bar = 250 nm.

Fig. 5.

Swimming-speed diminution in Bug22p-depleted cells. Shown is a track recording of paramecia swimming under a dark-field microscope equipped with a 10× objective during a 650-ms pause. (A) Control Nd7p-depleted cells. (B) Double RNAi targeting all four BUG22 genes. (C and D) Simple RNAi targeting BUG22a and BUG22b, and BUG22c and BUG22d, respectively. Bar = 100 μm. (E) Histograms of measured speeds under all conditions showing quantization of the cumulative effects of double RNAi compared to simple RNAi. The error bars indicate standard errors.

Fig. 6.

Complementation of the effects of Bug22p depletion by expression of the human gene in Paramecium. To test possible rescue of Bug22p-depleted paramecia by expression of human Bug22p, cells were first transformed by a plasmid containing the human BUG22 cDNA under Paramecium regulatory sequences and then subjected to RNAi against the BUG22 endogene (PtBUG22), the human BUG22 transgene (HsBUG22), or both (HsBUG22 and PtBUG22) to follow the phenotypes. Transformation of Paramecium nd7-1 cells (unable to undergo trichocyst exocytosis) was performed by microinjection of a mixture of two plasmids, one driving the expression of the wild-type ND7 gene to monitor the efficiency of the transformation and the other for expression of the human BUG22 cDNA. The clones derived from transformed cells were cultured and submitted to RNAi, with the ND7 gene used as a target for control RNAi. The phenotypic analysis consisted of monitoring the growth rate by cell counting and the swimming speed by automated analysis of recorded tracks. Each measurement was performed on accumulations of tracks of ca. 20 cells during 1-min acquisitions and was repeated 3 times on independent samples. Shown are the swimming speed (histograms and standard errors) and growth rate (values below the histograms) of an HsBug22 transformant (B) compared to a control clone (A). This transformant appeared to be resistant to RNAi targeting the endogenous Paramecium BUG22 genes (PtBUG22 RNAi) compared to the control clone under the same conditions, as shown by the heights of the bars labeled PtBUG22 RNAi in both histograms, and to be resistant to the human BUG22 transgene (HsBUG22 RNAi), as was the control clone, but not to double RNAi (Hs/PtBUG22 RNAi). This indicates that expression of the human cDNA is able to replace the depleted Paramecium protein, thus showing complementation ability, but does not represent a nonspecific effect, since the double RNAi inactivates any BUG22 function in these cells. Altogether, six independently transformed clones have been extensively analyzed with similar results. Some variation was observed in the levels of the complementation: PtBUG22- depleted cells swam at 10% of the normal speed, while cells rescued by transformation swam at 35% to 90% of the normal speed. Such variation is routinely observed in transformation of Paramecium, depending on the amount of transforming DNA introduced via microinjection. In three cases where the amount of injected DNA was optimal, the rescue was abolished when both Paramecium and human sequences were targeted together by RNAi.

Fig. 7.

Effects of Bug22p depletion on ciliary beating viewed by high-speed video microscopy. (A and B) Images of control Nd7p- and Bug22p-depleted cells, respectively, captured from high-speed video microscopy movies (see Videos S1 and S2 in the supplemental material) showing the ciliary beating, on which are superimposed the representation of a single cilium (in red) plus the representation of the same cilium in the 8 adjacent frames (blue), representing one power stroke in the control (25 ms) and erratic beating in the Bug22p-depleted cell. (C to F) Quantization of four parameters manually extracted from the movies on 14 beating cilia for BUG22 RNAi and 12 cilia for the control, with standard deviations. (C) Frequency of beating. (D) Angle between extreme positions taken by beating cilia. (E) Length of the trajectory of the tips of beating cilia. (F) Triangular surface swept by beating cilia. Bug22p depletion increased the frequency but decreased all the other parameters.

Fig. 8.

Ciliary defects in Bug22p-depleted paramecia observed in electron microscopy. (A to C) Observation of RNAi after 24 h. Magnification, ×22,000. (D and E) Observations of RNAi after 48 h. Magnification, ×16,000. (A and D) Control ND7 RNAi. (B, C, and D) BUG22 RNAi. Note the unusual bending of cilia in Bug22p-depleted cells, as viewed in the longitudinal axis of cilia in panels B and E and with contiguous longitudinal and cross sections of the same cilium in panel C. Images of exaggerated ciliary bending were never observed earlier and were detected for the first time in Bug22p-depleted cells. Among 140 longitudinal sections of cilia from 15 different cells, 43 cilia displayed such exaggerated bending, whereas none were observed in a similar number of sections in the control.