A microtubule-dynein tethering complex regulates the axonemal inner dynein f (I1).

Motility of cilia/flagella is generated by a coordinated activity of thousands of dyneins. Inner dynein arms (IDAs) are particularly important for the formation of ciliary/flagellar waveforms, but the molecular mechanism of IDA regulation is poorly understood. Here we show using cryoelectron tomography and biochemical analyses of Chlamydomonas flagella that a conserved protein FAP44 forms a complex that tethers IDA f (I1 dynein) head domains to the A-tubule of the axonemal outer doublet microtubule. In wild-type flagella, IDA f showed little nucleotide-dependent movement except for a tilt in the f β head perpendicular to the microtubule-sliding direction. In the absence of the tether complex, however, addition of ATP and vanadate caused a large conformational change in the IDA f head domains, suggesting that the movement of IDA f is mechanically restricted by the tether complex. Motility defects in flagella missing the tether demonstrates the importance of the IDA f-tether interaction in the regulation of ciliary/flagellar beating.

INTRODUCTION

Cilia and flagella are conserved organelles that generate fluid flow by their beating movement (Gibbons, 1981; Hirokawa ). In humans, defects in the assembly and motility of cilia and flagella result in multiple disorders (Pazour and Rosenbaum, 2002; Afzelius, 2004; Satir and Christensen, 2007). Ciliary beating is driven by the dynein motor proteins. The microtubule-sliding activity of dynein requires a series of conserved domains: the ring-shaped “head” consisting of six AAA ATPase units; a coiled-coil “stalk,” which binds to the microtubule in an ATP-dependent manner; and the amino-terminal “tail,” which works as a lever arm to generate the power stroke (King, 2000; Carter ; Cho and Vale, 2012; Kon ; Schmidt ). The axonemal dyneins are divided into two groups, the outer and the inner dynein arms (ODAs and IDAs). The ODAs and IDAs play distinct roles in ciliary movement; the ODAs produce forces for the control of beating frequency (Kamiya and Okamoto, 1985; Mitchell and Rosenbaum, 1985), while the IDAs determine the bend amplitude. There are seven (a–g) major and four minor subspecies of IDAs (Kagami and Kamiya, 1992; Bui ), but the functional differences among these IDA subspecies are poorly understood (Yagi ; Bui ; Kamiya and Yagi, 2014). IDA subspecies f (IDA f  ), also known as I1 dynein, is the only dimeric IDA and it is thought to be involved in the phototactic behavior of Chlamydomonas cells (King and Dutcher, 1997; Okita ). IDA f is also unique in that its two heavy chains (HCs), fα and fβ, interact with a large intermediate chain–light chain (IC-LC) complex composed of three intermediate chains (IC140, IC138, and IC97), five light chains (LC7a, LC7b, LC8, Tctex1, and Tctex2b), and an accessory subunit FAP120 (Piperno ; Smith and Sale, 1991; Porter ; Myster ; Harrison ; Perrone , 2000; Myster ; DiBella , b; Hendrickson ; Ikeda ; Toba ). Among these ICs and LCs, IC138 has been extensively studied (Habermacher and Sale, 1997; King and Dutcher, 1997; Yang ; Bower ; Wirschell ). Phosphorylation of IC138 by casein kinase1 or protein kinase A suppresses the microtubule sliding activity of axonemal dyneins, and dephosphorylation of IC138 by protein phosphatase 1 or 2A restores axonemal motility (Wirschell ; Gokhale ; Elam ). The radial spoke is thought to be involved in this IC138-mediated regulation of the dynein activity because absence of the radial spoke induces hyperphosphorylation of IC138, and an A-kinase anchoring protein is localized at the base of the radial spoke (Gaillard , 2006; Hendrickson ; VanderWaal ). In addition, the modifier of inner arms (MIA) structure, which bridges between the IC-LC complex and the nexin-dynein regulatory complex (N-DRC), suggests a functional communication between IC138 and the N-DRC (Yamamoto ). Despite these studies on the regulatory subunits, the poor in vitro microtubule-sliding activity of IDA f has precluded understanding of its mechanical properties and function in the axoneme (Smith and Sale, 1991; Kagami and Kamiya, 1992; Kotani ; Toba ). As the microtubule-sliding activity of IDA c is suppressed in the presence of IDA f in vitro, it has been proposed that IDA f acts as a drag on microtubule-sliding (Kotani ). A unique role for IDA f is also suggested by its appearance in the cryoelectron tomograms. The fα and fβ heads are located closer to the A-tubule surface of the outer doublet microtubule (DMT) compared with other IDA heads, and the ring of the fα is tilted relative to the longitudinal axis of the DMT (Bui ). Moreover, the fα and fβ heads appear to be anchored to the A-­tubule via a large structure designated the “tether,” which is continuous with a bump, termed “tether head,” on the fα head (Heuser ). These tether structures are conserved in Chlamydomonas flagella, Tetrahymena cilia, sea urchin sperm, and human respiratory cilia (Pigino ; Lin ).

In this report, we investigated the structure and the function of the IDA f using a mutant lacking the tether. Cryoelectron tomography revealed that the IDA f undergoes a large conformational change in the absence of the tether. The structural and functional defects in the tether-less mutant suggest that the tether locks the IDA f heads in a fixed position and converts the motor power into a modulating force for axonemal bending rather than microtubule-sliding motion.

RESULTS

Characterization of fap44 mutant

The large collection of insertional mutants available from the Chlamydomonas Library Project (CLiP) has enabled researchers to easily conduct reverse genetics on the green algae (Li ), which is an important model organism for studying cilia and flagella. In combination with the proteomic analysis of Chlamydomonas flagellar proteins (Pazour ), we searched for conserved axonemal proteins that had not been characterized. Among several candidates, we selected FAP44 (a homologue of mammalian CFAP44/WDR52 [ Tang ]), because it grouped with known IDA proteins in a cluster analysis of proteins in flagellar fractions (Supplemental Figure S1A) (Pazour ). Thus, we obtained a CLiP strain, LMJ.SG0182.019151 (fap44), as a possible IDA-defective mutant.

The fap44 mutant has an insertion of the paromomycin-resistance gene immediately after the start codon (Figure 1A), and the absence of FAP44 protein in the mutant was confirmed by immunoblotting (Figure 1B). The swimming speed is slower than that of wild type (Figure 1C), and the beating motion had a slight reduction in the bend amplitude (Figure 1D). The fap44 cells showed normal phototactic behavior, while the IDA f-missing mutant ida1 showed no phototaxis (Supplemental Figure S1B). Although the motility defect of fap44 is mild compared with ODA- or IDA-deficient mutants (e.g., oda2, ida1, and ida5), the immotile phenotype of the oda2fap44 double mutant indicates a significant defect in the motility-generating machinery of fap44. Similarly, sliding disintegration assays showed a significant reduction in the microtubule-sliding activity in the fap44 axoneme (Figure 1E).

FIGURE 1:

Characterization of the fap44 mutants. (A) Schematic illustrations of the genomic DNA sequences of FAP44, FAP43, and FAP244 genes. The boxes and the lines represent the exons and the introns/untranslated regions, respectively. The arrowheads indicate the position of the paromomycin-resistance gene cassettes inserted for the mutagenesis. (B) Immunoblots showing the absence of FAP44 protein in the fap44 axoneme. The fap43 mutant retained the FAP44 protein. (C) Swimming velocities of the wild-type and the mutant strains. The fap44 cells showed reduced motility. Asterisks indicate statistically significant differences (p < 0.01, Student’s t test). No swimming cells were observed in the fap44oda2 double mutant. Means ± SD for the swimming velocities were calculated from 20 cells. (D) Waveforms of the wild-type and the mutant strains. The fap44 flagella showed a slight reduction in the bend amplitude. (E) Sliding disintegration assay of the axoneme. Microtubule sliding velocities were measured by observing the sliding of protease-treated and ATP-activated axonemes. Means ± SEM for the sliding velocities were calculated from indicated number of sliding events. The fap44 axoneme showed reduced microtubule sliding activity. Asterisks indicate the corresponding probability values calculated by Student’s t test. Although there is no statistically significant difference between the sliding velocities of ida5 and fap44ida5 axonemes, few sliding events were observed in fap44ida5 (thus, n is only 5), suggesting a motility defect in the double mutant.

Cryoelectron tomography of the DMT structure of fap44 showed a large structural defect adjacent to the IDA f (Figure 2), while the central pair structure was normal (Supplemental Figure S1C). The missing structures were previously designated as the tether and the tether head (Figure 2, A–C, red) (Heuser ). Hereafter, we call the tether and the tether head collectively “tether.” Although the IDA f and the tether appeared to be associated with each other, the absence of the tether does not affect the assembly of IDA f and its IC-LC complex on the DMT. As FAP44 and the tether are present in the IDA f-missing mutant (Supplemental Figure S1D) (Heuser ; Yamamoto ), the axonemal assembly of IDA f and the tether are completely independent of each other. Because structural defects in IDA f can result in hyperphosphorylation of IC138 (VanderWaal ), we examined the phosphorylation state of IC138 in the fap44 mutant (Supplemental Figure S1E). However, we did not observe a shift in the mobility of the IC138 band in fap44. Thus, the motility defect in fap44 is not associated with IC138 phosphorylation.

FIGURE 2:

fap44 axoneme lacks the tether. (A) Cross-sectional views of the three-dimensional structure of the axoneme. The upper left inset is the base-to-tip view of the 9+2 structure. The red box indicates the position of the enlarged DMTs on the right. Red: tether; yellow: fα head; orange: fβ head; green: stalks. The directions of the views in B and C are indicated. RS: radial spoke; A and B: A-tubule and B-tubule, respectively. (B, C) Longitudinal views of the IDA f. The tether in red is missing in the fap44 axoneme. Red broken line indicates the shift in the fβ head toward the ODA. IC-LC: the IC-LC complex of the IDA f; a and b: the IDA subspecies a and b, respectively. The DMT structures are oriented with the distal ends on the right.

Nucleotide-dependent conformational change in IDA f

To investigate the reason for the motility defect in the fap44 mutant, we analyzed the nucleotide-dependent conformational change in IDA f (Figure 3 and 4). In ODA and other IDAs, binding of ATP and its hydrolysis cause a release and rebinding of the microtubule-binding domain of the stalk, and a tilt and a shift in the head toward the minus end of the microtubule (Goodenough and Heuser, 1982; Ueno , 2014; Lin ). The tilt and the shift in the head collectively cause an 8-nm-sliding per ATP hydrolysis cycle (Movassagh ; Lin ). In wild-type IDA f, the head and the stalk of fα showed a tilt, but the fβ head and stalk did not show evident shift or tilt (Figures 3A and 4A). The direction of the tilt in the fα head and stalk is perpendicular to microtubule-sliding direction (Figures 3A and 4A), suggesting that the tilt in the fα head and stalk does not contribute to the microtubule sliding along the longitudinal axis (Bui ). It is noteworthy that both the fα and fβ stalks point toward the gap between protofilaments B8 and B9 (Figure 4A, green), while other IDA stalks bind to the B-tubule between protofilaments B9 and B10 (Figure 4C) (Bui ; Lin ).

FIGURE 3:

Nucleotide-dependent conformational change of IDA f in three dimensions. The IDA f structures in the apo and the ATP plus vanadate (ATP+Va) states. (A) Wild type and (B) fap44 mutant. Red: tether; yellow: fα head; orange: fβ head; green: stalks. (A) The head and the stalk of the fβ show a tilt vertical to the long axis. (B) There are three states (I, II, and III) in the conformation of the IDA f in the presence of ATP+Va. Both fα and fβ heads show a large displacement away from the ODA.

FIGURE 4:

Two-dimensional slice representations of the nucleotide-dependent conformational change in IDA f. (A) Wild type and (B) fap44 mutant. Cross-sections are in base-to-tip views. Schematics showing the DMT, heads, and stalks are displayed on the left of the corresponding cross-sections. Yellow: fα head; orange: fβ head; green: stalk. Longitudinal sections, with the distal ends on the right, are placed next to the cross-sections. The orientations of the longitudinal sections are indicated with blue lines. Protofilaments in green indicate the expected position of the microtubule binding domain of the stalk. (C) Head and stalk orientations of the ODA and IDA e are shown in cross- and longitudinal sections.

In the apo structure of the fap44 axoneme, the orientations of the stalk domains were largely the same as in wild type (Figure 4B), although there is a slight shift toward the ODA in the fβ head domain (Figure 2B, red broken line). However, the fap44 mutant showed a large nucleotide-dependent conformational change in IDA f (Figures 3B and 4B). As the averaged IDA f structure in ATP-plus-vanadate form was unclear due to the great degree of structural heterogeneity, we carried out a principal component analysis to visualize the head and stalk domains (Heumann ). There are roughly three structural states in fap44 IDA f with ATP and vanadate; ∼45% of the particles belongs to state I, 30% belongs to state II, and the rest of the particles are in state III, which is similar to the apo state. In both states I and II, the fβ head moves toward the ODA and resides in the position that is occupied by the tether structure in wild-type axoneme (Figures 3B and 4B). The displacement of the fβ head is larger in state II relative to state I. As the fα head lies between fβ and the B-tubule, it is highly unlikely that the fβ stalk in states I and II can reach the B-tubule. Therefore, the nucleotide-dependent conformational change in wild-type IDA f appears to be mechanically restricted by the tether structure.

FAP43 and FAP244 are components of the tether

The large volume of the missing density in the fap44 structure suggested that the tether contains proteins in addition to FAP44. Thus, we conducted quantitative mass spectrometry analyses to compare the protein compositions of wild-type and fap44 axonemes (Table 1). In addition to FAP44, we found that several other proteins, including FAP43, FAP102, FAP244, Cre10.g452250, and a smaller protein, MOT7, are missing or greatly reduced in the fap44 axonemes. MOT7, FAP102, and Cre10.g452250 will not be discussed further here. Interestingly, FAP43, FAP44, and FAP244 are predicted to share the same domain organization: two tryptophan-aspartic acid (WD)-repeat domains in the amino-terminal half and several coiled-coil domains in the carboxy-terminal half (Figure 5A). Therefore, we obtained two more CLiP strains LMJ.SG0182.005221 (fap43) and LMJ.RY0402.124246 (fap244), which have disruptions in the FAP43 and FAP244 genes, respectively (Figure 1A).

TABLE 1:

Known or predicted flagellar proteins missing or greatly reduced in the fap44 axoneme.

	Predicted mass (kDa)	Fold change	Top 3 Precursor Intensity in fap44	Top 3 Precursor Intensity in WT
FAP44	222	0	0	1.79E+09
MOT7	26	0	0	7.81E+08
FAP244	267	0	0	5.53E+08
FAP43	199	0.02	2.11E+07	1.27E+09
Cre10.g452250	68	0.03	2.36E+07	8.87E+08
FAP102	156	0.08	4.05E+07	4.99E+08

“Top 3 Precursor Intensity” is a measure of protein abundance and is a unit-less number (Silva ). Cre10.g452250 is a Phytozome ID in the Chlamydomonas reinhardtii Joint Genome Institute v5.5 genome assembly (Goodstein ); the NCBI accession number for this protein is XP_001698522. No significant mammalian homologues were found for Cre10.g452250 and FAP102.

FIGURE 5:

Functional redundancy between FAP43 and FAP244. (A) Schematic illustrations showing the domain organizations of FAP44, FAP43, and FAP244. DAB: Duffy antigen-binding domain; WD: WD domain (polygons represent the number of WD repeats); CC: coiled-coil domain. “WD?” in FAP244 indicates low confidence in the assignment of the WD domain due to the degeneration of the motif sequences. Blue arrowheads indicate the positions of the BCCP tags. (B) Three-dimensional structure of the DMT. fap43 and fap244 retain the tether structure (red), while fap43fap244 mutant lacks the tether. Expression of BCCP-tagged FAP43 restored the tether in fap43fap244. (C) Immunoblots showing that the fap43fap244 double mutant lacks fap44.

To our surprise, the DMT structures of both fap43 and fap244 axonemes appeared the same as the wild type with normal tether structures (Figure 5B). We examined the interspecies conservation of the two proteins and found that the FAP43 gene is conserved across species including mammals (CFAP43/WDR96), while the FAP244 gene is conserved only among Chlorophyceae. The weak sequence similarity between FAP43 and FAP244 proteins suggested that FAP43 and FAP244 are functionally redundant. To test this hypothesis, we made a double mutant fap43fap244 and found that it is an exact phenocopy of fap44; it lacks the tether structures and FAP44 protein (Figure 5, B and C). We expressed the FAP43 protein tagged with a biotin carboxyl carrier protein (BCCP) at the carboxy-terminus in the fap43fap244 mutant and confirmed the recovery of FAP44 protein (Supplemental Figure S2A, FAP43C) and the tether structure (Figure 5B, fap43fap244:fap43fap244:FAP43-C-BCCP). These results indicate that the tether structure contains FAP44 and FAP43/FAP244 proteins and that FAP43 and FAP244 can compensate for each other.

Interestingly, the immunostaining with an anti-FAP43 antibody revealed that FAP43 is missing in the proximal one-fifth of the wild-type axoneme (Figure 6, right, red arrowheads). In contrast, FAP43 localizes along the whole length of the fap244 axoneme. These results suggest a functional differentiation between the two proteins: FAP43 and FAP244 localize to distal and proximal parts of the axoneme, respectively, but in the absence of FAP244, FAP43 replaces it in the proximal part. Moreover, labeling by the anti-FAP44 antibody staining was diminished at the distal tip of the fap43 flagella (Figure 6, left, white arrowheads). This result suggests that FAP244 may not be able to substitute for FAP43 at the distal tip of the fap43 axoneme and thus fails to recruit FAP44 in this region.

FIGURE 6:

Immunolocalization of FAP44 and FAP43 within the axoneme. The nucleoflagellar apparatus were stained with anti-FAP44 or anti-FAP43 antibody. White arrowheads indicate the distal ends of the flagella. FAP44 staining in the fap43 flagella is diminished near the distal ends. Red arrowheads indicate the proximal regions of the wild-type flagella lack FAP43 staining. In contrast, the whole lengths of the fap244 flagella were stained with anti-FAP43 antibody.

Structural configuration and composition of the tether

To investigate the molecular configuration of the tether structure, we rescued the fap44 and fap43fap244 mutants with constructs designed to express FAP44 and FAP43 fused to the BCCP tag (Oda and Kikkawa, 2013). To avoid disruption of functional domains such as WD domains and coiled-coils (Figure 5A), we inserted the BCCP-tag immediately after Ser43 (FAP44N) or Thr2089 (FAP44C) of FAP44 and at the carboxy-terminus of FAP43 (FAP43C). Ser43 of FAP44 is located within the predicted Duffy antigen-binding domain, but this domain is predicted to be largely disordered (Jones and Cozzetto, 2015). We then isolated the axonemes and labeled them with streptavidin and biotinylated cytochrome c (Oda ; Oda, 2017).

Cryoelectron tomography of the labeled axonemes revealed that FAP44 has an elongated conformation spanning from the fβ head to the RS3 stump (Figure 7, left). The carboxy-terminus of FAP43 also is located near the RS3 stump (Figure 7, right). These results suggest that the amino-terminal domain of FAP44, and probably the amino-terminal domain of FAP43, comprises a large part of the tether structure, while the carboxy-terminal coiled-coils of FAP44 and FAP43 form an anchor to the microtubule. The positions of the carboxy-termini of both FAP44 and FAP43 coincide with the interprotofilament gap, along which the 96-nm ruler complex runs (Oda ). It is possible that the position of the tether complex in the 96-nm repeat is determined by the interaction between the ruler complex and the carboxy-terminal coiled-coils of FAP44 and FAP43.

FIGURE 7:

Three-dimensional localizations of the termini of FAP44 and FAP43. Top, Tip-to-base view; bottom, distal end on the right. Tag densities were visualized by comparing the wild-type and labeled DMT structures. The broken circle indicates the RS stump (Pigino ). Colored densities correspond to positions of the labels indicated by arrowheads of the same color in Figure 5A.

IDA f, FAP44, and FAP43 form a complex

To confirm the interaction between IDA f and the tether complex, we purified IDA f, FAP44, and FAP43 using ion exchange chromatography and performed an immunoprecipitation assay. All three components comigrate on the anion exchange column (Figure 8A), and IDA f and FAP43 were coimmunoprecipitated with BCCP-tagged FAP44 (Figure 8B). These results indicate that FAP44 forms a complex with IDA f and FAP43.

FIGURE 8:

Biochemical characterization of the tether complex. (A) Chromatographic separation of the axonemal salt extracts from FAP44C strain using an UnoQ anion-exchange column. The dynein species of each peak was identified according to the previous study (Furuta ). FAP44 and FAP43 were detected using immunoblotting. Comigration of IDA f with FAP44 and FAP43 was clearly observed. (B) Immunoprecipitation of the FAP44 protein. Fractions containing purified IDA f, FAP44, and FAP43 were collected and BCCP-tagged FAP44 proteins were immunoprecipitated (IP) using streptavidin-agarose. As a negative control (biotin block), the streptavidin-agarose was blocked with 1 mM biocytin (biotin-lysine). IDA f was detected using the IC140 antibody. IDA f and FAP43 were co-immunoprecipitated with FAP44. (C) Sarkosyl fractionation of axonemal proteins. Axonemal proteins were extracted with various concentration of sarkosyl. Supernatant (sup) and precipitate (ppt) after centrifugation were analyzed by immunoblotting. Although the phenotypes of fap43 and fap244 are virtually wild type, extraction patterns showed destabilization of the tether complex.

To characterize the association of FAP44 and FAP43 with the axoneme, we performed fractionations of axonemal proteins using various concentrations of sarkosyl (Figure 8C) (Witman ). We found that some of FAP44 and FAP43 dissociated from the wild-type axoneme even without the detergent treatment, although the majority of the proteins required 0.2% sarkosyl for extraction. Even more of the proteins were extracted without sarkosyl in fap43 and fap244 axonemes, and the majority of the proteins were extracted with 0.1% sarkosyl in these mutants. These results suggest that the tether structures are relatively unstable or susceptible to spontaneous degradation and that the absence of FAP43 or FAP244 further destabilizes the complex even though FAP43 and FAP244 are thought to compensate each other.

DISCUSSION

The IDA f is a mysterious dynein; its in vitro microtubule gliding activity is quite low, and it appears to impede the microtubule sliding of other IDAs, whereas the IDA f-missing mutant exhibits a slow-swimming phenotype and IDA f-missing axonemes show reduced microtubule sliding (Figure 1, D and E) (Smith and Sale, 1991; Kagami and Kamiya, 1992; Kotani ; Toba ). One possible explanation for this discrepancy is that the IDA f acts as a “brake” on microtubule sliding and modulates the bending waves. Another possibility is that the manifestation of the full motor activity of IDA f requires the whole complex, including the tether and the IC-LC complex, to be aligned in the right position and configuration, which is unlikely to happen in the in vitro assay. Our structural results partially support the first hypothesis: in the presence of ATP and vanadate, the fα and fβ heads did not show a significant shift but the fα stalk changed its microtubule-binding site from the B8-B9 to the B9-B10 protofilaments (Figures 4A and 9). This nucleotide-dependent conformational change in IDA f does not seem to contribute to the microtubule sliding in the longitudinal direction, but it is possible that the circumferential tilt of the fα head and stalk modulates the direction of the forces that other dyneins generate. What is the function of the fβ, then? The in situ microtubule sliding activity measured by the axonemal disintegration assay (Toba ) suggests that the fβ mostly accounts for the motor activity of the IDA f. In our structural data, the fβ showed a large conformational change in the absence of the tether structures. This result suggests a “jostling” model that, in the wild-type axoneme, the fβ attempts to exert a power stroke, but the tether structures constrain the fβ to be stationary. The stationary fβ does not necessarily mean that the motor is wasting energy, but it can be involved in other activities, such as applying tension to the IC-LC complex or the MIA structure or direct collision with the adjacent IDA. Similarly, the fα head may be involved in the collision-based motility regulation because the apparent connection between the fα and IDA a was missing in fap44 mutant structure (Figure 2B). This result suggests that the tether mediates a functional interaction between the fα and IDA a. Further structural and biochemical studies are necessary to identify the precise function of the IDA f.

FIGURE 9:

Schematic models of the IDA f-tether complex. Yellow ring: fα head; orange ring: fβ head; green strings: stalks; green sphere: stalk heads (microtubule binding domains of IDA f ); red rings: WD domains of FAP44 and FAP43; red strings: coiled-coil domains of FAP44 and FAP43; gray tubes: protofilamens of the axonemal microtubules. The tether complex is thought to be a heterodimer of FAP44 and FAP43 bundled via their carboxy-terminal coiled-coil domains, which are anchored to the A-tubule or to the 96-nm ruler complex. The tether complex holds both fα and fβ heads, and the nucleotide-dependent movement of the IDA f is restricted to a tilt in the fα head and stalk, which brings the microtubule binding domain to the adjacent protofilament.

Recently, Tang reported that biallelic mutations in the mammalian homologue of FAP44 and FAP43 (CFAP44 and CFAP43, respectively) cause morphological abnormality in human sperm flagella, resulting in male infertility. In Chlamydomonas fap44 flagella, we did not observe flagellar length defect or axonemal disorganization. The severe phenotype in human sperm flagella can be attributed to the function of a CFAP44 splice variant, which is reportedly involved in the regulation of cytoplasmic microtubule dynamics (Tanaka and Tanaka, 2017). The immotile phenotype of fap44oda2 mutant (Figure 1C) may be related to the severe structural defect in mammalian sperms lacking FAP44. It is known that simultaneous loss of multiple axonemal structures, such as ODA, IDA, and radial spokes, causes short and immotile flagella (Piperno ; Kubo ; Lin ). It is possible that the simultaneous loss of the tether and the ODA in fap44oda2 mutant destabilizes the axonemal microtubules. We also found that fap44ida7 (Table 1; ida7 lacks the whole IDA f complex) showed immotile phenotype. These results suggest that the tether structure is important for the stabilization of the axonemal microtubules, especially in mammalian sperm.

The characteristic domain organization of FAP44 and FAP43, the WD domains and the coilded-coils, is similar to the structure of another dynein regulator, Lis1 (McKenney ; Huang ; Rompolas ). Lis1 has a seven-propeller WD domain at its carboxy-terminus and forms a homodimer through the amino-terminal coiled-coil. Lis1 binds to the AAA5 domain of the cytoplasmic dynein head via the WD domain (Toropova ). Binding of Lis1 to the dynein head inhibits the power stroke by mechanically blocks the swing in the linker region of the tail domain. In contrast to Lis1, which binds to the dynein near the root of the stalk, the tether appears to bind to the fα head at the opposite side to the stalk (Figures 2 and 3). This position corresponds to the AAA1 domain, which is the major catalytic domain of the ATP hydrolysis (Gibbons and Gibbons, 1987; Kon ). It is an interesting question whether the tether regulates the motor activity of IDA f by modulating the motion of the tail like Lis1 or altering the ATP hydrolysis in the AAA1. To investigate the molecular mechanism of the tether-mediated regulation, it is necessary to further improve the resolution of the three-dimensional structure of the IDA f-tether complex.

MATERIALS AND METHODS

Strains and reagents

Chlamydomonas reinhardtii cells were grown in Tris-acetate-phosphate (TAP) medium. To screen transformants, cells were grown on TAP agar supplemented with hygromycin B (20 μg/ml; Nacalai Tesque, Kyoto, Japan). The fap44 (LMJ.SG0182.019151), fap43 (LMJ.SG0182.005221), and fap244 (LMJ.RY0402.124246) mutant strains were obtained from the CLiP Library (Li ). The CliP strains were back-crossed with the wild-type strain at least five times before the experiments. Chlamydomonas reinhardtii strains used in this study are listed in Table 2. For generation of anti-FAP44 and FAP43 antibodies, cDNA sequence encoding the amino acids 1-271 of FAP44 and 689-1312 of FAP43 were each inserted into the pGEX-6p-2 plasmid and polypeptides were expressed in Escherichia coli cells. Anti-FAP44 and FAP43 rabbit polyclonal antibodies were then raised against the purified proteins. Anti-DRC3 and anti-IC140 antibodies were generated in the previous studies (Oda , 2015). Anti–human influenza hemagglutinin (HA) antibody 4B2 was purchased from Wako Pure Chemical Industries (Tokyo, Japan). Anti-IC138 antisera was a kind gift from W. S. Sale (Emory University, Atlanta, GA).

TABLE 2:

Strains used in this study.

Strains	Abbreviations	Mutated genes	Missing structures	References
Wild-type CC-125
oda1		DC2	ODA, ODA-DC	Kamiya, 1988
oda2		ODA γ	ODA	Kamiya, 1988; Koutoulis et al., 1997
ida1		DHC1	IDA f	Kamiya et al., 1991
ida4		p28	IDA a, c, d	Kagami and Kamiya, 1992; LeDizet and Piperno, 1995
ida5		actin	IDA a, c, d, e	Kato et al., 1993
ida7		IC140	IDA f	Perrone et al., 1998
pf18			Central pair	Adams et al., 1981
LMJ.SG0182.019151	fap44	FAP44	Tether	CLiP
fap44oda2				This study
fap44ida5				This study
fap44ida7				This study
fap44:FAP44-N-BCCP	FAP44N			This study
fap44:FAP44-C-BCCP	FAP44C			This study
LMJ.SG0182.005221	fap43	FAP43		CLiP
LMJ.RY0402.124246	fap244	FAP244		CLiP
fap43fap44			Tether	This study
fap43fap244:FAP43-C-BCCP	FAP43C			This study

Preparation of axonemes

Chlamydomonas cells were deflagellated with dibucaine-HCl (Nacalai Tesque), and axonemes were collected by centrifugation (Piperno ). Flagella were demembranated with 1% Nonidet P-40 (NP-40) in HMDENa buffer or HMDEK buffer composed of 30 mM HEPES-NaOH, pH 7.2, 5 mM MgCl2, 1 mM dithiothreitol (DTT), 1 mM ethylene glycol tetraacetic acid, 50 mM NaCl (HMDENa), or 50 mM CH3COOK (HMDEK), and 1× Protease inhibitor cocktail (Nacalai Tesque). To observe the phosphorylation state of IC138, cells were deflagellated with acetic acid, and trichloroacetic acid (final 5% vol/vol) was immediately added to fix the phosphorylation state of the axonemal proteins (Wakabayashi and King, 2006). Cells bodies were removed by differential centrifugation, and flagella were pelleted and washed three times with acetone. Dried flagella were resuspended in urea–SDS sample buffer (50 mM Tris-HCl, pH 6.8, 2% SDS, 8 M urea, and 2% 2-­mercaptoethanol) and boiled for 10 min. Axonemal proteins were resolved by SDS–PAGE on 5–15% polyacrylamide gradient gels (Nacalai Tesque) and then blotted onto polyvinylidene difluoride membranes. Blots were probed with the indicated primary antibodies.

Construction of the expression vectors

Fragments spanning from the start codon to immediately before the stop codon for genes encoding FAP44 and FAP43 were amplified with genomic PCR using genomic DNA from the wild-type strain CC-125 and were then inserted into pIC2 plasmids (Oda ). We inserted the tag sequence corresponding to amino acids 141–228 of Chlamydomonas BCCP in the middle of the sequences of FAP44 between Asp42 and Ser43 (FAP44-N-BCCP) or between Ala2099 and Thr2089 (FAP44-C-BCCP) or after the last residue of FAP43, Thr1950 (FAP43-C-BCCP). Expression of the tagged proteins were confirmed by immunoblotting (Supplemental Figure S2A) and immunocytochemistry (Supplemental Figure S2B).

Fluorescence microscopy detection of axonemes

For fluorescence staining using streptavidin, demembranated axonemes were attached to glass slides and blocked with 1 mg/ml bovine serum albumin (BSA) in HMDEK buffer. Axonemes were incubated with 1 μg/ml Alexa Fluor 546-conjugated streptavidin (Invitrogen, Carlsbad, CA) for 1 min. Labeled axonemes were washed three times with HMDEK buffer before observation. For immuno­staining using antibodies, nucleoflagellar apparatuses prepared by the previously described method (Taillon and Jarvik, 1995) were fixed with 2% formaldehyde for 10 min at room temperature, followed by treatment with cold acetone and methanol (–20°C). Fixed samples were immunostained by the previously described method (Sanders and Salisbury, 1995). Stained axonemes were observed using a fluorescence microscope (BX53; Olympus, Tokyo, Japan). Images were recorded using a charge-coupled device (CCD) camera (ORCA-Flash4.0; Hamamatsu PhotonICs, Hamamatsu, Japan).

Measurement of swimming velocity

The swimming velocity of Chlamydomonas cells was recorded using a BX53 microscope (Olympus) at a total magnification of ×100. A red filter with a cutoff wavelength of 630 nm was inserted before the condenser lens to suppress the cellular response to light.

Waveform analysis

Chlamydomonas cells were observed using a dark-field microscope (BX53, Olympus), and images were captured using a high-speed digital camera (EXILIM EX-F1; Casio, Tokyo, Japan) at 600 frames/s. Cells whose flagella were well in focus were selected, and the shapes of flagella were manually traced using Adobe Illustrator.

Sliding disintegration of the axoneme

To measure the sliding velocity, suspension of flagella was sonicated using a Q125 sonicator (Qsonica LLC, Newtown, CT). Sonication was repeated until the flagella were sheared in half lengthwise on average. The flagella were then demembranated with 1% NP-40 in HMDEK buffer and centrifuged. Axonemes were resuspended in HMDEKP (HMDE with 25 mM potassium acetate and 1% polyethylene glycol [20,000 MW]) buffer and absorbed onto a glass slide. The solution was covered with a coverslip with its two sides sealed with strips of double-stick tape. Axonemes were activated with HMDEKP with 1 mM ATP. Sliding disintegration was initiated by washing the flow chamber with HMDEKP with 1 mM ATP and 2 μg/ml nagarase (Sigma-Aldrich). Sliding of DMTs was observed using a dark-field microscope (BX53, Olympus) equipped with a ×50 oil-immersion objective lens and a 100-W mercury lamp. The image sequences were recorded using a CCD camera (M-3204C, Olympus).

Sample preparation for cryoelectron tomography

Streptavidin-cytochrome c labeling of BCCP-tagged axonemes was carried out as described previously (Oda ) with minor modifications. Demembranated axonemes were incubated with 0.05 mg/ml streptavidin for 15 min at 4°C in HMDENa buffer in the presence of 1 mg/ml BSA. Next, axonemes were washed and incubated with 0.05 mg/ml biotinylated cytochrome c. This sequence was repeated, so that the biotinylated Fab fragments were labeled with three rounds of streptavidin and two rounds of cytochrome c. Labeled or unlabeled axonemes were resuspended in HMDENa buffer at a concentration of 0.02 mg/ml and mixed with cationic 15-nm colloidal gold (BBI Solutions, Cardiff, UK). Both side of homemade holey carbon grids were glow-discharged for 20 s. Suspended axonemes plus colloidal gold (5 μl) was loaded onto the grids and plunge-frozen in liquid ethane at –180°C with a Vitrobot Mark IV automated plunge-freezing device (FEI, Hillsboro, OR).

Image acquisition

Grids were transferred to a JEM-3100FEF or JEM-2100F transmission electron microscope (JEOL, Tokyo, Japan) with a Gatan 914 (3100FEF) or 626 (2100F) high-tilt liquid nitrogen cryotransfer holder (Gatan, Pleasanton, CA). Tilt series images were recorded at -180°C using Gatan K2 summit direct detector (3100FEF) or a TemCam-F216 CMOS camera (2100F; TVIPS, Gauting, Germany) and automated acquisition was performed using SerialEM software (3100FEF) (Mastronarde, 2005) or EM-TOOLs program (2100F; TVIPS). The angular range of the tilt series was from -60° to 60° (3100FEF) or -45° to 45° (2100F) with 2.0° increments. The total electron dose was limited to ∼100 e–/Å2. Images were recorded at 300 keV, with 6–9 μm defocus (3100FEF) or 200 keV with 3- to 5-μm defocus (2100F), at a magnification of ×7100 and a pixel size of 7 Å (3100FEF) or ×19,500 and a pixel size of 8 Å (2100F). With JEM-3100FEF, an in-column omega energy filter was used to enhance image contrast in the zero-loss mode with a slit width of 40 eV.

Image processing

Image processing for subtomogram averaging of DMT structures was carried out as described previously (Oda and Kikkawa, 2013; Oda ). Tilt series images were aligned and back-projected to reconstruct three-dimensional tomograms using the IMOD software package (Kremer ). Tomograms of intact axonemes with a high signal-to-noise ratio were selected and used for subtomogram averaging of the 96-nm repeats of DMTs. Alignment and averaging of subtomograms were conducted using custom Ruby-Helix scripts (Metlagel ) and the PEET software suite (Nicastro ). The numbers of DMT subtomograms averaged are summarized in Table 3. The effective resolutions determined by Fourier shell correlation with a cutoff value of 0.5 were ∼4 nm (3100FEF) and ∼6 nm (2100F) (Supplemental Figure S2C). For principal component analysis, a mask volume enclosing two heads of the IDA f was imposed and top three features were selected for clustering into four classes.

TABLE 3:

Summary of averaged subtomograms.

	# of tomograms	# of particles	Microscope+camera
Wild-type apo	10	1460	3100FEF+K2
Wild-type ATP+Va	9	1503	3100FEF+K2
fap44 apo	17	2457	3100FEF+K2
fap44 ATP+Va	12	2250 State I: 765 State II: 499 State III: 436	3100FEF+K2
fap43	8	1216	2100F+F216
fap244	10	1224	2100F+F216
fap43fap244	13	1449	2100F+F216
FAP44N	13	2277	3100FEF+K2
FAP44C	9	1197	2100F+F216
FAP43C	17	2070	2100F+F216
Wild-type central pair	10	688	3100FEF+K2
fap44 central pair	17	966	3100FEF+K2

Surface renderings were generated using UCSF Chimera (Pettersen ). The EM maps of averaged DMT are available at the EM Data Bank (www.emdatabank.org) under the accession numbers EMD-6866-6874.

Statistical analysis

To identify statistically significant differences, we applied Student’s t test to compare wild-type and streptavidin-labeled axonemes as described previously (Oda and Kikkawa, 2013; Oda ). First, wild-type and streptavidin-labeled subtomograms were randomly divided into three data sets. Subtomograms for each data set were aligned and averaged, and a total of six averaged subtomograms were created. We calculated the t value for each voxel and presented it as a single t-value map. The isosurface threshold values were t > 7.17, with a one-tailed probability of <0.1%.

Label-free quantitation by mass spectrometry

Sample preparation.

The proteins in 20 µg of axonemes from wild-type and fap44 cells were separated on 4–15% SDS–polyacrylamide gels. Each lane was cut into five fractions, and each fraction was processed separately as described. Gel fractions were cut into 1 × 1-mm pieces and placed in 1.5-ml Eppendorf tubes with 1 ml of water for 30 min. The water was removed and 200 µl of 250 mM ammonium bicarbonate was added. For reduction 20 µl of a 45 mM solution of DTT was added and the samples were incubated at 50°C for 30 min. The samples were cooled to room temperature and then for alkylation 20 µl of a 100 mM iodoacetamide solution was added and allowed to react for 30 min. The gel slices were washed 2× with 1-ml water aliquots. The water was removed, 1 ml of 50:50 (50 mM ammonium bicarbonate: acetonitrile) was placed in each tube, and samples were incubated at room temperature for 1 h. The solution was then removed and 200 µl of acetonitrile was added to each tube, at which point the gels slices turned opaque white. The acetonitrile was removed, and gel slices were further dried in a Speed Vac (Savant Instruments). Gel slices were rehydrated in 100 µl of 4 ng/µl of sequencing grade trypsin (Sigma) in 0.01% ProteaseMAX Surfactant (Promega): 50 mM ammonium bicarbonate. Additional bicarbonate buffer was added to ensure complete submersion of the gel slices. Samples were incubated at 37°C for 18 h. The supernatant of each sample was then removed and placed in a separate 1.5-ml Eppendorf tube. Gel slices were further extracted with 200 µl of 80:20 (acetonitrile: 1% formic acid). The extracts were combined with the supernatants of each sample. The samples were then completely dried down in a Speed Vac.

LC/MS/MS analysis.

Tryptic peptide digests were reconstituted in 20 µl 5% acetonitrile containing 0.1% (vol/vol) trifluoroacetic acid and separated on a NanoAcquity (Waters) UPLC. In brief, a 4.0-µl injection was loaded in 5% acetonitrile containing 0.1% formic acid at 4.0 µl/min for 4.0 min onto a 100-µm I.D. fused-silica precolumn packed with 2 cm of 5-µm (200 Å) Magic C18AQ (Bruker-Michrom) and eluted using a gradient at 300 nl/min onto a 75-µm I.D. analytical column packed with 25 cm of 3-µm (100 Å) Magic C18AQ particles to a gravity-pulled tip. The solvents were A, water (0.1% formic acid); and B, acetonitrile (0.1% formic acid). A linear gradient was developed from 5% solvent A to 35% solvent B in 60 min. Ions were introduced by positive electrospray ionization via liquid junction into a Q Exactive hybrid mass spectrometer (Thermo). Mass spectra were acquired over m/z 300–1750 at 70,000 resolution (m/z 200) and data-dependent acquisition selected the top 10 most abundant precursor ions for tandem mass spectrometry by high-energy collisional dissociation fragmentation using an isolation width of 1.6 Da, collision energy of 27, and a resolution of 17,500.

Data analysis.

Raw data files were peak processed with Proteome Discoverer (version 1.4, Thermo) prior to database searching with Mascot Server (version 2.4) against the Chlamy Index of NCBInr database. Search parameters included trypsin specificity with two missed cleavages or no enzymatic specificity. The variable modifications of oxidized methionine, pyroglutamic acid for N-terminal glutamine, N-terminal acetylation of the protein, and a fixed modification for carbamidomethyl cysteine were considered. The mass tolerances were 10 ppm for the precursor and 0.05 Da for the fragments. Search results were then loaded into the Scaffold Viewer (Proteome Software) for peptide/protein validation and label-free quantitation.

To identify proteins specifically missing or greatly reduced in fap44 axonemes, all the identified proteins were further filtered to select only proteins with confidence above 95%, with at least five peptides identified and with peptide confidence above 95%. This reduced the total to 625 proteins. These proteins were further filtered to those whose normalized abundances in the mutant were less than 0.2 that of the same proteins in wild type and whose abundances in wild-type axonemes were equal to or greater than ¼ that of FAP44 using Top 3 Precursor Intensity as a quantitative value (Silva ). These criteria reduced the list to 10 proteins, four of which are extracellular matrix proteins. The latter are common contaminants in flagellar preparations, and it is not unexpected that they would be greatly reduced in the fap44 sample since the fap44 mutant originated from a cell-wall–less strain.

Immunoprecipitation of purified FAP44, FAP43, and IDA f

Dyneins were extracted from the axonemes with 0.6 M NaCl in HMDE (HMDENa without NaCl) buffer. The extract was first clarified by centrifugation at 40,000 × g for 20 min, diluted fivefold with HMDE buffer, applied onto a UnoQ anion exchange column (Bio-Rad, Hercules, CA), and eluted with a linear gradient of 130–350 mM NaCl in HMDE buffer. Peak fractions containing IDA f were combined and diluted 2.5-fold (final ∼100 mM NaCl) with HMDE buffer. Streptavidin-agarose (Sigma-Aldrich) preequilibrated with HMDE plus 100 mM NaCl was added to the IDA f-solution and incubated for 2 h at 4°C. The agarose beads were washed three times with HMDE plus 100 mM NaCl, and the bound proteins were released by boiling in the sample buffer. For a negative control, streptavidin-agarose was incubated with 1 mM biocytin (Sigma-Aldrich) for 1 h at 4°C before incubation with the IDA f-solution.

Click here for additional data file.