We report here the complete sequence of the γ dynein heavy chain of the outer arm of the Chlamydomonas flagellum, and partial sequences for six other dynein heavy chains. The γ dynein heavy chain sequence contains four P-loop motifs, one of which is the likely hydrolytic site based on its position relative to a previously mapped epitope. Comparison with available cytoplasmic and flagellar dynein heavy chain sequences reveals regions that are highly conserved in all dynein heavy chains sequenced to date, regions that are conserved only among axonemal dynein heavy chains, and regions that are unique to individual dynein heavy chains. The presumed hydrolytic site is absolutely conserved among dyneins, two other P loops are highly conserved among cytoplasmic dynein heavy chains but not in axonemal dynein heavy chains, and the fourth P loop is invariant in axonemal dynein heavy chains but not in cytoplasmic dynein. One region that is very highly conserved in all dynein heavy chains is similar to a portion of the ATP-sensitive microtubule-binding domain of kinesin. Two other regions present in all dynein heavy chains are predicted to have high α-helical content and have a high probability of forming coiled-coil structures. Overall, the central one-third of the γ dynein heavy chain is most conserved whereas the N-terminal one-third is least conserved; the fact that the latter region is divergent between the cytoplasmic dynein heavy chain and two different axonemal dynein heavy chains suggests that it is involved in chain-specific functions.

Dyneins are molecular motors involved in many different types of microtubule-based intracellular movements, including flagellar movement, organelle transport, mitotic spindle placement, and possibly certain types of chromosomal movement during cell division. The best characterized dyneins are the outer arms of flagella, which consist of two or three different heavy chains (depending on species) together with several intermediate and light chains (Witman, 1992; Witman et al., 1994), and cytoplasmic dynein, which consists of two copies of a single heavy chain plus intermediate and light chains (Paschal et al., 1987). In both cases the heavy chains contain the ATP-hydrolytic site(s) and are responsible for force production. However, axonemal dyneins differ from cytoplasmic dyneins in a number of important characteristics (Shpetner et al., 1988). Moreover, recent cell biological and genetic studies indicate that the different heavy chains of outer-arm dynein have different functions (Moss et al., 1992a,b; Sakakibara et al., 1991, 1993).

An understanding of the underlying bases for the various activities of the different dynein heavy chains (DHCs) will require detailed structural information on the individual chains. To date, full sequences are available only for the β DHC from sea urchin (Gibbons et al., 1991; Ogawa, 1991), and for the cytoplasmic DHC from Dictyostelium (Koonce et al., 1992) and rat (Mikami et al., 1993). The midregions of both chains contain four glycine-rich domains, which conform more or less to the P-loop consensus sequence GXXXXGKS/T that has been shown to be involved in nucleotide binding in many.proteins (Walker et al., 1982; Saraste et al., 1990). In the outer-arm β DHC, one of the P loops appears to correspond to the V1 site, which is cleaved by UV light in the presence of ATP and vanadate; this site is presumed to be the ATP-hydrolytic site (Lee-Eiford et al., 1986). This P loop is highly conserved between the outer-arm β and cytoplasmic DHCs. The other P loops are less highly conserved and their function is unknown. It has been suggested that one may correspond to the β DHC’s single V2 site, which is cleaved by UV light in the presence of oligomeric vanadate and the absence of ATP (Tang and Gibbons, 1987). In addition, the β DHC contains a fifth P-loop motif near its N terminus, which is not present in the cytoplasmic DHC. Sequence information for additional flagellar DHCs will be necessary to determine if any of these sites is highly conserved among the axonemal dyneins and therefore likely to be particularly important in the generation of flagellar movement.

There are also regions of overall structural similarity and divergence between the outer-arm β and cytoplasmic DHCs. There are two short regions of high α-helical content in the C-terminal third of both chains, which may be important in the formation of tertiary or quaternary structure (Koonce et al., 1992; Mikami et al., 1993). In contrast, the N-terminal one-third of the chains exhibit considerable sequence divergence, leading to speculation that this region may be important in the performance of axonemal vs cytoplasmic dynein functions (Koonce et al., 1992; Mikami et al., 1993). Again, information on additional axonemal DHCs will be important for determining which regions are conserved among all dyneins, which are conserved only between axonemal dyneins, and which are specific to individual DHCs.

Ultimately, identification of functional domains in DHCs will benefit from in vitro mutagenesis of sites of interest, followed by introduction of the altered gene into a cell having a null mutation for that gene. For cytoplasmic dynein, such experiments could be carried out most readily in yeast as long as the null mutation is not lethal. For flagellar dynein, the organism most amenable to such an approach is the flagellated green alga Chlamydomonas. Thirteen different loci have been identified that affect the structure or assembly of the outer arm of the Chlamydomonas flagellum (Kamiya, 1988; Sakakibara et al., 1991). Several of these loci encode structural components of the outer arm, including the α and β DHCs (Sakakibara et al., 1991, 1993). A physical map of the Chlamydomonas genome has been developed, allowing the tentative identification of cloned genes with specific mutations by RFLP mapping (Ranum et al., 1988). Chlamydomonas is efficiently transformed (Kindle et al., 1989), so cloned genes can be unequivocally identified with loci by complementation of mutants (Diener et al., 1990; Mitchell and Kang, 1991), and genes carrying specific mutations could be introduced into the nuclear genome and expressed in vivo. The outer-arm dynein of Chlamydomonas can be readily isolated and its subunits separated from each other (Pfister et al., 1982; Pfister and Witman, 1984), so the properties of individual dynein subunits having specific defects can be examined in vitro.

Here we present the complete sequence of the γ DHC of Chlamydomonas. Our approach allowed unequivocal identification of the P loop corresponding to the V1 site of this chain. A comparison of the sequence with other available DHC sequences confirms that this P loop is highly conserved among all dyneins. Two other P loops are conserved among cytoplasmic DHCs but not axonemal DHCs, whereas the fourth P loop is invariant among axonemal but not cytoplasmic DHCs. The comparison also reveals a region that is highly conserved among all cloned dyneins and has structural similarity to a portion of the ATP-dependent microtubule-binding domain of kinesin. Finally, the N-terminal one-third of the γ DHC is divergent from that of the sea urchin β and cytoplasmic DHCs, suggesting that this region carries out chain-specific functions rather than cytoplasmic vs axonemal functions. These results provide the basic structural information necessary for functional analysis of these regions using molecular genetic techniques.

cDNA cloning

RNA was isolated from vegetative Chlamydomonas (strain 1132D−) 30 minutes after pH-induced deflagellation (Witman et al., 1972) as follows: 10 ml of packed cells (concentrated by centrifugation at 10,000 g for 5 minutes) were vortexed to make a paste and then added slowly and with rapid stirring to 100 ml of 5% SDS, 20 mM Tris-Cl, pH 8.0, 20 mM EDTA and 1 mg/ml proteinase K at 50°C. The resulting mixture was incubated at 50°C for 4 hours. One-fifth volume of 7.5 M ammonium acetate, pH 7.5, was added and the solution was extracted with an equal volume of phenol/chloroform (50:50). An equal volume of isopropanol was added to precipitate the RNA. RNA was separated from DNA by LiCl precipitation. Poly-A+ RNA was prepared by oligo-dT chromatography. To enrich for large mRNAs such as would encode DHCs, 100 μg of the poly-A+ RNA was size-fractionated on a methyl mercury hydroxide sucrose gradient (Sambrook et al., 1989). Fractions of this gradient were used to prepare a northern blot, which was subsequently probed with a γ DHC genomic clone. Fractions containing γ DHC sequences with sizes in the 12-14 kb range were pooled and ethanol precipitated. Double-stranded cDNA was prepared using a random 9-mer as the primer (Gubler and Hoffman, 1984). Synthetic EcoRI adapters were added and the cDNA was ligated to EcoRI and alkaline phosphatase-digested λgt10 DNA. The ligation mixture was packaged into λ phage particles using Stratagene’s Gigapack Gold in vitro packaging system (Stratagene, La Jolla, CA).

cDNA clones were isolated by screening with a genomic clone picked from a genomic expression library using the monoclonal antibody 12γB. The first cDNA clone isolated, pcγ1, was then used to rescreen the size-selected cDNA library. Phage isolated in this manner vere subcloned into pBluescript SK+. Primers from the T7 and T3 RNA polymerase promoter regions of this plasmid were used to obtain sequence data from the ends of these clones. From the size of a cDNA clone and the sequence of its ends the amount of overlap was calculated. From each round, clones that contained the most new sequence were used to rescreen the cDNA library. This was repeated until the entire coding region of the γ DHC mRNA was obtained. A combination of ExoIII-generated nested deletions and synthetic oligonucleotide primers was used to obtain the sequence of both strands of the clones shown in Fig. 3. The sequencing was accomplished using the 7-deaza dGTP Sequenase Kit from USB (United States Biochemical, Cleveland, OH).

Peptide sequence

The γ DHC was obtained from a high-salt extract of Chlamydomonas axonemes and purified by sucrose density gradient centrifugation (King et al., 1986) followed by ion exchange chromatography on a MonoQ column (Pharmacia, Piscataway, NJ) (Goodenough et al., 1987). The sample was reduced, alkylated and subsequently cleaved with cyanogen bromide in 70% formic acid. Peptides were separated by reverse-phase chromatography on a Vydac C18 column. The peptides were sequenced using an Applied Biosystems 477A protein sequencer.

Northern and Southern blots

For northern blots, 10 μg of poly-A+ RNA were separated on formaldehyde-containing agarose gels and blotted to nylon membranes. Blots were probed at 65°C overnight with 32P-labeled DNA in 7% SDS, 0.25 M disodium phosphate, pH 7.2, and 1 mM EDTA. The blots were washed to a final stringency of 0.2× SSC and 0.1% SDS at 70°C. For Southern blots, 5 μg of genomic DNA were digested overnight with the appropriate restriction enzyme and electrophoresed on a 0.8% agarose gel in TBE buffer. Subsequent manipulations were as described for northern blots.

PCR cloning

The following primers were used in obtaining new DHC clones directly from the size-selected cDNA library:

  1. CGCGAATTC(CG)GC(CTG)GG(CT)AC(CT)GG(CT)AA, derived from the protein sequence PAGTGK;

  2. GCGCGAATTCTGCTT(CT)GA(CT)GAGTT(CT)AACCG derived from the protein sequence CFDEFNR; and

  3. GCGCCTCGAGCC(GCA)GGGTTCAT(GA)AA derived from the protein sequence ITMNPG.

A 1 μl sample of a phage lysate from the size-selected cDNA library was used in a 100 μl reaction mixture containing 20 mM Tris-Cl, pH 8.8, 10 mM KCl, 6 mM (NH4)2SO4, 1.5 mM MgCl2, and 0.1% Triton X-100. The reaction mixture was boiled for 5 minutes and quenched on ice. The tube containing the reaction mixture was placed in a Coy thermal cycler (Coy Laboratory Products, Inc., Ann Arbor, MI) that was at 96°C. Four units of Pfu DNA polymerase were added and 30 cycles of a three-step program (1 minute at 96°C, 1 minute at 30°C, 1 minute at 75°C) were executed. A final cycle of 10 minutes at 75°C was performed to ensure that most of the product was double-stranded. The PCR products were extracted with phenol/chloroform (50:50), ethanol precipitated, digested with EcoRI and XhoI, and ligated to pBluescript SK+.

Computational methods

The GCG suite of programs (Devereux et al., 1984) was used for sequence assembly, dot plot comparisons and protein structure predictions. The ALIMAT and FILTER programs (Vingron and Argos, 1991) were used to identify regions of similarity between dyneins and kinesins. The PHYLIP suite of programs (Felsenstein, 1989) were used to generate parsimony trees. PHD (Rost and Sander, 1992) was used to predict secondary structure of the potential microtubulebinding domains. The program NEWCOILS (Lupas et al., 1991) was used to predict regions of coiled-coil structure.

In previous studies of the γ DHC from Chlamydomonas, we had mapped the location of its V1 cleavage site, which is believed to correspond to the chain’s functional ATP-hydrolytic site, as well as the locations of two V2 cleavage sites, which are of unknown function but possibly related to ATP binding (Fig. 1) (King and Witman, 1988). Because we ultimately wished to unambiguously identify the V1 cleavage site within the dynein sequence, we used monoclonal antibody 12γB, which binds within 16 kDa of the V1 cleavage site (Fig. 1) (King and Witman, 1988), to screen a genomic expression library constructed in λgt11. A clone (p12γB3) producing an immunoreactive fusion protein was obtained and used to select several large cDNA clones from a λgt10 library prepared from size-selected mRNA isolated from cells that were regenerating flagella. One of the clones, pcγ1 (corresponding to amino acid residues 1659-2132 in the full sequence), hybridized to an ∼13 kb mRNA, which is of an appropriate size to encode a DHC (Fig. 2A). Southern blotting of genomic DNA cleaved with PvuII, PstI, NotI and HindIII revealed that the sequence of this clone is present once in the Chlamydomonas genome (Fig. 2B). Clone pcγ1 was used to select overlapping cDNA clones until the complete coding sequence was obtained. The relationship of these clones is shown in Fig. 3. Each of these overlapping clones was sequenced and the derived protein sequence is shown in Fig. 4. The deduced sequence contains 4486 amino acids and predicts a polypeptide mass of 513 kDa. The N terminus was identified as the first methionine following a region with stop codons in all three reading frames; the C terminus corresponds to the next stop codon, beyond which there are no long open reading frames.

Fig. 1.

Linear map of the γ DHC. V1, V2a, V2b, sites cleaved by UV light in the presence of vanadate; Box with B, region containing the epitope recognized by monoclonal antibody 12γB. The scale is in kDa and is based on SDS-PAGE analysis (King and Witman, 1988).

Fig. 1.

Linear map of the γ DHC. V1, V2a, V2b, sites cleaved by UV light in the presence of vanadate; Box with B, region containing the epitope recognized by monoclonal antibody 12γB. The scale is in kDa and is based on SDS-PAGE analysis (King and Witman, 1988).

Fig. 2.

(a) Northern blot prepared with 10 μg of poly-A+ RNA isolated from Chlamydomonas cells that were regenerating their flagella. The probe pcγ1 hybridized with a single band of ∼13.5 kb. (b) Southern blot probed with pcγ1. Single bands were detected in Chlamydomonas DNA (5 μg) cleaved with, in order left to right, PvuII, PstI, NotI and HindIII. Size standards (kb) are shown to the right of each blot.

Fig. 2.

(a) Northern blot prepared with 10 μg of poly-A+ RNA isolated from Chlamydomonas cells that were regenerating their flagella. The probe pcγ1 hybridized with a single band of ∼13.5 kb. (b) Southern blot probed with pcγ1. Single bands were detected in Chlamydomonas DNA (5 μg) cleaved with, in order left to right, PvuII, PstI, NotI and HindIII. Size standards (kb) are shown to the right of each blot.

Fig. 3.

Relationship of cDNA clones used to generate the sequence of the γ DHC. The numbers represent the positions of the end of each clone in the full nucleotide sequence.

Fig. 3.

Relationship of cDNA clones used to generate the sequence of the γ DHC. The numbers represent the positions of the end of each clone in the full nucleotide sequence.

Fig. 4.

Deduced amino acid sequence of the γ DHC. The seven amino acid sequences obtained from cyanogen bromide fragments of the γ DHC are indicated by single underlines (the underline from residues 3163 to 3199 represents two cyanogen bromide fragments). The region that includes the epitope for monoclonal antibody 12γB is indicated by filled circles. The four P-loop motifs are indicated by asterisks. The potential microtubule-binding region is indicated by the double underline.

Fig. 4.

Deduced amino acid sequence of the γ DHC. The seven amino acid sequences obtained from cyanogen bromide fragments of the γ DHC are indicated by single underlines (the underline from residues 3163 to 3199 represents two cyanogen bromide fragments). The region that includes the epitope for monoclonal antibody 12γB is indicated by filled circles. The four P-loop motifs are indicated by asterisks. The potential microtubule-binding region is indicated by the double underline.

Confirmation that this sequence was that of the outer-arm γ DHC was obtained by comparison with peptide sequences obtained directly from cyanogen bromide fragments of the purified γ DHC. In Fig. 4 the single underlines show seven sequences (two of the sequences are contiguous) that were an exact match for those obtained from direct protein sequencing. These matching sequences were spread throughout virtually the entire γ DHC.

The immunoreactive fusion protein produced by the original genomic clone p12γB3 ended at a stop codon in an intron. The sequence from the λgt11 EcoRI cloning site to the donor splice site for that intron encoded only 24 amino acids, which are shown with filled circles in Fig. 4. Therefore, the epitope recognized by monoclonal antibody 12γB is contained within the region bound by Q1735 and Q1758.

Sixty-one amino acids (∼7 kDa) towards the C terminus from the region containing the 12γB epitope is a P-loop consensus sequence GPAGTGKT. Because our previous mapping data showed that the 12γB epitope was located within 16 kDa of the V1 site, this P loop almost certainly corresponds to the V1 cleavage site and therefore the ATP-hydrolytic site. Three other P-loop consensus sequences are present in the γ DHC, with the next one being 32 kDa towards the C terminus from the first P loop. This second P loop is very near the position mapped for the V2a site, which was estimated to be ∼35 kDa from the V1 cleavage site (King and Witman, 1988). The third and fourth P loops are 69 and 112 kDa from the first P loop. Although biochemical studies indicated that the γ DHC’s V2b site was located ∼70 kDa from the V1 site, the discrepancy between the mass of the chain estimated by SDS-PAGE (415 kDa) and the actual mass calculated from the sequence (a discrepancy noted for all DHCs whose sequences have been determined) precludes unequivocal identification of the third or fourth P loop as the V2b site.

In the initial screen for γ DHC cDNAs, a clone (pcDHC5) was isolated and sequenced and found to be similar to but clearly different from the γ DHC sequence. Subsequently it was learned that this new sequence is identical to that of a portion of the α DHC of the Chlamydomonas outer arm (D. Mitchell, SUNY Health Science Center, Syracuse, NY; personal communication). Comparison of the new sequence (not shown) with that of the γ DHC and the published sequence for the sea urchin β DHC allowed us to identify regions of similarity between these three outer-arm DHCs. The sequence around the first P loop was very highly conserved, as would be expected if this is the ATP-hydrolytic site.

Another very highly conserved region began with the sequence FITMNP (double underline in Fig. 4). These and several adjacent residues are nearly invariant in all DHCs whose sequences have been published (Fig. 5). Visual inspection suggested that this region was similar to a portion of the microtubule-binding domain of kinesin. To test the significance of this similarity between the dyneins and kinesin, three axonemal dyneins were compared with four members of the kinesin superfamily using an objective alignment procedure that relies on dot-plot filtering and is useful for comparing distantly related sequences (Vingron and Argos, 1991). Due to the size of the dynein peptide and the limitations of computer memory only the regions containing the first two P-loop motifs from the DHC sequences were compared. The technique aligned five short regions between the two groups of proteins (Fig. 6A). As expected, the two P loops in the dyneins were aligned with the single P loop in the kinesins. Two additional conserved sequences (FEY and FNC) on either side of the first P loop in the dyneins were aligned with an invariant sequence (FAY) contiguous to the P loop in the kinesin sequences. Since the region of similarity is short and the surrounding regions are not conserved between dyneins and kinesins, these aligments are unlikely to be of functional significance. The fifth alignment was between the conserved dynein sequence FITMNP and a similar sequence within the ATP-sensitive microtubule-binding motif of the kinesin superfamily. Examination of the surrounding region revealed that the invariant glutamic acid and leucine in the kinesin superfamily are also present in the same positions in the axonemal dyneins (Fig. 5), although the glutamic acid is replaced by an asparagine in the Dictyostelium cytoplasmic DHC. Several other residues appear to represent conservative substitutions between the dyneins and members of the kinesin superfamily. A similar secondary structure, consisting of a β sheet, loop and an α helix (Fig. 6B), is predicted for this region for all of the proteins. The loop includes those residues that exhibit the greatest variation between the dyneins and kinesins; this may reflect lack of structural constraints in such regions. When those residues predicted to be part of an α helix are displayed using the GCG program HELICALWHEEL, hydrophilic and hydrophobic residues are clustered on opposite sides of the helix (Fig. 6C,D,E), indicating that the α helix would be amphiphilic (Kaiser and Kézdy, 1984) in both the dyneins and kinesins. This region also has one of the highest surface probabilities in the entire γ DHC (not shown).

Fig. 5.

Alignment of the potential microtubulebinding domains of the Chlamydomonas outer-arm α DHC (αDHC), the Chlamydomonas outer-arm β DHC (βDHC) (Mitchell and Brown, 1993), the sea urchin outer-arm β DHC (βS.U.) (Gibbons et al., 1991; Ogawa, 1991), the Chlamydomonas outerarm γ DHC (γDHC), the Dictyostelium cytoplasmic DHC (cytoDHC) (Koonce et al., 1992), Drosophila kinesin (kinesin) (Yang et al., 1989), Drosophila ncd (ncd) (Endow et al., 1990; McDonald and Goldstein, 1990), Saccharomyces KAR3 (kar3) (Meluh and Rose, 1990), Aspergillus bimC (bimc) (Enos and Morris, 1990), and Schizosaccharomyces cut7 (cut7) (Hagan and Yanagida, 1990). Residues are grouped according to Branden and Tooze (1991). The asterisks indicate amino acids that vary in at most one sequence.

Fig. 5.

Alignment of the potential microtubulebinding domains of the Chlamydomonas outer-arm α DHC (αDHC), the Chlamydomonas outer-arm β DHC (βDHC) (Mitchell and Brown, 1993), the sea urchin outer-arm β DHC (βS.U.) (Gibbons et al., 1991; Ogawa, 1991), the Chlamydomonas outerarm γ DHC (γDHC), the Dictyostelium cytoplasmic DHC (cytoDHC) (Koonce et al., 1992), Drosophila kinesin (kinesin) (Yang et al., 1989), Drosophila ncd (ncd) (Endow et al., 1990; McDonald and Goldstein, 1990), Saccharomyces KAR3 (kar3) (Meluh and Rose, 1990), Aspergillus bimC (bimc) (Enos and Morris, 1990), and Schizosaccharomyces cut7 (cut7) (Hagan and Yanagida, 1990). Residues are grouped according to Branden and Tooze (1991). The asterisks indicate amino acids that vary in at most one sequence.

Fig. 6.

(A) Filtered dot plots of three axonemal DHCs and four members of the kinesin superfamily. The program ALIMAT (Vingron and Argos, 1991) was used to generate pairwise comparisons between amino acids 1660 through 2130 of the γ DHC and comparable regions of the Chlamydomonas α DHC and the sea urchin β DHC. This region included the first and second P-loop motifs of these proteins. Pairwise comparisons were also generated between these partial sequences and the full sequences of Drosophila kinesin, Aspergillus bimC, Saccharomyces KAR3, and Drosophila ncd. Finally, pairwise comparisons were generated among the 4 members of the kinesin superfamily. Comparisons between dynein pairs and between kinesin pairs were performed using a cutoff of six standard deviations; comparisons between a dynein and a kinesin were performed with the cutoff at three standard deviations. The program FILTER was then used to extract the consistently alignable regions of the dot plots. All plots contained the same points after filtration. The plot shown is that for the Chlamydomonas γ DHC and Aspergillus bimC. (B) Secondary structure prediction using the program PHD (Rost and Sander, 1992). The line labeled γ is the amino acid sequence of the potential microtubule-binding domain of the Chlamydomonas outer-arm γ DHC. The lines labeled 1-6 are the secondary structure predictions for the Chlamydomonas γ DHC, the sea urchin β DHC, the Dictyostelium cytoplasmic DHC, Aspergillus bimC, Saccharomyces KAR3, and Drosophila kinesin, respectively. E, residues predicted to be part of a β sheet; H, residues predicted to be part of an α helix; −, residues predicted to be part of a loop region. (C-E) The GCG program HELICALWHEEL was used to display a 10-amino acid segment predicted to be α helical in the putative microtubule-binding domains of the Chlamydomonas γ DHC, Drosophila kinesin, and the Dictyostelium cytoplasmic DHC. The program illustrates how the amino acids would be positioned if they were in an α helix. In each case, the hydrophobic residues (boxed) are clustered on one side of the helix.

Fig. 6.

(A) Filtered dot plots of three axonemal DHCs and four members of the kinesin superfamily. The program ALIMAT (Vingron and Argos, 1991) was used to generate pairwise comparisons between amino acids 1660 through 2130 of the γ DHC and comparable regions of the Chlamydomonas α DHC and the sea urchin β DHC. This region included the first and second P-loop motifs of these proteins. Pairwise comparisons were also generated between these partial sequences and the full sequences of Drosophila kinesin, Aspergillus bimC, Saccharomyces KAR3, and Drosophila ncd. Finally, pairwise comparisons were generated among the 4 members of the kinesin superfamily. Comparisons between dynein pairs and between kinesin pairs were performed using a cutoff of six standard deviations; comparisons between a dynein and a kinesin were performed with the cutoff at three standard deviations. The program FILTER was then used to extract the consistently alignable regions of the dot plots. All plots contained the same points after filtration. The plot shown is that for the Chlamydomonas γ DHC and Aspergillus bimC. (B) Secondary structure prediction using the program PHD (Rost and Sander, 1992). The line labeled γ is the amino acid sequence of the potential microtubule-binding domain of the Chlamydomonas outer-arm γ DHC. The lines labeled 1-6 are the secondary structure predictions for the Chlamydomonas γ DHC, the sea urchin β DHC, the Dictyostelium cytoplasmic DHC, Aspergillus bimC, Saccharomyces KAR3, and Drosophila kinesin, respectively. E, residues predicted to be part of a β sheet; H, residues predicted to be part of an α helix; −, residues predicted to be part of a loop region. (C-E) The GCG program HELICALWHEEL was used to display a 10-amino acid segment predicted to be α helical in the putative microtubule-binding domains of the Chlamydomonas γ DHC, Drosophila kinesin, and the Dictyostelium cytoplasmic DHC. The program illustrates how the amino acids would be positioned if they were in an α helix. In each case, the hydrophobic residues (boxed) are clustered on one side of the helix.

Secondary structure analysis of the entire γ DHC predicts short regions of α helix and β sheet scattered throughout the γ DHC (not shown). The only features that stand out are two regions predicted to be almost exclusively α helix (residues ∼3025-∼3150 and ∼3200-∼3350); these correspond to the similar regions previously noted for both the β DHC of sea urchin and the cytoplasmic DHC (Koonce et al., 1992; Mikami et al., 1993). The NEWCOILS program (Lupas et al., 1991) predicts several short regions of high coiled-coil probability in both the N-terminal and C-terminal thirds of the chain; two of the latter correspond to the regions of high α-helicity noted above. A similar coiled-coil structure is predicted for the equivalent region of the cytoplasmic DHC (Koonce et al., 1992; Mikami et al., 1993) and the β DHC of both sea urchin and Chlamydomonas (Mitchell and Brown, 1993). The fact that this pattern of predicted structure is retained throughout the DHC family suggests that these regions are important in some aspect of dynein assembly or function.

Pairwise comparison between the γ DHC and the sea urchin β DHC is shown in Fig. 7B. The greatest degree of homology is in the middle one-third of the protein, which contains the P-loop motifs and the potential microtubule-binding region. The N-terminal one-third is the least homologous region. Fig. 7A shows a comparison of the Chlamydomonas γ DHC with the rat cytoplasmic DHC. Again, the middle of the chains are most conserved and the N-terminal regions least conserved.

Fig. 7.

Dot-matrix analysis of (A) the Chlamydomonas γ DHC and the Dictyostelium cytoplasmic DHC and (B) the Chlamydomonas γ DHC and the sea urchin β DHC using the GCG program COMPARE with a window size of 50 and a stringency of 30.

Fig. 7.

Dot-matrix analysis of (A) the Chlamydomonas γ DHC and the Dictyostelium cytoplasmic DHC and (B) the Chlamydomonas γ DHC and the sea urchin β DHC using the GCG program COMPARE with a window size of 50 and a stringency of 30.

Using the conserved sequences from the first P loop (PAGTGKT) and the site similar to the microtubule-binding domain of kinesin (ITMNPG), we designed degenerate oligonucleotides, which were used in the polymerase chain reaction to amplify additional DHC sequences from our λgt10 library prepared from size-selected mRNA. Three new sequences - pcr 1, pcr 2, and a sequence identical to that of a portion of the β DHC (Mitchell and Brown, 1993) – were obtained (Fig. 8A). An additional sequence, pcr 3, was obtained from the same library using degenerate oligonucleotides based on the conserved sequences CFDEFNR and ITMNPG. A fifth new sequence, pcr 4, was obtained from a genomic library using the latter primers. Our success in obtaining new DHC sequences indicates that the FITMNP sequence is a common one in DHCs. An alignment of all the Chlamydomonas DHC sequences is shown in Fig. 8A. Because the Pfu polymerase used in the polymerase chain reaction has a 3’ to 5’ exonuclease activity, the primer can be digested and then extended to reveal the sequence of the DNA under the original primer. This has occurred in pcr 1 where Y has replaced F in the FITMNP sequence, and in pcr 4 where V has replaced I in the same sequence. Both of these replacements should preserve the secondary structure and perhaps the function of this region.

Fig. 8.

(A) Alignment of Chlamydomonas dynein heavy chain PCR products. Seven different PCR products have been isolated and are shown aligned by the GCG program BESTFIT. Asterisks indicate amino acids that are invariant in these sequences. Dashed lines represent gaps inserted to align the sequences. (B) Unrooted parsimony tree for two PCR products and the corresponding regions of the Chlamydomonas α, β, and γ DHCs, the sea urchin β DHC and the yeast (Li et al., 1993), rat and Dictyostelium cytoplasmic DHCs. One hundred randomly resampled alignments generated using the bootstrap algorithm of the program SEQBOOT were analyzed using the program PROTPARS. The average of the results was determined with the program CONSENSE. The program DRAWTREE was used to generate the unrooted tree diagram. The number at each node represents the percentage of bootstrapped data sets that produced identical branch nodes.

Fig. 8.

(A) Alignment of Chlamydomonas dynein heavy chain PCR products. Seven different PCR products have been isolated and are shown aligned by the GCG program BESTFIT. Asterisks indicate amino acids that are invariant in these sequences. Dashed lines represent gaps inserted to align the sequences. (B) Unrooted parsimony tree for two PCR products and the corresponding regions of the Chlamydomonas α, β, and γ DHCs, the sea urchin β DHC and the yeast (Li et al., 1993), rat and Dictyostelium cytoplasmic DHCs. One hundred randomly resampled alignments generated using the bootstrap algorithm of the program SEQBOOT were analyzed using the program PROTPARS. The average of the results was determined with the program CONSENSE. The program DRAWTREE was used to generate the unrooted tree diagram. The number at each node represents the percentage of bootstrapped data sets that produced identical branch nodes.

Fig. 8B shows an unrooted parsimony tree of the α, β and γ outer-arm DHCs from Chlamydomonas, pcr 1 and pcr 2 from Chlamydomonas, the β outer-arm DHC from sea urchin, and three cytoplasmic DHCs. The tree is divided into three major branches representing: (1) the yeast cytoplasmic DHC; (2) the rat and Dictyostelium cytoplasmic DHCs; and (3) the outerarm DHCs plus pcr 1 and pcr 2. Within the latter branch, pcr 1 and pcr 2 are most closely associated with each other. The fact that pcr 1 and pcr 2 are grouped with known axonemal DHCs further suggests that they represent inner-arm DHCs. As was expected from their biochemical similarities (Pfister and Witman, 1984), the Chlamydomonas β DHC and the sea urchin β DHC are grouped together.

In this study we have obtained the complete sequence of the γ heavy chain of Chlamydomonas outer-arm dynein, and the partial sequence of the α heavy chain of the same dynein. The γ DHC sequence, like that of the cytoplasmic DHC (Koonce et al., 1992; Mikami et al., 1993) and the β DHC from Chlamydomonas (Mitchell and Brown, 1993), lacks the P-loop motif present near the N terminus of the sea urchin β DHC (Gibbons et al., 1991; Ogawa, 1991). Therefore, this P loop is unlikely to be necessary for generic dynein function. The four remaining P loops (here termed P1-P4) in the midregion of the chain are present in all DHCs sequenced to date (Fig. 9).

Fig. 9.

Sequence alignment of the four P loops of the Chlamydomonas α, β and γ DHCs, the sea urchin β DHC, and the cytoplasmic DHC of rat and Dictyostelium. Asterisks indicate amino acids that are invariant in these sequences. The data for P loops 3 and 4 of the Chlamydomonas α DHC and P loops 1-4 of the Chlamydomonas β DHC are courtesy of Dr D. Mitchell (SUNY Health Science Center, Syracuse, NY).

Fig. 9.

Sequence alignment of the four P loops of the Chlamydomonas α, β and γ DHCs, the sea urchin β DHC, and the cytoplasmic DHC of rat and Dictyostelium. Asterisks indicate amino acids that are invariant in these sequences. The data for P loops 3 and 4 of the Chlamydomonas α DHC and P loops 1-4 of the Chlamydomonas β DHC are courtesy of Dr D. Mitchell (SUNY Health Science Center, Syracuse, NY).

P1 is absolutely conserved in all available DHC sequences. This P loop was previously proposed to correspond to the ATP-hydrolytic site of the sea urchin β DHC because both it and the β DHC’s V1 site were located ∼70 kDa C-terminal to a tryptic cut site (Gibbons et al., 1991; Ogawa, 1991). Our finding that in the γ DHC this P loop and the V1 site are both located within 16 kDa of the epitope recognized by monoclonal antibody 12γB provides even stronger evidence that P1 and the V1 site are identical. This assignment is consistent with detailed biochemical evidence that UV-induced vanadate-dependent photocleavage occurs in the P loops of both myosin and adenylate kinase (Cremo et al., 1989, 1992).

P2 and P3 are highly conserved in the cytoplasmic DHC sequences reported to date. However, neither of these P loops are highly conserved in axonemal DHCs. In P2 of the Chlamydomonas α DHC, an R is substituted for the K in the consensus sequence. Since the K is important for the conformation of the P loop and may interact directly with the β- and γ-phosphates of the bound nucleotide (Saraste et al., 1990), this P loop may not function in ATP hydrolysis. Moreover, the region around this P loop is not particularly well conserved, suggesting that the region does not have a critical function in all DHCs. In P3 of the Chlamydomonas γ DHC, an A is substituted for the second G of the consensus sequence, suggesting that this P loop also may be incapable of hydrolysing ATP.

In contrast to P2 and P3, P4 is absolutely invariant in the axonemal DHCs sequenced to date. Interestingly, this sequence differs from the generally accepted P-loop motif in that a Q has replaced the consensus S or T. Ogawa (1991) noted that the P4 sequence resembles the ATP-binding sequence of some adenylate kinases. However, it does not conform to the consensus sequence (GXPGXGKGT) for that family of proteins (Saraste et al., 1990). Therefore, it is unclear whether this P loop could be involved in ATP binding or hydrolysis. Nevertheless, the fact that P4 is so highly conserved in different axonemal DHCs, and in the same axonemal DHC from different species, indicates that it must have a very important role in outer-arm DHC function. Consistent with this, the region around this P loop is also very highly conserved between axonemal dyneins.

The cytoplasmic DHC differs from the axonemal DHCs in that all four of its P loops conform to the canonical consensus sequence. Moreover, in the cytoplasmic DHC, P1 and P3 are identical between species, with the regions around P1, P3 and P4 being highly conserved. This is in contrast to the situation for the axonemal dyneins, where P1 and P4 are invariant and the regions around those two sites are much more highly conserved than the regions around the other two sites. Analysis of the functions of the individual P loops will be necessary to determine if these differences are responsible for some of the distinctive characteristics of axonemal vs cytoplasmic dynein.

Earlier studies comparing cytoplasmic and axonemal DHC sequences noted that the N-terminal one-third of the cytoplasmic DHC was conserved between species but had little similarity to the N-terminal one-third of the sea urchin β DHC (Koonce et al., 1992; Mikami et al., 1993). It was suggested that this region of the DHC specified functions unique to cytoplasmic vs axonemal dyneins. In light of our finding that the outer-arm β and γ DHCs are also most divergent in this region, it seems likely that the region carries out functions that are chain-specific rather than class-specific. One such function would be chain:chain interaction. The region probably allows each DHC to bind a different set of dynein subunits and to interact with different DHCs. In support of this, in the Chlamydomonas mutant oda4-s7, a truncated β DHC lacking the C-terminal two-thirds of the molecule is still able to bind dynein intermediate and light chains, and to associate with the α and γ DHCs (Sakakibara et al., 1993). Therefore, the N-terminal one-third of the molecule must contain the binding sites for these interactions. Because cytoplasmic dynein contains two identical DHCs that interact with each other in a parallel orientation, it might be expected that the rate of evolutionary change in the region of interaction would be slower than in the equivalent regions of axonemal dynein, where a change in one DHC could be compensated for by a change in another DHC.

We have also identified a region that is very highly conserved among DHCs and that has some similarity to the ATP-dependent microtubule-binding domain of kinesin (Yang et al., 1989; Stewart et al., 1993). Although there are substantial amino acid differences between the dyneins and kinesins in this region, most of the amino acids that are highly conserved in the kinesins are also present in the dyneins, and many of the other amino acids appear to be conservative substitutions. Moreover, the predicted secondary structure of this region - a beta sheet, loop, and amphiphilic α helix - is very similar in both the dyneins and kinesins. Finally, this region is in the same direction and at approximately the same distance from the ATP-hydrolytic site in all DHCs and kinesins examined. Although one cannot conclude from such comparisons alone that this region is involved in ATP-dependent microtubule binding in dynein, the fact that the region is so highly conserved among DHCs indicates that it has an important role in a generic dynein function. It would be of considerable interest to substitute this region of the dynein sequence for the equivalent region of kinesin clones and determine if the bacterially expressed protein is still able to bind and move microtubules (Yang et al., 1990). The kinesin heavy chain generates force in the opposite direction from dynein and ncd, so such chimeric molecules might reveal how the polarity of force generation is determined.

Using the polymerase chain reaction and primers based on highly conserved regions of the known DHCs, we have obtained PCR and subsequently cDNA clones for several additional DHCs. Two of these (pcr 1 and pcr 2) group with known axonemal DHCs in a parsimony tree, suggesting that they represent inner arm dyneins. Recent biochemical studies indicate that the Chlamydomonas axoneme may contain as many as eight different inner arm DHCs (Kagami and Kamiya, 1992), and it has been proposed that some of these inner arm DHCs are differentially distributed along the axoneme (Piperno and Ramanis, 1991). However, the lack of specific antibodies for distinguishing between the different inner-arm dyneins has made both biochemical and localization studies difficult. It now should be possible to generate antibodies to chain-specific peptides deduced from these PCR sequences, and to use these antibodies to determine where the corresponding DHCs are located in the axoneme.

In collaboration with Drs C. Silflow and P. Lefebvre (University of Minnesota), we have used restriction fragment length polymorphism (RFLP) analysis to map the γ DHC gene within the Chlamydomonas genome. Preliminary results indicate that the gene encoding the γ DHC is located on linkage group XI within a few map units of the locus defined by the mutation pf28/oda2, which results in loss of the outer dynein arm (Mitchell and Rosenbaum, 1985; Kamiya, 1988). Therefore, we have tentatively assigned the γ DHC gene to the pf28/oda2 locus. We are currently trying to confirm this assignment by rescuing the pf28/oda2 mutant by transformation with the wild-type γ DHC gene. If pf28/oda2 is a null mutant for the γ DHC, it should be possible to carry out in vitro mutagenesis of the P loops and the potential microtubule-binding domain, and then introduce the altered gene into the mutant by transformation to determine how modification of the sites affects the functioning of the dynein both in vivo and in vitro.

We thank Dr D. Mitchell for sending us his sequence of the Chlamydomonas β DHC prior to publication. This work was supported by NIH grant GM30626 and a grant from the W. M. Keck Foundation for the WFEB Protein Chemistry Facility.

Brandon
,
C.
and
Tooze
,
J.
(
1991
).
Introduction to Protein Structure
.
Garland Publishing, Inc
.,
New York
.
302
pp.
Cremo
,
C. R.
,
Grammar
,
J. C.
and
Yount
,
R. G.
(
1989
).
Direct chemical evidence that serine 180 in the glycine-rich loop of myosin binds to ATP
.
J. Biol. Chem
.
264
,
6608
6611
.
Cremo
,
C. R.
,
Loo
,
J. A.
,
Edmonds
,
C. G.
and
Hatlelid
,
K. M.
(
1992
).
Vanadate catalyzes photocleavage of adenylate kinase at proline-17 in the phosphate-binding loop
.
Biochemistry
31
,
491
497
.
Devereux
,
J.
,
Haeberli
,
P.
and
Smithies
,
O.
(
1984
).
A comprehensive set of sequence analysis programs for the VAX
.
Nucl. Acids Res
.
12
,
387
395
.
Diener
,
D. R.
,
Curry
,
A. M.
,
Johnson
,
K. A.
,
Williams
,
B. D.
,
Lefebvre
,
P. A.
,
Kindle
,
K. L.
and
Rosenbaum
,
J. L.
(
1990
).
Rescue of a paralyzed flagella mutant of Chlamydomonas by transformation
.
Proc. Nat. Acad. Sci. USA
87
,
5739
5743
.
Endow
,
S. A.
,
Henikoff
,
S.
and
Niedziela
,
L. S.
(
1990
).
Mediation of meiotic and early mitotic chromosome segregation in Drosophila by a protein related to kinesin
.
Nature
345
,
81
83
.
Enos
,
A. P.
and
Morris
,
N. R.
(
1990
).
Mutation of a gene that encodes a kinesin-like protein blocks nuclear division in A. nidulans
.
Cell
60
,
1019
1027
.
Felsenstein
,
J.
(
1989
).
PHYLIP -- Phylogeny Inference Package (Version 3.2)
.
Cladistics
5
,
164
166
.
Gibbons
,
I. R.
,
Gibbons
,
B. H.
,
Mocz
,
G.
and
Asai
,
D. J.
(
1991
).
Multiple nucleotide-binding sites in the sequence of dynein β heavy chain
.
Nature
352
,
640
643
.
Goodenough
,
U. W.
,
Gebhart
,
B.
,
Mermall
,
V.
,
Mitchell
,
D. R.
and
Heuser
,
J. E.
(
1987
).
High pressure liquid chromatography fractionation of Chlamydomonas dynein extracts and characterization of inner arm dynein subunits
.
J. Mol. Biol
.
194
,
481
494
.
Gubler
,
U.
and
Hoffman
,
B. J.
(
1984
).
A simple and very efficient method for generating cDNA libraries
.
Gene
25
,
263
269
.
Hagan
,
I.
and
Yanagida
,
M.
(
1990
).
Novel potential mitotic motor protein encoded by the fission yeast cut7+ gene
.
Nature
347
,
563
566
.
Kagami
,
O.
and
Kamiya
,
R.
(
1992
).
The translocation and rotation of microtubules caused by multiple species of Chlamydomonas inner-arm dynein
.
J. Cell Sci
.
103
,
653
664
.
Kaiser
,
E. T.
and
Kézdy
,
F. J.
(
1984
).
Amphiphilic secondary structure: design of peptide hormones
.
Science
223
,
249
255
.
Kamiya
,
R.
(
1988
).
Mutations at twelve independent loci result in absence of outer dynein arms in Chlamydomonas reinhardtii
.
J. Cell Biol
.
107
,
2253
2258
.
Kindle
,
K. L.
,
Schnell
,
R. A.
,
Fernandez
,
E.
and
Lefebvre
,
P. A.
(
1989
).
Stable nuclear transformation of Chlamydomonas using the Chlamydomonas gene for nitrate reductase
.
J. Cell Biol
.
109
,
2589
2601
.
King
,
S. M.
,
Otter
,
T.
and
Witman
,
G. B.
(
1986
).
Purification and characterization of Chlamydomonas flagella dyneins
.
Meth. Enzymol
.
134
,
291
306
.
King
,
S. M.
and
Witman
,
G. B.
(
1988
).
Structure of the γ heavy chain of the outer arm dynein from Chlamydomonas flagella
.
J. Cell Biol
.
107
,
1799
1808
.
Koonce
,
M. P.
,
Grissom
,
P. M.
and
McIntosh
,
J. R.
(
1992
).
Dynein from Dictyostelium: primary structure comparisons between a cytoplasmic motor enzyme and flagellar dynein
.
J. Cell Biol
.
119
,
1597
1604
.
Lee-Eiford
,
A.
,
Ow
,
R. A.
and
Gibbons
,
I. R.
(
1986
).
Specific cleavage of dynein heavy chains by ultraviolet irradiation in the presence of ATP and vanadate
.
J. Biol. Chem
.
261
,
2337
2342
.
Li
,
Y.-Y.
,
Yeh
,
E.
,
Hays
,
T.
and
Bloom
,
K.
(
1993
).
Disruption of mitotic spindle orientation in a yeast dynein mutant
.
Proc. Nat. Acad. Sci. USA
90
,
10096
10100
.
Lupas
,
A.
,
Van Dyke
,
M.
and
Stock
,
J.
(
1991
).
Predicting coiled coils from protein sequences
.
Science
252
,
1162
1164
.
McDonald
,
H. B.
and
Goldstein
,
L. S. B.
(
1990
).
Identification and characterization of a gene encoding a kinesin-like protein in Drosophila
.
Cell
61
,
991
1000
.
Meluh
,
P. M.
and
Rose
,
M. D.
(
1990
).
KAR3, a kinesin-related gene required for yeast nuclear fusion
.
Cell
60
,
1029
1041
.
Mikami
,
A.
,
Paschal
,
B. M.
,
Mazumdar
,
M.
and
Vallee
,
R. B.
(
1993
).
Molecular cloning of the retrograde motor cytoplasmic dynein (MAP1C)
.
Neuron
10
,
787
796
.
Mitchell
,
D. R.
and
Rosenbaum
,
J. L.
(
1985
).
A motile Chlamydomonas flagellar mutant that lacks outer dynein arms
.
J. Cell Biol
.
100
,
1228
1234
.
Mitchell
,
D. R.
and
Kang
,
Y.
(
1991
).
Identification of oda6 as a Chlamydomonas dynein mutant by rescue with the wild-type gene
.
J. Cell Biol
.
113
,
835
842
.
Mitchell
,
D. R.
and
Brown
,
K.
(
1993
).
Sequence analysis of Chlamydomonas dynein heavy chain genes
.
J. Cell Sci
.
107
,
635
644
.
Moss
,
A. G.
,
Gatti
,
J.-L.
and
Witman
,
G. B.
(
1992a
).
The motile β/IC1 subunit of sea urchin sperm outer arm dynein does not form a rigor bond
.
J. Cell Biol
.
118
,
1177
1188
.
Moss
,
A. G.
,
Sale
,
W. S.
,
Fox
,
L. A.
and
Witman
,
G. B.
(
1992b
).
The α subunit of sea urchin outer arm dynein mediates structural and rigor binding to microtubules
.
J. Cell Biol
.
118
,
1189
1200
.
Ogawa
,
K.
(
1991
).
Four ATP-binding sites in the midregion of the β heavy chain of dynein
.
Nature
352
,
643
645
.
Paschal
,
B. M.
,
Shpetner
,
H. S.
and
Vallee
,
R. B.
(
1987
).
MAP1C is a microtubule-activated ATPase which translocates microtubules in vitro and has dynein-like properties
.
J. Cell Biol
.
105
,
1273
1282
.
Pfister
,
K. K.
,
Fay
,
R. B.
and
Witman
,
G. B.
(
1982
).
Purification and polypeptide composition of dynein ATPases from Chlamydomonas reinhardtii
.
Cell Motil
.
2
,
525
547
.
Pfister
,
K. K.
and
Witman
,
G. B.
(
1984
).
Subfractionation of Chlamydomonas 18S dynein into two unique subunits containing ATPase activity
.
J. Biol. Chem
.
259
,
12072
12080
.
Piperno
,
G.
and
Ramanis
,
Z.
(
1991
).
The proximal portion of Chlamydomonas flagella contains a distinct set of inner dynein arms
.
J. Cell Biol
.
112
,
701
709
.
Ranum
,
L. P. W.
,
Thompson
,
M. M.
,
Schloss
,
J. A.
,
Lefebvre
,
P. A.
and
Silflow
,
C. D.
(
1988
).
Mapping flagellar genes in Chlamydomonas using restriction fragment length polymorphisms
.
Genetics
120
,
109
122
.
Rost
,
B.
and
Sander
,
C.
(
1992
).
Jury returns on structure prediction
.
Nature
360
,
540
.
Sakakibara
,
H.
,
Mitchell
,
D. R.
and
Kamiya
,
R.
(
1991
).
A Chlamydomonas outer arm dynein mutant missing the α heavy chain
.
J. Cell Biol
.
113
,
615
622
.
Sakakibara
,
H.
,
Takada
,
S.
,
King
,
S. M.
,
Witman
,
G. B.
and
Kamiya
,
R.
(
1993
).
A Chlamydomonas outer arm dynein mutant with a truncated β heavy chain
.
J. Cell Biol
.
122
,
653
661
.
Sambrook
,
J.
,
Fritsch
,
E. F.
and
Maniatis
,
T.
(
1989
).
Molecular Cloning: a Laboratory Manual
.
Wiley-Interscience
,
New York
.
Saraste
,
M.
,
Sibbald
,
P. R.
and
Wittinghofer
,
A.
(
1990
).
The P-loop -- a common motif in ATP- and GTP-binding proteins
.
Trends Biochem. Sci
.
15
,
430
434
.
Shpetner
,
H. S.
,
Paschal
,
B. M.
and
Vallee
,
R. B.
(
1988
).
Characterization of the microtubule-activated ATPase of brain cytoplasmic dynein (MAP1C)
.
J. Cell Biol
.
107
,
1001
1009
.
Stewart
,
R. J.
,
Thaler
,
J. P.
and
Goldstein
,
L. S. B.
(
1993
).
Direction of microtubule movement is an intrinsic property of the motor domains of kinesin heavy chain and Drosophila ncd protein
.
Proc. Nat. Acad. Sci. USA
90
,
5209
5213
.
Tang
,
W.-J. Y.
and
Gibbons
,
I. R.
(
1987
).
Photosensitized cleavage of dynein heavy chains: cleavage at the V2 site by irradiation at 365 nm in the presence of oligovanadate
.
J. Biol. Chem
.
262
,
17728
17734
.
Vingron
,
M.
and
Argos
,
P.
(
1991
).
Motif recognition and alignment for many sequences by comparison of dot-matrices
.
J. Mol. Biol
.
218
,
33
43
.
Walker
,
J. E.
,
Saraste
,
M.
,
Runswick
,
M. J.
and
Gay
,
N. J.
(
1982
).
Distantly related sequences in the α- and β-subunits of ATP synthase, myosin, kinases and other ATP-requiring enzymes and a common nucleotide binding fold
.
EMBO J
.
1
,
945
951
.
Witman
,
G. B.
,
Carlson
,
K.
,
Berliner
,
J.
and
Rosenbaum
,
J. L.
(
1972
).
Chlamydomonas flagella. 1. Isolation and electrophoretic analysis of microtubules, membranes, matrix and mastigonemes
.
J. Cell Biol
.
54
,
507
539
.
Witman
,
G. B.
(
1992
).
Axonemal dyneins
.
Curr. 0pin. Cell Biol
.
4
,
74
79
.
Witman
,
G. B.
,
Wilkerson
,
C. G.
and
King
,
S. M.
(
1994
).
The biochemistry, genetics and molecular biology of flagellar dynein
.
In Microtubules
. (ed.
J. S.
Hyams
and
C. W.
Lloyd
), pp.
229
249
.
Wiley-Liss, Inc
.,
New York
.
Yang
,
J. T.
,
Laymon
,
R. A.
and
Goldstein
,
L. S. B.
(
1989
).
A three-domain structure of kinesin heavy chain revealed by DNA sequence and microtubule binding analyses
.
Cell
56
,
879
889
.
Yang
,
J. T.
,
Saxton
,
W. M.
,
Stewart
,
R. J.
,
Raff
,
E. C.
and
Goldstein
,
L. S. B.
(
1990
).
Evidence that the head of kinesin is sufficient for force generation and motility in vitro
.
Science
249
,
42
47
.