|Home | About | Journals | Submit | Contact Us | Français|
The highly conserved LC8/DYNLL family proteins were originally identified in axonemal dyneins and subsequently found to function in multiple enzyme systems. Genomic analysis uncovered a third member (LC10) of this protein class in Chlamydomonas. The LC10 protein is extracted from flagellar axonemes with 0.6 M NaCl and cofractionates with the outer dynein arm in sucrose density gradients. Furthermore, LC10 is specifically missing only from axonemes of those strains that fail to assemble outer dynein arms. Previously, the oda12-1 insertional allele was shown to lack the Tctex2-related dynein light chain LC2. The LC10 gene is located ~2 kb from that of LC2 and is also completely missing from this mutant but not from oda12-2, which lacks only the 3′ end of the LC2 gene. Although oda12-1 cells assemble outer arms that lack only LC2 and LC10, this strain exhibits a flagellar beat frequency that is consistently less than that observed for strains that fail to assemble the entire outer arm and docking complex (e.g., oda1). These results support a key regulatory role for the intermediate chain/light chain complex that is an integral and highly conserved feature of all oligomeric dynein motors.
The LC8 protein (now termed DYNLL1 and DYNLL2 in mammals; Pfister et al., 2005 ) was originally identified using biochemical methods as a component of the Chlamydomonas outer dynein arm (Piperno and Luck, 1979 ; Pfister et al., 1982 ). Molecular cloning revealed that it defines a very highly conserved family of proteins sharing ~90% sequence identity that are present in a wide variety of eukaryotes ranging from algae to mammals, including those that lack motile cilia and flagella (King and Patel-King, 1995 ). Subsequent studies demonstrated that LC8/DYNLL proteins are also integral components of, or are associated with, many distinct cellular systems including cytoplasmic dynein (King et al., 1996 ), the dynein that powers retrograde intraflagellar transport (Pazour et al., 1998 ; Rompolas et al., 2007 ), myosin V (Espindola et al., 2000 ), neuronal nitric oxide synthase (Jaffrey and Snyder, 1996 ), flagellar radial spokes (Yang et al., 2001 ), the Bim proapoptotic factor (Puthalakath et al., 1999 ), Rabies virus P protein (Raux et al., 2000 ; Poisson et al., 2001 ), Drosophila swallow (Schnorrer et al., 2000 ), and many others. More distant homologues are present in higher plants (King and Patel-King, 1995 ) and have been found to associate with sireviruses (Havecker et al., 2005 ). LC8/DYNLL proteins form symmetric dimers with two identical grooves into which short segments of the target proteins bind (Liang et al., 1999 ; Fan et al., 2001 ); the monomer-dimer transition can be modulated by altering the pH and by phosphorylation (Liang et al., 1999 ; Nyarko et al., 2005 ; Song et al., 2008 ).
In Drosophila, complete lack of LC8 (ddlc1) function is embryonic lethal due to the induction of apoptosis, whereas partial loss-of-function alleles show pleiotropic defects in bristle and wing development, female sterility, and altered neuronal development (Dick et al., 1996a ; Phillis et al., 1996 ). The Aspergillus LC8/DYNLL orthologue (NudG) is required for nuclear distribution along hyphae presumably due to its role in cytoplasmic dynein activity (Beckwith et al., 1998 ). However, in both Saccharomyces cerevisiae (Dick et al., 1996b ) and Schizosaccharomyces pombe (Miki et al., 2002 ) null alleles have only very minor phenotypes. Similarly, a null mutant for LC8 in Chlamydomonas grows at wild-type rates; however, it does exhibit marked defects in retrograde intraflagellar transport and forms only short flagellar stubs that are deficient for both outer and inner dynein arms, radial spokes, and the projections within the B-tubules of the axonemal outer doublet microtubules (Pazour et al., 1998 ).
In the Chlamydomonas outer dynein arm, LC8 is a component of the intermediate chain/light chain (IC/LC) complex that is located at the base of the soluble dynein particle and is involved in attachment of the motor to its target site within the flagellar axoneme (King et al., 1991 ; Wilkerson et al., 1995 ; and see King and Kamiya, 2008 for recent review). In addition, a LC8-related protein termed LC6 that shares 41% sequence identity is also present (Pfister et al., 1982 ; King and Patel-King, 1995 ); intriguingly, orthologues of LC6 have yet to be identified in other organisms, suggesting that its role in outer arm function may be Chlamydomonas-specific. Lack of LC6 leads to a subtle decrease in flagellar beat frequency but does not affect outer arm dynein assembly within the axonemal superstructure (Pazour and Witman, 2000 ; DiBella et al., 2005 ).
Here we demonstrate that Chlamydomonas expresses a third member of the LC8 family (termed LC10) that is an integral component of the outer dynein arm. We find that LC10 is necessary for wild-type motor function but is not required for dynein assembly. These observations suggest there are fundamental differences in the roles played by the various members of this ubiquitous class of proteins. Proteins closely related to LC10 are found across a broad phylogenetic spectrum in organisms with motile cilia, and we also present evidence to support the assignment of DNAL4 as the mammalian orthologue of Chlamydomonas LC10.
The Chlamydomonas reinhardtii strain 1132D was used as wild type. The mutant strains used were fla14, ida1, ida4, oda1, oda2, oda3, oda4, oda4-s7, oda4-s7 oda11, oda5, oda6, oda6-r75, oda6-r88, oda7, oda8, oda9, oda12-1, oda12-2, oda13, and pf28 pf30 ssh1 (also known as the WS4 isolate). All strains except oda6-r75, oda6-r88, and the oda4-s7 oda11 double mutant may be obtained from the Chlamydomonas Center (http://www.chlamy.org/). Chlamydomonas were grown in R medium with continuous bubbling with 5% CO2 95% air under a 15 h/9 h light/dark cycle (Witman, 1986 ). Beat frequency was determined using the population-based fast Fourier transform method of (Kamiya, 2000 ) as described previously (Wakabayashi and King, 2006 ).
Flagella were detached from Chlamydomonas strains by treatment with dibucaine and were isolated using standard methods (King, 1995 ). Subsequently, the membrane and matrix components were solubilized with 1% IGEPAL CA-630 in 30 mM HEPES, pH 7.4, 5 mM MgSO4, 0.5 mM EDTA, 25 mM KCl, and the resulting axonemes were treated with 0.6 M NaCl to extract the dynein arms. Dyneins were then purified by sedimentation in 5–20% sucrose density gradients using a Beckman SW-55 rotor (Fullerton, CA).
The LC10 coding sequence was first assembled from three expressed sequence tags (ESTs) and a probe encompassing the protein coding sequence obtained from first strand cDNA using the PCR. Subsequently a full-length cDNA was isolated from a λZapII library made from mRNA derived from wild-type Chlamydomonas that were actively regenerating their flagella (Wilkerson et al., 1995 ) using our standard methods (King and Patel-King, 1995 ).
The LC10 coding region was obtained using the PCR and subcloned into the pMAL-c2 vector across the XmnI/XbaI restriction sites. This resulted in the fusion of LC10 to the C-terminus of maltose-binding protein (MBP) via a hydrophilic linker that incorporated a Factor Xa proteolytic cleavage site. The full-length fusion protein was used as the immunogen for antibody production in rabbit CT247, and specific antibody was subsequently obtained from serum by blot purification with recombinant LC10. The MBP-LC6 and MBP-LC8 fusion proteins were described previously (King and Patel-King, 1995 ). Using a similar protocol, we also raised antibody CT254 against a MBP fusion with the murine LC10-related protein DNAL4; the DNAL4 coding sequence was obtained using the PCR from expressed sequence tag BQ895415.
All samples were separated by electrophoresis in 5–15% acrylamide gradient gels and were either stained with Coomassie blue, or blotted to nitrocellulose for immunostaining using standard methods. Nitrocellulose blots were first stained for total protein using 0.05% Reactive Brown 10 (aq.) before incubation with 5% nonfat dry milk 0.05% Tween in Tris-buffered saline. Primary antibody reactivity was detected with a peroxidase-conjugated goat anti-rabbit secondary antibody and enhanced chemiluminescence (ECL; GE Healthcare, Little Chalfont, United Kingdom).
Five hundred-milliliter cultures of Chlamydomonas were grown to late-log phase and harvested by low speed centrifugation to yield 1.5–2 ml of packed cells. Cells were resuspended in 6–10 ml of 10 mM HEPES, pH 7.4, containing protease inhibitor cocktail (P8340; Sigma, St. Louis, MO), 1 mM DTT, and 1 mM phenylmethylsulfonyl fluoride and deflagellated. After three washes with 10 mM HEPES, pH 7.4, cells (~15 ml) were then disrupted by passage three times through a French press to ensure complete lysis. Subsequently, the lysate was subject to ultracentrifugation using a TLA100.2 rotor at 33,000 rpm for 2 h at 4°C. After centrifugation, the lysate supernatant was removed and concentrated to ~400 μl using an Amicon Ultra (10,000 mol. wt. cutoff) ultrafiltration unit (Beverly, MA). The upper 200 μl of the concentrate was layered onto a 5-ml 5–20% sucrose gradient made in 30 mM HEPES, pH 7.4, 5 mM MgSO4, 0.5 mM EDTA, and 25 mM K acetate. The sample was spun for 10 h at 30,000 rpm in a SW55Ti rotor and fractionated. Note that the lower ~200 μl of the concentrate was discarded because it contained a dense pigment that caused the sample to sink into the sucrose gradient.
Sections of murine lung and testis were stained for the presence of DNAL4 and counterstained with hematoxylin by the University of Connecticut Health Center (UCHC) histology core facility using standard methods. Images of murine embryos and brain probed for expression of various dynein components by in situ hybridization were obtained from the BGEM database, which is freely available for use (Magdaleno et al., 2006 ; http://www.stjudebgem.org).
Analysis of the Chlamydomonas genome was performed at http://genome.jgi-psf.org/Chlre3/Chlre3.home.html. Searches of the nonredundant databases were made using BLAST (http://www.ncbi.nlm.nih.gov/blast/Blast.cgi). Protein sequences were aligned using CLUSTALW and the output processed with BOXSHADE. The residues that are not conserved between Chlamydomonas LC8 and LC10 were mapped onto the rat DYNLL1 structure (pdb accession 1F3C) and molecular surface using MOLMOL (Koradi et al., 1996 ).
Analysis of the genomic region surrounding the LC2 gene that is deleted in the oda12-1 mutant revealed a second potential dynein gene (here termed LC10) located within ~2 kb (Figure 1a). The LC10 coding sequence was initially assembled from overlapping ESTs BI722005, BE337027, and BI719072. Subsequently, the entire 736-base pair cDNA was obtained from a λZapII library using a portion of the coding region obtained by PCR from first-strand cDNA as a probe. Several full-length clones were sequenced to confirm the EST assembly (this sequence is available under accession no. EU448319). Analysis of the Chlamydomonas genome (ver. 3.0) and Southern blotting (Figure 1b) demonstrated that there is one LC10 gene in Chlamydomonas. Northern analysis (Figure 1c) revealed a single message of ~1.3 kb that was greatly up-regulated after deflagellation as is characteristic of flagellar proteins.
Toc 2 is a ~1.2-kb repeated sequence in the Chlamydomonas genome that contains imperfect terminal repeats of 14 base pairs and is related to class II transposable elements (Day, 1995 ). One copy of this element, in reverse orientation to the published sequence (X84663), starts at base pairs 670 of the LC10 gene. One of the LC10 cDNA clones we obtained contained this entire element at the 3′ end of the LC10 coding region. However, we presume this transcriptional read-through represents a rare event, as we did not observe a second message by Northern blot analysis (Figure 1c).
Previously, two insertional alleles at the ODA12 locus that are defective for LC2 were described (Pazour et al., 1999 ). The oda12-1 allele lacks the entire LC2 gene, and Southern blotting demonstrated that this strain is also a complete null for LC10 (Figure 1d). In contrast, oda12-2 lacks only the 3′ end of the LC2 gene and retains the entire LC10 gene. The oda12-1 motility phenotype may be essentially completely rescued by transformation with a λ clone (λODA12#20) that is now known to contain both LC2 and LC10 genes; use of an ~3-kb HindIII fragment (pBD14) containing only the LC2 gene (Figure 1, a and d) also complemented the mutation (Pazour et al., 1999 ).
The full-length LC10 clone encodes a protein of 103 residues with a mass of 12,086 Da and a calculated pI of 5.28. There is one in-frame stop codon located within the 5′ untranslated region and a perfect copy of the Chlamydomonas polyadenylation signal within the 3′ untranslated region. The LC10 protein shares 33% identity (45% similarity) with Chlamydomonas LC8 and 45% identity (61% similarity) with the sea urchin (Anthocidaris crassispina) outer arm dynein LC4 protein. A CLUSTALW alignment and phylogenetic analysis of various members of this protein family revealed three distinct groupings (Figure 2,a and b). Chlamydomonas LC10, sea urchin LC4, and the human DNAL4 protein form a subgroup that is distinct from both canonical LC8/DYNLL proteins and the related Chlamydomonas outer arm component LC6. In this analysis, the Tctex1-related outer arm dynein protein LC9, which is structurally related to LC8 although it shares little sequence homology (DiBella et al., 2005 ; Wu et al., 2005 ), was used as the out-group.
The residues that differ between LC8 and LC10 were mapped onto the rat LC8/DYNLL1 molecular surface (PDB accession 1F3C; Figure 2c). This revealed considerable variation between the two proteins spread over all external faces including the intermonomer groove involved in binding dynein ICs and the α-helical face that contains most of the sequence differences between mammalian DYNLL1 and DYNLL2; note that for a particular isoform this region is highly conserved between species.
To define the flagellar location of LC10, we raised a rabbit polyclonal antiserum (CT247) against the full-length protein expressed as a C-terminal fusion with MBP and blot-purified the antibody against recombinant LC10. Immunoblot analysis revealed that this antibody preparation specifically recognized LC10 and importantly did not react with recombinant LC6 or LC8 (Figure 3a). The CT247 antibody recognized a single protein band of Mr ~12,000 in whole flagella samples (Figure 3b). After detergent treatment to remove the flagellar membrane, essentially all of this protein was associated with the axoneme and was solubilized by 0.6 M NaCl (Figure 3c).
Analysis of mutant axonemes lacking various substructures revealed that LC10 was specifically missing only in those strains that lack the outer dynein arm (oda1, oda2, oda6, oda9, and pf28 pf30 ssh1) or are null for LC10 (oda12-1; Figure 4). This protein was present in mutants defective for inner arms (ida1 and ida4) and radial spokes (pf14). Additional outer arm mutants defective for the α HC (oda11), the β HC motor domain (oda4-s7), and the oda4-s7 oda11 double mutant all contained LC10 (Figure 4). Previously, we found that oda6-r88, which contains an altered segment of 23-residues within the N-terminal region of IC2 (Mitchell and Kang, 1993 ), does not incorporate LC2, LC6, and LC9 into the outer arm, whereas a second IC2 mutant (oda6-r75) altered in a neighboring region did (DiBella et al., 2005 ). Immunoblot analysis revealed that both oda6-r75 (not shown) and oda6-r88 (Figure 4) retain LC10, although the amount present in oda6-r88 appears significantly reduced compared with wild type (Figure 4).
To confirm that LC10 is indeed an outer arm component, we purified this motor from the 0.6 M NaCl axonemal extract by sucrose density gradient centrifugation in the presence of Mg2+ and at low hydrostatic pressure; conditions designed to retain association of the entire outer dynein arm and docking complex (Takada et al., 1992 ). Most LC10 comigrated with known outer dynein arm components in fractions 3–7 at ~23 S (Figure 5). A minor amount of LC10 was also present higher in the sucrose gradient and sedimented at ~8 S (fractions 12 and 13); this likely represents a small pool of LC that dissociated from the main complex, as we have observed similar levels of dissociation products with other dynein LCs (e.g., LC7b; DiBella et al., 2004 ).
Previously, the swimming velocity of the oda12-1 strain rescued with a large genomic clone containing both the LC2 and LC10 genes was restored to wild-type levels (Pazour et al., 1999 ). To further test whether the lack of LC10 altered dynein function, we examined the flagellar beat frequency of wild type, oda1, oda12-1, oda12-2, and oda12-1 rescued for LC2 (but not LC10) with pBD14 and oda13 using the population-based fast Fourier transform method (Figure 6); to ensure that the power spectra were directly comparable all strains were grown and analyzed under identical conditions. The oda1 strain, which lacks the entire outer arm and docking complex, had a beat frequency of ~25 Hz, whereas that of oda12-1 was consistently lower (~15–20 Hz); oda12-2 was less impaired and showed a single peak in the power spectrum centered at ~30 Hz, with a broad shoulder extending toward 50 Hz. The oda13 mutant that lacks only LC6 had a peak frequency slightly less than wild type as reported previously (DiBella et al., 2005 ). The oda12-1 strain that had been rescued for LC2 with the pBD14 plasmid exhibited a broad peak from ~50–60 Hz that was subtly distinct from wild type. Thus, the lack of LC10 alone has only a minor effect but in combination with an LC2 defect results in a beat frequency that is less than that of strains that lack the entire outer arm. To further confirm this result, we compared the beat frequency of oda12-1 to that of oda1-oda9 when all these strains were grown under identical conditions and assayed at the same time of day. We obtained a mean beat frequency of 16 Hz for oda12-1 compared with 20 Hz (oda1), 23 Hz (oda2), 19 Hz (oda3), 21 Hz (oda4), 20 Hz (oda5), 23 Hz (oda6), 19 Hz (oda7), 20 Hz (oda8), and 19 Hz (oda9).
To assess whether these LCs are involved in stabilizing dynein complexes during assembly, we fractionated cytoplasm from wild-type Chlamydomonas and from strains specifically lacking LC6 or LC8 (Figure 7a). Immunoblot analysis (Figure 7b) revealed that the lack of LC8 in fla14 had a significant effect on the stability of some dynein components (Figure 7c) such that IC1, LC6, and LC9 were undetectable, and LC2 and LC10 migrated near the top of the gradient; the γ HC-associated LC1 protein was unaffected. In contrast, lack of LC6 in oda13 had essentially no effect on the migration of IC1 but did result in dissociation of LC2/LC10 from the IC complex under these conditions. Thus lack of LC8 leads to instability of some dynein components and the failure of outer arm assembly, whereas lack of LC6 results in complex dissociation in sucrose gradients but does not inhibit axonemal incorporation.
The DNAL4 protein expressed in mammals shares 50% identity (65% similarity) with Chlamydomonas LC10 (Figure 2a). To assess whether DNAL4 is also present in cilia and flagella, we generated a polyclonal antibody (CT254) against a MBP-DNAL4 fusion. After blot purification, this antibody was found to exhibit weak cross-reactivity with LC10 but did not recognize either LC6 or LC8 (Figure 8a). DNAL4 could be detected in a variety of murine tissue homogenates including testis, ovary, and brain (Figure 8b), and histological analysis revealed that it is highly expressed in the ciliated epithelial layer lining the bronchi of the lungs and in sperm flagella and spermatids within testis (Figure 8, c and d). These observations suggest that DNAL4 is associated with motile cilia and flagella. However, it is possible that DNAL4 plays additional roles in mammals as in situ hybridization studies indicate that this protein is highly expressed throughout the developing mouse embryo as is the ubiquitous DYNLL1 (aka LC8) protein (Figure 9). In contrast, other exclusively axonemal dynein components such as DNAHC5 and DNAIC1 exhibit much lower expression. Furthermore, DNAL4 is also expressed in adult mouse brain and is especially prominent in the hippocampus (Figure 9).
Here we report the identification of a third member of the LC8/DYNLL family in Chlamydomonas. LC10 orthologues are missing in nematodes and higher plants, suggesting that this protein is present only in those organisms that have motile cilia/flagella. LC10 copurifies with the outer dynein arm and is completely absent from axonemes of Chlamydomonas strains lacking only this motor, indicating that it is specifically associated with outer arm dynein and is not also present in the inner arm system as are LC7a, LC7b, and LC8. A Chlamydomonas LC10-null mutant exhibits subtle defects in motility indicating that this protein is indeed necessary for wild-type motor function and the combined LC2/LC10 deficiency is highly deleterious. However, LC10 does not appear necessary for dynein arm assembly per se. In contrast, lack of LC8 results in the complete failure of outer arm assembly, whereas the loss of LC6 has only a very minor effect on motility but does alter cytoplasmic stability of dynein subcomplexes. These results indicate that the three members of the LC8/DYNLL family in Chlamydomonas are functionally distinct.
Analysis of 35S-labeled αβ HC dimer from the Chlamydomonas outer arm dynein using 2D nonequilibrium pH gradient electrophoresis revealed nine LC spots in addition to the two LCs associated with the γ HC (Pfister et al., 1982 ). With the identification of LC10 as an integral component of the IC/LC complex, we can now account for all the LC components that copurify with this dynein (Table 1 and see Figure 7C). The IC/LC complex consists of two WD-repeat ICs (IC1 and IC2) associated with dimers of the three LC8/DYNLL-related proteins (LC6, LC8, and LC10), two members of the LC7/Roadblock/DYNLRB family (LC7a and LC7b), the dimeric Tctex1/DYNLT1 homologue (LC9), and the apparently monomeric LC2 protein that is related to murine Tctex2. The other LCs that have distinct regulatory functions interact directly with the N-terminal regions of the HCs (LC3, LC4, and LC5) or with the AAA+ domains of the γ HC motor unit (LC1).
Structural analysis has revealed that LC8/DYNLL proteins contain two α helices and five β strands and form symmetric dimers. The interface consists of strand-switched β sheets and leads to formation of two identical grooves into which short segments of the target protein bind (Liang et al., 1999 ; Fan et al., 2001 ). The α helical face of each monomer protrudes away from this interaction site and no associations via this region have yet been reported, even though these surfaces have been very highly conserved. The monomer-dimer transition is modulated by pH in vitro (Liang et al., 1999 ; Nyarko et al., 2005 ) and has been reported to be sensitive to phosphorylation (Song et al., 2008 ). The LC10 sequence aligns readily with that of LC8, suggesting that it has a similar overall architecture. Although both LC6 and LC10 have N-terminal extensions of 11–13 residues, the LC6 protein more closely resembles LC8 with the exception of an enlarged loop region between α2 and β2 elements and an extended C-terminal segment. In each LC8 monomer, the β strands interact to form an antiparallel β sheet that mediates dimerization. However, there are significant differences between LC8 and LC10 in the β3 strand, which is switched out to form the strand-switched dimer interface, suggesting that these two proteins are unlikely to form heterodimers. Furthermore, the properties of the molecular surface of LC8 and LC10 are quite distinct and are thus predicted to mediate distinct interactions within the dynein complex.
Previously, we described the comparative analysis of oda12-1 and oda6-r88 strains both of which lack LC2 (DiBella et al., 2005 ). We found that oda12-1 has a beat frequency of <~20 Hz, whereas that of oda6-r88 was 30–35 Hz. This observation was surprising in that although both strains lack LC2, oda6-r88 is in addition missing LC6 and LC9. Our current finding that oda12-1 is also a complete null for LC10, whereas this protein is present in oda6-r88, suggests that the combined lack of LC2 and LC10 is more detrimental to outer arm function than is the LC2, LC6, and LC9 combination. Furthermore, we have now found that when assayed under identical conditions, the oda12-1 strain exhibits a consistently lower beat frequency than do the oda1-oda9 strains that lack the docking complex and/or the entire outer arm. Even though the amount of γ HC is reduced from wild type, all the motor and other regulatory components are present in oda12-1. Thus, the combined lack of LC2/LC10 has a further negative effect on flagellar function. This might be caused by dysfunction of the remaining outer arm motor units providing in essence a braking effect or potentially by negative feedback onto the inner arm system. Intriguingly, although dynein function is abrogated in oda12-1, outer arm assembly can still occur although not to wild-type levels. Only in the oda6-r88 oda12-1 double mutant that is missing LCs 2, 6, 9, and 10 did we observe a complete lack of dynein arm incorporation into the flagellum (DiBella et al., 2005 ).
The lack of LC6 alone or in combination with LC2 and LC9 results in flagellar beat frequencies that are either little altered from wild type or remain greater than that of outer arm-less mutants. In contrast, LC8 null cells are completely unable to assemble outer dynein arms into the short flagella stubs that form in this mutant. Thus, there appears to be a very distinct difference between the roles played by the three members of this dynein LC family within the same dynein holoenzyme.
Structural studies of the LC8(DYNLL1) and Tctex1(DYNLT1) dimers in complex with a small region of the cytoplasmic dynein N-terminal domain revealed an extended conformation with the LCs packed close together between the two IC strands (Williams et al., 2007 ). In the Chlamydomonas outer arm, alteration of residues 31–54 within IC2 leads to the failure of LC2, LC6, and LC9 assembly and a marked reduction in LC10 levels. This IC region is predicted to form an extended β strand. We found previously using chemical cross-linking that LC2 and LC6 are within 9.2 Å of each other (DiBella et al., 2001 ) and that the LC9 dimer directly associates with both IC1 and IC2. Thus it is likely that these LCs bind in a sequential manner similar to that observed in cytoplasmic dynein. Alteration of the adjoining residues 54–64 of IC2 had no effect on LC incorporation (DiBella et al., 2005 ), further suggesting that these four LCs all bind IC2 within the N-terminal 54 residues.
The location at which LC8 associates with the outer arm ICs is uncertain. However, the sea urchin orthologue of IC1 contains a perfect copy of the canonical LC8 binding motif K/RxTQT and sequence alignment consequently allows for the tentative identification of the motif 184RETFT188 as the likely site in Chlamydomonas IC1. However, neither of these motifs is present in Chlamydomonas IC2 or its sea urchin equivalent. Similarly, the sites at which the LC7a/b proteins bind the ICs also remain to be defined, although based on analysis of cytoplasmic dynein (Susalka et al., 2002 ) and inner arm I1/f (Hendrickson et al., 2004 ), the interaction site is likely close to the C-terminal WD-repeat β-propellers.
The human DNAL4 gene maps to chromosome 22q13.1 and encodes a protein that is 50% identical to Chlamydomonas LC10. Immunostaining demonstrated that DNAL4 is highly concentrated in the apical cilia of epithelial cells lining the bronchi of the lungs and in sperm flagella and spermatids. These observations support the conclusion that DNAL4 is likely associated with axonemal dyneins. Surprisingly, however, examination of the BGEM in situ hybridization database (Magdaleno et al., 2006 ) revealed that in murine embryos DNAL4 is ubiquitously expressed at a very high level (as is DYNLL1) in contrast to known axonemal dynein components such as the DNAHC5 dynein HC. Furthermore, DNAL4 is also highly expressed in brain and especially the hippocampus. Thus, it is likely that DNAL4 plays additional roles in mammals beyond acting as an outer arm dynein LC.
In conclusion, we have demonstrated that the Chlamydomonas outer dynein arm contains three members of the LC8/DYNLL family that have distinct roles in motor assembly and activity. These data also provide strong evidence for the essential role of the IC/LC complex in controlling dynein motor function. Furthermore, we have found that DNAL4 represents the mammalian orthologue of Chlamydomonas LC10 and that its expression patterns suggest an additional role in processes other than ciliary/flagellar motility.
We thank Dr. Nada Zecevic (UCHC) for discussion of embryonic anatomy. This study was supported by Grants GM51293 (to S.M.K.) and GM060992 (to G.J.P.) from the National Institutes of Health.
This article was published online ahead of print in MBC in Press (http://www.molbiolcell.org/cgi/doi/10.1091/mbc.E08-04-0362) on June 25, 2008.