Cytoplasmic Dynein ATPase Activity Is Regulated by Dynactin-dependent Phosphorylation*

Cytoplasmic dynein is a microtubule-associated motor that utilizes ATP hydrolysis to conduct minus-end directed transport of various organelles. Dynactin is a multisubunit complex that has been proposed to both link dynein with cargo and activate dynein motor function. The mechanisms by which dynactin regulates dynein activity are not clear. In this study, we examine the role of dynactin in regulating dynein ATPase activity. We show that dynein-microtubule binding and ATP-dependent release of dynein from microtubules are reduced in dynactin null mutants, Δro-3 (p150Glued) and Δro-4(Arp1), relative to wild-type. The dynein-microtubule binding activity, but not the ATP-dependent release of dynein from microtubules, is restored by in vitro mixing of extracts from dynein and dynactin mutants. Dynein produced in a Δro-3 mutant has ∼8-fold reduced ATPase activity relative to dynein isolated from wild-type. However, dynein ATPase activity from wild-type is not reduced when dynactin is separated from dynein, suggesting that dynein produced in a dynactin mutant is inactivated. Treatment of dynein isolated from the Δro-3mutant with λ protein phosphatase restores the ATPase activity to near wild-type levels. The reduced dynein ATPase activity observed in dynactin null mutants is mainly due to altered affinity for ATP. Radiolabeling experiments revealed that low molecular mass proteins, particularly 20- and 8-kDa proteins, that immunoprecipitate with dynein heavy chain are hyperphosphorylated in the dynactin mutant and dephosphorylated upon λ protein phosphatase treatment. The results suggest that cytoplasmic dynein ATPase activity is regulated by dynactin-dependent phosphorylation of dynein light chains.

intermediate chains (ϳ55 kDa), and light chains (8 -23 kDa) (5)(6)(7). The C-terminal two-thirds of each heavy chain folds to form a large globular head domain that interacts with microtubules and is the site of ATP binding and hydrolysis (8,9). The N-terminal one-third of the heavy chains allow for dimerization of the heavy chains and interaction with additional dynein subunits to form a large globular base (2,10).
Dynactin, an additional multisubunit complex, has been proposed to link dynein with membranous cargo and also activate the dynein motor (11). Actin-related protein 1 (Arp1) is the most abundant dynactin subunit with 8 -10 Arp1 monomers polymerizing to generate a 37-nm-long filament (12). p150 Glued , the largest dynactin subunit, along with the p24 and p50/ dynamitin subunits, forms a projecting side arm from the Arp1 filament and mediates interaction of dynactin with dynein through contacts with the dynein intermediate chains (3,(13)(14)(15). The N-terminal domain of p150 Glued binds to microtubules, whereas a C-terminal domain has been shown to interact with the Arp1 filament (14). Additional dynactin subunits bind to either end of the Arp1 filament (16 -18).
Genetic studies in the yeast Saccharomyces cerevisiae, the filamentous fungi Neurospora crassa and Aspergillus nidulans, and in Drosophila have provided strong support to the hypothesis that dynactin is required for cytoplasmic dynein-based motility (19 -26). However, the mechanisms by which dynactin activates dynein motor activity and hence translocation of cargo are not well understood. A number of studies have shown that the phosphorylation state of dynein and dynactin subunits varies with alterations in dynein-dependent transport in vivo (27)(28)(29)(30)(31)(32). Analysis of dynein in rat kidney fibroblasts revealed that increased phosphorylation of dynein heavy chain correlated with activated dynein-dependent motor function (27). Treatment of rat cells with okadaic acid and activators of protein kinase A and protein kinase C resulted in increased vesicular transport and enhanced phosphorylation of p150 Glued (28). In axons, the phosphorylation state of dynein has been characterized, and it has been proposed that increased phosphorylation of dynein heavy chain enhances dynein-based transport in vivo (29). In contrast, in vitro studies have suggested that the increased phosphorylation of dynein heavy chain and dynein intermediate chains results in inhibition of dynein ATPase activity (32). The intermediate chains of dynein and p150 Glued of dynactin have been found to be hyperphosphorylated during M phase in HeLa cells, suggesting that dynein intermediate chains and p150 Glued phosphorylation may positively regulate mitotic processes or negatively regulate interphase processes such as minus-end directed membrane trafficking (31).
Dynein light chains have also been proposed to play a role in regulating dynein activity. Supportive evidence for a regulatory role of light chains comes primarily from work on axonemal dynein. To date, eight light chains have been identified for axonemal dynein and three light chains for cytoplasmic dynein (5,33,34). Activated motility of sperm, primarily by activated axonemal dynein, has been correlated with phosphorylation of the dynein light chain LC2 (35). In Paramecium tetraurelia, cAMP-stimulated phosphorylation of an axonemal polypeptide (29 kDa) that copurifies with 22 S dynein arm has been shown to activate microtubule translocation velocity and swimming speed (36). A direct interaction of light chains and motor domain of axonemal dynein heavy chain has also been shown (37). The C terminus of LC1, a homologue of the cytoplasmic dynein light chain Tctex1, interacts with the motor domain of axonemal dynein heavy chain (37). The role of the three cytoplasmic dynein light chains in regulating dynein motor activity is unknown; however, the Tctex1 light chain has been shown to be involved in cargo binding (7, 33, 38 -42).
We have developed a genetic screen that allows the isolation of hundreds of N. crassa mutants (referred to as ropy mutants) that are defective in cytoplasmic dynein/dynactin function (24,43). We have shown that the N. crassa ro-1 gene encodes cytoplasmic dynein heavy chain, whereas the ro-2, ro-3, ro-4, ro-7, and ro-12 genes encode the dynactin subunits p62, p150 Glued , Arp1, Arp11, and p25, respectively (24,25,43,44). 1,2 Determining the specific roles of these various subunits requires biochemical characterization of dynein isolated from these mutants. Recently, we reported the isolation and characterization of dynein ATPase from wild-type N. crassa (45). In this paper, we have isolated dynein from dynactin null mutants (⌬ro-3 and ⌬ro-4) to understand the mechanism by which dynactin regulates cytoplasmic dynein activity. Our results indicate that dynein ATPase is regulated by dynactin-dependent phosphorylation that appears to involve 20-and 8-kDa dynein light chains.

EXPERIMENTAL PROCEDURES
Strains and Growth Conditions-Wild-type N. crassa (74-OR23-1VA; FGSC 2489) was obtained from the Fungal Genetics Stock Center, Department of Microbiology, University of Kansas Medical Center, Kansas City, KS. The ⌬ro-3 and ⌬ro-4 mutants were as described (24,25). Media were as described (46). Mycelia were harvested from wildtype grown for 18 h and from ⌬ro-3 or ⌬ro-4 mutants grown for 24 h in liquid media inoculated with 1 ϫ 10 6 conidia/ml. The mycelia were harvested by filtration, frozen in liquid nitrogen, and kept at Ϫ80°C before use. For the 32 P-radiolabeling experiment, mycelia were grown in a low phosphate medium consisting of 0.1 mg/ml potassium phosphate and 5 mCi of [ 32 P]orthophosphate.
Preparation of Tubulin from Bovine Brain-Tubulin was purified from bovine brain white matter by cycles of polymerization and depolymerization as described (47). Subsequently, microtubule-associated proteins were removed by using an S15 cation exchanger membrane as described (48). One bovine brain (100 g) yielded about 20 mg of microtubule-associated protein-free tubulin. Final tubulin preparations contained negligible ATPase activity, which did not interfere with purification of N. crassa dynein or dynein ATPase assays.
Binding and ATP-dependent Release of Dynein from Microtubules-Frozen mycelia (1 g) were suspended in 1.5 ml of extraction buffer (EB 3 ; 50 mM PIPES (pH 7.0), 50 mM HEPES, 2 mM MgCl 2 , 1 mM EDTA, 1 mM DTT, and protease inhibitors; 1 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, 10 g/ml N␣-p-tosyl-L-arginine methyl ester, 1 g/ml pepstatin A, and 10 g/ml soybean trypsin inhibitor) (49). Zirconium beads (1 g) were added, and mycelia were ground using a mortar and pestle. Ground hyphae were centrifuged at 7,000 ϫ g for 10 min and then 12,000 ϫ g for 10 min to remove zirconium beads, unbroken cells, and insoluble material. The supernatant was further centrifuged at 200,000 ϫ g in a Beckman Ti 100.3 rotor for 60 min to clear the extract of membrane and membrane-associated proteins. Exogenous microtu-bules (0.2 mg/ml), apyrase (2 units), and taxol (20 M) were added to 1 ml of cell extract containing 5 mg/ml soluble protein. The solution was incubated for 60 min, overlaid over a 7.5% sucrose cushion, and then centrifuged at 60,000 ϫ g for 30 min in a Beckman 100.3 rotor. The supernatant was removed, and the pellet was resuspended in 0.5 ml of EB containing 3 mM GTP and 20 M taxol. The resuspended pellet was incubated for 15 min prior to centrifugation at 60,000 ϫ g for 30 min. The supernatant was removed, and the pellet was resuspended in 0.5 ml of EB, containing 20 M taxol and 10 mM ATP, and incubated for 2 h. The resuspended pellet was centrifuged at 100,000 ϫ g in a Beckman Ti 100.3 rotor for 30 min. Proteins contained within supernatant and pellet after each step were analyzed by silver staining (50), and cytoplasmic dynein heavy chain (RO1) and p150 Glued (RO3) were detected by Western blotting (51).
Isolation of Dynein ATPase from N. crassa-Large scale preparations of dynein/dynactin from N. crassa were performed using a published protocol, designed for the purification of mammalian cytoplasmic dynein, with slight modifications (49). Using the protocol described above for dynein-microtubule binding and ATP-dependent release, extracts were prepared from 10 g of frozen mycelia resuspended in 25 ml of EB. All other procedures were as described above, except that no apyrase was added while binding to microtubules, and no taxol was added during the ATP release step to ensure near 100% recovery from microtubules. The supernatant containing ATP-released cytoplasmic dynein was used for sucrose density gradient fractionation. One ml of supernatant, containing ATP-released dynein, was loaded onto a 10-ml 5-20% sucrose gradient in fractionation buffer (20 mM Tris-HCl, pH 7.6, 50 mM KCl, 5 mM MgSO 4 , 0.5 mM EDTA, and 1 mM DTT) as described (49). Centrifugation of the sucrose gradient was performed at 125,000 ϫ g in a Beckman SW 41 rotor for 16 h. Eleven 1-ml fractions were collected from the bottom of the tube, and 60 l of each fraction was analyzed by SDS-PAGE followed by silver stain and Western blot analysis. All operations were performed at 4°C. An alternate method was also used to isolate dynein ATPase activity by gel filtration (45). In this method, high speed cell extracts were loaded onto a Sepharose CL-4B (bed volume, 100 ml) column, pre-equilibrated with 150 ml of fractionation buffer (9) without DTT. One-ml fractions were collected, and each fraction was assayed for ATPase activity. All the operations were performed at 4°C.
ATPase Assay of Cytoplasmic Dynein Fractions-ATPase assays were performed in 50-l reaction mixtures containing 20 mM Tris-HCl (pH 7.6), 50 mM KCl, 5 mM MgSO 4 , 0.5 mM EDTA, and 1 mM DTT. In a standard assay reaction, 10 l of enzyme fractions and 4 mM ATP were incubated with assay buffer at 37°C for 40 min. Reactions were stopped using highly acidic malachite green reagent as described, and absorbance was read at 660 nm (52). The results shown are the average of three independent experiments. The amount of inorganic phosphate released in the enzymatic reaction was calculated using a standard calibration curve generated with inorganic phosphate. The control in this assay contained all ingredients of the reaction mixture, but the reaction was stopped immediately. Microtubule-stimulated ATPase activity was determined by the addition of 12.5 g of microtubules/reaction.
Protein Phosphatase Treatment-Treatment of dynein/dynactin containing fractions with phosphatase were performed in 50-l reaction mixtures containing 20 mM Tris-HCl (pH 7.6), 2 mM MnCl 2 , 5 mM DTT, and 400 units of phosphatase. The control contained all the ingredients except phosphatase, and the mixture was incubated for 30 min at room temperature. Following treatment with phosphatase, ATPase activities were measured as described above.
Radiolabeling and Immunoprecipitation-Wild-type and ⌬ro-3 mycelia (1.0 g each), grown in the presence of 5 mCi of [ 32 P]phosphoric acid, were suspended in 2.5 ml of extraction buffer containing 20 mM Tris-HCl (pH 7.6), 5 mM MgCl 2 , 0.5 mM EDTA, and 50 mM KCl with protease inhibitors as described (38). Cell extracts were made as described above. Dynein ATPase was isolated by a gel fitration method as described previously (45). Samples (0.5 ml) were taken from the peak fractions for the ATPase activity and treated with 2000 units of phosphatase as described above and then assayed for ATPase activity. Nonidet P-40 was added to final concentration of 1% to the phosphatase-treated and untreated peak fraction and then 1 ml of NET-gel buffer (50 mM Tris-HCl, pH 7.5, 0.1% Nonidet P-40, 1 mM EDTA, 0.25% gelatin, and 0.02% sodium azide) was added. Ten l of affinity-purified anti-RO1 antibody was added to the above solution and incubated for 2 h at 4°C. Afterward, 100 l of protein A-Sepharose was added and continued to incubate for an additional 3 h at 4°C. The pellet was isolated by centrifugation at 2000 rpm for 20 s and rinsed twice with wash buffer (10 mM Tris-HCl, pH 7.5, and 0.1% Nonidet P-40). The pellet was resuspended in 50 l of phosphate-buffered saline and 20 l of 4ϫ sample buffer and then subjected to SDS-PAGE (5-20%). Dynein heavy chain was detected by Western blotting using anti-RO1 antibody, and 32 P-labeled dyneinassociated proteins were identified using a Storm PhosphorImager.

Microtubule Binding of Cytoplasmic Dynein and Its ATP-dependent Release from Microtubules Is Reduced in Dynactin
Null Mutants-As a first step in the investigation of how dynactin mutations affect cytoplasmic dynein activity, we have examined cytoplasmic dynein-microtubule binding and its ATPdependent release from microtubules using extracts from wildtype and the dynactin null mutants ⌬ro-3 (p150 Glued ) and ⌬ro-4 (Arp1). N. crassa cell extracts were incubated with taxol-stabilized microtubules, and the microtubules were pelleted. The microtubules were then incubated with GTP to release kinesin and kinesin-related proteins, followed by incubation with 10 mM ATP to release microtubule-associated cytoplasmic dynein/ dynactin (45,48). Approximately 50% of the cytoplasmic dynein and dynactin contained in wild-type cell extracts bound to microtubules, and both dynein and dynactin were released to the same extent (ϳ50%) following ATP addition in the presence of taxol (Fig. 1, lanes 1-5). In contrast, reduced cytoplasmic dynein-microtubule binding was observed (ϳ25%) from extracts of the ⌬ro-3 and ⌬ro-4 dynactin mutants (Fig. 1, lanes 1  and 2). ATP-dependent release of dynein from microtubules was much reduced in the dynactin mutants relative to wildtype at 10 mM ATP (Fig. 1, lanes 4 and 5). The amounts of inorganic phosphate produced in these extracts were ϳ30% that of wild-type (data not shown), and this suggested that there was very low ATP hydrolysis in these mutants. Addition of higher amounts of ATP to the dynein-microtubule complexes from the dynactin mutants resulted in about 30% and 60% of dynein released at 25 and 50 mM ATP, respectively (data not shown). This result suggested that the affinity of ATP for dynein ATPase was reduced in dynactin mutants.
Dynein-Microtubule Binding, but Not ATP-dependent Release from Microtubules, Is Rescued by in Vitro Complementation-Reduced microtubule binding and ATP-dependent release of cytoplasmic dynein isolated from dynactin mutants could be due to either the absence of dynactin or to the modification of cytoplasmic dynein when produced in a dynactin mutant. To explore this, in vitro complementation tests were conducted by mixing an equal amount of protein from ro-1(B15) and ⌬ro-3 cell extracts, which lack cytoplasmic dynein heavy chain (RO1) and p150 Glued (RO3), respectively. Following the mixing of extracts, microtubule-binding, GTP wash and ATPdependent release experiments were carried out as before. The ro-1(B15) and ⌬ro-3 cell extracts were treated in parallel experiments to serve as controls. Dynactin from the ro-1(B15) mutant did not bind to microtubules in the absence of cytoplasmic dynein (Fig. 1, lanes 1 and 2). In contrast, microtubulebinding of dynein was improved when extracts were mixed (ro-1(B15) ϩ ⌬ro-3) (Fig. 1, lanes 1 and 2). However, ATP-dependent release of cytoplasmic dynein and dynactin from microtubules at 10 mM ATP was not improved when the cell extracts were mixed (Fig. 1, lanes 4 and 5). Only at a high concentration of ATP (50 mM) was dynein released from microtubules at a level similar to that of wild type (data not shown). These results indicate that cytoplasmic dynein microtubule binding activity, but not ATP-dependent release from microtubules, can be rescued in vitro. This suggests that both dynein and dynactin are required for dynein-microtubule binding, but not for ATP hydrolysis. From these results, it appears that ATP hydrolysis is reduced in dynactin mutants due to decreased affinity for its substrate.
Dynein ATPase Activity Is Reduced in a Dynactin Null Mutant-To assay dynein-specific ATPase activity, the microtubule-released proteins were subjected to fractionation on a 5-20% sucrose gradient. In this experiment, dynein was released from microtubules in the absence of taxol to ensure ϳ90% release from wild-type and dynactin mutant. ATPase activity was measured in sucrose gradient fractions in the absence and presence of taxol-stabilized exogenous microtubules (Fig. 2). Two peaks of ATPase activity were observed in wild-type and ⌬ro-3 fractions, but only the activity of the first peak was stimulated by the addition of microtubules (Fig. 2). The ATPase activity of the first peak was due to cytoplasmic dynein (45), and this activity was about 15-fold less in the ⌬ro-3 mutant relative to wild-type (Fig. 2). The ATPase activity in the second peak represents some other microtubule-associated protein having ATPase activity that is not affected by either dynein or dynactin null mutations (6,45). The 15-fold lower activity in the first peak of ⌬ro-3 is due to both reduced yield of dynein and a reduction in the ATPase activity of dynein. The yield of dynein from the ⌬ro-3 mutant was estimated to be 2-fold less than the yield from wild-type due to ϳ50% loss of dynein at the initial microtubule-binding step. Therefore, the real reduction in ATPase activity was estimated as ϳ8-fold in the ⌬ro-3 mutant relative to wild-type. In contrast, microtubule stimulation of dynein ATPase activity was not affected by the ⌬ro-3 mutation. Similar fractionation patterns and corresponding ATPase activities were observed when dynein was purified from the ⌬ro-4 mutant and from the mixed extracts (ro-1(B15) ϩ ⌬ro-3; data not shown).
To examine the nature of reduction in dynein ATPase activity in the dynactin mutant, the effect on kinetic parameters (K m(ATP) and V m ) of dynein ATPase in ⌬ro-3 and mixed (ro-1(B15) ϩ ⌬ro-3) were calculated. The K m(ATP) and V m of dynein ATPase in peak fraction 4 were determined in the presence and absence of taxol-stabilized microtubules ( Table I). The results showed that there is a significant increase in K m(ATP) (10 -15fold), but only a small decrease in V m (ϳ2-fold) for the dynein ATPase from ⌬ro3 and mixed extracts (ro-1(B15) ϩ ⌬ro-3) relative to dynein from wild-type. The change in affinity for ATP (K m(ATP) ) was accounted for the absolute reduction in the ATPase activity.
Dynein ATPase Activity Isolated from Wild-type Is Independent of Dynein-Dynactin Physical Contact-Two simple models could explain the reduction in affinity of dynein for ATP observed with the ⌬ro-3 mutant versus the wild-type: 1) the physical absence of dynactin leads to reduced dynein ATPase; or 2) dynein, produced in a dynactin null mutant, is in an inactive state. To test the first possibility, we determined dynein ATPase activity in a fraction from wild-type containing almost no dynactin (Fig. 3, fraction 3) and a fraction containing high levels of dynactin (Fig. 3, fraction 4) as well as the corresponding fractions from ⌬ro-3. The results showed that the dynein ATPase activities correlated well with the relative abundance of RO1 protein in the respective fractions of wildtype and ⌬ro-3, irrespective of the presence or absence of RO3. Therefore, the low dynein ATPase activity observed in the ⌬ro-3 mutant was not due to the simple absence of dynactin. The observation that fraction 4 has more than 15-fold lower activity in ⌬ro-3 than the same fraction of wild-type, even though the RO1 abundance is only ϳ2-fold less, suggests that the reduced activity is due to inactivation/inhibition of the dynein ATPase. In addition, the sedimentation velocity of dy-nein from ⌬ro-3 is not significantly different from that of dynein from a wild-type strain (Fig. 3A). This suggests that the gross structure of dynein from the two strains is similar and that inactivation of dynein ATPase activity in the ⌬ro-3 mutant is likely the result of a relatively small conformational change.
Dynein ATPase Activity Isolated from a Dynactin Null Mutant Is Increased by Phosphatase Treatment-To determine whether the reduced dynein ATPase activity observed in dynactin null mutants is due to hyper/hypophosphorylation of dynein, we treated the dynein fractions (Fig. 3, fraction 4) with phosphatase (a general phosphatase) and then measured ATPase activities. Dynein ATPase activity from wild-type increased ϳ1.25-fold following phosphatase treatment; however, ATPase activity increased ϳ10-fold for dynein isolated from the ⌬ro-3 mutant (data not shown). This result suggests that reduced dynein ATPase activity observed in dynactin null mutants is due to phosphorylation of one or more dynein subunits. The K m(ATP) and V m of dynein ATPase isolated from the ⌬ro-3 mutant were determined after treatment with phosphatase in the presence and absence of taxol-stabilized microtubules ( Table I). The results showed that there is a significant decrease in K m(ATP) (8 -10-fold), following phosphatase treatment, suggesting that the affinity of dynein ATPase for ATP is enhanced when dynein is dephosphorylated.
There Is Increased Phosphorylation of Low Molecular Weight Dynein-associated Proteins in a Dynactin Null Mutant-Our observation that dephosphorylation enhances dynein ATPase activity from the dynactin mutant led us to investigate the proteins(s) that are responsible for activation/inhibition of dynein ATPase activity by phosphorylation. N. crassa wild-type and ⌬ro-3 mutant were grown in 32 P supplemented media, extracts were made, and both dynein ATPase activities and the phosphorylation states of dynein-associated polypeptides were determined. For these experiments, we needed to simultaneously isolate dynein ATPase activity and look at the phosphorylation state of dynein-associated proteins. Therefore, dynein was isolated using a previously described protocol that employs gel filtration (45). Peak fractions of dynein from wildtype and the dynactin mutant were treated with phosphatase and then assayed for ATPase activity. As above, dynein ATPase  -1 ϩ ⌬ro-3), and protein phosphatase-treated ⌬ro-3 The ATP saturation experiments were performed using the peak fraction of 5-20% sucrose gradient. The kinetic constants were calculated by double-reciprocal plot and were verified by using Eadie-Hofstee plot. Data represent the mean of three independent determinations. MT, microtubules; PP, protein phosphatase. activity from wild-type increased slightly upon treatment with phosphatase, whereas dynein ATPase activity from the dynactin mutant increased ϳ10-fold (Fig. 4A). To determine the extent of phosphorylation of dynein-associated proteins in these phosphatase-treated and untreated samples, dynein was immunoprecipitated and the precipitated proteins were resolved by SDS-PAGE. Western analysis showed that approximately equal amounts of RO1 were precipitated from all samples (Fig. 4B). RO1 was found to be phosphorylated in wild-type and ⌬ro-3 mutant (Fig. 4C). To a lesser extent, a 50-kDa protein (putative dynein light intermediate chain) was also phosphorylated in wild-type and the dynactin mutant. The phosphorylation states of RO1 and the putative dynein light intermediate chain were unaffected by the dynactin mutation, and neither protein was dephosphorylated when samples were treated with phosphatase. In contrast, a high level of phosphorylation was observed for four small proteins (48, 32, 20, and 8 kDa) in the dynactin mutant relative to wild-type (Fig.  4C). These proteins were dephosphorylated when samples from either wild-type or the dynactin mutant were treated with phosphatase (Fig. 4C). The identity of these proteins was not determined, but the 20-and 8-kDa proteins may represent dynein light chains (35,38). Increased phosphorylation of these putative dynein light chains correlates with reduced dynein ATPase activity from the ⌬ro-3 mutant (Fig. 4, compare A and  C). DISCUSSION Current evidence indicates that dynactin is required for all dynein-based activities (53,54). Dynactin has been proposed to function as an adapter for linking dynein to cargo and as an activator of motor function. The adapter function of dynactin has been proposed to operate by interaction of the Arp1 filament with a spectrin-like cytoskeleton associated with Golgi membrane (54 -56). The motor activator function of dynactin is not well understood. Recently, dynactin was shown to be required for increased processivity of dynein motor along microtubules; however, dynein ATPase activity was found to be unaffected by the presence or absence of dynactin (57). Consistent with this study, we report here that dynein ATPase activity, isolated from wild-type N. crassa, is unaffected by the amount of co-fractionating dynactin. However, our analysis of dynein produced in a dynactin null mutant suggests that dynein ATPase activity is regulated by dynactin-dependent phosphorylation of dynein light chains. Data from several biochemical experiments support this statement. First, ATP-dependent release of dynein from microtubules is reduced in a dynactin null mutant. Second, ATPase activity of dynein is reduced in a dynactin null mutant and treatment with phosphatase restores dynein ATPase activity to that of wild-type. Third, there is increased phosphorylation of dynein-associated polypeptides (putative dynein light chains) when dynein is immunoprecipitated from a dynactin mutant. Our results are consistent with two recent studies that have also demonstrated that dynein can be activated by dephosphorylation (32,58). Treatment of rat hepatocytes with phosphatase inhibitors results in inhibition of dynein ATPase activity; however, in this study loss of ATPase activity correlated with increased phosphorylation of dynein heavy and intermediate chains (32). Similarly, dephosphorylation of dynein isolated from Xenopus melanophore extracts results in activation of dynein motility, but activation occurs via increased dynein-microtubule binding (58).
A major question raised by our findings is how are kinase/ phosphatase activities coordinated with dynactin function to properly regulated dynein motor activity. Given that one function of dynactin is to link dynein to cargo and subsequently activate dynein motor (54 -56), it is likely that the phosphatase activity responsible for activating dynein ATPase in N. crassa is coupled with cargo attachment. Recently, we showed that cytoplasmic dynein of N. crassa, like that of more complex eukaryotes, is required for retrograde transport of membranous cargo (59). As part of this transport cycle, it is likely that cytoplasmic dynein and dynactin link with cargo at hyphal tips and form a translocation competent complex. At distal regions of hyphae, cargo is released and cytoplasmic dynein and dynactin are transported back to hyphal tips, presumably in an inactive state. We propose that, in N. crassa, dynein ATPase is activated at hyphal tips by a dynactin-dependent phosphatase, whereas at distal cargo-release sites, dynein ATPase is inactivated by a protein kinase. These antagonistic activities could be localized specifically to hyphal tips and distal regions, respectively. Alternatively, they may be localized throughout the length of hyphae, but their ability to act on dynein may be regulated by dynactin along with the presence or absence of membranous cargo.
We have also observed phosphorylation of dynein heavy FIG. 4. Examination of cytoplasmic dynein ATPase activity and the phosphorylation state of dynein-associated proteins in wild-type (WT) and ⌬ro-3 mutant. A, effect of phosphatase treatment on cytoplasmic dynein ATPase activity. Wild-type and ⌬ro-3 mutant were grown in 32 P-supplemented media, dynein ATPase was isolated by gel filtration (45), and the peak dynein-containing fractions from wild-type and the ⌬ro-3 mutant were treated with protein phosphatase ( PP) followed by measurement of dynein ATPase activity (see "Experimental Procedures"). The open bars represent dynein ATPase activity in controls (i.e. without protein phosphatase), and closed bars represent dynein ATPase activity after treatment with protein phosphatase. B, Western blot analysis of RO1 protein immunoprecipitated from protein phosphatase treated and untreated samples of wild-type and ⌬ro-3 mutant. Immunoprecipitated proteins were electrophoresed on a 5-20% SDS-polyacrylamide gel. C, detection of 32 Plabeled proteins that co-immunoprecipitate with cytoplasmic dynein heavy chain from protein phosphatase treated and untreated sample of wild-type and a ⌬ro-3 mutant. Arrows indicate dynein heavy chain (RO1) and lower molecular weight dynein-associated proteins. chain and what may be dynein light intermediate chain in wild-type and dynactin null mutant (Fig. 4C). The phosphorylation state of these proteins is unaffected by a dynactin null mutation, unlike the smaller dynein-associated proteins. The significance of these phosphorylation events are unknown. However, observations by several groups suggest that there may be more that one protein kinase and phosphatase involved in phosphorylation/dephosphorylation of subunits of the dynein and dynactin complexes (28 -32, 60). Variable phosphorylation of different dynein/dynactin subunits may be one of the mechanisms used to independently regulate specific steps of dynein/ dynactin-dependent transport.
An additional question raised by our study is how is dynein ATPase activity inhibited by phosphorylation of what are apparently dynein light chains. Although we have not shown that the 20-and 8-kDa proteins represent dynein light chains, fungi have been found to have two of the three known cytoplasmic dynein light chains, Tctex1 (14 kDa) and the 8-kDa dynein subunit (Ref. 33; accession no. AF196291). Dynein isolated from either wild-type N. crassa or a dynactin null mutant shows the same sedimentation coefficient, and it is unlikely that phosphorylation of these putative subunits leads to a large conformational change or disassembly of the dynein complex. It is more likely that the phosphorylation of one or more of these putative dynein subunits leads to a relatively small, but significant, conformational change in the dynein heavy chain. Tctex1 and the 8-kDa subunit have been shown to interact with the dynein intermediate chain, and are part of the large globular base of the dynein motor (41,61). Although this places the dynein light chains relatively far from the very large dynein motor domain (Ͼ300 kDa), it has been found that dynein heavy chain ATPase activity is very sensitive to conformational change and can be easily perturbed by mutational changes that are far from the ATP binding pocket (9). Therefore, phosphorylation of dynein light chains may cause a small conformational change in the motor domain of dynein heavy chain either by direct interaction with dynein heavy chain or through dynein intermediate chain, that in turn inactivates dynein ATPase activity. In addition, analysis of dynein-dynactin interaction in vivo has suggested that multiple pools of cytoplasmic dynein exist with distinct light chain compositions (41). It is also possible that interaction of light chains with the dynein complex is affected by phosphorylation and it is the presence or absence of specific light chains that controls dynein ATPase activity. Finally, it is important to note that Tctex1 has also been shown to be a cargo-binding subunit of dynein (42). Variable phosphorylation of Tctex1 could regulate both dynein ATPase activity and cargo attachment, thereby providing a mechanism to couple these two processes.
Although dynein ATPase activity is strongly affected by dynactin null mutations, the ability of microtubules to stimulate dynein ATPase activity is not. The phosphorylation events that reduce the affinity of dynein for ATP are apparently separate from the ability of microtubules to stimulate ATPase activity. In contrast, dynactin null mutations do reduce the ability of dynein to bind and release from microtubules relative to dynein isolated from wild-type. Results from in vitro complementation experiments indicate that microtubule binding is restored but ATP-dependent release is not rescued for dynein isolated from the dynactin mutants. These results suggest that dynactin is required for dynein-microtubule binding and are also consistent with the finding that dynactin enhances both dynein-microtubule binding and motor processivity (57). The reduction in release of dynein from microtubules observed with the dynactin null mutants is due to the reduced ATPase activity.
We have developed an extensive collection of cytoplasmic dynein and dynactin mutants 1,2 (24,25,43,51,62). Some of these dynactin mutants result in accumulation of dynein and dynactin along microtubules at hyphal tips, whereas other mutants result in accumulation of dynein and dynactin at the minus ends of microtubules located in distal regions of hyphae. 2 Biochemical characterization of cytoplasmic dynein/dynactin from these various mutants may help in defining the dynein/dynactin subunits and the specific domains of these subunits that are required for regulation of dynein ATPase activity with respect to specific steps in the membranous cargo transport cycle. In addition, there are available N. crassa mutants defective in specific kinases and phosphatases, which should also be helpful in identifying the protein kinases/phosphatases involved in regulating dynein function (63).