Differential expression and phosphorylation of the 74-kDa intermediate chains of cytoplasmic dynein in cultured neurons and glia.

The 74-kDa intermediate chains (IC74) of the cytoplasmic dynein complex are believed to be involved in the association of dynein with membranous organelles. While each dynein molecule is thought to have two or three IC74 subunits, at least six different IC74 protein isoforms were found in dynein from brain. Therefore we investigated the relationships of the brain cytoplasmic dynein IC74 isoforms and their association in the dynein complex at the cellular level. We found that cultured cortical neurons and glia express distinct IC74 isoforms. The IC74 isoform pattern observed in dynein from cortical neurons was generally similar to that found in dynein from adult brain, indicating that there are different populations of cytoplasmic dynein in neurons. Two IC74 isoforms were observed on two-dimensional gels of dynein from glia, while a single glial IC74 mRNA was detected. Metabolic labeling of glial dynein with P followed by treatment of the isolated dynein with phosphatase in vitro demonstrated that one of the glial IC74 isoforms is the product of the single glial IC74 mRNA and that the other is its phosphoisoform. A single mRNA product and its phosphoisoform are therefore sufficient for constitutive dynein function and regulation in glial cells.

Cytoplasmic dynein is a ubiquitous minus end-directed microtubule-based motor protein (1,2). In neurons, it is believed to be the motor for retrograde transport of membranous organelles from the synapse to the cell body (3,4). Retrograde transport is essential for the movement of neurotrophic factors from the synapse to the cell body and for the recycling and degradation of cellular components (5,6). The cytoplasmic dynein complex is composed of two 530-kDa heavy chains, which make up two globular heads, as well as intermediate chains of ϳ74 kDa (IC74) 1 and 53-59 kDa (7). Molecular analyses indicate that portions of the cytoplasmic dynein IC74 are related to the ICs from the outer arm dynein of flagella (8 -11). In addi-tion, the antibody 74.1 directed against the IC74 subunit of cytoplasmic dynein cross-reacts with a polypeptide of similar molecular mass in mammalian cilia and flagella (8). It is known that the two IC subunits from the outer dynein arm of Chlamydomonas flagella are located at the base of the molecule (12,13) and that one, IC78, participates in binding the outer arm to its cargo, the A tubule of the axoneme outer doublet microtubules (12). The similarity of flagellar and cytoplasmic dynein ICs makes it likely that the cytoplasmic dynein IC74 is also located at the base of the two stalks and is involved in binding to cargo, such as membranous organelles.
While quantitative analyses of the polypeptide composition of the cytoplasmic dynein complex indicate that it contains only two or three IC74 subunits per molecule (4,7,14,15), we recently identified at least six different IC74 isoforms in dynein immunoprecipitated from adult rat brain (16). Therefore, there must be populations of cytoplasmic dynein in brain that differ in the composition of their IC74 subunits. Since many of the brain IC74 isoforms were labeled in vivo with 32 P, some of the IC74 diversity is presumably generated by posttranslational modification of a more limited number of polypeptides, and the role of phosphorylation in the regulation of cytoplasmic dynein is under active investigation (16 -18). Recently two cytoplasmic dynein ic74 genes, and five alternative splice mRNA variants were identified in brain (19) which suggested that at least five distinct IC74 polypeptides can be synthesized. Therefore, we sought to understand the circumstances which generate the multiple IC74 isoforms and to determine if different populations of cytoplasmic dynein could be identified. As brain tissue is made up of many specialized cell types, the IC74 isoforms of two brain-specific cell types, cultured cortical neurons and glia, were investigated at the protein and mRNA levels. In addition, the role of phosphorylation in generating the IC74 isoforms in brain and the cultured cells was investigated. We report that neurons and glia express different IC74 isoforms. Two IC74 isoforms were found when cytoplasmic dynein from glia was analyzed by two-dimensional gel electrophoresis. One isoform is the product of the single dynein IC74 mRNA transcribed in glia, and the other is its phosphoisoform. Three additional ic74 gene products and their phosphoisoforms are found in neurons. These results demonstrate that there are cell specific differences in the expression of the cytoplasmic dynein IC74. Furthermore, given the number of ic74 gene products identified in cortical neurons, neurons must contain populations of cytoplasmic dyneins which have different IC74 polypeptide compositions and differences in polypeptide phosphorylation. However, all the cytoplasmic dynein molecules in glia have the same IC74 polypeptide and differ only in the extent of its posttranslational modification. These results have important implications for the mechanism of cytoplasmic dynein function.

Immunoprecipitation, Electrophoresis, and in Vitro Phosphatase
Treatment of Cytoplasmic Dynein-Immunoprecipitation and electrophoresis were performed as described previously (16). Western blotting was as described previously, except a poly(vinylidene difluoride) membrane was used instead of nitrocellulose (16). Mouse monoclonal antibody 74.1 (16) and rabbit polyclonal anti-IC74-2 antibody were used to screen the blots. Anti-IC74-2 is specific for the common COOH-terminal 19 amino acids of the ic74-2 gene products (19). Horseradish peroxidase-conjugated goat anti-rabbit and goat anti-mouse antibodies were used as secondary antibodies and were detected by Renaissance ECL (DuPont, Wilmington DE). The incorporation of radioisotope into the dynein IC polypeptides was analyzed with a PhosphorImager (Molecular Dynamics, Sunnyvale, CA) using storage phosphor screen autoradiography and quantified as described previously (16). Immunoprecipitated dynein bound to protein A beads was treated with 200 units of phosphatase (New England Biolabs, Beverly, MA) for 40 min at 30°C. The beads were then washed two times and eluted with IEF sample buffer (16).
Cell Culture-Cultures of cortical neurons were prepared as described by Bartlett and Banker (20) for hippocampal neurons with some modifications. Briefly, rat fetuses were harvested from anesthetized female rats on the 18th day of gestation. The cortexes from 10 -15 fetuses were dissected from the brains and treated with 0.25% trypsin for 15 min at 37°C. The tissue was washed in magnesium-and calciumfree Hank's balanced salt solution, dissociated by repeated passage through a constricted Pasteur pipette, and filtered through a 70-m Nitex filter. Cells were diluted in DMEM (Sigma, St. Louis, MO) containing 10% bovine calf serum (Hyclone, Logan UT) and plated on poly-L-lysine-treated six-well plates at a density of 10 6 cells/well. After 2 h at 36°C the medium was replaced by glia-conditioned DMEM with 10% calf serum. The next day, cytosine-arabinoside (Calbiochem) (5 M) was added to inhibit glial proliferation. Cultures were maintained in glial-conditioned DMEM medium and used 3 days after plating. Cultured glia, primarily astroglia, were prepared as described in Goslin and Banker (21). Briefly, the cortexes from three to five newborn rats were dissected from brains and treated with 0.25% trypsin for 20 min at 37°C. The tissue was washed in magnesium-and calcium-free Hank's balanced salt solution, dissociated by repeated passage through a pipette, and filtered through a 70-m Nitex filter. Cells were counted and diluted in minimal essential medium containing 10% fetal bovine serum. Cells were plated on 60-mm diameter tissue culture dishes at a density of 5 ϫ 10 5 cells/dish and used when they reached confluence in ϳ1 month.
Radiolabeling of Cytoplasmic Dynein in Cultured Cells and Adult Brain in Vivo-Cultured cortical neurons and glia were labeled for 1-18 h with 0.15-0.35 mCi/ml of either [ 32 P]orthophosphoric acid in a 9:1 mixture of phosphate-free DMEM (Sigma) supplemented with 10 M phosphate, sodium pyruvate, and 5% dialyzed calf serum (Life Technologies, Inc.), and DMEM, 10% bovine calf serum; or Trans 35 Slabel (ICN) in a 9:1 mixture of methionine-and cysteine-free DMEM (Life Technologies, Inc.) supplemented with sodium pyruvate, 5% dialyzed calf serum, and DMEM, 10% bovine calf serum. Rat brain cytoplasmic dynein was labeled in vivo by injection of 32 P into the cerebrospinal fluid of the ventricles as described in Dillman and Pfister (16).
RT-PCR of IC74 -RNA from cultured cells or rat tissue was isolated using the QIAGEN (Chatsworth, CA) RNeasy kit, according to the manufacturer's directions, from 30 -100 mg of tissue. The first strand cDNA was made using a SuperScript preamplification system (Life Technologies, Inc.) according to the manufacturer's directions, with oligo(dT) as the primer. First strand synthesis was followed with an RNase H treatment. The first strand cDNA was diluted 1:5 with distilled H 2 O, and 1 l of cDNA per 20 l of reaction was used in the PCR. The PCR conditions and reagent concentrations were as per the manufacturer's directions for PCR amplification of first strand cDNA using a hot start protocol: 94°C for 5 min, then 30 cycles of 94°C for 45 s, 55°C for 45 s, 72°C for 60 s. Six oligonucleotide primers designed to distinguish the two genes and five alternative splice variants of IC74 (19) were synthesized (Life Technologies, Inc.): G1/S ϭ 5Ј-GAAGCTG-GAAGCCAAGACGATC-3Ј, G1/AS ϭ 5Ј-ACCTGGGTAACCTTTGA-CACGC-3Ј, G2/S1 ϭ 5Ј- The oligonucleotides were used in the following four combinations: G1/S and G1/AS, G2/S1 and G2/AS1, G2/S1 and G2/AS2, and G2/S2 and G2/AS1. To distinguish the different PCR products generated by these primers, the products of each PCR were separated on 8% acrylamide gels, 1 ϫ TBE (89 mM Tris, 89 mM boric acid, 25 mM sodium EDTA) buffer. A HaeIII digest of pBR322 DNA (Marker V, Boehringer Mannheim) was used to determine the sizes of the PCR products.

Modification of the Adult Brain IC74 Isoform Pattern by
Phosphatase Treatment-At least six isoforms of the adult brain cytoplasmic dynein IC74 subunits can be resolved on two-dimensional gels (Fig. 1, A and B) (16). The isoforms are arranged in two arcs of three spots each. We refer to the acidic arc of spots on the IEF gradient as the A, A1, and A2 spots, and the basic arc of spots as the B, B1, and B2 spots. In order to clearly resolve and visualize all of the spots, the A2 and B2 spots are generally fainter and less well resolved than the other spots, and they tend to appear as "tails" of the A1 and B1 spots. Often the A spot is not well resolved from the A1 spot ( Fig. 2A). This pattern of spots was not unique to a specific region of the adult brain, as the same pattern was found in dynein immunoprecipitated from the cerebral cortex, cerebellum, medulla, and pons (not shown). When brain cytoplasmic dynein, labeled with 32 P in vivo, was analyzed on two-dimensional gels, label was found over all the spots, except the B spot (16). However, the extent to which this posttranslational modification contributed to the diversity of the IC74 isoforms seen in adult brain was unknown. To identify and characterize the unphosphorylated IC74 isoforms, we treated immunoprecipitated dynein with phosphatase in vitro and analyzed the resulting IC74 isoform pattern on twodimensional gels (Fig. 2). To demonstrate the effectiveness of the phosphatase treatment, brain cytoplasmic dynein was first labeled in vivo with 32 P, then immunoprecipitated and resolved on two-dimensional gels. As previously reported (16), label was observed over all the spots except the B spot (Fig. 2, A and B).
Phosphatase treatment of the 32 P-labeled brain dynein removed all of the detectable 32 P (Fig. 2D). Coincident with the removal of the 32 P, the A2 and B2 spots were no longer seen (Fig. 2C). Therefore, some of the adult rat brain IC74 isoform diversity is solely the result of posttranslational modification of other isoforms. While four distinct polypeptides corresponding to the A, A1, B, and B1 spots remain after dephosphorylation, the A2 and B2 spots are generated by phosphorylation of the other polypeptides.
Comparison of the IC74 Isoforms of Adult Rat Brain and Primary Cell Cultures-Brain tissue is a mixture of cell types. To identify IC74 isoforms at the cellular level we investigated cytoplasmic dynein from two primary cell cultures, glia and cortical neurons. As shown in Fig. 3A, the IC74 isoform patterns of 35 S-labeled dynein isolated from glia and neurons are very distinctive. The similarity of the cortical neuron (Fig. 3A, FIG. 1. The cytoplasmic dynein IC74 isoforms of adult rat brain. An adult rat brain was homogenized in lysis buffer. The dynein was immunoprecipitated and analyzed by two-dimensional gel electrophoresis, and the gel was silver-stained, as described previously (16). A, the IC74 region of the two-dimensional gel, showing the six resolved isoforms. B, diagram identifying the IC74 isoforms. The arrow points to the region known as the B1 spot, the arrowhead points to the A1 spot. The acidic and basic ends of the IEF pH gradient are indicated with a "ϩ" and "Ϫ" on the top of Panel A.
center panel) and brain IC74 isoform patterns suggested that the observed cortical neuron spots corresponded to B and B1, as well as poorly resolved A and A1 spots. Similar results were also obtained with cultured hippocampal neurons (not shown). To more precisely correlate the cultured cell isoforms with those identified in brain, 35 S-labeled glial dynein was mixed with an excess of unlabeled dynein from brain and analyzed by two-dimensional gel electrophoresis. When the stained gel and autoradiograph were compared, we found that the major and minor glial IC74 isoforms co-migrate with the brain dynein B and B1 spots, respectively (Fig. 3B).
To directly compare the neuronal and glial isoforms, approximately equal counts/min of metabolically labeled glial and neuronal dynein were mixed, and the combined sample was resolved by two-dimensional gel electrophoresis (Fig. 3A, MIXED). Two major spots of approximately equal intensity, as well as a small amount of a slightly more acidic and faster migrating spot, are identified on the autoradiograph of the combined samples. This demonstrated that the glial and neuronal IC74 isoform patterns overlap. From our analysis of 35 S-labeled glial dynein, as well as the general pattern of the spots in the combined sample, we conclude that the majority of the B spot was contributed by the glial dynein and that the faint A spots and majority of the B1 spot were derived from the cortical neuron dynein.
In these initial studies, the glia and neurons were labeled for a relatively short time period. When glia were labeled for longer times (see Fig. 5C), the same IC74 isoform pattern was observed. Furthermore, the relative amounts of the glial IC74 isoforms seen in all the metabolic labeling experiments closely resembled the results observed when the gels were stained for protein (Fig. 4A). However, when cortical neurons were labeled for longer time periods the B2 spot was clearly identified, and the relative amount of the A isoforms appeared to increase (Fig.  5E). Nevertheless, dynein from cultured neurons still had a FIG. 2. Phosphatase treatment of the cytoplasmic dynein eliminates the IC74-A2 and -B2 spots. Adult brain cytoplasmic dynein was labeled in vivo as described under "Materials and Methods." The radiolabeled dynein was immunoprecipitated and divided into two aliquots. One aliquot was treated with the general protein phosphatase, phosphatase, as described under "Materials and Methods," and the other aliquot was treated in the same manner, except that phosphatase was not added to the reaction mixture. The IC74 isoforms were resolved by two-dimensional gel electrophoresis, visualized by silver staining, and then analyzed by autoradiography. Only the IC74 portion of the gel is shown. A, silver-stained gel of control dynein immunoprecipitated from adult brain; B, autoradiograph of 32 P-labeled control dynein; C, silver-stained gel of dynein treated with phosphatase; D, autoradiograph of 32 P-labeled dynein treated with phosphatase. The arrows point to the B1 spot and the arrowheads point to the A1 spot. After phosphatase treatment, the A2 and B2 spots are not found, and no spots are labeled with 32 P.

FIG. 3. Cultured glia and cortical neurons have different IC74
isoforms. Cultured rat glia and cortical neurons were metabolically labeled with 1 mCi of Trans 35 S-label for 1 h as described under "Materials and Methods." Cytoplasmic dynein from the radiolabeled cultures was immunoprecipitated, resolved by two-dimensional gel electrophoresis, and analyzed by autoradiography. A, the IC74 region of three autoradiographs showing cytoplasmic dynein from glia (left), neurons (center), and a mixture of equal counts/min of dynein from glia and neurons (right). B, 35 S-labeled dynein from glia was mixed with cold carrier dynein from rat brain, and the combined sample was analyzed by two-dimensional gel electrophoresis. The gel was stained with Coomassie Blue and then analyzed by autoradiography. Left, Coomassie Blue-stained gel; right, autoradiograph of the gel showing the position of 35 S-labeled glial dynein IC isoforms. The arrows point to the B1 spot, and the arrowheads point to the poorly resolved A1 spot. The glial spots co-migrate with the B and B1 spots. Cortical neurons, but not glia, have A spots. lesser amount of the A spots compared to dynein from adult brain. These results demonstrate that different cell types from brain have different cytoplasmic dynein IC74 isoforms.
Phosphorylation of Cortical Neuron and Glial Cytoplasmic Dynein IC74 -To further understand the basis for the different IC74 isoforms in neurons and glia, and their relation to the isoforms found in brain, we analyzed the in vivo phosphorylation of the IC74 polypeptides from the cultured cells (Fig. 4). Glia and cortical neurons were metabolically labeled with 32 P, and the cytoplasmic dynein was immunoprecipitated and then analyzed by two-dimensional gel electrophoresis and autoradiography. Analysis of the glia protein gel and corresponding autoradiograph demonstrated that the glial B1 spot was labeled with 32 P (Fig. 4A). There was no 32 P associated with the B spot. Furthermore, no B2 spot or "tail" was seen in cytoplasmic dynein immunoprecipitated from glia. In dynein from cortical neurons, 32 P labeling was observed over the B1 spot, and heavy labeling extended into the region corresponding to the B2 spot (Fig. 4B). Faint labeling was also observed over the A spots of dynein from cortical neurons.
To determine if the IC74 isoform resolved at the B1 spot was a phosphoisoform of the polypeptide resolved at the B spot, glial cytoplasmic dynein was treated with phosphatase in vitro. Cytoplasmic dynein was immunoprecipitated from cultures metabolically labeled with either [ 32 P]orthophosphate or 35 S-labeled amino acids. One half of each of the labeled dynein samples was treated with phosphatase. Then the control and treated samples were analyzed by two-dimensional gel electrophoresis. The results are shown in Fig. 5. When the extent of 32 P-labeling of the glial dynein IC74 subunit before and after phosphatase treatment was quantified, we found that the phosphatase removed ϳ95% of the 32 P from the immunoprecipitated glial dynein (Fig. 5, A and B). Next, the 35 S-labeled IC74 polypeptides were analyzed. Following phosphatase treatment, only the B spot was observed. No B1 spot was found (Fig. 5, C  and D). This shows that, in glial cells, the cytoplasmic dynein polypeptide migrating with the B1 spot is a phosphorylated form of the polypeptide migrating with the B spot. Therefore, in contrast to brain, glial cells have only one unmodified IC74 subunit. Quantitation of the relative amounts of 35 S-label in the B and B1 spots of dynein from cultured glia indicates that only ϳ20% of the glial IC74 polypeptide is phosphorylated at any one time.
When cytoplasmic dynein from cortical neuron cultures was treated with phosphatase the IC74 isoform pattern resembled that of similarly treated dynein from brain (Fig. 5F). Four distinct spots were observed, B, B1, A, A1. Interestingly the spots are more clearly resolved than those observed for the untreated control (Fig. 5E). It is likely that this enhanced resolution results when the protein spread out in the A2 and B2 spots migrates with the other spots after the removal of the phosphate. Similar changes in the resolution of the A spots are observed when embryonic brain dynein, which also has relatively little of the A spots, is treated with phosphatase. 2 Taken together, the phosphatase treatment of cytoplasmic dynein from brain, glia, and neurons indicates that two polypeptides co-migrate with the B1 spot, a phosphate-modified form of the polypeptide resolved at the B spot and an additional polypeptide.
Identification of the Protein in the B Spots as Products of ic74 Gene 2-Two genes for the cytoplasmic dynein IC74 subunit, ic74-1 and ic74-2, and five alternative splice variants of those two genes have recently been identified (19). We therefore sought to correlate the observed two-dimensional gel protein isoforms with the IC74 mRNAs. First, we used an antibody specific for the carboxyl terminus of the ic74-2 gene products to screen blots of brain cytoplasmic dynein resolved on two-dimensional gels. As is seen in Fig. 6A, the anti-IC74-2 antibody reacted with only a single arc of spots, which from their shape appeared to be the B spots. The blot was then probed with the 74.1 antibody. With this antibody an additional IC74 polypep- 2 6. Identification of the IC74 isoform B spots as products of the ic74-2 gene. Cytoplasmic dynein was immunoprecipitated from adult brain, resolved by two-dimensional gel electrophoresis, transferred to poly(vinylidene difluoride) membrane, and probed first with anti-IC74-2 (A), and then without stripping the membrane, reprobed with antibody 74.1 (B). IC-2 reacts well with the B spots, including the B2 spot, but not with the A spots. The A1 spot is detected when the blot is reprobed with 74.1. To maximize the resolution of the spots when the blot was reprobed with antibody 74.1, the time the blot was exposed to film during ECL was kept to a minimum. Therefore, while the A1 spot is clearly visible, it appears smaller in size than on protein gels. The arrows point to the position of the B1 spot, the arrowheads to the A1 spot. The proteins resolved in the B spots are products of gene ic74-2.
tide at the position of the A1 spot was observed (Fig. 6B). Therefore, the IC74-2-specific antibody reacts with the polypeptides of the B arc of spots, not with the A spots. We conclude that the protein resolved in the B arc of spots is a product of gene ic74-2 and its splice variants.
Identification of the IC74 mRNA Transcribed in Cultured Glia and Cortical Neurons Using RT-PCR-To determine which of the ic74 gene messages are expressed in the cultured cells, an RT-PCR procedure based on the alternative splice differences of the five IC74 mRNA sequences was used. Oligonucleotide primers were designed for both gene ic74-1 and gene ic74-2. When used in appropriate combinations in PCR, the primers generate short DNA fragments whose sizes are diagnostic for each of the known alternatively spliced mRNAs (Fig. 7, A and B). To identify the two mRNAs of the ic74-1 gene, one set of primers from each side of the alternative splice region was prepared. As indicated in Fig. 7B, with these two primers the mRNA IC74-1A, which encodes the entire message, will yield a product of 140 bp, while the splice variant IC74-2B gives a product of 77 bp. To identify the three different IC74-2 mRNAs, which are generated by two alternative splice sites, three combinations of primers were utilized. The first set of primers encompasses both of the alternative splice regions. Therefore when they are used in PCR, a different size product for each IC74-2 mRNA is produced; a product of 348 bp results from the full-length IC74-2A mRNA, a product of 330 bp is produced from IC74-2B, which has region 5 spliced out, and a product of 270 bp results from IC74-2C, which has both regions 5 and 7 spliced out. The second and third reaction mixtures use primers that are specific for each of the alternative splice regions, and in combination with the results of the first reaction mixture confirm the identification of the IC74-2 mRNAs. Therefore, the presence of each of the known IC74-2 mRNAs in a particular tissue or cell line can be determined.
The RT-PCR procedure was first used to analyze mRNA isolated from adult rat brain. As seen in Fig. 7C, products of 140 and 77 bp are produced in the reaction designed to identify the gene ic74-1 mRNAs. This is the result predicted for the IC74-1A and IC74-1B mRNAs, respectively, and demonstrates that, as expected, they are present in RNA isolated from adult . Therefore, with this analysis, each IC74 mRNA is uniquely identified. C, ethidium-stained gel of the PCR products produced from adult rat brain mRNA isolated as described under "Materials and Methods," using the primers described in A and B. Lane 1, using the probes for IC74-1, products of 140 and 77 bp are obtained, diagnostic of mRNA IC74-1A and IC74-1B, respectively. The band at Ͼ348 bp is unidentified and presumably an artifact. Lane 2, using probes for IC74-2, products of 348, 330, and 270 bp are obtained, diagnostic of the mRNAs, IC74-2A, -2B, and -2C. The products of lanes 3 and 4 confirm the interpretation of the lane 2 results. mRNA from adult rat brain has messages from each of the five known IC74 splice variants. rat brain. Similarly, when the procedure was used to probe for the IC74-2 messages, products diagnostic of all three IC74-2 mRNAs are found in mRNA from adult brain. Therefore, when the procedure was used to screen adult brain mRNA, PCR products of the size predicted for each of the five known messages for the ic74 genes were obtained, verifying the utility of this procedure for identifying the five known mRNAs for IC74.
When RNA from glia was analyzed by the RT-PCR procedure (Fig. 8A) no message for either IC74-1 mRNA was readily detected. However, a product of the size predicted for mRNA IC74-2C was observed in the analysis with the general IC74-2 primers. Further analysis with the primer reactions specific for each of the alternatively spliced regions confirmed the identification of IC74-2C mRNA in glia. mRNA from cortical neurons contained messages for IC74-1A, IC74-1B, IC74-2B, and IC74-2C (Fig. 8B). The amount of product diagnostic for mRNA IC74-2C was very weak and did not reproduce well photographically. However, it was clearly visible upon further cycles of amplification (not shown). No specific products diagnostic for mRNA IC74-2A were detected using mRNA isolated from neurons. DISCUSSION The cytoplasmic dynein protein complex is believed to contain two or three IC74 polypeptides per molecule (4, 7, 14, 15). However, we previously reported that at least six IC74 isoforms are resolved on two-dimensional gels of dynein isolated from adult rat brain (16). Furthermore, five alternatively spliced IC74 mRNAs have been identified (19). We therefore initiated this study to investigate the relationships of the various IC74 isoforms and their association in the dynein complex. Our analysis has identified several factors which account for the two-dimensional gel IC74 spot pattern. First, individual cell types express different IC74 mRNAs and their polypeptide products. One mRNA, IC74-2C, is expressed in glia, while four mRNAs are found in neurons. Second, a fraction of the molecules in each IC74 polypeptide pool is phosphorylated. In brain and neurons, the A2 and B2 spots originate by the phosphorylation of the other isoforms. Finally, in at least one instance, different IC74 isoforms co-migrate at the same spot on twodimensional gels. This is the first report of differences at the cellular level in the subunits of cytoplasmic dynein from the same species and tissue and demonstrates the existence of different pools of cytoplasmic dynein in neurons.
The cytoplasmic dynein IC74 isoforms from cultured glia and cortical neurons are remarkably different. One IC74 mRNA is transcribed in glia, the IC74-2C message. The polypeptide product of this mRNA migrates at the B spot. We will hereafter refer to this polypeptide as the IC74-2C isoform. Previously we reported that the isoelectric point of the B spot was pH 4.9 (16). This is in reasonable agreement with the value of pH 5.16 calculated by a computer program for the product of the IC74-2C mRNA (Expasy Server, University of Geneva). Approximately 20% of the IC74-2C isoform is phosphorylated in vivo. This produces a phospho-IC74-2C isoform that co-migrates with the B1 spot. Cultured glia show no evidence of the phosphoisoform that migrates with the B2 spot. The two-dimensional gel IC74 spot pattern of dynein from cultured cortical neurons is generally similar to that of dynein from adult brain. Treatment of cytoplasmic dynein from cortical neurons with phosphatase in vitro yielded the same four discrete IC74 spots observed in phosphatase treated brain dynein. However, a lesser amount of the proteins migrating in the A spots was found when cytoplasmic dynein is immunoprecipitated from cultured cortical neurons than was found in adult brain. This is not an artifact of labeling cultured cells with 35 S-labeled amino acids. Rather, the expression levels of several of the neuronal IC74 polypeptides which resolve at the A, A1, B, and B1 spots change during brain development. 2 The relative amounts of the IC74 isoforms of cytoplasmic dynein isolated from cultured cortical neurons, prepared from cerebral cortexes obtained on the 18th day of gestation, are very similar to those of dynein immunoprecipitated from 18th day of gestation brains or cortexes. The RT-PCR assay demonstrated that four of the five known brain IC74 mRNAs are expressed in the cultured cortical neurons. The IC74-2A mRNA was not found in mRNA prepared from cultured cortical neurons. Interestingly, it was also not found in mRNA prepared from embryonic or newborn brain through P5, 2 indicating that the expression of this IC74 isoform is also developmentally regulated, although with a different time period than that of the other neuronal IC74 isoforms.
Analysis of these results allowed us to deduce a correlation between the IC74 mRNA products and the two-dimensional gel spot pattern (summarized in Table I). As shown above, the product of IC74-2C migrates at the B spot, and its phosphoisoform, which is eliminated by phosphatase treatment, migrates at the B1 spot. In contrast, the B1 spot of cytoplasmic dynein isolated from neurons and adult brain is still observed after phosphatase treatment. This indicates that some IC74 isoforms co-migrate on two-dimensional gels. The B1 spot observed FIG. 8. Analysis of glial and neuronal 74-kDa IC mRNA by RT-PCR. Glia has mRNA only for IC74-2C. As described under "Materials and Methods," mRNA was isolated from cultured cells and RT-PCR with oligo(dT) was used to produce cDNA from the RNA. PCR using the oligonucleotides described in Fig. 7 was used to generate different sized products, diagnostic for each of the mRNAs. A, RNA from cultured glia. Lane 1, analysis of IC74-1 products, none are discernable. when dynein is isolated from adult rat brain or cultured neurons must contain an unphosphorylated polypeptide as well as the phospho-IC74-2C isoform. Since the antibody specific for ic74-2 gene products recognizes all the B spots, this neuronal IC74 polypeptide co-migrating with the B1 spot is most likely the product of an IC74-2 mRNA. Two IC74 mRNAs are expressed in cortical neurons, IC74-2C and IC74-2B. Therefore, it is likely that the unphosphorylated polypeptide co-migrating with the B1 spot is the product of the IC74-2B mRNA. Evidence supporting this conclusion comes from the demonstration that the IC74-2B message is found in all tissues that have a B1 spot after phosphatase treatment, including those with no IC74-1 messages. 2 We were unable to identify the B2 spot in analyses of either 35 S-labeled or 32 P-labeled cytoplasmic dynein from glia. This suggests that phosphorylation of the IC74-2C isoform does not generate the B2 spot. The B2 spot was observed in cytoplasmic dynein from neurons. It therefore appears likely that the B2 spot is the phospho-IC74-2B isoform. However, the possibility that a cortical neuron kinase, not present in glia, generates the B2 spot by hyperphosphorylating the IC74-2C isoform cannot be ruled out. The observation that polypeptides migrating in the A arc of spots are not detected by the IC74-2-specific antibody suggests that the polypeptides resolved in the A arc of spots are the products of the two IC74-1 mRNAs. Given that cortical neurons have both the IC74-1A and IC74-1B messages, it appears likely that one corresponds to the protein of the A spot and the other to the A1 spot. This was confirmed by further study. 2 The polypeptide resolved at the A spot is the IC74-1B isoform and the A1 spot is the IC74-1A isoform.
The best studied dynein ICs are the outer arm dynein ICs of Chlamydomonas flagella, which serve as a model for understanding the role of the cytoplasmic dynein ICs (8,9,12,13,22). There are two ICs per outer arm dynein, IC68 and IC79 (22,23). These are the products of distinct, although related genes (9 -11). The two ICs associate with one another are located at the cargo binding end of the dynein molecule and contribute to the assembly of the dynein complex (12,13,22). The flagellar ic69 and ic78 genes are related to the ic74 genes of cytoplasmic dynein (8,9). The flagellar IC78 binds the dynein complex to its sole cargo, the A microtubule (12). Interestingly, only one isoform of each IC is observed on two-dimensional gels (24,25). Since the cargo of cytoplasmic dynein includes a variety of membranous organelles, it seemed possible that the different IC74 isoforms were involved in binding cytoplasmic dynein to specific organelles. However, glia express only the IC74-2C mRNA. Therefore, the presence of only the IC74-2C polypeptide in the dynein complex is sufficient for the constitutive dynein-powered movement of membranous organelles including mitochondria, endosomes, and endoplasmic reticulum-Golgi traffic in cultured cells (26 -30). The results presented here further demonstrate that, unlike flagellar dynein, the products of both ic74 genes are not necessary for general cytoplasmic dynein function.
The observation that 20% of the single IC74 polypeptide found in glia, the IC74-2C isoform, is phosphorylated is consistent with our previous identification of differences in the phosphorylation of cytoplasmic dynein associated with anterograde organelles and that of the whole cell pool (16). Since cytoplasmic dynein from glia has a single IC74 polypeptide, the study of glial dynein should simplify further investigations into the role of phosphorylation of the IC74 subunit on the functional properties of dynein. Recent studies suggest that phosphorylation of the kinesin heavy and light chains in vivo correlates with the association of kinesin with membranous organelles (31,32). Genetic studies on the outer arm flagellar dynein support the hypothesis that the IC subunits regulate dynein function (22). Furthermore, it was recently realized that the conserved portions of the flagellar and cytoplasmic genes are a series of WD repeats in the COOH-terminal portion of all three molecules (9,10). WD repeats are believed to be important for subunit-subunit interactions in protein complexes and, interestingly, all the other proteins with WD repeats are regulatory proteins (33).
Since dynein-based membrane traffic in a glial cell functions with only the ic74-2c gene product, the roles of the three additional neuronal IC74 polypeptides and their phosphoisoforms in cytoplasmic dynein function remain to be determined. Interestingly, multiple isoforms of both the heavy and light chains of bovine brain kinesin have also been identified on one-and two-dimensional gels (34). In an analysis of kinesin from rat brain, Cyr et al. (35) identified a single kinesin light chain gene and three alternative splice variants and suggested that the alternative isoforms may be involved in binding to different organelles. Subsequent work on the light chains from other species has yielded similar results (36 -38). Elluru et al. (39) have examined the anterograde axonal transport of the two kinesin heavy chain isoforms. They find that one kinesin heavy chain isoform is predominantly associated with anterogradely moving synaptic vesicles, while the other is associated with mitochondria. While two of the neuronal IC74 polypeptides are found in at least one other tissue, one is specific to neurons. 2 Cytoplasmic dynein has two to three IC74 subunits per molecule. Therefore, there must be distinct populations of cytoplas-  Fig. 7 and Vaughan and Vallee (19). Each IC74 mRNA produces a protein product which is given the same name and its phosphoisoform. The alphanumeric designations of the IC74 2D gel spots are from Fig. 1B and Dillman and Pfister (16). The alignment of the IC74 -2B and -2C mRNA protein products and their phosphoisoforms with the indicated two-dimensional gel spots is described in the text. The identification of the other mRNA products with two-dimensional gel spots is from unpublished results, 2 except that the position of the IC74 -2A protein product is derived from a computer calculation (Expasy Server, University of Geneva) of its pI and molecular mass from the IC74 -2A mRNA sequence. mic dynein in neurons. It is tempting to speculate that the additional IC74 isoforms may confer a regulatory or functional specificity on the dynein complex needed in neurons and some other cells, but not glia. Neurons are specialized for long distance axonal transport, and the presence of additional IC74 isoforms may be related to the specialized cargo moved in retrograde axonal transport, or the regulation of retrograde axonal transport. Interestingly, we find that all the neuronal dynein IC74 isoforms are present in axons (16). However, understanding the role of the various neuronal cytoplasmic dynein IC74 isoforms remains a challenge. The demonstration of basic differences in the IC74 subunits of cytoplasmic dynein from glia and neurons should also raise a cautionary note concerning biochemical studies of dynein from mammalian brain, as dynein from this source is a mixed population of the motor protein.