Role of Cytoplasmic in Melanosome Transport in Human

H. Randolph Byers, Mina Yaar, Mark S. Eller, Nicole L. Jalbert, and Barbara A. Gilchrest Department of Dermatology, Boston University School of Medicine, Boston, Massachusetts, U.S.A.

Cytoplasmic dynein is a -associated transport of melanosomes from the body into the retrograde-directed motor molecule for transport of dendrites, whereas melanosome distribution was not membrane-bound . To determine whether affected in sense-treated melanocytes. To determine cytoplasmic dynein is expressed in melanocytes, we whether irradiation modi®es cytoplasmic performed reverse transcriptase polymerase chain dynein expression, cultures were reaction using melanocyte cDNA and primers com- exposed to increasing doses of solar-simulated irra- plementary to human brain cytoplasmic dynein diation, equivalent to a mild to moderate sunburn heavy chain. A polymerase chain reaction product of exposure for intact skin. Within 24 h, doses of 5 and the expected molecular size was generated and the 10 mJ per cm2 induced cytoplasmic dynein protein, identity was con®rmed by sequence analysis. whereas doses of 30 mJ per cm2 or more were asso- Western blotting of total melanocyte proteins reacted ciated with decreased levels of cytoplasmic dynein with an anti-intermediate chain cytoplasmic dynein compared with sham-irradiated controls. Our data antibody identi®ed the appropriate 74 kDa band. To show that cytoplasmic dynein participates in retro- determine whether cytoplasmic dynein plays a role grade melanosomal transport in human melanocytes in melanosome transport, duplicate cultures were and suggest that the altered melanosomal distribu- treated with cytoplasmic dynein antisense or sense tion in skin after sun exposure is due, at least in (control) oligodeoxynucleotides and the cells were part, to decreased cytoplasmic dynein levels result- observed by high-resolution time-lapse microscopy, ing in augmented anterograde transport. Key words: which allows visualization of melanosomal aggre- cytoplasmic dynein/melanocytes/melanosome/ultraviolet gates and individual melanosomes. Antisense-treated irradiation. J Invest Dermatol 114:990±997, 2000 melanocytes demonstrated a strong anterograde

kin color and ultraviolet (UV) photoprotection are largely an important motor in actin-®lament-based transport of organelles dependent on content and distribution. Melano- (Mehta et al, 1999). Mutations of myosin V in dilute mice (Mercer cytes synthesize melanin within membrane-bound et al, 1991) and in patients with Griscelli PrunieÂras syndrome organelles called melanosomes that are transported from (Griscelli et al, 1978; Pastural et al, 1997) result in altered skin and Sthe perinuclear region, their site of origin, into the hair due to reduced melanosome transport into melanocyte melanocyte dendrites and eventually into the adjacent dendrites (Provance et al, 1996; Wu et al, 1997, 1998; (Mottaz and Zelickson, 1967; Cohen and Szabo, 1968; Klaus, Lambert et al, 1998). The majority of the family motor 1969; Wolff et al, 1974; Jimbow et al, 1976). As with all membrane- proteins are plus-end directed microtubule-activated ATPases that bound organelles, the transport of melanosomes is mediated by the are important in fast axoplasmic or anterograde transport of net sumof centripetal and centrifugal cytoskeletal forces acting organelles in neurons (Brady et al, 1982; Brady, 1985; Schnapp et al, upon the . Intracellular transport of organelles in all cells is 1985; Vale et al, 1985; Bloom et al, 1988). Kinesin functions in all directed by two major polarized cytoskeletal macromolecular cells to transport , mitochondria, and other cargo polymers, actin ®laments and , composed of actin membrane-bound organelles (Hirokawa, 1998), and recent work monomers and a:b-tubulin dimers respectively. Organelles are from our laboratory suggests a role for kinesin in melanosome shuttled along actin ®laments and microtubules by certain members transport in melanocytes (Hara et al, 2000). Cytoplasmic dynein, of three major families of adenosine triphosphatase (ATPase) another microtubule-activated motor protein, is believed to chemokinetic motor proteins: (Hirokawa, 1998), transport organelles such as and multivesicular bodies (Hirokawa, 1998), and myosins (Mermall et al, 1998). Myosin V is in the opposite direction to kinesin in neurons (retrograde transport) (Lye et al, 1987; Paschal et al, 1987; Schnapp and Reese, Manuscript received November 2, 1999; revised January 21, 2000 1989; Schroer et al, 1989). The transport of organelles in all cells is accepted for publication February 1, 2000 thus believed to be controlled in part by the net balance of the Reprint requests to: Dr. H. Randolph Byers, Department of myosin V actin-®lament-based transport, the kinesin micro- Dermatology, Boston University School of Medicine, 609 Albany St, tubule-based motors, and the opposing cytoplasmic dynein Boston, MA 02118. Email: [email protected] microtubule-based motor. The focus of our current investigation Abbreviations: Dyh1, DHC1, cytoplasmic dynein heavy chain 1; Dyh2, DHC2, cytoplasmic dynein heavy chain 2; DHC3, cytoplasmic dynein is to determine the role of cytoplasmic dynein in melanosome heavy chain 3. transport in melanocytes.

0022-202X/00/$15.00 ´ Copyright # 2000 by The Society for Investigative Dermatology, Inc. 990 VOL. 114, NO. 5 MAY 2000 CYTOPLASMIC DYNEIN AND MELANOSOME TRANSPORT 991

Cytoplasmic dynein is a multimeric high-molecular mass Diego, CA). Digital images were taken every 5 s to every 30 s for a duration complex of two heavy chains, three intermediate chains, and four of several minutes at 0, 5, 10, 24, and 48 h following sense or antisense light intermediate chains. The ATPase mechanochemical transduc- oligonucleotide treatment or 24 h after solar-simulated UV exposure. The tion residues are found in two cytoplasmic dynein heavy chains of 5±30 s intervals permitted study of bidirectional individual melanosome migration and the hour intervals permitted evaluation of net displacement approximately 530 kDa each (Holzbaur and Vallee, 1994; Schroer, of melanosomes. Video frames were captured with an LG-3 scienti®c frame 1994). Molecular cloning and sequencing of the most abundant, grabber card (Scion, Frederick, MD) and analyzed for transport rates and ubiquitous cytoplasmic dynein heavy chain (Koonce et al, 1992; pigment distribution using IP lab spectrum software (Scanalytics, Mikami et al, 1993; Zhang et al, 1993) has determined it to be Fairfax, VA). relatively conserved from amoeba to mammalian brain. Partial sequence data and biochemical evidence has led to the identi®ca- Northern blot analysis Rat brain cytoplasmic dynein heavy chain has tion of at least three other cytoplasmic heavy chain dyneins (Tanaka been cloned and sequenced by two independent investigators (Mikami et al, et al, 1995; Vaisberg et al, 1996; Criswell and Asai, 1998). Recent 1993; Zhang et al, 1993). A 400 nucleotide partial sequence of the cytoplasmic dynein heavy chain from human fetal brain was obtained from evidence indicates that the 74 kDa intermediate chains and the light Genbank (Accession T05469 or NCBI GI: 316619). This nucleotide intermediate chains (approximately 55 kDa) participate in the sequence exhibits overlap with the cloned rat brain or cytoplasmic dynein binding to the protein complex dynactin, a macromolecular heavy chain 1 (DHC1) isoformbut not the cytoplasmicdynein heavy chain complex linked to the membranes of transported organelles (Gill 2 (DHC2) or cytoplasmic dynein heavy chain 3 (DHC3) isoforms et al, 1991; Schafer et al, 1994; Karki and Holzbaur, 1995; Vaughan (Vaisberg et al, 1996). Total RNA was isolated fromcultured human and Vallee, 1995). Cytoplasmic dynein functions as a chemo- melanocytes using the Tri-reagent (Gibco-BRL), and RNA concentrations mechanical motor that ratchets along the a- and b-tubulin dimers were determined as described previously by absorbance at 260 nm within microtubules toward the minus-end, pulling along the (Goukassian et al, 1999). The purity of the RNA was determined by the dynactin linkage mechanism that is attached to the membrane- ratio of 260/280 readings and was consistently over 1.8 (Chomczynski and Sacchi, 1987). The cDNA was generated using the Molony leukemia virus bound organelle. reverse transcriptase (Pharmacia LKB Biotech, Piscataway, NJ). Degenerate Based on research in a variety of eukaryotic cells, transport of primers ¯anking a 600 bp region in the non-microtubule binding domain membrane-bound pigment granules or melanosomes into melano- and non-conserved region near the 5¢ end in the rat brain sequence were cyte dendrites is hypothesized to be controlled in part by the two prepared. Polymerase chain reaction (PCR) was then performed with opposing motors dynein and kinesin prior to pigment transfer to human melanocyte cDNA. Primers were also constructed framing the keratinocytes. For example, Western blots of melanosomes isolated 400 bp sequence of human fetal brain cytoplasmic dynein and PCR was from ®sh melanophores show enrichment of cytoplasmic dynein performed with human melanocyte cDNA. The ampli®ed products and kinesin-like proteins compared with whole lysates of cultured obtained were puri®ed and sequenced to con®rmidentity. Then the melanophores (Rogers et al, 1997); and microinjection of anti- cDNA was radiolabeled with 32P as previously (Eller et al, 1992). Brie¯y, radiolabeling of cDNA was carried out using [32P]dCTP (New England dynein (Nilsson and Wallin, 1997) and anti-kinesin (Rodionov et al, Nuclear, Boston, MA) and the oligonucleotide primed DNA labeling 1991) antibody inhibits respectively minus-end and plus-end system (Pharmacia Fine Chemicals, Piscataway, NJ). The unincorporated microtubule-directed melanosome transport. In order to investigate [32P]dCTP was separated fromthe labeled DNA by centrifugation through the possible role of cytoplasmic dynein in melanosome transport in a mini-spin column (G50; Worthington Biochemical, Freehold, NJ). human melanocytes, we used northern blots, Western blots and Electrophoresis of 20 mg of melanocyte RNA was performed on 1% agarose immuno¯uorescent techniques to detect cytoplasmic dynein and its gels containing 2.2 M formaldehyde as previously described (Yaar et al, intermediate chain and to elucidate its distribution in melanocytes. 1991). Similarly, foreskin human ®broblast mRNA was harvested and We used sense and antisense oligonucleotides in time-lapse electrophoresed as controls, as published northern blots have shown that a functional assays to test their effect on melanosome transport. variety of non-neural tissues and tissue culture cells including foreskin and cultured ®broblasts exhibit cytoplasmic dynein (Vaisberg et al, 1996). RNA Finally we performed studies on whether a component of was transferred to nylon membranes (Hybon, Amersham, Arlington cytoplasmic dynein important for linkage to organelle membranes Heights, IL) and then immobilized by short wave UV irradiation (UV is modulated by solar-simulated UV light exposure. Strata Linker 1800; Stratagene, La Jolla, CA) and ®nally hybridized with the 32P-labeled cytoplasmic dynein cDNA probe. Hybridization, autoradio- MATERIALS AND METHODS graphy with XAR ®lm(EastmanKodak, Rochester, NY), and development after overnight exposure at ±70°C was performed as Melanocyte culture Human melanocytes were cultured from neonatal described previously (Yaar et al, 1994). foreskin as described previously (Gilchrest et al, 1984; Park et al, 1993) with modi®cations. Brie¯y, after cutting the foreskin into 1 mm wide strips and Western blot analysis Total cellular proteins of human melanocytes incubating for 45 min at 37°C and then overnight in 0.25% trypsin at 4°C, were harvested in a lysis buffer containing 20 mM tri(hydroxymethyl)- the was separated from the dermis. Cells were initially established aminomethane (Tris), 2 mM ethylenediamine tetraacetic acid, 0.5 mM in Medium199 (Gibco BRL, Grand Island, NY) supplementedwith 10 mg dithiothreitol, 10 mg aprotinin per ml, 50 mM okadaic acid, and 1 mM epidermal growth factor per ml (Bethesda Research Laboratories, phenylmethylsulfonyl ¯uoride. Similarly, cutaneous human ®broblasts were Gaithersburg, MD), 10±9 M triiodothyronine (Sigma Chemical, St Louis, harvested for controls as published immunoblots have shown that a variety MO), 10 mg transferrin per ml (Sigma), 1.4 3 10±6 M hydrocortisone of non-neural tissue and tissue culture cells including ®broblasts express the (Calbiochem-Behring, La Jolla, CA), 10 mg insulin per ml (Sigma), 10 mg 74 kDa intermediate cytoplasmic dynein chain (Lin and Collins, 1993; Lin basic ®broblast growth factor per ml (Amgen, Thousand Oaks, CA), 10±9 M et al, 1994; P®ster et al, 1996a). Twenty or 50 mg protein were loaded per cholera toxin (List Biological, Campbell, CA), and 5% fetal bovine serum lane, separated by 10% sodiumdodecyl sulfate polyacrylamidegel (FBS). Post-primary cultures were maintained in a similar medium but with electrophoresis, and transferred to nitrocellulose membrane at 100 V for low calcium (<0.03 mM) to selectively support melanocyte growth 1 h. To con®rmequal loading of the lanes in the UV experimentswhere (Naeyaert et al, 1991). Sub-con¯uent melanocytes at passages 3 to 6 were 50 mg protein were loaded per lane, the lower portion of the membrane used for all experiments. For video time-lapse image analysis or solar was cut off and stained with amido black solution (0.5% amido black, 45% simulator studies cells were seeded on 35 mm petri dishes and for methanol, 97% acetic acid) for 1 min. It was then destained with 45% immuno¯uorescent studies cells were seeded on eight-chamber slides methanol, 9% acetic acid, and 46% water for 1 h by agitation at room (Permanox Tissue Tek, Nunc, Naperville, IL). temperature. Following overnight blocking with 5% nonfat dry milk/ bovine serumalbumin(BSA) in Tris±phosphate-buffered saline (PBS) Video time-lapse and image analysis Intracellular motility of (0.5% Tween-20 in PBS) at 4°C, the membranes or upper portion of the melanosomes was analyzed by time-lapse computerized microscopy as membranes in the UV experiments were incubated overnight with a described previously, with modi®cations (Byers et al, 1991). Brie¯y, monoclonal antibody against the cytoplasmic dynein 74 kDa intermediate melanocytes were viewed with the Nikon MicroPhot inverted microscope chain (monoclonal 74.1 Chemicon, Temecula, CA) (Dillman and P®ster, equipped with a Plexiglas housing, Nikon incubator, and CO2±air¯ow 1994) at a dilution of 1:1000. After incubation, the membrane was brie¯y mixer to maintain constant pH and temperature (37°C). Intracellular washed in PBS, then three times with Tris±PBS, and incubated with a transport of melanosomes was recorded using the 403 objective under secondary goat antimouse antibody at 1:1000 dilution (Pierce, Rockford, bright ®eld illumination with a Cohu high performance CCD camera (San IL) in 1% milk/0.2% BSA for 1 h. After extensive washing the membrane 992 BYERS ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY was incubated for 1 min with the enzyme-linked chemiluminescence Western Blotting Detection Reagent (Amersham, Buckinghamshire, U.K.), and bands were detected on pre¯ashed Kodak X-Omat ®lm and developed after several seconds of exposure.

Immuno¯uorescent staining and confocal laser scanning microscopy Melanocytes cultured for several days in eight-chamber slides (described above) were ®xed with 10% formalin for 15 min at room Figure 1. Schematic representation of the protein binding regions temperature, permeabilized for 2 min with 0.2% Triton X-100, and then of the cytoplasmic dynein heavy chain. The organelle linkage protein incubated with 5% normal goat serum for 10 min to block nonspeci®c region represented by the dark shaded region (A) on the amino end is binding sites. The cells were incubated for 1 h at 37°C with the monoclonal speci®c to cytoplasmic dyneins. Within this region was the site of PCR antibody against the intermediate chain (74 kDa; see above) at a dilution of ampli®cation. This cDNA probe construction site corresponded to the 1:200 in PBS with 0.5% BSA. After several rinses with PBS, goat antimouse human brain cytoplasmic dynein cDNA sequence (Genbank Accession ¯uorescein isothiocyanate (FITC) (Sigma) was added at a dilution of 1:100 T05469 or NCBI GI: 316619). Within this site two antisense in PBS and 0.5% BSA for 1 h at 37°C. Control preparations included the oligonucleotides were prepared in order to inhibit cytoplasmic dynein FITC-conjugated goat antimouse IgG alone without the primary antibody mRNA translation in cultured melanocytes. Sense oligonucleotides were and isotype-matched unrelated primary antibody (CD5, Sigma). We also also prepared to serve as controls. The 74 kDa intermediate chain involved used polyclonal antibody to cytoplasmic dynein (gift of C. Collins) (Lin and in linkage to dynactin±membrane-bound organelles binds within the dark Collins, 1992, 1993; Lin et al, 1994) and confocal immuno¯uorescence shaded regions of the brain cytoplasmic dynein heavy chain. Antibodies microscopy to assess the effect of antisense and sense DNA on cytoplasmic against this intermediate chain were used to characterize the cytoplasmic dynein expression. After washing, the cells were mounted with cover slips dynein-linked organelle distribution in human melanocytes. The light using the slow-fade kit (Molecular Probes, Eugene, OR). The preparations shaded region (B) on the carboxyl end represents the microtubule binding were examined with a Leica confocal laser scanning microscope equipped region common to both cytoplasmic and axonemal (in ¯agella and cilia) with an argon ion laser with an output power of 2±50 mW and a 253 dyneins. The lightest region between (A) and (B) contains four ATP (n.a. = 0.75) objective. Digitized images were stored on an optical disk. binding sequences.

Preparation of sense and antisense oligonucleotides Oligo- nucleotides were obtained fromQuality Controlled Biochemicals (Hopkinton, MA). Sulfur-modi®ed phosphagen linkages stabilized the oligonucleotides. Two antisense oligonucleotides were made based on the human fetal brain cytoplasmic heavy chain sequence (Genbank Accession T05469 or NCBI GI: 316619): 5¢-TTTCAGCTTGCATATCCCA- TAA-3¢, corresponding to the complement of nucleotides 149±170, and 5¢-TGCATTGTCAAAGGTTCCTCT-3¢, corresponding to the com- plement of nucleotides 239±259. These nucleotide positions correspond to nucleotides 4512±4533 and 4602 and 4622 respectively of the fully sequenced 15,500 nucleotides coding for the rat brain cytoplasmic dynein heavy chain (Zhang et al, 1993). Sense oligonucleotides were also ordered representing the exact RNA sequence.

Delivery of oligonucleotides in culture and functional analysis Twelve to 24 h after plating onto 35 mm dishes, melanocytes were given serum-free melanocyte growth medium and supplemented with 50 mM sense or antisense oligonucleotides. Every 12 h an additional 25 mMof oligonucleotide was added to the medium. Time-lapse studies on melanosome transport and dendrite changes were recorded from 5 to 30 s intervals with duration of observation for several minutes at 0, 5, 24, and 48 h. The distribution of melanosomes was recorded in digital images and the perinuclear and peripheral cytoplasmwas analyzed by densitometry in multiple cells (80 cells treated with antisense and 80 cells treated with sense oligonucleotides; see statistical method section below). Figure 2. Primers for human brain cytoplasmic dynein amplify the UV irradiation The solar simulator consisted of an irradiation source appropriately sized reverse transcriptase PCR product using (LPS 255HR with Universal arc-lamp power supply; Spectral Energy, melanocyte (Mc) cDNA. The expected 400 bp band was identi®ed Westwood, NJ). Adjustment of irradiance to 2.6 3 10±4 W per cm2 was using agarose electrophoresis and a 100 bp incremental standard (std). metered at 285 6 5 nmwith a UVB probe and radiometer(detector SSE 240, radiometer model Il11700A; International Light, Newburyport, MA) that provides a spectral output highly similar to sunlight (Werninghaus brain cytoplasmic dynein heavy chain has been cloned and et al, 1991). Sub-con¯uent melanocytes from several different donors were sequenced (Mikami et al, 1993; Zhang et al, 1993), and a partial either shamirradiated or exposed to 5, 10, 20, 30, or 40 mJper cm 2. Cells were irradiated in PBS to avoid formation of toxic photoproducts in the sequence isolated fromhumanfetal brain tissue (Genbank medium and then were re-fed the original medium. Cells were harvested Accession T05469 or NCBI GI: 316619) shows high homology 12 or 24 h later and Western immunodetection for cytoplasmic to rat brain cytoplasmic dynein. Primer selection was made in the intermediate chain was performed as described above. non-microtubule binding region in the intermediate chain and presumptive organelle linkage protein binding region (Fig 1) Statistical analysis The confocal and treatment data were subjected to (Holzbaur and Vallee, 1994; Schroer, 1994; Hirokawa, 1998) statistical analysis using Statview (SAS Institute, Cary, NC). Western blot near the N-terminus. This region is speci®c for cytoplasmic dynein densitometry data and perinuclear and peripheral pigment distributions heavy chain, in contrast with the microtubule-binding domain, were analyzed using Student's t test, and pigment distribution and confocal which is shared by 12 ¯agellar and cilial dyneins (Vaisberg et al, ¯uorescence for cytoplasmic dynein was analyzed by regression analysis. 1996; Milisav, 1998). Degenerate primers from the rat brain sequence framing the 600 bp sequence failed to amplify a product RESULTS from human melanocyte cDNA, whereas speci®c primers to the Identi®cation of human brain cytoplasmic dynein heavy human brain sequences ampli®ed a single band at the expected chain in human melanocytes To determine whether molecular size of 400 bp (Fig 2). cytoplasmic dynein heavy chain is expressed in melanocytes, In order to con®rm that human melanocytes express cytoplasmic cDNA probes speci®c to cytoplasmic dynein were prepared. Rat dynein heavy chain mRNA, the 400 bp band was puri®ed and VOL. 114, NO. 5 MAY 2000 CYTOPLASMIC DYNEIN AND MELANOSOME TRANSPORT 993

Figure 3. Sequence of ampli®ed 400 bp frag- ment obtained from human melanocytes exhibits homology with brain cytoplasmic dynein. The original rat brain cytoplasmic dynein sequence (top rows) is compared with human brain cytoplasmic dynein (middle rows) and the human melanocyte cytoplasmic dynein (bottom rows). The light shaded bars indicate sequence differences between rat and human cytoplasmic dynein whereas the open bars indicate new sequence data of human cytoplasmic dynein obtained from sequencing the melanocyte cytoplasmic dynein. The human melanocyte sequence (Genbank Accession: BankIt320710 AF234785) and rat sequence do not have an extra nucleotide (dark bar) reported in the human brain sequence that spuriously encodes a premature stop codon.

Figure 5. The intermediate chain of cytoplasmic dynein is present in cultured human melanocytes. Immunoblot analysis of lysate from Figure 4. Northern blot analysis of total melanocyte (Mc) RNA and cultured melanocytes (Mc) reveals a single band at a relative mobility (Mr) the 400 bp cytoplasmic dynein probe detects the presence of of 74 kDa corresponding to the 74 kDa intermediate chains of cytoplasmic melanocyte cytoplasmic dynein. The 32P radiolabeled probe hybridized dynein. Lanes 1 and 2 were loaded with 50 mg and 20 mg respectively of to cultured Mc RNA at a relative position of 15 kb, the expected size of the human melanocyte protein. Control ®broblast (Fb) lysate known to contain cytoplasmic dynein transcript. Control ®broblast (Fb) RNA known to 74 kDa intermediate chains of cytoplasmic dynein also shows a single band contain cytoplasmic dynein transcripts also shows the probe hybridizes at at the expected relative mobility of 74 kDa. the expected relative position of 15 kb. sequenced. Comparison with the rat brain cytoplasmic dynein human melanocytes, we performed Western immunoblot on heavy chain partial sequence and the human fetal brain homologous lysates of human melanocytes using a monoclonal antibody sequence revealed 85% nucleotide homology (Fig 3) and 98% against the 74 kDa intermediate chain of cytoplasmic dynein deduced amino acid homology, respectively. There was 99% (P®ster et al, 1996b). A single band at 74 kDa was identi®ed (Fig 5), nucleotide homology between the fetal brain sequence and the con®rming that human melanocytes express the 74 kDa melanocyte PCR product, demonstrating that melanocytes express intermediate chain of cytoplasmic dynein. Control ®broblasts also the cytoplasmic dynein heavy chain sequence. exhibit a single 74 kDa band as reported previously (Lin and The 400 bp human melanocyte probe was radiolabeled and total Collins, 1993; Lin et al, 1994; P®ster et al, 1996a). human melanocyte RNA from melanocyte cultures was isolated and analyzed by northern blotting. A strong, single band at 15 kb The intermediate chain of cytoplasmic dynein in human was identi®ed, consistent with the reported cytoplasmic dynein melanocytes has a punctate cytoplasmic distribution In heavy chain mRNA (Fig 4). Control ®broblast RNA also revealed order to determine the distribution of cytoplasmic dynein in human an expected 15 kb band as reported previously (Vaisberg et al, 1996) melanocytes, we performed immuno¯uorescent confocal scanning when the radiolabeled 400 bp probe was used. microscopy. Monoclonal antibodies to the 74 kDa intermediate chain revealed punctate localization throughout the melanocyte Cytoplasmic dynein intermediate chains are expressed in , whereas control preparations with an unrelated isotype- human melanocytes Three cytoplasmic dynein 74 kDa matched antibody and secondary ¯uorescent labeled antibody did intermediate chains are believed to indirectly link the cytoplasmic not show speci®c staining (Fig 6). Staining intensity for the dynein heavy chains to the dynactin complex that is attached to the intermediate chain of cytoplasmic dynein was greater in the membrane of the organelle undergoing transport (Hirokawa, perinuclear region, but punctate distribution was also detected 1998). In order to determine whether the 74 kDa intermediate throughout the melanocyte dendrites (Fig 6a, inset). Similarly chains, which formpart of the linkage mechanismbetween polyclonal antibodies against cytoplasmic dynein also demonstrated cytoplasmic dynein heavy chain and organelles, are expressed in punctate staining throughout the melanocyte cytoplasm with 994 BYERS ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY increased perinuclear staining (data not shown). The increased processes than sense-treated controls (Fig 7). High-resolution perinuclear staining with both antibodies correlated (p < 0.05) time-lapse microscopy of lightly pigmented cells permitted with the increased perinuclear distribution of melanosomes visualization of the movement of individual melanosomes. At 12 as determined by confocal ¯uorescence image analysis, to 24 h, sense-treated cells showed normal perinuclear densitometry of bright ®eld microscopy, and regression analysis concentration of melanosomes whereas antisense-treated cells (data not shown). with either of the two antisense sequences demonstrated progressive loss of perinuclear melanosome concentration and net Cytoplasmic dynein plays a role in retrograde melanosomal anterograde transport of melanosomes throughout the cytoplasm transport To determine the role of cytoplasmic dynein in and towards the cell periphery (Fig 7). Surprisingly, saltatory, retrograde (minus-directed) melanosome transport, melanocytes bidirectional melanosome motility (anterograde and retrograde) were maintained and recorded by time-lapse microscopy at 5 and continued regardless of treatment, although the melanosomes were 30 s intervals for several minutes at 0, 5, 10, 24, and 48 h after generally more dispersed in antisense-treated cells compared with treatment with diluent alone (control), sense (control) or diluent and sense-treated cells. Melanocyte cultures were then two different antisense cytoplasmic dynein heavy chain ®xed, immuno-stained with the antibody against cytoplasmic oligonucleotides. Cell morphology and distribution of melano- dynein, and analyzed by confocal laser scanning microscopy. somes were analyzed by time-lapse microscopy (Fig 7). During Confocal laser scanning microscopy and ¯uorescent quanti®cation time frames up to 10 h after stimulation, diluent-treated and sense- of cytoplasmic dynein labeling showed that ¯uorescent intensity treated cells showed normal perinuclear aggregates whereas was reduced more than 80% 24 h after antisense treatment antisense-treated cells showed dissolution of perinuclear pigment compared with sense- or diluent-treated cells. Densitometric aggregates and net anterograde transport of pigment into the analysis of bright ®eld digitized images was also performed to peripheral dendrites (data not shown). No signi®cant changes in quantify pigment distribution in perinuclear versus peripheral cell morphology were detected up to 12 h; however, after 24 h cytoplasm. Eighty cells in each treatment category were analyzed. antisense-treated melanocytes appeared to exhibit fewer cell Antisense treatment signi®cantly reduced perinuclear pigment localization and signi®cantly increased peripheral pigment distribution compared with sense-treated controls (Fig 8), indicating a net increase in anterograde displacement of melanosomes towards the cell periphery. UV irradiation modulates cytoplasmic dynein intermediate chain expression UV irradiation is known to increase skin pigmentation by increasing melanocyte pigment production and melanosome transfer to keratinocytes (Jimbow et al, 1976; Friedmann and Gilchrest, 1987). To determine the effect of UV irradiation on cytoplasmic dynein levels we examined the protein level for the 74 kDa intermediate chain of cytoplasmic dynein known to link cytoplasmic heavy chain and the dynactin complex to membrane-bound organelles (Steffen et al, 1997), presumptively melanosomes in melanocytes. At lower UV doses of 5 and 10 mJ Figure 6. Intermediate chain of cytoplasmic dynein is detected per cm2, equivalent to mild to moderate sunburn exposures throughout cultured melanocyte cytoplasm in a punctate pattern (Friedmann and Gilchrest, 1987), we consistently observed and is increased in the perinuclear region. (A) Confocal immuno- ¯uorescent microscopy reveals punctate intermediate chain staining upregulation of cytoplasmic dynein 3 d after irradiation. In 2 throughout the cytoplasmwith greatest intensity in the perinuclear region contrast, at higher UV doses of 30 mJ per cm or more, protein (inset). Note that punctate staining is also detected distally in dendrites. (B) level decreased (Fig 9) and was as low as 30% of sham-irradiated Control confocal immuno¯uorescent staining with secondary antibody controls. Response to intermediate UV doses was variable, as alone shows no signi®cant staining of the cytoplasm. Scale bar:50mm. expected for a biphasic response in multiple donors.

Figure 7. Antisense DNA complementary to cytoplasmic dynein sequence disperses peri- nuclear melanosomes. Sense DNA treated melanocytes show normal perinuclear distribution of melanosomes at 12 h (a) and 24 h (c). In con- trast, antisense DNA treated cells show a peri- nuclear melanosome-free zone (arrow) with increased peripheral membrane pigment distribu- tion at 12 h (b), and loss of perinuclear melano- somes and marked net anterograde dispersion of melanosomes throughout the cytoplasm at 24 h (d). Scale bar:5mm. VOL. 114, NO. 5 MAY 2000 CYTOPLASMIC DYNEIN AND MELANOSOME TRANSPORT 995

DISCUSSION the concentrations tested. These ®ndings suggest that at least the majority of microtubules in melanocytes are oriented with their Human skin pigmentation is dependent on the transport of plus ends towards the cell periphery, similar to nerve axons (Baas melanosomes within melanocytes from the area of synthesis near et al, 1988) and ®sh melanophores (McNiven and Porter, 1986). the cell center to the peripheral dendrites before transfer to This orientation would promote a steady-state bidirectional but net keratinocytes (Mottaz and Zelickson, 1967; Cohen and Szabo, outward transport of melanosomes as plus-end directed motors are 1968; Klaus, 1969; Wolff et al, 1974). The net transport of observed to override minus-end motors in axons (Muresan et al, melanosomes will be dependent on the balance of the cell-center- 1996). directed versus peripheral-directed forces on the melanosomes. Dynein is a chemomechanical transducer molecule or motor that Myosin V (Provance et al, 1996; Wu et al, 1998) and kinesin (Hara converts ATP-stored energy into kinetic energy following binding to et al, 2000) appear to effect peripheral transport of melanosomes. the microtubule. The dynein superfamily consists of the 12 member Our identi®cation of cytoplasmic dynein mRNA and inhibition by family of axonemal dyneins generating the shearing force in cilia and antisense nucleotides provides evidence that cytoplasmic dynein ¯agella (Brokaw, 1994) and cytoplasmic dyneins which appear to plays an important role in the cell-center-directed transport of participate in microtubule-dependent fast axonal retrograde (minus- melanosomes, similar to the microtubule-dependent fast axonal end directed) organelle transport (Schroer et al, 1989; Muresan et al, retrograde (minus-end microtubule-directed) organelle transport in 1996). Since the initial cloning and sequencing of the most abundant, neurons (Schroer et al, 1989; Muresan et al, 1996). ubiquitous cytoplasmic dynein heavy chain Dyh1, also called It is unknown whether human melanocyte dendrites resemble MAP1c, DHC1, and DHC1a (Koonce et al, 1992; Mikami et al, neuronal axons or dendrites with regard to microtubule polarity. In 1993; Vaisberg et al, 1993; Zhang et al, 1993), and the onset of our human neurons, for example, axons contain only microtubules studies, two cytoplasmic dynein family members have been identi®ed oriented with their plus ends directed towards the axonal growth in rat brain (Tanaka et al, 1995), three in ®broblasts and Hela cells cone, whereas dendrites contain microtubules with mixed polarity (Vaisberg et al, 1996), and possibly four in rat testis (Criswell and Asai, (Baas et al, 1988). It is important to emphasize that our antisense 1998). Specialization of organelle binding may reside in the various experiments resulted in a net anterograde (outward) movement of isoforms of cytoplasmic dynein heavy chains, including fast axonal melanosomes over several hours, and that we noted bidirectional transport of organelles, polarized cell secretion, organization of the motility over minute intervals throughout the period of exposure at mitotic spindle, and migration of chromosomes or nuclei during mitosis (Holzbaur and Vallee, 1994; Schroer, 1994; Tanaka et al, 1995; Vaisberg et al, 1996). The discovery of new cytoplasmic dynein isoforms, however, does not detract from our functional studies as our antisense oligonucleotides represent human melanocyte Dyh1. Although limited sequence data for the other isoforms are available, in the human fetal brain cytoplasmic dynein sequence that we used to construct primers, sequences are shared by Dyh1 cloned by Vaisberg et al (1996), indicating that we cloned and sequenced a partial sequence of human melanocyte Dyh1. This ubiquitous isoform appears to be very important, as Dyh1 is the only isoform found in some lower eukaryotic organisms. Furthermore, the Dyh1 knockout in mice is lethal and cultured blastocyte cells from this knockout show loss of perinuclear endosomes and lysosomes with fragmentation of the throughout the cytoplasm(Harada et al, 1998). Cytoplasmic dynein also appears responsible for the slow anterograde component of axoplasm and may be essential for movement of microtubules into axons and the maintenance of cell anisometry or dendricity (Dillman et al, 1996; Ahmad et al, 1998). Indeed, longer treatment of melanocytes with the antisense oligonucleotides appeared to disrupt the melanocyte processes (see Fig 7). Given the Figure 8. Antisense DNA complementary to cytoplasmic dynein previous work on neuron dendricity (Dillman et al, 1996; Ahmad et al, sequences redistributes cultured melanocyte pigment from peri- 1998), a reasonable explanation for this morphologic change in the nuclear to peripheral cytoplasm. Densitometric analysis of pigment distribution in 160 cells shows signi®cant reduction of perinuclear pigment antisense-treated melanocytes is a reduction of the cytoplasmic- distribution in antisense-compared with sense-treated melanocytes (black dynein-mediated slow transport of microtubules into the cell bars). Similarly, a signi®cant increase in the peripheral pigment distribution processes. This reduction in microtubule transport eventually leads is seen in the same cells (gray bars). *p < 0.001. to instability and collapse of the cell processes.

Figure 9. Modulation of cytoplasmic dynein intermediate chain by UV irradiation. Wes- tern immunodetection of 74 kDa intermediate chain in cultured melanocytes in three donors 24 h after exposure to increasing doses of UV (mJ per cm2 metered at 285 6 5 nmusing solar simula- tor). Donor 1 (left panel) shows increase in inter- mediate chain after 5 and 10 mJ per cm2. Donor 2 (middle panel) con®rms increased intermediate chain at 10 mJ per cm2. Donors 2 and 3 (right pa- nel) show decrease in intermediate chain at high UV doses (>30 mJ per cm2). Donor variability is seen at 20 mJ per cm2. Amido-black-stained bot- tomthree panels show protein loading of gels above. Fifty milligrams of melanocyte protein was loaded onto each lane. 996 BYERS ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Functional redundancy among the cytoplasmic isoforms may intermediate chain is present preferentially in the perinuclear exist in the human melanocyte and may explain the continued region and coincident with melanosome distribution. We provide bidirectional melanosome motility and lack of complete peripheral the ®rst functional evidence that Dyh1 participates in retrograde dispersion of melanosomes in our antisense studies. For example, melanosomal transport and that the levels of the intermediate chain cytoplasmic dynein heavy chain 2 isoform Dyh2 (also known as protein known to be involved in linkage of Dyh1 to membrane- cDHC2, DHC1b and DLP4) (Tanaka et al, 1995; Criswell et al, bound organelles are modulated by UV irradiation. Taken 1996; Vaisberg et al, 1996; Criswell and Asai, 1998; Lee et al, 1999) together, these ®ndings indicate an important role of cytoplasmic is also localized to the Golgi apparatus. Microinjection of anti- dynein in melanosome transport in human melanocytes and suggest Dyh2 antibodies induces dispersal of the Golgi apparatus (Vaisberg that it participates in UV-induced pigmentation of human skin. et al, 1996), analogous to the Dyh1 knockout ®ndings. On the other hand, another cytoplasmic dynein isoform (DHC3) appears REFERENCES not to be associated with the Golgi or any other clearly de®ned Ahmad FJ, Echeverri CJ, Vallee RB, Baas PW: Cytoplasmic dynein and dynactin are membranous cytoplasmic compartment (Vaisberg et al, 1996), and required for the transport of microtubules into the axon. JCell Biol 140:391± 401, 1998 therefore appears not to have a redundant function with Dyh1 or Baas PW, Deitch JS, Black MM, Banker GA: Polarity orientation of microtubules in Dyh2. The sequences of these isoforms are similar near the hippocampal neurons: uniformity in the axon and nonuniformity in the phosphate-binding region, but different in the organelle-binding dendrite. Proc Natl Acad Sci USA 85:8335±8339, 1988 region near our antisense sequences for Dyh1. Nonetheless, some BloomGS, Wagner MC, P®ster KK, Brady ST: Native structure and physical properties of bovine brain kinesin and identi®cation of the ATP-binding functional redundancy of Dyh1 and Dyh2, acting on melanosomes, subunit polypeptide. Biochemistry 27:3409±3416, 1988 may explain the continued bidirectional melanosome motility and Brady ST: A novel brain ATPase with properties expected for the fast axonal lack of complete peripheral dispersion of melanosomes in our transport motor. Nature 317:73±75, 1985 Brady ST, Lasek RJ, Allen RD: Fast axonal transport in extruded axoplasmfrom antisense studies. It is also possible that limited delivery of antisense giant axon. Science 218:1129±1131, 1982 oligonucleotides and incomplete disruption of cytoplasmic dynein Brokaw CJ: Control of ¯agellar bending: a new agenda based on dynein diversity. protein synthesis prevented complete inhibition of inward-directed Cell Motil Cytoskel 28:199±204, 1994 melanosome transport in our experiments. Alternatively, some Byers HR, Etoh T, Doherty JR, Sober AJ, MihmMC, Jr: Cell migrationand actin organization in cultured human primary, recurrent cutaneous and metastatic kinesin family members show microtubule minus-end-directed melanoma. Time-lapse and image analysis. Am JPathol 139:423±435, 1991 motility (Hirokawa, 1998) and might also contribute to the Chomczynski P, Sacchi N: Single-step method of RNA isolation by acid bidirectional motility and perinuclear positioning of some melano- guanidiniumthiocyanate-phenol-chloroformextraction. Anal Biochem somes in the cytoplasmic dynein antisense-treated melanocytes. 162:156±159, 1987 Cohen J, Szabo G: Study of pigment donation in vitro. Exp Cell Res 50:418±434, 1968 Recent microinjection of antibodies against Dyh1, however, was Criswell PS, Asai DJ: Evidence for four cytoplasmic dynein heavy chain isoforms in shown to disrupt the central aggregation of melanosomes in ®sh rat testis. Mol Cell Biol 9:237±247, 1998 melanophores and enhance peripheral cytoplasmic pigment Criswell PS, Ostrowski LE, Asai DJ: A novel cytoplasmic dynein heavy chain: expression of DHC1b in mammalian ciliated epithelial cells. JCell Sci distribution, supporting evidence of our ®ndings that treatment 109:1891±1898, 1996 with antisense oligonucleotides to Dyh1 mRNA disrupts peri- Dillman JF, 3rd, P®ster KK: Differential phosphorylation in vivo of cytoplasmic nuclear melanosomes and promotes net dispersion of melanosomes. dynein associated with anterogradely moving organelles. JCell Biol 127:1671± Our ®ndings thus extend the role of cytoplasmic dynein to 1681, 1994 Dillman JF, 3rd, Dabney LP, P®ster KK: Cytoplasmic dynein is associated with slow retrograde melanosome transport in human melanocytes. axonal transport. Proc Natl Acad Sci USA 93:141±144, 1996 In order to provide another line of functional evidence for a role Eller MS, Oleksiak MF, McQuaid TJ, McAfee SG, Gilchrest BA: The molecular of cytoplasmic dynein in melanosome transport other than antisense cloning and expression of two CRABP cDNAs fromhumanskin. Exp Cell Res experiments, we used antibodies against the cytoplasmic dynein 198:328±336, 1992 Friedmann PS, Gilchrest BA: Ultraviolet radiation directly induces pigment 74 kDa intermediate chain (P®ster et al, 1996b) to characterize the production by cultured human melanocytes. JCell Physiol 133:88±94, 1987 distribution and expression of this protein in normal melanocytes Gilchrest BA, Vrabel MA, Flynn E, Szabo G: Selective cultivation of human and UV-irradiated melanocytes. The cytoplasmic dynein 74 kDa melanocytes from newborn and adult epidermis. JInvest Dermatol 83:370±376, intermediate chain is an important functional linkage protein 1984 Gill SR, Schroer TA, Szilak I, Steuer ER, Sheetz MP, Cleveland DW: Dynactin, a involved in binding the entire motor complex to membranous conserved, ubiquitously expressed component of an activator of vesicle motility organelles via dynactin and its ``p150 glued'' subunit (Karki and mediated by cytoplasmic dynein. JCell Biol 115:1639±1650, 1991 Holzbaur, 1995; Vaughan and Vallee, 1995; Steffen et al, 1997). In Goukassian DA, Eller MS, Yaar M, Gilchrest BA: Thymidine dinucleotide mimics support of a functional role of cytoplasmic dynein in transport of the effect of solar simulated irradiation on p53 and p53-regulated proteins. JInvest Dermatol 112:25±31, 1999 melanosomes, we found that cytoplasmic dynein intermediate Griscelli C, Durandy A, Guy-Grand D, Daguillard F, Herzog C, Prunieras M: A chain immunolocalization correlated with the distribution of syndrome associating partial and immunode®ciency. Am JMed melanosomes. Recent evidence suggests that the intermediate 65:691±702, 1978 chain does not associate with the Dyh2 isoformas they fail to co- Hara M, Yaar M, Byers HR, Goukassian D, Fine RE, Gonzalves J, Gilchrest BA: Kinesin participates in melanosomal movement along melanocyte dendrites. purify whereas it co-puri®es with Dyh1 (Criswell and Asai, 1998). JInvest Dermatol in press: 2000 This ®nding, in addition to the co-localization of Dyh1 with the Harada A, Takei Y, Kanai Y, Tanaka Y, Nonaka S, Hirokawa N: Golgi vesiculation intermediate chain in other cells and not with the DHC3 isoform, and dispersion in cells lacking cytoplasmic dynein. JCell Biol 141:51± makes the 74 kDa intermediate chain useful to study the transport 59, 1998 Hirokawa N: Kinesin and dynein superfamily proteins and the mechanism of of melanosomes. The extensive localization of the intermediate organelle transport. Science 279:519±526, 1998 chain and melanosomes in the perinuclear region supports the Holzbaur EL, Vallee RB: DYNEINS: molecular structure and cellular function. Ann antisense data indicating an important role of cytoplasmic dynein in Rev Cell Biol 10:339±372, 1994 maintaining the ``inward'' pull on melanosomes towards the cell Jimbow K, Quevedo WC, Jr, Fitzpatrick TB, Szabo G: Some aspects of melanin biology: 1950±1975. JInvest Dermatol 67:72±89, 1976 center to counteract the overriding outward ``pull'' of the kinesin Karki S, Holzbaur EL: Af®nity chromatography demonstrates a direct binding motor proteins. Interestingly, we show for the ®rst time a between cytoplasmic dynein and the dynactin complex. JBiol Chem modulating effect of UV irradiation on cytoplasmic dynein 270:28806±28811, 1995 intermediate chain protein levels. We found that UV irradiation Klaus SN: Pigment transfer in mammalian epidermis. Arch Dermatol 100:756±762, 1969 at suf®ciently high doses promotes the outward transport of Koonce MP, Grissom PM, McIntosh JR: Dynein from Dictyostelium: primary melanosomes by reducing this cytoplasmic dynein linkage protein structure comparisons between a cytoplasmic motor enzyme and ¯agellar required for binding membrane-bound organelles. dynein. JCell Biol 119:1597±1604, 1992 In summary, we show that Dyh1, the major isoform of Lambert J, Onderwater J, Vander Haeghen Y, Vancoillie G, Koerten HK, Mommaas AM, Naeyaert JM: Myosin V colocalizes with melanosomes and cytoplasmic dynein heavy chain, is expressed in human melano- subcortical actin bundles not associated with stress ®bers in human epidermal cytes. Furthermore, the associated organelle linkage 74 kDa melanocytes. JInvest Dermatol 111:835±840, 1998 VOL. 114, NO. 5 MAY 2000 CYTOPLASMIC DYNEIN AND MELANOSOME TRANSPORT 997

Lee S, Wisniewski JC, Dentler WL, Asai DJ: Gene knockouts reveal separate Rodionov VI, Gyoeva FK, Gelfand VI: Kinesin is responsible for centrifugal functions for two cytoplasmic dyneins in Tetrahymena thermophila. Mol Cell movement of pigment granules in melanophores. Proc Natl Acad Sci USA Biol 10:771±784, 1999 88:4956±4960, 1991 Lin SX, Collins CA: Immunolocalization of cytoplasmic dynein to lysosomes in Rogers SL, Tint IS, Fanapour PC, Gelfand VI: Regulated bidirectional motility of cultured cells. JCell Sci 101:125±137, 1992 melanophore pigment granules along microtubules in vitro. Proc Natl Acad Sci Lin SX, Collins CA: Regulation of the intracellular distribution of cytoplasmic USA 94:3720±3725, 1997 dynein by serumfactors and calcium. JCell Sci 105:579±588, 1993 Schafer DA, Gill SR, Cooper JA, Heuser JE, Schroer TA: Ultrastructural analysis of Lin SX, Ferro KL, Collins CA: Cytoplasmic dynein undergoes intracellular the dynactin complex: an actin-related protein is a component of a ®lament redistribution concomitant with phosphorylation of the heavy chain in that resembles F-actin. JCell Biol 126:403±412, 1994 response to serumstarvation and okadaic acid. JCell Biol 127:1009±1019, 1994 Schnapp BJ, Reese TS: Dynein is the motor for retrograde axonal transport of Lye RJ, Porter ME, Scholey JM, McIntosh JR: Identi®cation of a microtubule-based organelles. Proc Natl Acad Sci USA 86:1548±1552, 1989 cytoplasmic motor in the nematode C. elegans. Cell 51:309±318, 1987 Schnapp BJ, Vale RD, Sheetz MP, Reese TS: Single microtubules from squid McNiven MA, Porter KR: Microtubule polarity confers direction to pigment axoplasm support bidirectional movement of organelles. Cell 40:455±462, 1985 transport in . JCell Biol 103:1547±1555, 1986 Schroer TA: Structure, function and regulation of cytoplasmic dynein. Curr Opin Cell Mehta AD, Rock RS, Rief M, Spudich JA, Mooseker MS, Cheney RE: Myosin-V is Biol 6:69±73, 1994 a processive actin-based motor. Nature 400:590±593, 1999 Schroer TA, Steuer ER, Sheetz MPL; Cytoplasmic dynein is a minus end-directed Mercer JA, Seperack PK, Strobel MC, Copeland NG, Jenkins NA: Novel myosin motor for membranous organelles. Cell 56:937±946, 1989 heavy chain encoded by murine dilute coat colour locus [published erratum Steffen W, Karki S, Vaughan KT, Vallee RB, Holzbaur EL, Weiss DG, appears in Nature 1991 Aug 8;352(6336):547]. Nature 349:709±713, 1991 Kuznetsov SA: The involvement of the intermediate chain of cytoplasmic Mermall V, Post PL, Mooseker MS: Unconventional myosins in cell movement, dynein in binding the motor complex to membranous organelles of Xenopus membrane traf®c, and signal transduction. Science 279:527±533, 1998 oocytes. Mol Biol Cell 8:2077±2088, 1997 Mikami A, Paschal BM, Mazumdar M, Vallee RB: Molecular cloning of the Tanaka Y, Zhang Z, Hirokawa N: Identi®cation and molecular evolution of new retrograde transport motor cytoplasmic dynein (MAP 1C). Neuron 10:787±796, dynein-like protein sequences in rat brain. JCell Sci 108:1883±1893, 1995 1993 Vaisberg EA, Koonce MP, McIntosh JR: Cytoplasmic dynein plays a role in Milisav I: Dynein and dynein-related genes. Cell Motil Cytoskel 39:261±272, 1998 mammalian mitotic spindle formation. JCell Biol 123:849±858, 1993 Mottaz JH, Zelickson AS: Melanin transfer: a possible phagocytic process. JInvest Vaisberg EA, Grissom PM, McIntosh JR: Mammalian cells express three distinct Dermatol 49:605±610, 1967 dynein heavy chains that are localized to different cytoplasmic organelles. JCell Muresan V, Godek CP, Reese TS, Schnapp BJ: Plus-end motors override minus-end Biol 133:831±842, 1996 motors during transport of squid axon vesicles on microtubules. JCell Biol 135:383±397, 1996 Vale RD, Reese TS, Sheetz MP: Identi®cation of a novel force-generating protein, Naeyaert JM, Eller M, Gordon PR, Park HY, Gilchrest BA: Pigment content of kinesin, involved in microtubule-based motility. Cell 42:39±50, 1985 cultured human melanocytes does not correlate with tyrosinase message level. Vaughan KT, Vallee RB: Cytoplasmic dynein bind dynactin through a direct Br JDermatol 125:297±303, 1991 interaction between the intermediate chains and p150Glued. JCell Biol Nilsson H, Wallin M: Evidence for several roles of dynein in pigment transport in 131:1507±1516, 1995 melanophores. Cell Motil Cytoskel 38:397±409, 1997 Werninghaus K, Handjani RM, Gilchrest BA: Protective effect of alpha-tocopherol Park HY, Russakovsky V, Ohno S, Gilchrest BA: The beta isoformof protein kinase in carrier liposomes on ultraviolet-mediated human epidermal cell damage C stimulates human melanogenesis by activating tyrosinase in pigment cells. in vitro. Photodermatol Photoimmunol Photomed 8:236±242, 1991 JBiol Chem 268:11742±11749, 1993 Wolff K, Jimbow K, Fitzpatrick TB: Experimental pigment donation in vivo. JUltrastr Paschal BM, Shpetner HS, Vallee RB: MAP 1C is a microtubule-activated ATPase Res 47:400±419, 1974 which translocates microtubules in vitro and has dynein-like properties. JCell Wu X, Bowers B, Wei Q, Kocher B, Hammer JA, 3rd: Myosin V associates with Biol 105:1273±1282, 1987 melanosomes in mouse melanocytes: evidence that myosin V is an organelle Pastural E, Barrat FJ, Dufourcq-Lagelouse R, et al: Griscelli disease maps to motor. JCell Sci 110:847±859, 1997 chromosome 15q21 and is associated with mutations in the myosin-Va gene. Wu X, Bowers B, Rao K, Wei Q, Hammer JA, 3rd: Visualization of melanosome Nature Genet 16:289±292, 1997 dynamics within wild-type and dilute melanocytes suggests a paradigm for P®ster KK, Salata MW, Dillman JF, 3rd, Torre E, Lye RJ: Identi®cation and myosin V function in vivo. JCell Biol 143:1899±1918, 1998 developmental regulation of a neuron-speci®c subunit of cytoplasmic dynein. Yaar M, Grossman K, Eller M, Gilchrest BA: Evidence for nerve growth Mol Biol Cell 7:331±343, 1996a factor-mediated paracrine effects in human epidermis. JCell Biol 115:821± P®ster KK, Salata MW, Dillman JF, 3rd, Vaughan KT, Vallee RB, Torre E, Lye RJ: 828, 1991 Differential expression and phosphorylation of the 74-kDa intermediate chains Yaar M, Eller MS, DiBenedetto P, et al: The trk family of receptors mediates nerve of cytoplasmic dynein in cultured neurons and glia. JBiol Chem 271:1687± growth factor and neurotrophin-3 effects in melanocytes. JClin Invest 94:1550± 1694, 1996b 1562, 1994 Provance DW, Jr, Wei M, Ipe V, Mercer JA: Cultured melanocytes form dilute Zhang Z, Tanaka Y, Nonaka S, Aizawa H, Kawasaki H, Nakata T, Hirokawa N: mutant mice exhibit dendritic morphology and altered melanosome The primary structure of rat brain (cytoplasmic) dynein heavy chain, a distribution. Proc Natl Acad Sci USA 93:14554±14558, 1996 cytoplasmic motor enzyme. Proc Natl Acad Sci USA 90:7928±7932, 1993