Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Invest Ophthalmol Vis Sci. Author manuscript; available in PMC 2010 November 1.
Published in final edited form as:
PMCID: PMC2777535

Retrograde Intraflagellar Transport by Cytoplasmic Dynein-2 Is Required for Outer Segment Extension in Vertebrate Photoreceptors but Not Arrestin Translocation



Anterograde intraflagellar transport (IFT) is essential for photoreceptor outer segment formation and maintenance, as well as for opsin trafficking. However, the role of retrograde IFT in vertebrate photoreceptors remains unclear. The purpose of this study was to evaluate zebrafish photoreceptors lacking the retrograde IFT motor, cytoplasmic dynein-2.


Morpholino oligonucleotides against the heavy chain (dync2-h1), light intermediate chain (dync2-li1), and intermediate chain (dync2-i1) subunits of cytoplasmic dynein-2 were injected into zebrafish embryos. Retinas and ciliated cells of these zebrafish morphants were analyzed by immunohisto-chemistry and transmission electron microscopy. Whole-field electroretinograms (ERGs) were performed on dynein morphants at 5 to 6 days after fertilization (dpf).


Zebrafish lacking cytoplasmic dynein-2 function exhibited small eyes, kidney cysts, and short photoreceptor outer segments, some of which were disorganized with accumulated vesicles. Morphant photoreceptor connecting cilia were swollen, but neither opsin nor arrestin was mislocalized, although IFT88 accumulated in the distal region of the connecting cilium. Nasal cilia were shortened and displayed cytoplasmic swelling along the axoneme. Loss of cytoplasmic dynein-2 function resulted in a significant reduction in the amplitude of ERG a-, b-, and d-waves but no change in threshold response.


Retrograde IFT is essential for outer segment extension and IFT protein recycling in vertebrate photoreceptors. The results show, for the first time, that the dync2-i1 subunit of cytoplasmic dynein-2 is necessary for retrograde IFT. In addition, arrestin translocation does not require retrograde IFT. Finally, the ERG results indicate that loss of cytoplasmic dynein-2 reduces the photoreceptor light response.

Vertebrate photoreceptor outer segments are modified sensory cilia,1 and as is true of all other cilia, lack the cellular machinery for protein synthesis. The proteins required for phototransduction and that make up the outer segment, such as opsins, must pass through the connecting cilium. In addition to its function in phototransduction, opsin plays an important structural role in the outer segment2 and the effective transport of proteins through the connecting cilium is critical to both the function and survival of photoreceptors. Failure of the efficient transport of opsins to the outer segment leads to photoreceptor degeneration and death.3,4

Intraflagellar transport (IFT) transports cargo through the connecting cilium to the outer segment.5,6 Studies of the zebrafish ift57, ift88, and ift172 mutants have shown that the initial stages of outer segment formation require IFT and that IFT is also necessary for outer segment maintenance.7-9 Similarly, loss of IFT causes mislocalization of opsins and photoreceptor cell death. These studies, along with similar results in mouse Ift88/polaris mutants and kinesin-II conditional knockout mutants,5,6 reinforce the premise that the outer segment is, in fact, a modified cilium.6 As several diseases affecting cilia and ciliary trafficking result in blindness and retinal degeneration, understanding the role of IFT in vertebrate photoreceptors is of great importance.

In photoreceptor cell biology, anterograde IFT has received the most scrutiny, while the role of retrograde IFT remains largely unexplored.5,9,10 Evidence from several species indicates a conserved mechanism whereby kinesin-II11 and osm-312 cooperate to drive anterograde IFT, whereas cytoplasmic dynein-213 mediates retrograde IFT. The retrograde IFT motor is composed of an ATP-dependent homodimeric heavy chain, Dync2-h1 (previously termed DHC-1b)13-16, a cytoplasmic dynein-2 light intermediate chain, Dync2-li1,17-19 an intermediate chain, Dync2-i1,20 and light chain known as LC8.20,21 Mutation of Dync2-h1, Dync2-li1, and LC8 all result in phenotypes consistent with defective retrograde IFT, with cells displaying shortened cilia that accumulate IFT particles. Dync2-i1 was shown biochemically to associate with cytoplasmic dynein-2; however, the phenotypic consequence of dync2-i1 dysfunction is unknown. Of importance, Dync2-h1 and Dync2-li1 localize to the connecting cilium of bovine photoreceptors,18 suggesting a role for retrograde IFT in photoreceptors.

Retrograde IFT recycles IFT proteins and other ciliary components by returning them to the basal body from the tip of the cilium.13,22 As photoreceptors shed approximately 10% of the outer segment material from the tips each day and most proteins move through the connecting cilium in a unidirectional manner (e.g., opsins), a role for retrograde IFT in photoreceptors has been sought. Retrograde IFT is necessary for other vertebrate cilia, as Dync2-h1 and Dync2-li1 mouse mutants have stumpy nodal cilia with swollen ciliary tips that contain disorganized microtubules, IFT proteins and cellular debris,23,24 reflecting the inability to return cargo to the ciliary base. Does retrograde IFT function in photoreceptors?

Although it is assumed that IFT components require retrograde IFT for recycling, arrestin and transducin also exhibit retrograde movement (reviewed in Calvert et al.25). During light adaptation, arrestin moves from the inner segment to the outer segment through the connecting cilium, while transducin moves in a complementary fashion. During dark adaptation, arrestin translocates back to the inner segment, while transducin returns into the outer segment.26-28 Translocation of arrestin and transducin still occurs after ATP depletion of photoreceptors in retinal explants, suggesting that these proteins move via passive diffusion.29 The kinetics of arrestin movement during light adaptation are 100 to 1000 times that of rhodopsin trafficking, suggesting that an active transport mechanism could not move arrestin to the outer segment so quickly.25 However, the retrograde movement of arrestin during dark adaptation is much slower, raising the possibility that anterograde and retrograde arrestin movement are mediated by different processes. Furthermore, both the actin and microtubule cytoskeletons are required for the retrograde movements of arrestin and transducin, suggesting that motors could be involved.30 As cytoplasmic dynein-2 is a microtubule minus-end motor located in the connecting cilium, IFT may mediate retrograde arrestin translocation through the photoreceptor connecting cilium.

We used morpholino oligonucleotides to eliminate the function of dync2-h1, dync2-li1, and dync2-i1 in zebrafish embryos to understand the function of retrograde IFT in vertebrate photoreceptors. Retrograde IFT was necessary for outer segment extension, organization, and recycling of IFT proteins in photoreceptors. In addition, we describe robust light-dependent translocation of visual arrestin in larval zebrafish rods. Finally, we show that dynein morphant retinas exhibit reduced ERG a- and b-wave amplitudes, indicating that visual function requires cytoplasmic dynein-2.

Materials and Methods

Fish Maintenance and Breeding

Wild-type zebrafish of the AB strain were housed, bred, and staged according to standard procedures.31 The zebrafish were treated in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research.

Cloning of Dynein Genes and Phylogenetic Analysis of dync2-h1

5′ RACE (rapid amplification of cDNA ends) was performed on retinal cDNAs to clone the 5′ end of dync2-h1 using the primer dync2-h1 RACE1R: 5′ TGCAGCAGGACGGGGCTGTAGACCTGA 3′ and the 5′ end of dync2-i1 using the nested primers dync2-i1 RACE 1R: 5′ TTTGGCCGCTCTGGTCTTGTGTTTG 3′ and dync2-i1-nested-2R: 5′ATCGCCATCATCCACACGGCCAAAT3′. 5′ RACE was performed (Gene Racer Kit; Invitrogen, Eugene, OR) according to the manufacturer's instructions. Initial dync2-h1 sequence data were obtained from accession numbers: XR_029028 and XR_029028.2. The dync2-h1 sequence was aligned against genomic contig NW_001514387 to annotate intron-exon boundaries for the 5′ 1.3kb of the coding sequence for morpholino design. Sequence data for dync2-i1 and dync2-li1 was obtained from Ensembl, numbers ENDSARG00000057635 and ENSDAR00000039770, respectively.32 Phylogenetic analysis was performed with MEGA software ( Molecular Evolutionary Genetics Analysis, developed by S. Kumar, K. Tamura, and M. Nei and provided in the public domain by the Center for Evolutionary Functional Genomics, The Biode-sign Institute, Tempe, AZ). GenBank accession numbers (; provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD) and Ensembl numbers are as follows: Chlamydomonas reinhardtii dhc1b (XP_001696428); Caenorhabditis elegans DHC1 (NP_491363.1), CHE-3 (NP_492221.2); Danio rerio dync1-h1 (NP_001036210.1); Mus musculus Dync1-h1 (NP_084514.2), Dync2-h1 (NP_084127.2); Rattus norvegicus dync1-h1 (ENSNOG00000006178), dync2-h1 (NP_075413.1); Gallus gallus dync1-h1 (ENGALT00000037512), dync2-h1 (ENSGALP00000027700); Oryzias latipes dync1-h1 (ENSORLP00000015944), dync2-h1 (ENSORLT00000013451); Gasterosteus aculeatus dync1-h1 (ENSGALP00000014094), dync2-h1 (ENSGALP00000010733); and Takifugu rubripes dync1-h1 (ENSTRUP00000031697), dync2-h1 (ENSTRUP00000010356).

Morpholino Microinjection and Design

One- to two-cell-stage zebrafish embryos were injected as described elsewhere.33 Morpholinos were synthesized by Gene Tools, LLC (Philomath, OR). The sequences were: dync2-h1MO1: 5′CTGCTGTGAATGCATGTACCTGCTG3′; dync2-h1MO2: 5′GACAGGTGTGCAGTCTGACCTCCTC3′; dync2-li1MO1sp: 5′GGACTGTCACTTTTACCGTATTATG3′; dync2-li1MO2atg: 5′GCTTACTTTCGGCATTATTTTACCG3′; dync2-i1MO1 5′ACAATAAACTGTTACCGGCCAAATC3′; and dync2-i1MO2 5′CCACACTAATCCCAGAAATACAACA3′.

In Situ Hybridization and Reverse Transcription-Polymerase Chain Reaction

Antisense riboprobes corresponding to dync2-h1, dync2-li1, and dync2-i1 were synthesized with SP6 RNA polymerase (Invitrogen) and in situ hybridization performed,34 and RT-PCR was performed.8

Histology, Immunohistochemistry, and Immunocytochemistry

Embryos were processed for histology and electron microscopy as described elsewhere.35 Wholemount immunostaining of zebrafish 30 hours after fertilization (hpf) was performed as previously described,36 with an anti-acetylated tubulin antibody (1:50; Sigma-Aldrich, St. Louis, MO). Immunohistochemistry and imaging of frozen sections was performed as described8 with the following antibodies: anti-acetylated tubulin (1:500), anti-IFT88 (1:5000),8 1d1 (monoclonal zebrafish rhodopsin antibody 1:100) (Fadool JM, et al. IOVS 1999;40:ARVO Abstract 1251), rabbit anti-zebrafish rhodopsin and rabbit anti-zebrafish blue cone opsin (both 1:200).37 Staining with the mouse monoclonal anti-arrestin antibody, xArl-6, (1:25) staining was performed.38

Antibody Generation

A rabbit polyclonal anti-dync2-h1 antibody was generated against the synthetic peptide MAPGTDDPRK corresponding to the N-terminus of zebrafish dync2-h1 and affinity-purified by Bethyl Laboratories (Montgomery, TX).

SDS-PAGE and Western Blot Analysis

Total protein from larval heads was generated,8 separated by electrophoresis on 5% to 15% gradient SDS-polyacrylamide gels (Bio-Rad, Hercules, CA), and transferred to PVDF membranes. Rabbit anti-dync2-h1 (1:5000) and anti-α-tubulin antibodies (Clone 12G10; 1:2000; Developmental Studies Hybridoma Bank, University of Iowa) were used and detected.8


Larval eyes were isolated with a tungsten wire loop39 and ERGs performed.40 Full-field light stimulation was prolonged for 500 ms under photopic conditions to ensure b- and d-wave separation. Light flashes were presented at intervals of 10 seconds to 3 minutes, depending on the recovery of the ERG from a test flash. Recordings from three sequential flashes were averaged to reduce noise and obtain the ERG trace. The ERG traces shown are representative of 5 to 10 separate trials of each morphant and wild-type fish. L-AP4 (L-(+)-2-amino-4-phosphonobutyric acid) and TBOA (DL-threo-b-benzyloxyaspartic acid) were purchased from ToCris Biosciences (Ellisville, MA). L-AP4 was dissolved in 0.1 M NaOH and TBOA dissolved in DMSO. Reagents were diluted in fish water to a final concentration of 0.2 mM TBOA and 0.4 mM L-AP4. Larval fish at 5 to 6 dpf were incubated in TBOA and L-AP4 for 30 minutes.


Phylogenetic Analysis of dync2-h1

To identify the zebrafish dync2-h1 orthologue, we performed a BLAST search of the zebrafish genome for sequences with homology to mouse Dync2-h1. Two partial mRNA sequences were recovered that represented 6.3 kb of the 5′ end and 11.4 kb of the 3′ end of dync2-h1 (accession numbers XR_029028 and XR_029028.2, respectively). These two sequences combined yielded a predicted protein product of 4266 amino acids, which was 69% identical with mouse Dync2-h1. To verify this as a true zebrafish dync2-h1 orthologue, we performed a phylogenetic analysis using the neighbor-joining method (Fig. 1) and found that the zebrafish sequence is most closely related to other vertebrate dync2-h1 genes. Sequence data for dync2-li1 and dync2-i1, the zebrafish orthologue of Chlamydomonas FAP133 were identified by using Ensembl.32

Figure 1
Phylogenetic analysis of zebrafish dync2-h1. A neighbor-joining tree was generated based on Clustal W protein sequence alignments of Cr, Chlamydomonas reinhardtii; Ce, Caenorhabditis elegans; Ga, Gasterosteus aculeatus; Gg, Gallus gallus; Hs, Homo sapiens ...

Cytoplasmic dynein-2 Genes in the CNS and Pronephros

As cytoplasmic dynein-2 genes have not been described in zebrafish, we determined the expression pattern of dync2-h1, dync2-li1, and dync2-i1 in 72 hpf embryos. Similar to the expression of other zebrafish IFT genes,41 expression of dync2-h1, dync2-li1, and dync2-i1 was observed in the CNS and pronephros (Figs. 2A-C). Although dync2-h1 and dync2-li1 have been found in photoreceptors,18 it has been suggested that dync2-i1 only functions in motile cilia.20 We sectioned embryos after in situ hybridization and observed that dync2-h1, dync2-li1, and dync2-i1 genes were expressed throughout the retina, including in photoreceptors (Figs. 2D-F).

Figure 2
Dynein gene expression and morphant phenotypes. (AC) Wholemount in situ hybridization performed on 72 hpf zebrafish revealed identical expression of all three dynein genes, with the strongest expression in the CNS and pronephros (arrowhead). ...

IFT Phenotypes of Dynein Morphants

To disrupt retrograde IFT, we injected morpholino oligonucleotides against the dync2-h1, dync2-li1, and dync2-i1 genes into one-cell zebrafish embryos. Zebrafish IFT mutants exhibit specific phenotypes due to ciliary dysfunction, including ventral curvature of the body axis, kidney cysts, and small eyes. Remarkably, dync2-h1 morphants looked very similar to previously described IFT mutants, such as ift88oval and ift57hippi.9,42 The dync2-h1 morphants had a distinctive ventral curvature of the body axis, small eyes, and 100% penetrance of kidney cysts at 4 days post fertilization (dpf) (Fig. 2G). The dync2-li1 morphants similarly exhibited small eyes and a high penetrance of kidney cysts (Fig. 2H), although they did not exhibit ventral curvature of the body. Knockdown of dync2-i1 (Fig. 2I) yielded morphants with small eyes, pronounced swelling in the hindbrain and a penetrance of kidney cysts similar to that of dync2-li1 morphants (Table 1). Some dync2-i1 morphants had a ventral curvature of the body while others did not, and both dync2-li1 and dync2-i1 morphants were shorter than age-matched control fish. Of importance, knockdown experiments with nonoverlapping morpholinos designed against each gene resulted in identical phenotypes, strongly suggesting that the dynein morphant phenotypes are specific (data not shown). These phenotypes are consistent with cytoplasmic dynein-2 function as the retrograde IFT motor.

Table 1
Frequency of Kidney Cysts in 4-dpf Dynein Morphants

Western blot analysis performed on heads of 4- and 5-dpf larvae, with a polyclonal antibody against zebrafish Dync2-h1, showed a high-molecular-weight band that was depleted in morphants at 4 and 5 dpf (Fig. 2J, black arrowhead). RT-PCR analysis of dync2-h1 morphants showed similar results (Fig. 2J′). Because of the absence of suitable antibodies against dync2-li1 and dync2-i1, we performed RT-PCR to assess the efficacy of morpholinos to interfere with splicing. Morpholinos against both dync2-li1 and dync2-i1 were effective at all time points assayed (Figs. 2K, 2L)

Features of Dynein Morphant Eyes

To determine how loss of cytoplasmic dynein-2 affects retinal organization, we analyzed histologic sections of 4-dpf morphants. Similar to that described in IFT mutants, retinal lamination was normal in all three dynein morphants but the eyes were smaller than control eyes. We did not observe pyknotic nuclei or other evidence of cell death within the photoreceptor layer of dynein morphants, unlike previously described IFT mutants8,9 (Fig. 3).

Figure 3
Transverse histologic sections through 4-dpf wild-type and dynein morphant retinas (AD). The wild-type zebrafish retina (A) is highly organized by 4 dpf with all cellular layers readily observable, including the ganglion cell layer (GCL), inner ...

IFT plays a critical role in outer segment formation and maintenance, and we hypothesized that outer segment integrity would require retrograde transport. At 4 dpf, transmission electron microscopy (TEM) revealed that wild-type photoreceptors have robust outer segments that extend into the retinal pigment epithelium (Figs. 4A, 4B). In contrast, all the dynein morphants exhibited very short outer segments, with some photoreceptors completely lacking these structures (Figs. 4C, 4E, 4G). Higher magnification images revealed that connecting cilia in all dynein morphants were swollen, perhaps because of the accumulation of IFT particles and cellular debris (Figs. 4D, 4F, 4H; arrows). Disc stacking was essentially normal in dync2-h1 and dync2-i1 morphants. In contrast, many dync2-li1 morphant photoreceptors had outer segments composed of disordered membranes, some of which also contained large vesicles (Figs. 4E, 4F). These structures were highly abnormal and were never observed in wild-type photoreceptors. We occasionally observed vertically stacked membranes within the outer segments of dync2-li1 and dync2-i1 morphants (Figs. 4F, 4H). These data show that normal extension of the outer segments requires retrograde IFT and suggest that recycling of IFT proteins may be a limiting factor in outer segment growth.

Figure 4
Transmission electron microscopy of 4-dpf wild-type and dynein morphant retinas. (A, C, E, F) Lowmagnification electron micrographs of wild-type and dynein morphant retinas. (A) Wild-type photoreceptor outer segments (OS), inner segments (IS), nuclei ...

Ciliogenesis in Other Tissues in Dynein Morphants

Retrograde IFT is essential for cilium formation and maintenance in Chlamydomonas, C. elegans, and mouse. Kidney cysts occurred with high penetrance in dync2-h1, dync2-li1, and dync2-i1 morphants, suggesting that pronephric cilia were impaired (Table 1). To directly evaluate the effect of cytoplasmic dynein-2 dysfunction on pronephric cilia, we performed wholemount immunolabeling on 30 hpf zebrafish embryos with an antibody against acetylated tubulin, a marker for cilia. The anterior zebrafish pronephros is highly ciliated, with numerous monocilia extending into the pronephric lumen (Fig. 5A). Most cilia were missing in dync2-h1 morphants and those present were significantly shorter than wild-type cilia (Fig. 5B). In dync2-li1 morphants, cilia were shorter and present at lower density than in the wild-type (Fig. 5C). Cilia density was also lower in dync2-i1 morphants, although cilia lengths were comparable to those in wild-type animals (Fig. 5D). These data correlate with the penetrance of kidney cysts observed at 4 dpf (Table 1).

Figure 5
Analysis of nonphotoreceptor cilia. (AD) The 30-hpf wild-type and dynein morphant embryos were immunolabeled with an antiacetylated tubulin antibody, a marker for cilia. Images were collected of the anterior pronephros Anterior is to the left ...

The canonical ciliary phenotype of disrupted retrograde IFT is swollen ciliary axonemes in which accumulated cytoplasm, IFT proteins, and cellular debris are present.13,15,23 The zebrafish nasal epithelium is highly ciliated and provides an accessible tissue for analysis. TEM of the nasal epithelium of 4-dpf wild-type larvae revealed robust cilia (Fig. 5E). As expected, all three dynein morphants exhibited very short nasal cilia with swollen axonemes (Figs. 5F-H). Combined, these data show that loss of dync2-h1, dync2-li1, and dync2-i1 compromised retrograde IFT, which is essential for nasal cilium integrity.

Rod and Cone Opsins in Dynein Morphants

Rod and cone opsins mislocalize in kinesin-II, ift88, and ift57 mutants, implying that opsins are cargo of IFT in photoreceptors.5,6,8,9 As opsins and many other known outer segment proteins move unidirectionally and do not return to the inner segment, it is unclear whether the loss of retrograde IFT would affect opsin trafficking. We performed immunohistochemistry on 4-dpf dynein morphants using antibodies directed against zebrafish rhodopsin and blue cone opsin. At 4 dpf, both rod and blue cone opsins localized nearly exclusively to the outer segment (Fig. 6A). Although outer segments were shorter, opsin trafficking was normal in dync2-h1, dync2-li1, and dync2-i1 morphants (Figs. 6B–D). We observed subtle mislocalization of opsins in some morphants, but this phenotype was observed only in the most severely affected photoreceptors and probably was a secondary consequence of severely shortened outer segments.

Figure 6
Immunohistochemical analysis of wild-type and dynein morphant photoreceptors. (AD) Immunohistochemical labeling of rhodopsin (red) and blue cone opsin (green) in 4-dpf wild-type and dynein morphant photoreceptors. (A) Wild -type photoreceptors ...

IFT88 Protein Distribution along Photoreceptor Connecting Cilia with Dynein Dysfunction

As disruption of retrograde IFT causes accumulation of IFT particles within cilia,13-15,21 we speculated that the debris we observed in photoreceptor cilia was IFT proteins. To investigate, we immunolabeled 4-dpf retinal sections with antibodies against IFT88 and acetylated tubulin. In wild-type photoreceptors, IFT88 colocalized with acetylated tubulin at the connecting cilium (Fig. 6E). Acetylated tubulin stained the connecting cilium and axoneme, and the staining often extended distally beyond the IFT88 staining (Fig. 6E). In the dync2-h1 morphants, however, IFT88 and acetylated tubulin staining often colocalized in a small puncta and failed to elongate in the proximal-distal axis (Fig. 6F, arrowhead). In dync2-li1 and dync2-i1 morphants, IFT88 labeling was biased distally relative to acetylated tubulin, suggesting that IFT88 accumulates near the distal end of the connecting cilium (Figs. 6G, 6H, arrows). This mislocalization of IFT88 is consistent with the role cytoplasmic dynein-2 in IFT protein recycling and suggests that the axonemal swelling seen in the dynein morphants reflects accumulated IFT particles.

Effect of Disrupted Retrograde IFT on Arrestin Translocation

Although cytoplasmic dynein-2 does not participate in anterograde opsin trafficking, studies of mammalian photoreceptors have shown that transducin and arrestin move in a retrograde fashion through the connecting cilium. As an independent test of the previous results,29 we sought to directly examine the role of cytoplasmic dynein-2 in retrograde movement of phototransduction components. Zebrafish cone transducin does not move to the inner segment during light adaptation,43 but arrestin translocation has not been tested in zebrafish. To test light-dependent arrestin movement in zebrafish rods, we performed immunohistochemistry on 5-dpf zebrafish larvae with a monoclonal antibody against Xenopus arrestin.38 After dark adaptation, arrestin was located predominantly in the rod inner segments (Fig. 7A). We observed robust light-dependent arrestin movement to the rod outer segment, with outer segment localization peaking after 15 minutes of light adaptation (Fig. 7B). With continued light exposure, arrestin progressively migrated back to the inner segment (Fig. 7C), with migration nearly complete after 60 minutes of light exposure (Fig. 7D). This is similar to the pattern of arrestin translocation in Xenopus.38 To evaluate retrograde arrestin translocation, 5-dpf larvae were dark adapted for 3 hours, light adapted for 15 minutes, and then returned to the dark where arrestin translocation was analyzed by immunohistochemistry. Nearly all the arrestin translocated to the rod inner segments within 15 minutes of dark adaptation in 5-dpf wild-type larvae. Of note, arrestin translocation occurred normally in dynein morphant rods (Figs. 7E–L). These data indicate that cytoplasmic dynein-2 and, by extension, retrograde IFT do not mediate retrograde arrestin movement during dark adaptation.

Figure 7
Immunohistochemical analysis of arrestin translocation in 5-dpf wild-type zebrafish. (AD) Arrestin moves to the outer segment during light adaptation. Animals were dark adapted for 3 hours, light adapted on a benchtop in a well-lit laboratory ...

Effect of Compromised Dynein Function on Retinal Electrophysiology

To examine visual function after loss of retrograde IFT, we recorded full-field electroretinograms on 5- and 6-dpf dynein morphants. Wild-type larvae have an ERG response characterized by a corneal negative a-wave, and a corneal positive b- and d-wave, which reflect the response of photoreceptors and on and off-bipolar cells respectively. The dynein morphant ERGs were variable, with some morphants exhibiting no response, whereas others had decreased b- and d-wave amplitudes (Fig. 8A). Inhibiting the photoreceptor transmission to ON-bipolar cells using the pharmacologic agents L-Ap4 and TBOA revealed that dynein morphants also displayed reduced photoreceptor potential (Fig. 8B). However, the threshold responses of dynein morphants were comparable to wild-type (Fig. 8A). These data show that visual function is impaired in dynein morphants.

Figure 8
Electrophysiological analysis of dynein morphants. ERGs of wild-type (wt), dync2-h1, dync2-li1, and dync2-i1 morphant 5- to 6-dpf larvae (n = 5–10) under photopic conditions. (A) ERGs at increasing light intensities (B) ERGs after incubation in ...


We investigated the function of retrograde IFT in vertebrate photoreceptors and nonphotoreceptor cilia through morpholino-mediated gene knockdown of three cytoplasmic dynein-2 subunits. Photoreceptors lacking cytoplasmic dynein-2 exhibit different phenotypes than those described in anterograde IFT mutants. First, ift88 and ift172 mutants fail to form outer segments,9 whereas ift57 mutants produce short outer segments.7,8 Although the outer segments in dynein morphants appear shorter than those described for ift57 mutants, they are not accompanied by significant opsin mislocalization or cell death, as was observed in ift57 mutants. Second, the outer segments formed in ift57 mutants had surprisingly normal ultrastructure, whereas we often observed disorganized disc membranes with accumulated vesicles and debris in dynein morphant outer segments.

Our results represent the first direct evidence that retrograde IFT plays an important role in vertebrate photoreceptors. The outer segment sheds roughly 10% of its volume from the distal tips each day and it was assumed that most, if not all, active transport was unidirectional to the outer segment. The outer segment is nevertheless a modified sensory cilium44 and several aspects of axoneme maintenance and dynamics require retrograde IFT. Recycling kinesin motors and IFT particles to the inner segment would reduce the metabolic demands of synthesizing large numbers of new IFT components. Within the outer segment the microtubule array of the axoneme diverges into singlets and doublets44 and only extend approximately 80% of the length of toad outer segments.45 The mechanisms underlying cilia assembly and disassembly at microtubule tips, particularly of microtubule singlets like those seen in photoreceptors, remains poorly understood, but retrograde IFT most likely returns axonemal turnover products to the base. Retrograde IFT plays an important role in signaling pathways that use cilia.46,47 Retrograde IFT could function in signaling pathways that control outer segment length and outer segment trafficking.

The mechanism underlying arrestin translocation has been the subject of several studies.26,29,30,38,48,49 Although other investigators concluded that arrestin translocation occurs via passive diffusion,29 a direct test of retrograde IFT had not been performed. Although we observed several phenotypes consistent with disrupted retrograde IFT in dynein morphant photoreceptors, nasal cilia and pronephric cilia, translocation of rod arrestin occurred normally. Although consistent with previous reports,29 our data directly showed that retrograde IFT did not participate in this process. To our knowledge, we are the first to describe light-dependent translocation of rod arrestin in zebrafish. The pattern of arrestin localization is similar to that described in Xenopus, where arrestin migrates to the outer segment during light adaptation and returns during prolonged light exposure.38 This differs from murine systems where arrestin remains in the outer segment with continued light adaptation.26,28,48 This result suggests that retrograde arrestin translocation during constant light is common to teleosts and amphibians.

Testing the role of the accessory subunits of cytoplasmic dynein-2 is essential for a thorough understanding of the mechanism of retrograde IFT. Mutational studies in Chlamydomonas, C. elegans, and mouse have established a role for dync2-h1, dync2-li1 and LC8 in retrograde IFT.13-15,21,23,46 Our phenotypic analysis of dync2-i1 morphants is the first, however, to show that the cytoplasmic dynein-2 intermediate chain is necessary for cytoplasmic dynein-2 function. dync2-i1 morphant photoreceptors and nasal cilia were shorter than wild-type and exhibited swollen axonemes, consistent with defects in retrograde IFT. Although the length of pronephric cilia did not appear shorter in dync2-i1 morphants, the density of cilia was reduced and the morphants developed kidney cysts.

Finally, we observed reduced amplitude in dynein morphant ERG a-, b- and d-waves, which is most likely explained by the presence of short photoreceptor outer segments and the overall reduction of eye size. The impairment of visual function after loss of dynein function also suggests that genes encoding components of the retrograde pathway may underlie certain forms of human blindness.


Supported by National Institutes of Health Grant EY017037 (BDP).


Disclosure: B.L. Krock, None; I. Mills-Henry, None; B.D. Perkins, None


1. De Robertis E. Some observations on the ultrastructure and morphogenesis of photoreceptors. J Gen Physiol. 1960;43(6suppl):1–13. [PMC free article] [PubMed]
2. Lem J, Krasnoperova NV, Calvert PD, et al. Morphological, physiological, and biochemical changes in rhodopsin knockout mice. Proc Natl Acad Sci USA. 1999;96:736–741. [PubMed]
3. Grimm C, Wenzel A, Hafezi F, Yu S, Redmond TM, Reme CE. Protection of Rpe65-deficient mice identifies rhodopsin as a mediator of light-induced retinal degeneration. Nat Genet. 2000;25:63–66. [PubMed]
4. Portera-Cailliau C, Sung CH, Nathans J, Adler R. Apoptotic photoreceptor cell death in mouse models of retinitis pigmentosa. Proc Natl Acad Sci U S A. 1994;91:974–978. [PubMed]
5. Pazour GJ, Baker SA, Deane JA, et al. The intraflagellar transport protein, IFT88, is essential for vertebrate photoreceptor assembly and maintenance. J Cell Biol. 2002;157:103–113. [PMC free article] [PubMed]
6. Marszalek JR, Liu X, Roberts EA, et al. Genetic evidence for selective transport of opsin and arrestin by kinesin-II in mammalian photoreceptors. Cell. 2000;102:175–187. [PubMed]
7. Sukumaran S, Perkins BD. Early defects in photoreceptor outer segment morphogenesis in zebrafish ift57, ift88 and ift172 intraflagellar transport mutants. Vision Res. 2009;49:479–489. [PMC free article] [PubMed]
8. Krock BL, Perkins BD. The intraflagellar transport protein IFT57 is required for cilia maintenance and regulates IFT-particle-kinesin-II dissociation in vertebrate photoreceptors. J Cell Sci. 2008;121:1907–1915. [PMC free article] [PubMed]
9. Tsujikawa M, Malicki J. Intraflagellar transport genes are essential for differentiation and survival of vertebrate sensory neurons. Neuron. 2004;42:703–716. [PubMed]
10. Insinna C, Besharse JC. Intraflagellar transport and the sensory outer segment of vertebrate photoreceptors. Dev Dyn. 2008;237:1982–1992. [PMC free article] [PubMed]
11. Kozminski KG, Beech PL, Rosenbaum JL. The Chlamydomonas kinesin-like protein FLA10 is involved in motility associated with the flagellar membrane. J Cell Biol. 1995;131:1517–1527. [PMC free article] [PubMed]
12. Snow JJ, Ou G, Gunnarson AL, et al. Two anterograde intraflagellar transport motors cooperate to build sensory cilia on C. elegans neurons. Nat Cell Biol. 2004;6:1109–1113. [PubMed]
13. Pazour GJ, Dickert BL, Witman GB. The DHC1b (DHC2) isoform of cytoplasmic dynein is required for flagellar assembly. J Cell Biol. 1999;144:473–481. [PMC free article] [PubMed]
14. Porter ME, Bower R, Knott JA, Byrd P, Dentler W. Cytoplasmic dynein heavy chain 1b is required for flagellar assembly in Chlamydomonas. Mol Biol Cell. 1999;10:693–712. [PMC free article] [PubMed]
15. Signor D, Wedaman KP, Orozco JT, et al. Role of a class DHC1b dynein in retrograde transport of IFT motors and IFT raft particles along cilia, but not dendrites, in chemosensory neurons of living Caenorhabditis elegans. J Cell Biol. 1999;147:519–530. [PMC free article] [PubMed]
16. Zhang Z, Tanaka Y, Nonaka S, et al. The primary structure of rat brain (cytoplasmic) dynein heavy chain, a cytoplasmic motor enzyme. Proc Natl Acad Sci U S A. 1993;90:7928–7932. [PubMed]
17. Hou Y, Pazour GJ, Witman GB. A dynein light intermediate chain, D1bLIC, is required for retrograde intraflagellar transport. Mol Biol Cell. 2004;15:4382–4394. [PMC free article] [PubMed]
18. Mikami A, Tynan SH, Hama T, et al. Molecular structure of cytoplasmic dynein 2 and its distribution in neuronal and ciliated cells. J Cell Sci. 2002;115:4801–4808. [PubMed]
19. Perrone CA, Tritschler D, Taulman P, Bower R, Yoder BK, Porter ME. A novel dynein light intermediate chain colocalizes with the retrograde motor for intraflagellar transport at sites of axoneme assembly in Chlamydomonas and Mammalian cells. Mol Biol Cell. 2003;14:2041–2056. [PMC free article] [PubMed]
20. Rompolas P, Pedersen LB, Patel-King RS, King SM. Chlamydomonas FAP133 is a dynein intermediate chain associated with the retrograde intraflagellar transport motor. J Cell Sci. 2007;120:3653–3665. [PubMed]
21. Pazour GJ, Wilkerson CG, Witman GB. A dynein light chain is essential for the retrograde particle movement of intraflagellar transport (IFT) J Cell Biol. 1998;141:979–992. [PMC free article] [PubMed]
22. Qin H, Diener DR, Geimer S, Cole DG, Rosenbaum JL. Intraflagellar transport (IFT) cargo: IFT transports flagellar precursors to the tip and turnover products to the cell body. J Cell Biol. 2004;164:255–266. [PMC free article] [PubMed]
23. May SR, Ashique AM, Karlen M, et al. Loss of the retrograde motor for IFT disrupts localization of Smo to cilia and prevents the expression of both activator and repressor functions of Gli. Dev Biol. 2005;287:378–389. [PubMed]
24. Rana AA, Barbera JP, Rodriguez TA, et al. Targeted deletion of the novel cytoplasmic dynein mD2LIC disrupts the embryonic organiser, formation of the body axes and specification of ventral cell fates. Development. 2004;131:4999–5007. [PubMed]
25. Calvert PD, Strissel KJ, Schiesser WE, Pugh EN, Jr, Arshavsky VY. Light-driven translocation of signaling proteins in vertebrate photoreceptors. Trends Cell Biol. 2006;16:560–568. [PubMed]
26. Philp NJ, Chang W, Long K. Light-stimulated protein movement in rod photoreceptor cells of the rat retina. FEBS Lett. 1987;225:127–132. [PubMed]
27. Brann MR, Cohen LV. Diurnal expression of transducin mRNA and translocation of transducin in rods of rat retina. Science. 1987;235:585–587. [PubMed]
28. Whelan JP, McGinnis JF. Light-dependent subcellular movement of photoreceptor proteins. J Neurosci Res. 1988;20:263–270. [PubMed]
29. Nair KS, Hanson SM, Mendez A, et al. Light-dependent redistribution of arrestin in vertebrate rods is an energy-independent process governed by protein-protein interactions. Neuron. 2005;46:555–567. [PMC free article] [PubMed]
30. Broekhuyse RM, Tolhuizen EF, Janssen AP, Winkens HJ. Light induced shift and binding of S-antigen in retinal rods. Curr Eye Res. 1985;4:613–618. [PubMed]
31. Westerfield M. The Zebrafish Book. 3rd ed. University of Oregon Press; Eugene, OR: 1995.
32. Hubbard TJP, Aken BL, Beal K, et al. Ensembl 2007. Nucleic Acids Res. 2007;35:D610–D617. [PubMed]
33. Nasevicius A, Ekker SC. Effective targeted gene ‘knockdown’ in zebrafish. Nat Genet. 2000;26:216–220. [PubMed]
34. Jowett T, Lettice L. Whole-mount in situ hybridizations on zebrafish embryos using a mixture of digoxigenin- and fluorescein-labelled probes. Trends Genet. 1994;10:73–74. [PubMed]
35. Schmitt EA, Dowling JE. Early retinal development in the zebrafish, Danio rerio: light and electron microscopic analyses. J Comp Neurol. 1999;404:515–536. [PubMed]
36. Riley BB, Chiang M, Farmer L, Heck R. The deltaA gene of zebrafish mediates lateral inhibition of hair cells in the inner ear and is regulated by pax2.1. Development. 1999;126:5669–5678. [PubMed]
37. Vihtelic TS, Doro CJ, Hyde DR. Cloning and characterization of six zebrafish photoreceptor opsin cDNAs and immunolocalization of their corresponding proteins. Vis Neurosci. 1999;16:571–585. [PubMed]
38. Peterson JJ, Tam BM, Moritz OL, et al. Arrestin migrates in photoreceptors in response to light: a study of arrestin localization using an arrestin-GFP fusion protein in transgenic frogs. Exp Eye Res. 2003;76:553–563. [PubMed]
39. Wong KY, Adolph AR, Dowling JE. Retinal bipolar cell input mechanisms in giant danio. I. Electroretinographic analysis. J Neurophysiol. 2005;93:84–93. [PubMed]
40. Emran F, Rihel J, Adolph AR, Wong KY, Kraves S, Dowling JE. OFF ganglion cells cannot drive the optokinetic reflex in zebrafish. Proc Natl Acad Sci U S A. 2007;104:19126–19131. [PubMed]
41. Bisgrove BW, Snarr BS, Emrazian A, Yost HJ. Polaris and polycystin-2 in dorsal forerunner cells and Kupffer's vesicle are required for specification of the zebrafish left-right axis. Dev Biol. 2005;287:274–288. [PubMed]
42. Kramer-Zucker AG, Olale F, Haycraft CJ, Yoder BK, Schier AF, Drummond IA. Cilia-driven fluid flow in the zebrafish pronephros, brain and Kupffer's vesicle is required for normal organogenesis. Development. 2005;132:1907–1921. [PubMed]
43. Kennedy MJ, Dunn FA, Hurley JB. Visual pigment phosphorylation but not transducin translocation can contribute to light adaptation in zebrafish cones. Neuron. 2004;41:915–928. [PubMed]
44. Insinna C, Pathak N, Perkins B, Drummond I, Besharse JC. The homodimeric kinesin, Kif17, is essential for vertebrate photoreceptor sensory outer segment development. Dev Biol. 2008;316:160–170. [PMC free article] [PubMed]
45. Roof D, Adamian M, Jacobs D, Hayes A. Cytoskeletal specializations at the rod photoreceptor distal tip. J Comp Neurol. 1991;305:289–303. [PubMed]
46. Huangfu D, Anderson KV. Cilia and Hedgehog responsiveness in the mouse. Proc Natl Acad Sci U S A. 2005;102:11325–11330. [PubMed]
47. Tran PV, Haycraft CJ, Besschetnova TY, et al. THM1 negatively modulates mouse sonic hedgehog signal transduction and affects retrograde intraflagellar transport in cilia. Nat Genet. 2008;40:403–410. [PubMed]
48. McGinnis JF, Whelan JP, Donoso LA. Transient, cyclic changes in mouse visual cell gene products during the light-dark cycle. J Neurosci Res. 1992;31:584–590. [PubMed]
49. McGinnis JF, Matsumoto B, Whelan JP, Cao W. Cytoskeleton participation in subcellular trafficking of signal transduction proteins in rod photoreceptor cells. J Neurosci Res. 2002;67:290–297. [PubMed]