The roles of cytoplasmic dynein complex in various ocular disorders

Abstract

Background

The cytoplasmic dynein complex mediates retrograde transport of various intracellular components and plays a critical role in mitosis, nuclear migration, organelle positioning, vesicle trafficking, misfolded protein clearance, and intercellular signaling.

Main body

Mutations in human cytoplasmic dynein subunits and regulators have been directly linked to various neurological diseases and are increasingly recognized as contributors to ocular degenerative diseases. However, the precise mechanisms underlying ocular morphogenesis and degeneration remain poorly understood. Various animal models, including Drosophila, mouse, and zebrafish, have been established to investigate the pathogenesis of cytoplasmic dynein complex-related ocular disorders. Findings from these models indicate that dynein complex-related ocular pathologies often involve endoplasmic reticulum stress, impaired Notch signaling, and disrupted Sonic Hedgehog pathways. Systematically integrating gene functional data and molecular mechanism clues derived from different models can help refine the “gene-phenotype-mechanism” correlation network and may ultimately advance precision diagnosis and targeted therapy.

Conclusion

Emerging evidence has highlighted the cytoplasmic dynein complex as a key factor in ocular development and disease. Integrating findings across model systems may enable more precise diagnosis and the development of targeted interventions for dynein-related ocular disorders.

Background

Dyneins are large ATP-dependent motor protein complexes first discovered for their role in driving flagellar movement in Tetrahymena pyriformis (Gibbons and Rowe 1965). They are classified into two main categories: axonemal dyneins, which are anchored on doublet microtubules within cilia and function as molecular engines for ciliary beating (Walton et al. 2021), and cytoplasmic dyneins (hereafter referred to as dyneins), which move along microtubules and are responsible for the retrograde transport of various cellular components (Reck-Peterson et al. 2018; Roberts 2018).

Dynein-1 powers the minus-end-directed transport of various cargos within the cytoplasm, including membrane-bound organelles, protein complexes, RNAs, and viruses (Reck-Peterson et al. 2018). Dynein-2 is a ubiquitous motor responsible for retrograde intraflagellar transport (IFT) within the flagella and cilia (Roberts 2018). As such, dynein plays critical roles in mitosis, nuclear migration, organelle positioning, vesicle trafficking, misfolded protein clearance, and intercellular signaling (Eschbach and Dupuis 2011; Karki and Holzbaur 1999; Merdes et al. 1996).

Ocular degenerative diseases are the major blinding ophthalmic disorders worldwide, mainly including glaucoma, aged-related macular degeneration, retinitis pigmentosa, and diabetic retinopathy. While these diseases may have distinct etiologies and pathogenesis, they are all characterized by the progressive loss of photoreceptor cell (PRC) or retinal ganglion cell (RGC), which may be triggered by a combination of genetic, environmental, and aging factors (Wang and Cepko 2022; Cuenca et al. 2014).

Mutations in dynein subunits and their regulators have been directly linked to various neurological diseases (Lipka et al. 2013). Recently, they have emerged as important therapeutic targets for the treatment of ocular degenerative diseases. Although several studies have attempted to elucidate their involvement in ocular pathology, the precise mechanisms underlying ocular morphogenesis and degeneration remain poorly understood.

This review first outlines the structure and mechanics of dynein complexes, followed by a discussion of the genetic and clinical spectrum of dynein complex-related disorders. Finally, we examine the current animal models of the dynein complex to describe the main characteristics of their ocular disorders and their potential pathological mechanisms. These animal models provide valuable insights into dynein complex function and contribute to the development of targeted therapies for dynein complex-related ocular disorders.

Structure and mechanics of dynein complex

Dynein subunits

Dynein-1 is a dimeric complex comprising six subunit classes. It consists of heavy chains (DHC1; encoded by DYNC1H1) and a set of accessory components, including intermediate chains (DIC1/2; encoded by DYNC1I1/2), light intermediate chains (DLIC1/2; encoded by DYNC1LI1/2), and three light chains: Roadblock (ROBL1/2; encoded by DYNLRB1/2), LC8-1/2 (encoded by DYNLL1/2), and TCTEX1/3 (encoded by DYNLT1/3) (Reck-Peterson et al. 2018; Canty et al. 2021) (Fig. 1, Table 1)

Fig. 1.

Architecture of the dynein-1 complex (PDB 5NVU). The N-terminal third of the DHCs is referred to as the tail domain, which serves as a scaffold for other chains. These two DHCs are linked through the N-terminal dimerization domain. The DICs possess extended N-termini that associate with homodimers of ROBL1/2, LC8-1/2, and TCTEX1/3. The DLICs combine midway through the tail domain. The remaining motor domain forms a stacked structure with crossed coiled-coil stalks

Table 1.

Subunit composition of dynein and dynactin complexes

Complex	Type	Subunit	Gene
Dynein-1	Heavy	DHC1	DYNC1H1
	Intermediate	DIC1/2	DYNC1I1/2
	Light intermediate	DLIC1/2	DYNC1LI1/2
	Light	ROBL1/2	DYNLRB1/2
		LC8-1/2	DYNLL1/2
		TCTEX1/3	DYNLT1/3
Dynein-2	Heavy	DHC2	DYNC2H1
	Intermediate	WDR34 or DIC5	DYNC2I2
		WDR60 or DIC6	DYNC2I1
	Light intermediate	D2LIC or LIC3	DYNC2LI1
	Light	TCTEX1D2	DYNLT2B
Dynactin	ARP1 filament	ARP1	ACTR1A/B
		β-actin	ACTB
	Pointed end	p25	DCTN5
		p27	DCTN6
		p62	DCTN4
		ARP11	ACTR10
	Barbed end	CAPZ	CAPZA1/2, CAPZB
	Shoulder-like domain	p150Glued or p150 or p135	DCTN1
		Dynamitin or p50	DCTN2
		p22 or p24	DCTN3

The shared subunits between dynein-1 and dynein-2 are shown in bold

Dynein-2 comprises five subunit classes, including a heavy-chain homodimer (DHC2; encoded by DYNC2H1), two distinct intermediate chains (WDR34/60, also known as DIC5/6; encoded by DYNC2I2/1), a light intermediate chain (D2LIC, also known as LIC3; encoded by DYNC2LI1), a dynein-2-specific light chain (TCTEX1D2; encoded by DYNLT2B), in addition to sharing the three light chains with dynein-1 (ROBL1/2, LC8-1/2, and TCTEX1) (Vuolo et al. 20; Asante et al. 2014; Toropova et al. 2019) (Fig. 2, Table 1).

Fig. 2.

Architecture of the dynein-2 complex (derived from PDB 6SC2 and 8RGI). The two DHCs are dimerized by a heterodimer of two distinct intermediate chains (WDR34 and WDR60), which are attached by LIC3 and TCTEX1D2 and share three common light chains with dynein-1. Notably, the N-proximal regions of WDR34 and WDR60 are linked by heterodimers of TCTEX1 and TCTEX1D2, in contrast to other homodimers of the light chains

Dynein regulators

Isolated dynein adopts an autoinhibited conformation, known as the φ-particle, which is incapable of generating robust motility unless assembled with other accessory regulators that induce or enhance its motor processivity (Asante et al. 2014; Torisawa et al. 2014; Tirumala and Ananthanarayanan 2020).

Among these regulators, the dynactin complex is one of the best-characterized and serves as an essential partner of dynein-1 (Schroer 2004). However, it has not been found to be associated with dynein-2 (Asante et al. 2014). The core of the dynactin complex consists of an ARP1 filament, which contains ARP1 (encoded by ACTR1A and ACTR1B) and β-actin (encoded by ACTB). The pointed end of this filament is capped by p25 (encoded by DCTN5), p27 (encoded by DCTN6), p62 (encoded by DCTN4), and ARP11 (encoded by ACTR10), whereas the barbed end is capped by CAPZ (encoded by CAPZA1/2 and CAPZB). A shoulder-like domain extends from the ARP1 filament and contains p150^Glued (also known as p150 or p135; encoded by DCTN1), dynamitin (also known as p50; encoded by DCTN2), and p22 (also known as p24; encoded by DCTN3) (Reck-Peterson et al. 2018; Canty and Yildiz 2020) (Fig. 3, Table 1).

Fig. 3.

Architecture of the dynactin complex (PDB 5ADX). The core structure comprises a top protofilament of five ARP1 and a bottom protofilament of a β-actin and three ARP1. The pointed end is capped by p25, p27, p62, or ARP11. The barbed end is capped by CAPZ. Atop the ARP1 filament sits a shoulder-like domain containing p150.^Glued, dynamitin, and p22

Notably, dynactin has a low affinity for dynein-1 and exerts only a minor effect on its velocity and ATPase activity in vitro (Kardon et al. 2009; King et al. 2003; King and Schroer 2000). Dynein-1 assembles a tripartite complex with its principal regulator dynactin and a family of activating adaptors (dynein-dynactin-adaptor, DDA), which increases both processivity and velocity on the microtubule tracks (Urnavicius et al. 2015; Chaaban and Carter 2022). An expanding family of activating adaptors has been identified that form complexes with dynein-dynactin, thereby enhancing the motility of the motor complex. The best-characterized activators have been identified via in vitro reconstitution experiments and include HOOK3 (McKenney et al. 2014), Bicaudal D2 (BICD2) (Schlager et al. 2014), BICD-related protein 1 (BICDR1, also known as BICDL1) (Urnavicius et al. 2018), ninein (NIN) (Redwine, et al., 2017), ninein-like protein (NINL) (Redwine, et al., 2017), RAB45 (Wang et al. 2019), and CRACR2a (Wang et al. 2019). Additional activators, such as HOOK1 (Olenick et al. 2016), Rab family interacting protein 3 (RAB11-FIP3) (McKenney et al. 2014), and Spindly (SPDL1) (Gama et al. 2017), have been identified using crude cell lysates, but they have not yet been tested with purified components. Furthermore, several other molecules have been implicated as potential candidates, including BICD1 (Fumoto et al. 2006), HOOK2 (Dwivedi et al. 2019), Klarsicht/Anc-1/Syne homology domain protein 2 (KASH2, also known as Syne2 or Nesprin2) (Yu et al. 2011), KASH5 (Agrawal, et al., 2022), Sad1/UNC-84 domain protein 1 (SUN1) (Yu et al. 2011), trafficking kinesin-binding protein 1/2 (TRAK1/2) (Canty et al. 2023), c-Jun N-terminal kinase-interacting protein 3 (JIP3) (Celestino 2022), huntingtin-associated protein 1 (HAP1) (Li and Li 2005), Rab7-interacting lysosomal protein (RILP) (Johansson et al. 2007), nuclear mitotic apparatus protein (NUMA) (Merdes et al. 2000), and coiled-coil domain protein 88 A/B/C (CCDC88A/B/C) (Redwine, et al., 2017); Ham et al. 2015) (Fig. 4, Table 2).

Fig. 4.

Schematic representation of the activating adaptors and their associated cellular cargos for dynein-1. Adaptors that are confirmed through in vitro reconstitution experiments or in crude cell lysates are shown in regular font, while the hypothesized adaptors are shown in italic font

Table 2.

Activating adaptors and their associated cellular cargos for dynein-1

Adaptor	Cargo	Reference
Identified in vitro reconstitution experiments
HOOK3	Secretory vesicles	McKenney et al. 2014)
BICD2	Golgi-derived vesicles, cytoplasmic vesicles	Schlager et al. 2014)
BICDR1 or BICDL1	Secretory vesicles	Urnavicius et al. 2018)
NIN	Centrosome	Redwine, et al., 2017)
NINL	Centrosome	Redwine, et al., 2017)
RAB45	Endosomes	Wang et al. 2019)
CRACR2a	Endosomes	Wang et al. 2019)
Identified in crude cell lysates
HOOK1	Endosomes	Olenick et al. 2016)
RAB11-FIP3	Recycling endosomes	McKenney et al. 2014)
SPDL1	Kinetochore	Gama et al. 2017)
Potential candidates
BICD1	Golgi-derived vesicles, cytoplasmic vesicles	Fumoto et al. 2006)
HOOK2	Golgi-derived vesicles	Dwivedi et al. 2019)
KASH2 or Syne2 or Nesprin2	Chromosome	Yu et al. 2011)
KASH5	Chromosome	Agrawal, et al., 2022)
SUN1	Chromosome	Yu et al. 2011)
TRAK1/2	Mitochondria	Canty et al. 2023)
JIP3	Amyloid precursor protein vesicles, lysosomes, c-Jun N-terminal kinase	Celestino, et al., 2022)
HAP1	Autophagosomes, brain-derived neurotrophic factor-containing vesicles	Li and Li 2005)
RILP	Late endosomes, lysosomes	Johansson et al. 2007)
NUMA	Microtubules	Merdes et al. 2000)
CCDC88A/B/C	Endosomes, lytic granules	Redwine, et al., 2017; Ham et al. 2015)

Lissencephaly-1 (LIS1) functions as a key regulator adaptor that interacts with nuclear distribution element 1 (NDE1, also known as NudE) and NDE-like 1 (NDEL1, also known as Nudel) to regulate the activation of dynein-1. Under the facilitation of NDE1/NDEL1, LIS1 binds directly to the motor domain of dynein-1 (Huang et al. 2012). This interaction promotes the transition of the dynein-1 φ-particle into an open conformation, facilitating the assembly of DDA. Notably, once DDA is fully assembled, LIS1 dissociates and no longer participates in the subsequent motion process (Htet et al. 2020; Elshenawy et al. 2020).

Compared with dynein-1, the mechanochemistry and regulatory mechanism of dynein-2 remain significantly less well understood. Generally, the states of dynein-2 can be tightly coupled to the dynamics of the IFT trains. The core IFT machinery contains IFT-A and IFT-B complexes. First, the IFT-B complexes polymerize at the microtubule doublets at the base of cilia and act as a platform for the recruitment and polymerization of IFT-A complexes, forming a “double-decker” polymer. At the same time, dynein-2 is loaded onto these anterograde trains in the inactive φ conformation, oriented away from the microtubule, in which the two motor domains are stacked together, rendering it unable to bind or move along microtubules. The heterotrimeric kinesin-2 is then recruited and drives anterograde movement, with some membrane cargoes deposited along the axoneme (Jordan et al. 2018; Mitra et al. 2024; Hoek et al. 2022). At the ciliary tip, the anterograde train depolymerizes and remodels into a morphologically distinct retrograde train. This process involves stable conformational changes in IFT-A complexes, each of which recruits two IFT-B complexes. During remodeling, dynein-2 is released, then transitions into its open conformation and binds to the new retrograde train. The active dynein-2 reorients towards the microtubule and initiates retrograde transport (Chien, et al., 2017); Lacey and Pigino 2025; Lacey et al. 2024). Dynein-2 activation may involve distinct adaptors or regulation by post-translational modifications; however, the precise mechanisms remain to be elucidated (Liang et al. 2014).

Furthermore, certain adaptors may function as bidirectional molecular switches, enabling precise spatial–temporal regulation of cargo trafficking (Olenick and Holzbaur 2019). Integrative approaches involving in vitro reconstitution, targeted mutagenesis, and in-depth biophysical characterization will be essential to elucidate their precise roles and regulatory mechanisms in both normal physiology and disease progression.

Genetic and clinical spectrum of dynein complex-related disorders

Defects in the retrograde transport system play a critical role in the pathogenesis of neuronal degeneration in both the central and peripheral nervous systems. Therefore, dynein dysfunction represents a central hub of the pathophysiological mechanisms driving various neurological disorders (Eschbach and Dupuis 2011).

Mutations in human DYNC1H1 are associated with both neuromuscular disorders (NMD) and neurodevelopmental disorders (NDD). NMD manifestations are primarily restricted to the peripheral nervous system, including Charcot-Marie-Tooth disease 2O (CMT2O) (Weedon et al. 2011) and spinal muscular atrophy with lower extremity predominance 1 (SMALED1) (Harms et al. 2012), whereas NDD manifestations extend into the central nervous system and include intellectual disability (Willemsen et al. 2012) and cortical development malformation (Poirier et al. 2013). Hertecant et al. (Hertecant et al. 2016) reported the first DYNC1H1 variant in an infant presenting with cortical development malformation accompanied by cataract, implicating its possible role in ocular development. Möller et al. (Möller et al. 2025) subsequently reported that 48.83% of their patient cohort harboring the DYNC1H1 mutation exhibited multiple ophthalmological features, including strabismus, astigmatism, bilateral cataracts, and limited peripheral and night vision, further supporting the pathogenic impact of DYNC1H1 in ocular diseases.

DYNC2H1 mutations are among the major genetic causes of short-rib polydactyly (SRP) and Jeune asphyxiating thoracic dystrophy (JATD). SRP is typically associated with severe cardiorespiratory failure, resulting in an almost 100% mortality rate in affected neonates. Patients with JATD exhibit variable survival rates (40–80%); however, long-term survivors frequently develop retinal degeneration and other extraskeletal ciliopathies (Baujat et al. 2013; Dagoneau et al. 2009; Hokayem et al. 2012; Schmidts et al. 2013). Because these patients often present with severe multisystem defects and a short life expectancy, their retinal involvement may be underestimated because of clinical oversight. Moreover, several DYNC2H1 variants have been implicated in non-syndromic retinitis pigmentosa (Vig et al. 2020). Overall, DYNC2H1 plays significant roles in both syndromic and non-syndromic retinal degeneration.

WDR34 has been identified as the second most commonly mutated gene after DYNC2H1 in patients with JATD. In rare instances, WDR34 mutations were associated with rod-cone dystrophy (the most common form of retinitis pigmentosa), with or without additional ciliopathies (Solaguren-Beascoa et al. 2021; Schmidts et al. 2013). Similarly, WDR60 can cause ciliopathic symptoms characterized by retinal degeneration with polydactyly (Kakar et al. 2018).

Other genes encoding dynein regulators, including DCTN1, BICD2, and NDE1, have been implicated in a broad spectrum of neurological disorders; however, their potential retinal phenotypes have not yet been described (Lipka et al. 2013).

Overall, dynein and its regulators play pathogenic roles in ocular diseases; however, further investigation is warranted. A comprehensive understanding of the precise functions and molecular mechanisms of ocular morphogenesis and visual function will enhance clinical insights, identify novel candidate pathogenic genes, and facilitate the development of targeted therapies.

Current animal models of dynein complex-related ocular disorders

Gene modifications in animal models frequently produce phenotypes that resemble those of human diseases, offering valuable insights into the molecular basis of target genes. In this section, we summarize animal models involving various components of the dynein complex, with emphasis on their ocular manifestations.

Drosophila models mimicking dynein complex-related ocular disorders

Drosophila lacking dynactin

The Drosophila Glued gene (Gl) encodes a homologue of the vertebrate p150^Glued subunit of dynactin (Holzbaur et al. 1991; Gill et al. 1991). The Gl¹ mutation was initially identified through high-temperature mutagenesis screening (Plough and Ives 1935). Molecular analysis revealed a transposon insertion near the 3’ end of Gl, resulting in truncation of the C-terminal protein (Swaroop et al. 1985). Although the truncated Gl polypeptide cannot assemble into a functional dynactin complex, it retains its dynein-binding domain. Consequently, it acts as a dominant-negative “poison” subunit that disrupts dynein–cargo interactions (McGrail et al. 1995; Vaughan and Vallee 1995). The homozygous Gl¹ mutation results in late embryonic or larval lethality, indicating that Gl¹ is essential for cell viability (Harte and Kankel 1982). A heterozygous Gl¹ mutation (Gl¹/+) results in severe defects in visual system development. The Gl¹/+ mutants typically display a rough-eye phenotype, characterized by irregular ommatidia, fewer and misarrayed bristles, smaller rhabdomeres, and clear disruption of accessory cells (Fan and Ready 1997; Ma et al. 2009). These abnormalities arise from cumulative defects in developmental stages, including mitotic delay, nuclear mispositioning, impaired assembly and extension of adherens junctions, and disrupted axonal projections from the ommatidium to the optic lobe (Fan and Ready 1997; Meyerowitz and Kankel 1978; Fan 2004; Whited et al. 2004).

Drosophila lacking dynein-1

In Drosophila, the cytoplasmic dynein-1 heavy and intermediate chains have been designated as Dhc64C and Dic19C based on their genomic regions in the cytological map (Li et al. 1994; Boylan et al. 2000). Mutations in Dhc64C or Dic19C on a Gl¹/+ background further impair dynein-mediated transport, markedly exacerbating the rough-eye phenotype of Gl¹/+ (McGrail et al. 1995; Boylan et al. 2000; Gepner et al. 1996; Boylan and Hays 2002). Conversely, the introduction of a Dic transgene into a Gl¹/+ background restored the effective linkage between dynein and dynactin and suppressed the rough-eye phenotype (Boylan et al. 2000).

Drosophila lacking dynein-1 adaptors

Several adaptor proteins are essential for the apical migration of PRC nuclei, a process crucial for normal eye morphogenesis in Drosophila. For example, Klaroid (Koi, a SUN-domain protein on the inner nuclear membrane) interacts with Klarsicht (Klar, a KASH-domain protein on the outer nuclear membrane), forming the linker of nucleoskeleton and cytoskeleton (LINC) complex. Klar associates with the microtubule network to anchor the nucleus to the cytoskeleton, thereby facilitating its migration. Drosophila harboring single or double mutations in Koi and Klar can survive and reproduce, exhibiting only a morphological defect characterized by a rough-eye phenotype (Patterson et al. 2004; Kracklauer et al. 2007). Two other adaptors, Bicd and Lis1, have also been implicated in PRC nuclear migration; their mutant phenotypes are comparatively milder than those observed in Koi and Klar (Swan et al. 1999).

Mouse models mimicking dynein complex-related ocular disorders

Mouse lacking Dync1h1

Germline disruption of mouse Dync1h1 (cDHC^−/−, truncation after exon 1) results in embryonic lethality before embryonic day (E) 8.5 (Harada et al. 1998). Several mouse mutants, including legs at odd angles (Loa, harboring the Dync1h1^F580Y mutation) (Hafezparast et al. 2003; Ilieva et al. 2008), Cramping1 (Cra1, harboring the Dync1h1^Y1055C mutation) (Hafezparast et al. 2003; Dupuis et al. 2009), Sprawling (Swl, harboring the Dync1h1^{[GIVT]1040[A]} mutation) (Chen et al. 2007; Brook and Duchen 1990), and CMT2O mouse model (harboring the Dync1h1^H304R mutation) (Sabblah et al. 2018; Nandini et al. 2019), exhibit proprioceptive sensory or motor coordination defects. Nevertheless, the potential retinal phenotypes have not been assessed.

Conditional knockout of Dync1h1 in mouse retina using Six3Cre (^retDync1h1^−/−, truncation after exon 23) initiates gene ablation at E9, and DYNC1H1 is effectively silenced in the inner segment (IS) of PRC by postnatal day (P) 6 (Dahl et al. 2021). Phenotypically, ^retDync1h1^−/− mice exhibited varying degrees of microphthalmia, with rapid PRC degeneration within two postnatal weeks. Notably, Six3 increases Cre recombinase expression more strongly in the central retina than in the peripheral retina (Diacou et al. 2018). Consequently, the central retina of ^retDync1h1^−/− mice exhibited severe disorganization of the outer/inner nuclear layer (ONL/INL), an unrecognizable outer segment (OS), and near-complete absence of the microtubule cytoskeleton. In contrast, residual Dync1h1 expression in the peripheral retina enabled partial preservation of retinal lamination and OS elongation (Dahl et al. 2021; Gerstner 2022; Dahl and Baehr 2021).

The Prom1-CreER^T2 system enables tamoxifen-inducible deletion of Dync1h1 in adult PRC (^tamDync1h1^−/−, truncation after exon 23). Tamoxifen was administered intraperitoneally between P21 and P30, followed by daily injections for five consecutive days (Dahl et al. 2021; Jászai et al. 2007). The OS/IS/ONL of ^tamDync1h1^−/− mice began to shrink 3 weeks post-tamoxifen induction (wPTI) and were nearly eliminated by 4 wPTI. The efficiency of tamoxifen induction varied, potentially due to incomplete dissolution (Dahl et al. 2021).

Subretinal injection of AAV2/8-shDync1h1 in P1 mice selectively decreased Dync1h1 expression in rod, with Dync1h1 mRNA levels decreasing by 46% at 10 days post-transduction. Three weeks post administration, the rod nuclei were ectopically positioned within the outer plexiform layer (OPL) and INL, resulting in disrupted lamination of the ONL. Synaptic organization was also disrupted in these basally displaced cells, as evidenced by mislocalized presynaptic C-terminal binding protein 2 (CTBP2, also known as RIBEYE) and a reduced alignment with postsynaptic metabotropic glutamate receptor 6 (mGluR6) (Aghaizu et al. 2021).

Furthermore, iCre75 drove rod-specific deletion of Dync1h1 by P10, with protein levels of Dync1h1 reduced by P16 (^rodDync1h1^−/−, truncation after exon 23) (Dahl et al. 2021; Li et al. 2005). In ^rodDync1h1^−/− mice, rod degeneration began around P16 and was complete by P30. The OS/IS/ONL diminished concurrently during this period, consistent with the disrupted transport of the OS protein phosphodiesterase 6 (PDE6) and synaptic protein RIBEYE. Additionally, cone function remained stable under rod-specific knockout conditions until P21. However, with the gradual decline in rod-derived cone viability factor (RdCVF), the cone underwent secondary degeneration and ultimately became functionally inactive by P30 (Dahl et al. 2021).

In the mouse retinal pigment epithelium (RPE), melanosomes dynamically reposition in response to variable changes in light intensities. During dark adaptation, melanosomes migrate into the RPE cell body, whereas under light conditions, they extend into narrow apical processes, with their long axis parallel to the light path, thereby shielding the surrounding PRC OS (Futter et al. 2004). Jiang et al. investigated the role of Dync1h1 in RPE melanosome dynamics by subretinal injection and electroporation of Dync1h1 shRNA in P2–4 mice (Jiang 2020). In the transfected regions, Dync1h1 shRNA disrupted this physical movement; melanosomes were largely absent from the apical RPE and were redistributed throughout the cell body. The resulting loss of protection around the PRC OS likely increased their susceptibility to light-induced damage (Jiang 2020).

Mouse lacking Dlic1

Dlic1 is not essential for the viability or fertility of mice (Banks et al. 2011; Kong et al. 2013). The first reported Dlic1 mutant (Dlic1^N235Y, a gain-of-function mutation) exhibited increased anxiety-like behavior and changes in neuronal development, but no retinal phenotype was detected (Banks et al. 2011).

The Dlic1^−/− mouse (truncation after exon 4) exhibited gradual retinal degeneration. Loss of Dlic1 destabilized dynein subunits, partially impaired endoplasmic reticulum (ER) export, and disrupted Rab11-mediated vesicle trafficking from the Golgi apparatus to the basal body. Consequently, OS proteins such as rhodopsin and arrestin abnormally accumulated within the IS and OPL, impairing OS growth and ciliogenesis, and ultimately inducing PRC apoptosis. These pathological alterations were most pronounced between P20 and P30 and progressed at a much slower rate as age increased (Kong et al. 2013).

Mouse lacking dynein-1 adaptors

Sun1 (a SUN-domain protein on the inner nuclear membrane) and Syne2 (a KASH-domain protein on the outer nuclear membrane) are key components of the LINC complex that mediate PRC nuclear migration. Sun1^−/− and Syne2^−/− mice exhibited similar retinal phenotypes. During the early developmental stages, both exhibited mislocalized PRC nuclei within the OPL and INL, excessive apoptosis, reduced ONL thickness, and electrophysiological dysfunction. These findings highlight the essential role of Sun1 and Syne2 in early retinal development. Notably, these abnormalities resolved spontaneously in adult mice, possibly reflecting a reduced dependence on the LINC complex and compensatory activity from homologous proteins in adulthood (Yu et al. 2011; Razafsky et al. 2012).

Mouse lacking dynein-2

Dynein-2 subunits implicated in human retinal degeneration include DYNC2H1, WDR34, and WDR60. Homozygous mutations in these genes were embryonically lethal. Prior studies have already demonstrated that both Dync2h1 and Wdr34 are essential for axonemal extension, whereas Wdr60 is essential for dynein-2 assembly (Yan et al. 2023; Huangfu and Anderson 2005; Ocbina et al. 2011). A recent investigation further elucidated the critical roles of Wdr34 for PRC survival and function by generating rod- and cone-conditional knockout mice (Wdr34^RKO and Wdr34 ^HKO, truncation after exon 2) (Zou et al. 2026). The rod-specific promoter (iCre75) mediated excision became detectable at P7, and was completed by P18 (Li et al. 2005). Conditional knockout of Wdr34 in rod down-regulated a suite of genes related to axonemal integrity and microtubule-based transport, which disrupted the stability of ciliary axonemes, thereby impaired the efficient proteins supply of PRC OS, ultimately triggered OS shortening and rod degeneration with significant declines in scotopic electroretinography (Zou et al. 2026). Similarly, the cone-specific promoter (HRGP-Cre) mediated recombination occurred before P10 (Le et al. 2004), leading to impaired photopic electroretinography followed by cone death (Zou et al. 2026).

These findings imply complex associations between dynein-2 and the morphogenesis and functional maintenance of PRC; however, further analyses of ocular phenotypes and pathogenic mechanisms are required to establish more empirical evidence.

Zebrafish models mimicking dynein complex-related ocular disorders

Zebrafish lacking Dync1h1

The dync1h1^Y3102* mutation (dync1h1^mw20, truncated after exon 43) results in the cannonball (cnb) phenotype in zebrafish (Insinna et al. 2010). Homozygous mutants exhibit severe disorganization of PRC organelles, impaired post-Golgi vesicle trafficking, and defective OS morphogenesis. Rhodopsin was mislocalized within the ONL, and the nucleus was displaced toward the synaptic region. The embryos died within 6–8 days post-fertilization (dpf) due to rapid pan-cellular degeneration. Additionally, cnb embryos often exhibit a characteristic darkened appearance, reflecting the disrupted melanosome transport (Insinna et al. 2010; Fogerty et al. 2016).

Notably, teleost, which possesses a fixed pupil, cannot regulate the amount of incoming light via the pupillary response. To adapt to dramatic changes in illumination, they instead rely on retinomotor movements and morphological adjustments that reposition the PRC OS. In dark-adapted retina, cone elongates and becomes surrounded by the apical RPE, whereas rod contracts to detect dim light stimuli, thereby enabling scotopic vision. Conversely, in the light-adapted retina, rod elongates and becomes enveloped by the apical RPE for protection, whereas cone contracts to detect bright light and color stimuli, thereby mediating photopic and color vision. Concurrently, melanosomes within the RPE shift their positions between the cell body and the apical process based on the light cycle (as mentioned in 4.2.1) (Levinson and Burnside 1981; Lewis et al. 2018). The dynamic contraction and elongation of PRC depend on microtubule-concentrated myoid (Warren and Brunside 1978). During the elongation process, the dynein-1 complex transports additional polymerized microtubules to the end of the myoid and bundles with existing microtubules, leading to myoid elongation (Lewis et al. 2018). The heterozygous dync1h1^Y3102* mutation markedly impairs myoid elongation, supporting its role as a force-generating motor in microtubule shuttling (Lewis et al. 2018).

We previously employed a novel dync1h1^E42fs7* mutation (truncation at exon 1) to determine its precise roles in PRC morphogenesis and degeneration (dync1h1^−/− and dync1h1^±). Our data demonstrated that our dync1h1^−/− zebrafish closely resembled the defects in cnb. At 5 dpf, dync1h1^−/− zebrafish exhibited marked microphthalmia and severe disorganization of retinal lamination without polarized OS/IS structures. Phototransduction proteins failed to be transported to their designated OS domains and instead accumulated around the basal body, where they were degraded, a process that ultimately triggered ER stress-induced PRC apoptosis. In contrast to the severe damage observed in homozygotes, dync1h1^± zebrafish exhibited a mild but progressive phenotype, with no obvious rate difference during the indicated stages (Zhou et al. 2025).

Moreover, morpholino knockdown of Dync1h1 (dync1h1 MO) resulted in dose-dependent abnormalities. Embryos injected with a high dose of antisense morpholino phenocopied the cnb defects, whereas the low-dose embryos primarily exhibited OS abnormalities without overt IS polarity or trafficking defects (Insinna et al. 2010).

Zebrafish lacking Dlic

Knockout of dlic1 in zebrafish (dync1li1^−/−, harboring the dync1li1^N25fs16* mutation, truncation at exon 1) did not produce global defects, but rather resulted in a cone-dominant progressive degeneration phenotype. In dync1li1^−/− retina, Rab8 accumulated abnormally in the OPL, indicating disrupted post-Golgi vesicle trafficking and progressively altered expression and localization of cone opsins, particularly blue-cone opsins. Ectopic accumulation of opsins ultimately induced PRC apoptosis (Zhang et al. 2023). However, no abnormalities were detected in the rod OS of dync1li1^−/− zebrafish (Zhang et al. 2023), contrasting with the pronounced impairment of rod OS in Dlic1^−/− mouse (Kong et al. 2013). This discrepancy likely reflects interspecies differences in the ratio of cone to rod between teleosts and rodents. Teleost retina is cone-dominant, comprising approximately 92% cone in larvae and 60% cone in adults (Zang and Neuhauss 2021), whereas rodent retina is rod-dominant, comprising 97% rod and 3% cone at all ages (Carter-Dawson and LaVail 1979).

A dlic2 knockout zebrafish line was generated, but no defects were detected in dync1li2^−/− PRC at 1.5 years of age, indicating that dlic2 is not essential in the development and maintenance of the retina (Zhang et al. 2023).

Zebrafish lacking dynactin

Mutations in the zebrafish mikre oko (mok, also known as dctn1a), encoding a homologue of the vertebrate p150^Glued subunit of dynactin, were detected in some large-scale chemical mutagenesis screenings: mok^s632 and mok^s309 (the latter was formerly named bug^s309) (Malicki et al. 1996; Wehman et al. 2005), which introduced a premature stop codon at amino acids 799 and 867, respectively, completely deleting the C-terminal third of the protein (Tsujikawa et al. 2007; Bene et al. 2008). The mok PRC exhibited generalized degeneration, including loss of the characteristic elongated morphology, basal displacement of nuclei toward the synaptic terminus, defects in post-Golgi vesicle trafficking, and ectopic distribution of phototransduction proteins (Fogerty et al. 2016; Tsujikawa et al. 2007; Bene et al. 2008; Doerre and Malicki 2001). Furthermore, mutant retina exhibited excessive production of RGC and severe depletion of Müller glia cell (MGC) and bipolar cell (Bene et al. 2008; Doerre and Malicki 2001). This is because all retinal neurons arise from a common progenitor pool, and RGC is the first neuronal subtype born in the retina (Ohnuma et al. 2002; Burrill and Easter 1995). The mok mutation induces early cell cycle exit, thereby resulting in a relative overproduction of RGC, whereas other late-differentiated neurons, such as MGC and bipolar cell, become relatively underrepresented (Bene et al. 2008). Genetic mosaic analysis has demonstrated that the mok PRC phenotype primarily involves cell-nonautonomous interactions (Doerre and Malicki 2001), whereas the proportion of different retinal cell types depends on cell-autonomous interactions (Bene et al. 2008). Although both mutants exhibited no gross morphological defects, they nonetheless died before 10 dpf (Doerre and Malicki 2001).

Teleost has a unique and duplicated homologue, dctn1b or mok2, sharing 77% sequence identity with dctn1a and potentially providing a functional compensation in the mok mutant (Tsujikawa et al. 2007). Therefore, the complete impact of dctn1 on PRC was verified using morpholino-mediated depletion of both dctn1a and dctn1b (dctn1a/b MO). The retinal defects of dctn1a/b MO were more severe than those of dctn1a deletion alone, but much less than those of dync1h1 deletion. In dctn1a/b MO PRC, polarization defects in the IS and mislocalized rhodopsin in the cell body were detected; however, most OS appeared morphologically normal or only slightly shortened (Insinna et al. 2010). These findings support its auxiliary role in IS morphogenesis rather than OS.

Zebrafish ale oko (ako, also known as dctn2) encodes the zebrafish homologue of the vertebrate p50 subunit of dynactin. The ako^jj50 mutant (harboring the dctn2^T262ins2* mutation, truncation after exon 10) exhibited rapid degeneration in both PRC and MGC, which was more severe than that in the mok. Moreover, degeneration of MGC in ako^jj50 resulted in rapid loss of apical–basal polarity. The apical proteins Nagie oko (Nok, also known as MPP5a) and atypical protein kinase C (aPKC) were found to accumulate ectopically within the cell body, concomitant with the loss of apical processes and basal perikaryal displacement toward the vitreous surface. At later developmental stages, ako^jj50 RGC failed to form densely branched axonal processes in the optic tectum, instead developing sparse or irregular branches. Given that progressive degeneration of RGC is a key pathological hallmark of glaucoma, these findings suggest a potential correlation between dctn2 dysfunction and glaucoma pathogenesis (Jing and Malicki 2009).

Zebrafish lacking dynein-1 adaptors

Morpholino-mediated knockdown of ninl (ninl MO) in zebrafish results in dysmorphic and shortened OS, mispositioned organelles, and mislocalized rhodopsin, consequently impairing visual function. Under normal conditions, melanosomes are transported from the cell periphery to the perinuclear region upon epinephrine stimulation, resulting in a lighter color. However, this centripetal transport of melanosomes is disrupted in the ninl morphants, resulting in a significant delay in pigment retraction (Dona et al. 2015).

There are four syne homologs in zebrafish: syne1a, syne1b, syne2a, and syne2b, with syne2a being particularly enriched in PRC (Tsujikawa et al. 2007). Overexpression of the syne2a-KASH domain (69 C-terminal amino acids) or knockdown of syne2a (syne2a MO) results in severe basal displacement of the nucleus, accompanied by decreased survival of PRC and MGC (Tsujikawa et al. 2007; Bene et al. 2008).

Similarly, overexpression of the N-terminal region of lis1a (1–87 amino acids) or knockdown of lis1a (lis1a MO) in zebrafish disrupts nuclear migration and decreases PRC survival, resulting in pathological features resembling those of mok mutant or syne2a morphant (Tsujikawa et al. 2007).

Zebrafish lacking dynein-2

The functions of dynein-2 in the zebrafish PRC were examined through morpholino-mediated knockdown of the dync2h1, dync2i1, and dync2li1 genes (dync2h1 MO, dync2i1 MO, and dync2li1 MO). Each morphant exhibited identical ocular phenotypes, including microphthalmia, swollen connecting cilia, accumulated vesicles, shrunken and disorganized OS, and significantly impaired visual function. Notably, mild mislocalization of rhodopsin and blue-cone opsin was detected only in the most severely affected PRC, whereas light-dependent translocation of arrestin proceeded normally in rod (Krock et al. 2009). These findings indicate that dynein-2 is unlikely to serve as the primary motor responsible for the retrograde transport of phototransduction proteins.

Potential pathological mechanisms of dynein complex-related ocular disorders

ER stress pathway

Under normal protein homeostasis, the molecular chaperone-binding immunoglobulin protein (BiP, also known as GRP78) binds to ER stress sensors, maintaining them in an inactive state. Under ER stress, BiP exhibits a high affinity for unfolded or misfolded proteins, thereby dissociating from ER stress sensors to initiate an unfolded protein response (UPR). Ultimately, abnormal protein concentrations are alleviated through pathways such as ER-associated degradation (ERAD). However, under chronic or severe conditions, the UPR fails to restore homeostasis, resulting in cellular dysfunction and apoptosis (Oakes and Papa 2015). ER stress has been implicated in numerous ocular diseases, including retinitis pigmentosa, diabetic retinopathy, age-related macular degeneration, optic nerve degeneration, achromatopsia, glaucoma, and cataracts (Song et al. 2020; McLaughlin et al. 2022).

Most dynein complex animal models share a common feature: widespread accumulation of phototransduction proteins (cone opsins, rhodopsin, arrestin, and PDE6), apical polarity proteins (Nok and aPKC), and synaptic proteins (RIBEYE and mGluR6), which may trigger ER stress and disrupt protein homeostasis. Our previous findings demonstrated the activation of ER stress signaling in dync1h1^−/− zebrafish, as evidenced by the upregulation of BiP, activating transcription factor 4 (ATF4), and C/EBP homologous protein (CHOP, also known as GADD153), suggesting a direct correlation between dynein complex dysfunction and ER stress. Notably, PRC exhibits clear differential susceptibility to this ER stress response: blue-cone is the most vulnerable, followed by rod, whereas red/green-cone is the least affected (Zhou et al. 2025) (Fig. 5A-A’).

Fig. 5.

Schematic representation of the potential pathological mechanisms of dynein-related ocular disorders. A-A’: The ER stress pathway and dynein-related ocular disorders. Under normal protein homeostasis, BiP binds to ER stress sensors, maintaining them in an inactive state (A). Under conditions of dynein complex dysfunction, BiP exhibits a high affinity for unfolded or misfolded proteins (ectopic phototransduction proteins, apical polarity proteins, and synaptic proteins), thereby activating the ER stress pathway and triggering the BiP-ATF4-CHOP signaling-linked cell apoptosis (A’). B-B’: The Notch pathway and dynein-related ocular disorders. During normal mitosis, each RPC nucleus moves twice along the apical–basal gradient. When RPC nuclei approach the apical region, where Notch activity is increased, their daughter cells tend to remain proliferative. Conversely, when displaced more basally, where Notch activity is reduced, the daughter cells tend to differentiate into neurons (B). Under conditions of dynein complex dysfunction, the basal displacement of nuclei disrupts the balance between proliferation and differentiation (B’). C-C’: The SHH pathway and dynein-related ocular disorders. In the normally extended primary cilium, secreted protein SHH binds to its transmembrane receptor Ptch1, thereby releasing Ptch1’s inhibitory effect on Smo, followed by upregulation of transcription factor Gli-A (C). Under conditions of dynein complex dysfunction, which commonly have little or no axoneme extension, the SHH pathway is blocked due to impaired signaling protein transport (C’)

Notch pathway

The Notch pathway plays a crucial role in regulating the balance between neurogenic differentiation and proliferative division of postnatal retinal progenitor cell (RPC) (Furukawa et al. 2000; Scheer et al. 2001). In zebrafish retina, notch1a RNA is enriched in the apical region of the developing neuroretina, with its gradient decreasing toward the basal side (Bene et al. 2008). During mitosis, each RPC nucleus moves twice along the apical–basal gradient. When RPC nuclei approach the apical region, where Notch activity is increased, their daughter cells tend to remain proliferative. Conversely, when displaced more basally, where Notch activity is reduced, the daughter cells tend to differentiate into neurons (Bene et al. 2008; Murciano et al. 2002). Furthermore, Notch signaling is elevated during early development to sustain the proliferative ability of RPC. As development proceeds, Notch signaling gradually declines, directing the sequential generation of various types of distinct retinal neuron types (Bene et al. 2008; Murciano et al. 2002; Martins and Pearson 2008; Rapaport et al. 2004).

The characteristic basal displacement of nuclei observed in dynein complex animal models indicates that the dynein complex regulates retinal morphogenesis and neuronal composition by modulating RPC nuclear migration along the Notch spatial gradient. The spontaneous recovery of retinal abnormalities in Sun1^−/− and Syne2^−/− mice after adulthood may be attributed to the temporal variations in Notch signaling (Fig. 5B-B’).

Sonic Hedgehog (SHH) pathway

The SHH pathway is essential for embryonic tissue patterning, post-embryonic tissue regeneration, and tumorigenesis. The pathway is initiated by the binding of secreted SHH protein to its transmembrane receptor Patched1 (Ptch1). Subsequently, the transmembrane protein Smoothened (Smo) is activated and transmits signals to upregulate the downstream transcription factor Gli-A (Rohatgi et al. 2007; Gorojankina 2016). This signaling pathway is strictly dependent on the integrity of the primary cilium in vertebrates. Mutations in many core IFT components can lead to SHH signal dysfunction (Goetz and Anderson 2010; Zhang and Beachy 2023).

Mechanistic studies in dynein-2-related mouse models have demonstrated that loss of Dync2h1, WDR34, or WDR60 disrupts retrograde IFT, impairs signaling protein transport, blocks the SHH pathway, and ultimately results in severe systemic developmental abnormalities and embryonic death (Yan et al. 2023; Ocbina et al. 2011; Wu et al. 2017).

In the aforementioned dynein-2-related mouse model (Wdr34^RKO) and zebrafish models (dync2h1 MO, dync2i1 MO, dync2li1 MO), PRC ciliary structure is commonly defective, indicating that dynein-2 likely affects the signal transduction of the SHH pathway by controlling ciliogenesis during ocular development, thereby regulating ocular morphogenesis and maintaining visual function (Fig. 5C-C’).

Other pathways

Furthermore, several other pathways have also displayed intricate crosstalk with dynein, although the direct ophthalmic evidence is still lacking to date.

Nuclear factor-κB (NF-κB) is critical for immunity, inflammation, and oncogenesis. Dysregulated NF-κB signaling can lead to atherosclerosis, multiple sclerosis, rheumatoid arthritis, and can also promote tumor progression (Mao et al. 2025). LC8 has previously been identified as a common NF-κB inhibitor, which can alleviate the phosphorylation and degradation of the inhibitory protein IκBα, thereby reducing subsequent nuclear translocation of NF-κB (Jung et al. 2008; Shrum et al. 2009). Accumulated evidence suggests that boosting LC8 may be a promising therapeutic strategy for nonalcoholic steatohepatitis, osteoporosis, and arthritic bone diseases by inhibiting the NF-κB pathway (Kim et al. 2013; Lee et al. 2022). In addition, overexpression of WDR34 can inhibit IL-1β or LPS-induced, but not TNF-α-induced NF-κB activation, suggesting a potential role of the NF-κB pathway in WDR34-associated skeletal ciliopathies (Gao et al. 2009; Huber et al. 2013).

Wnt signaling plays essential roles in orchestrating embryonic development and tissue homeostasis by regulating cell fate specification, mitotic activity, and cell polarity. Dysregulation of the Wnt pathway is frequently associated with various cancers (Holzem et al. 2024; Rim et al. 2022). The TCTEX3 exerts an inhibitory effect on cervical cancer by suppressing the Wnt pathway (Zhang et al. 2022), whereas the LC8 is an oncogenic factor that facilitates the malignant characteristics of lung adenocarcinoma and hepatocellular carcinoma (HCC) via activating the Wnt pathway (Li et al. 2025; Shen et al. 2024). Moreover, WDR34 also plays an oncogenic role in the progression of HCC via activating the Wnt signaling, and its expression is negatively correlated with the survival of HCC patients (Luo et al. 2019). Thus, they have become promising prognostic biomarkers and potential therapeutic targets for specific cancers (Zhang et al. 2022; Li et al. 2025; Shen et al. 2024; Luo et al. 2019).

Furthermore, individual studies also pointed out that the IFN-γ-JAK-STAT pathway may be activated in the Dync1h1-induced non-small cell lung cancer metastasis (Pan et al. 2021), while the activation of PI3K-Akt pathway may be associated with WDR60-driven increased aggressiveness in pancreatic neuroendocrine tumor (Grassi et al. 2025).

Conclusions

Since the discovery of dynein, numerous studies have highlighted its pivotal role in the treatment of neurological diseases. Several mutations in dynein subunits and regulators result in central and peripheral nervous system degeneration, clearly underscoring their essential roles in neuronal survival. However, the loss of dynein complex components is typically lethal in the early developmental stages; consequently, ocular involvement may be relatively overlooked and remain insufficiently characterized. Diverse animal models have become valuable tools not only for identifying ocular phenotypes but also for elucidating possible pathological mechanisms.

Despite extensive research, the ocular morphology and developmental pattern of Drosophila differ markedly from those of vertebrates, substantially limiting their applicability as a model for visual system studies. Rodents remain the most widely used models, and the availability of advanced conditional knockout techniques helps circumvent the lethality associated with systemic gene deletions. However, there is still a gap in the composition of the PRC compared to that of humans. Zebrafish possess a cone-dominant retina and are highly amenable to efficient genetic manipulation, making them a powerful emerging model for ocular research. Integrating findings from diverse models will be key to understanding the molecular mechanisms and may have preclinical value for developing new therapies.

In conclusion, accumulating evidence indicates that components of the dynein complex are candidate contributors to ocular degenerative diseases and play crucial roles in the morphogenesis and maintenance of retinal neurons. Dynein complex-mediated retrograde transport is critically involved in numerous biological processes, including melanosome movement, cilium biogenesis, protein trafficking, organelle positioning, nuclear migration, and cell cycle control. Most of the animal models of dynein dysfunction exhibit consistent ocular defects, such as PRC degeneration, microtubule cytoskeleton disorganization, retrograde transport disruption, and visual function impairment. Among them, the dysregulation of melanosome movement can be captured through Adaptive Optics imaging, which can visually display the loss of melanosomes at the apical RPE and abnormal aggregation in the cell body, providing a potential examination indicator for clinical diagnosis.

However, despite the evolutionary conservation of dynein’s core roles, the ocular phenotypic manifestations of dynein dysfunction somewhat vary among different animal models. Molecular mechanistic investigations, especially cross-species comparative studies, remain relatively scarce. Generally, dynein components may exert their ophthalmic effects through multiple signaling pathways, including ER stress, Notch, and SHH. The NF-κB, Wnt, IFN-γ-JAK-STAT, and PI3K-Akt signaling are also related to dynein, but there is still a blank of their ocular research. It is worth noting that these pathways have not been systematically validated in a cross-species manner. The functional redundancy and crosstalk between these pathways may be variable in different animal models, making it unachievable to definitively interpret the species-specific molecular mechanisms at this stage. Given the current limitations, future investigations into other components and pathways of the dynein complex, as well as cross-species parallel validation, are needed to expand and deepen the “gene–phenotype–mechanism” correlation network.

Prospects

At present, there is no available treatment method for dynein-related ocular disorders, and systematic preclinical therapeutic intervention studies are still lacking in the published reports. In this review, we systematically summarize the advances in ocular research on dynein complex based on human, Drosophila, mouse, and zebrafish studies. To our knowledge, this is the first such comprehensive summary, which provides an essential foundation for the rational development of future therapeutic strategies.

Novel therapeutic strategies for inherited ocular degenerative diseases mainly consist of gene therapy, stem cell replacement therapy, and small molecule pharmaceutical therapy. Gene therapy is expected to fundamentally correct genetic defects, but it is still restricted by the high mutation specificity, limited vector packaging capacity, and potential immunogenicity. Stem cell replacement therapy provides a new idea for repairing damaged ocular cells, but it carries significant challenges in neural integration, immune rejection, and tumorigenic risk. Small molecule pharmaceutical therapy is a potential strategy, while related targeted compounds and mechanistic research are still in the initial stage.

In summary, further elucidation of the pathogenic molecular mechanisms of dynein complex-related ocular disorders and deeper exploration of therapeutic targets are the urgent scientific priorities and future core directions. Although translating current insights of the “gene-phenotype-mechanism” framework into clinical application remains a long-term and arduous task, continuous progress will lay a solid foundation for precise diagnosis and targeted therapy, ultimately contributing to the prevention of blinding diseases in humans.

Abbreviations

IFT

Retrograde intraflagellar transport

PRC

Photoreceptor cell

RGC

Retinal ganglion cell

DHC

Dynein heavy chain

DIC

Dynein intermediate chain

DLIC

Dynein light intermediate chain

ROBL

Roadblock

DDA

Dynein-dynactin-adaptor

BICD2

Bicaudal D2

BICDR1

BICD-related protein 1

NIN

Ninein

NINL

Ninein-like protein

RAB11-FIP3

Rab family-interacting protein 3

SPDL1

Spindly

KASH2

Klarsicht/Anc-1/Syne homology domain protein 2

SUN1

Sad1/UNC-84 domain protein 1

TRAK1/2

Trafficking kinesin-binding protein 1/2

JIP3

C-Jun N-terminal kinase-interacting protein 3

HAP1

Huntingtin-associated protein 1

RILP

Rab7-interacting lysosomal protein

NUMA

Nuclear mitotic apparatus protein

CCDC88A/B/C

Coiled-coil domain protein 88 A/B/C

LIS1

Lissencephaly-1

NDE1

Nuclear distribution element 1

NDEL1

NDE-like 1

NMD

Neuromuscular disorder

NDD

Neurodevelopmental disorder

CMT2O

Charcot-Marie-Tooth disease 2O

SMALED1

Spinal muscular atrophy with lower extremity predominance 1

SRP

Short-rib polydactyly

JATD

Jeune asphyxiating thoracic dystrophy

Gl

Glued

Koi

Klaroid

Klar

Klarsicht

LINC

Linker of the nucleoskeleton and cytoskeleton

E

Embryonic day

Loa

Legs at odd angles

Cra1

Cramping1

Swl

Sprawling

IS

Inner segment

P

Postnatal day

ONL

Outer nuclear layer

INL

Inner nuclear layer

OS

Outer segment

wPTI

Weeks post-tamoxifen induction

OPL

Outer plexiform layer

CTBP2

C-terminal binding protein 2

mGluR6

Metabotropic glutamate receptor 6

PDE6

Phosphodiesterase 6

RdCVF

Rod-derived cone viability factor

RPE

Retinal pigment epithelium

ER

Endoplasmic reticulum

cnb

Cannonball

dpf

Days post-fertilization

MO

Morpholino knockdown

mok

Mikre oko

MGC

Müller glia cell

ako

Ale oko

Nok

Nagie oko

aPKC

Atypical protein kinase C

BiP

Binding-immunoglobulin protein

UPR

Unfolded protein response

ERAD

ER-associated degradation

ATF4

Activating transcription factor 4

CHOP

C/EBP homologous protein

RPC

Retinal progenitor cell

SHH

Sonic Hedgehog

Ptch1

Patched1

Smo

Smoothened

NF-κB

Nuclear factor-κB

HCC

Hepatocellular carcinoma

Authors’ contributions

Xuebin Zhou conceptualized the review, conducted the literature search, and wrote the original draft. Jianan Xie offered guidance on the structure and content. Wanqing Tong contributed to the manuscript revisions and provided critical feedback. Jinling Fu reviewed the manuscript and provided funding support. All authors read and approved the final manuscript.

Funding

This work was supported by the Natural Science Foundation Project of Science and Technology Department of Jilin Province (20250601017RC).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.