Effect of Dync1h1 on Phototransduction Protein Transport and the Development and Maintenance of Photoreceptor Cells in Zebrafish

Abstract

Purpose

Retinitis pigmentosa (RP) is a hereditary retinal disease characterized by progressive degeneration of photoreceptor cells (PRCs). Identifying potential pathogenic genes and understanding the mechanisms of PRC degeneration are essential for improving diagnosis and treatment. Cytoplasmic dynein 1, responsible for retrograde axonal transport along microtubules, plays critical roles in neuronal function. This study utilized dync1h1-deficient zebrafish to investigate its roles in PRC morphogenesis and degeneration.

Methods

Heterozygous and homozygous dync1h1-deficient zebrafish were confirmed through Sanger sequencing. Morphological changes and retinal phenotypes were assessed through histological analysis. RNA sequencing and bioinformatics were used to explore molecular mechanisms in dync1h1^−/− zebrafish, with endoplasmic reticulum (ER) stress and apoptosis pathways validated in vivo.

Results

Dync1h1^−/− zebrafish exhibit severe developmental defects, including microphthalmia, disorganized retinal lamination, and PRC apoptosis. In dync1h1^+/−, mild but progressive growth retardation and PRC defects were observed. Dync1h1 loss impaired retrograde axonal transport, leading to defects in cilium biogenesis and transport disorder of phototransduction proteins in PRCs. This triggered ER stress, activating BiP-ATF4-CHOP signaling pathway and leading to PRC degeneration.

Conclusions

Dync1h1 is essential for maintaining retrograde axonal transport and proper trafficking of phototransduction proteins in PRCs. Non-resolving ER stress-induced PRC apoptosis is a key factor in DYNC1H1-associated retinal degeneration. This study provides important insights into the precise diagnosis and may help in the development of targeted therapies for retinal degenerative diseases.

Retinitis pigmentosa (RP) is a hereditary retinal degenerative disease primarily marked by progressive photoreceptor cell (PRC) death. Initially, RP presents as night blindness and tunnel vision due to rod cell involvement. As the disease progresses, cone cell degeneration leads to loss of central vision and eventual blindness. Globally, RP prevalence ranges from 1 in 7000 to 1 in 3000. Although >100 pathogenic genes involved in various biological pathways were identified, many RP-associated genes remain undiscovered.¹ Moreover, mutations affecting protein trafficking often lead to retinal degeneration.^2–4 The search for potential RP-associated genes and detailed investigation into pathological changes are vital for understanding disease mechanisms, leading to improved clinical diagnosis and treatments.

Cytoplasmic dynein 1, a large motor protein complex responsible for retrograde axonal transport along microtubules, plays critical roles in neuronal function and survival,^5–7 including mitosis, autophagy, organelle positioning, nuclear migration, vesicle trafficking, etc.^6,8 Dynein 1 comprises a pair of heavy chains (DHC1; encoded by DYNC1H1) and several light and intermediate chains.^6,9 Mutations in DYNC1H1 are linked to neuromuscular disorders and neurodevelopmental disorders.¹⁰ Recently, Möller et al. observed that 48.83% of their patients with DYNC1H1 exhibited multiple ophthalmological features,¹⁰ suggesting the roles of DYNC1H1 in ophthalmic diseases.

The correlation between DYNC1H1 and retinal degeneration has been previously explored. Disruption of Dync1h1 in the mouse causes early embryonic death,¹¹ whereas conditional knockout in the retina leads to severe retinal degeneration shortly after birth.^12,13 Rod-specific knockout causes rapid degeneration, with rod function lost by postnatal day 30 (P30) and cone function diminished shortly thereafter.¹⁴ Tamoxifen-induced conditional knockout in developed PRCs results in PRC degeneration,^14,15 with outer segments (OS) shortening by 3 weeks post-induction and disappearing by the fourth week. A zebrafish model with dync1h1^Y3102X mutation showed severe defects in PRC organelle organization and vesicle trafficking, leading to death within 6 to 8 days post-fertilization (dpf).^16,17 Morpholino knockdown of zebrafish dync1h1 causes dose-dependent abnormalities. Both the zebrafish models showed disrupted rhodopsin transport and mislocalization in the outer nuclear layer (ONL).¹⁷

However, the precise role of Dync1h1 in PRC morphogenesis and degeneration remains unclear. Here, we cultivated a dync1h1 knockout zebrafish model. Loss of dync1h1 disrupts retrograde axonal transport in PRCs, leading to abnormal accumulation and degradation of phototransduction proteins in cell body and inner segment (IS). This metabolic overload disrupts PRC morphogenesis, causing endoplasmic reticulum (ER) stress-induced apoptosis.

Methods

Zebrafish Strains and Husbandry

Zebrafish strains (China zebrafish Resource Center (CZRC), CZ479) were sourced from the CZRC and maintained under standard laboratory conditions: a 14-hour light/10-hour dark cycle at 28.5°C with circulating water. Embryos were cultured in E3 medium (5 mM NaCl, 0.33 mM CaCl₂, 0.33 mM MgSO₄, and 0.17 mM KCl) at 28.5°C. All animals were treated in accordance with the Guide for the Care and Use of Laboratory Animals and the ARVO Statement for Use of Animals in Ophthalmic and Vision Research and monitored with approval from the Institutional Animal Care and Use Committee (IACUC) of Jilin University.

Genotyping

Tail fin samples were extracted, and each zebrafish was isolated for genotyping. Each tail sample was incubated in 50 µL of 50 mM NaOH at 100°C for 15 minutes, then neutralized with 10 µL of 1 M Tris-HCl (pH 8.0) and centrifuged at 12,000 rpm for 5 minutes. Polymerase chain reaction (PCR) was performed using Taq DNA Polymerase (TaKaRa, R001A), with forward (5′-AGTCCGTGATACAGCTCGC-3′) and reverse (5′-ATAATAGCGGTGTCGGTGC-3′) primers to amplify 350 bp DNA fragments containing the mutation site. The PCR product was sequenced by Sangon Biotech Co. Ltd. (Shanghai, China).

Antibodies

Rabbit antibodies of anti-Dync1h1 (1:200 for immunofluorescence, 1:500 for Western blot; ProteinTech, 12345-1-AP), anti-nagie oko (Nok, 1:200; a gift from Zou Jian, PhD), anti-Red/Green Opsin (1:500; Sigma, AB5405), anti-Opn1sw2 (1:200; Abcepta, Azb21565b), anti-GRP78/BiP (1:2000; ProteinTech, 11587-1-AP), anti-ATF4 (1:500; ProteinTech, 10835-1-AP), anti-β-actin (1:50000; ABclonal, AC026), and anti-Gapdh (1:5000; Huabio, ET1601-4). Mouse antibodies of anti-ZO-1 (1:500; Invitrogen, 339100), anti-acetylated tubulin (1:3000; Sigma Millipore, T7451), and anti-Rhodopsin (1:200; Abcam, ab5417). Secondary antibodies of AlexaFluor-594-conjugated goat anti-rabbit and anti-mouse IgG (1:200; Invitrogen, A32740; Abcam, ab150116), AlexaFluor-488-conjugated goat anti-mouse IgG (1:200; Abcam, ab150113), and HRP-conjugated goat anti-rabbit IgG (1:10000; ProteinTech, SA00001-2). AlexaFluor-488-conjugated peanut agglutinin (PNA, 1:50; Invitrogen, L21409) was used for cone OS staining. 4′6-diamidino-2-phenylin-dole (DAPI)-containing antifade medium (Beyotime, P0131) for nuclear staining.

Hematoxylin and Eosin, Immunofluorescence, and TUNEL Staining

Zebrafish larvae and adult eyes were fixed in 4% paraformaldehyde overnight at 4°C. Paraffin-wax-embedded tissue were sectioned at 2 µm. For frozen sections, samples were dehydrated in 10%, 20%, and 30% sucrose, embedded in optimum cutting temperature (OCT) compound, and then cryo-sectioned at 5 µm. Antigen retrieval was performed using the Tris-EDTA Antigen Retrieval Solution (Beyotime, P0084) following the manufacturer's instructions. The sections were subsequently subjected to standard hematoxylin and eosin (H&E) or immunofluorescence staining. Apoptosis detection was performed using the TUNEL Kit (Abbkine, KTA2011) following the manufacturer's instructions, and images were acquired using Fluoview FV1000 confocal microscope (Olympus, Tokyo, Japan).

Transmission Electron Microscopy

Tissues were fixed in 2.5% glutaraldehyde at 4°C overnight, rinsed, re-fixed in 1% osmium tetroxide at room temperature for 2 hours, dehydrated with ethanol, and embedded in resin. Ultrathin sections (100 nm) were prepared and stained with uranyl acetate and lead citrate, and images were acquired using a transmission electron microscope (TEM; JEOL, Tokyo, Japan).

Western Blot Analysis

Embryos at 5 dpf were lysed in RIPA buffer (Beyotime, P0013B) containing phenylmethanesulfonyl fluoride (Beyotime, ST506), each sample consisting of 10 pooled embryos. Protein concentration was measured using the Enhanced BCA Protein Assay Kit (Beyotime, P0010). Samples were boiled for 5 minutes and stored at −20°C. Lysates were subjected to SDS-PAGE, transferred to polyvinylidene fluoride membranes, blocked with 5% skim milk for 1 hour, and incubated overnight with primary antibodies at 4°C. Following incubation with HRP-conjugated secondary antibodies, chemiluminescence was used for detection, and ImageJ was used for protein quantification. Experiments were repeated three times.

Quantitative Real-Time Polymerase Chain Reaction Analysis

Total RNA from 5 dpf embryos was extracted using the Total RNA Extraction Kit (Promega, LS1040). The cDNA synthesis was conducted using the cDNA Synthesis Kit (GeneCopoeia, QP057), and qRT-PCR was performed using SYBR Green qPCR mix 2.0 (GeneCopoeia, QP031). β-actin was used as the reference gene. Primer sequences are shown in Supplementary Table S1.

RNA-Sequencing and Bioinformatics Analysis

Total RNA from 5 dpf dync1h1^−/− and wild-type (WT) embryos was extracted using TRIeasy Reagent (Yeasen, 10606ES60), with each sample containing 20 pooled embryos. RNA sequencing was performed using Illumina NovaSeq 6000. Clean reads were aligned to the zebrafish reference genome (Ensembl, GRCz11) using Hisat2 (version 2.0.5), and mRNA expression quantified via featureCounts (version 1.5.0-p3). Differential expression analysis used DESeq2 (version 1.20.0) with a threshold of |log2 (FoldChange)|≥1 and adjusted P value ≤ 0.05. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were conducted using ClusterProfiler (http://www.genome.jp/kegg/).

Statistical Analysis

Data analyses were performed using GraphPad Prism 9, with results indicated as mean ± SD. The Shapiro–Wilk test was used to assess whether the data obeyed the normal distribution. If the data obeyed the normal distribution, statistical significance was calculated using Student's t-test or 1-way ANOVA analysis (3 or more groups, followed by Tukey's multiple comparisons for the post hoc test). If the data were not normally distributed, statistical significance was calculated using the Mann–Whitney U test or Kruskal–Wallis test (3 or more groups, followed by Dunnett's multiple comparisons for the post hoc test). The significance levels noted as not significant (ns) = P > 0.05; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Results

Dync1h1-Deficiency Affects the Survival and Overall Development of Zebrafish

To investigate the role of dync1h1 in zebrafish, we analyzed phenotypes of dync1h1-deletion strains with homozygous or heterozygous c.198_204del mutations (dync1h1^−/− and dync1h1^+/−). This mutation is expected to cause a frameshift leading to premature termination and protein dysfunction (p.Glu42Alafs*7). We screened WT and dync1h1 mutants using PCR and Sanger sequencing (Figs. 1A, 1B). Immunofluorescence and Western blot confirmed that dync1h1 was completely absent in dync1h1^−/− compared to WT (Figs. 1C, 1D). Prior to 5 dpf, we could not distinguish between mutant embryos and their WT siblings. At 5 dpf, WT and dync1h1^+/− zebrafish swam normally and began feeding, whereas dync1h1^−/− exhibited significantly reduced activity and sank to the tank's bottom. Specifically, these mutants displayed a darkened appearance. Pigmentation relies on the repositioning of melanosomes from the perinucleus synthesis area to the distal extremities. Dynein regulates centripetal movement of melanosomes by counteracting the activity of kinesin.¹⁸ Considering this function of dynein, we speculate that loss of dync1h1 hinders the retrograde transport of melanosomes, leading to diffuse distribution, which may contribute to the hyperpigmentation features. In addition, homozygotes mutants exhibited other defects, including pericardial edema, abdominal distension, underdeveloped organs, microphthalmia, and eyeball protrusion, and died between 6 and 8 dpf, whereas heterozygotes were viable and fertile (Figs. 2A, 2B).

Figure 1.

Identification of dync1h1 knockout zebrafish. (A) Schematic representation of the genomic region and exon/intron structure of dync1h1 in zebrafish. Knockout of dync1h1 results in a truncated protein, with deletion located in exon 1. (B) Direct sequencing comparison of WT, dync1h1^+/−, and dync1h1^−/− zebrafish lines. (C) Immunofluorescence staining showing dync1h1 protein expression in 5 dpf WT and dync1h1^−/− zebrafish. Nuclei are stained with DAPI. (D) Western blot analysis of dync1h1 protein expression in 5 dpf WT and dync1h1^−/− zebrafish, with β-actin as the endogenous control. +/−, dync1h1 heterozygote; −/−, dync1h1 homozygote. Scale bar = 50 µm. Data are shown as mean ± SD. **, P < 0.01, n = 3 biologically independent samples per group. DAPI, 4′,6-diamidino-2-phenylin-dole; dpf, days post fertilization; SD, standard deviation; WT, wild type.

Figure 2.

Impact of dync1h1 loss on zebrafish survival and overall development. (A) Kaplan-Meier survival curve for the progeny of self-crossed WT or dync1h1^+/− zebrafish. (B) External morphology of WT and dync1h1^+/− zebrafish at 5 dpf. (C) Schematic diagram illustrating measurement indicators at 5 dpf: body length (blue), axial length (red), long axial length (yellow), short axial length (green), and anterior chamber depth (white). (D) Quantitative analysis of binocular measurements as a percentage of body length in WT, dync1h1^+/−, and dync1h1^−/− zebrafish at 5 dpf. +/−, dync1h1 heterozygote; −/−, dync1h1 homozygote. Scale bar = 200 µm. Data were presented as mean ± SD. *, P < 0.05; **, P < 0.01; ****, P < 0.0001, n = 8 (WT), n = 10 (dync1h1^+/−), and n = 12 (dync1h1^−/−). dpf, days post fertilization; SD, standard deviation; WT, wild type.

To quantitatively assess eye development, we measured body length, axial length, long and short axial lengths, and anterior chamber depth at 5 dpf (Fig. 2C). Body length was defined as the distance from the mouth's anterior border to the caudal fin, the axial length was defined as the distance from cornea's anterior surface to the eyeball's posterior pole, the long axial length was defined as the distance between the nasal and temporal edges, the short axial length was defined as the distance between the ventral and dorsal edges, and the anterior chamber depth was defined as the distance between the cornea's anterior surface to the iris’ anterior surface. To minimize measurement errors from embryonic positioning and individual developmental differences, we averaged binocular measurements and calculated their percentage of body lengths, and found that dync1h1^−/− had shorter body lengths, increased anterior chamber depths, and obvious microphthalmia compared with WT, resulting in a relatively convex eye shape. This curvature may stem from a shortened radial distance, causing the lens to protrude and pushing of the cornea outward. Dync1h1^+/− also showed reduced body lengths compared with WT, suggesting that even partial dync1h1-deficiency impacts the overall development (Fig. 2D).

Dync1h1-Deficiency Results in Retinal Degeneration

We then examined the anatomic structure of the eyeball in 5 dpf embryos with dync1h1-deficiency. Histological examination using H&E staining revealed that dync1h1^−/− had a complete loss of PRC IS and OS, with the retinal pigment epithelium (RPE) exhibiting a disrupted morphology (Fig. 3A). To better evaluate the effects of dync1h1-deficiency on PRC ultrastructure, we conducted TEM analysis. In dync1h1^−/− retinas, RPE was dispersed, invading the inner layers, and the PRCs were mislocalized and displayed karyopyknosis, lacking polarized IS or OS. Outer limiting membrane (OLM) was not visible, and ONL showed significant vacuolization (Fig. 3B). Additionally, mitochondria were abnormally distributed, with reduced quantity and volume, and disordered cristae (Fig. 3C). ER and Golgi were expanded, surrounded by swollen vesicles (Figs. 3D, 3E), indicating vesicular transport disorders. We also performed immunofluorescence staining using anti-Nok^19,20 and PNA antibodies,^21,22 previously identified as markers for IS and OS in zebrafish. Consistent with H&E and TEM findings, dync1h1^−/− retinas displayed severe disruption in lamination, with no detectable IS/OS fluorescence signals (Fig. 3F). OLM integrity (stained with anti-ZO1 antibody) was entirely compromised in dync1h1^−/− (Fig. 3G). TUNEL assays showed extensive apoptotic signals throughout the retina, especially in ONL (Fig. 3H), confirming that dync1h1 is crucial for the morphogenesis and survival of PRCs.

Figure 3.

Retinal degeneration in dync1h1^−/− zebrafish at 5 dpf. (A) Representative H&E-stained images of WT and dync1h1^−/− retina. (B–E) TEM images of WT and dync1h1^−/− retina. (B) Overview of the outer retina. (C) Mitochondria indicated by white arrows and zoom-in pictures shown in the red rectangle. (D) ER (white arrow). (E) Golgi (white arrow). (F) Immunofluorescence-stained cryosections showing PRC IS/OS using anti-Nok and PNA; nuclei are stained by DAPI. (G) Immunofluorescence-stained cryosections showing OLM integrity using anti-ZO1; nuclei stained by DAPI. (H) TUNEL assay images with DAPI-stained nuclei. −/−, dync1h1 homozygote. Scale bar = 50 µm (A, F–H), 5 µm (B), 2 µm in (C), 500 nm (D, E). DAPI, 4′,6-diamidino-2-phenylin-dole; dpf, days post fertilization; ER, endoplasmic reticulum; H&E, hematoxylin and eosin; IS/OS, inner segment/outer segment; Nok, nagie oko; ONL, outer nuclear layer; PNA, peanut agglutinin; PRC, photoreceptor cell; WT, wild type.

The complete loss of dync1h1 in homozygous mutants leads to severe developmental defects and rapid retinal degeneration, limiting extensive analysis of PRC phenotypes at later stages. Consequently, we investigated heterozygous phenotypes during adult development. Immunofluorescence revealed no significant abnormalities in dync1h1^+/− retinal lamination at 5, 7, and 10-months post-fertilization (mpf; Fig. 4A). However, ONL thickness in dync1h1^+/− was significantly shorter than in WT at both 7 and 10 mpf (9.45 ± 0.18 µm vs. 10.74 ± 0.27 µm at 7 mpf, and 9.32 ± 0.86 µm vs. 10.80 ± 0.26 µm at 10 mpf; Fig. 4B). OLM in mutants was as integral as that in WT, with no detectable TUNEL signals throughout the retina (Figs. 5A, 5B). Further TEM observations at 20 mpf revealed well-developed mitochondria and Golgi apparatus in both WT and dync1h1^+/− retinas (Figs. 6A, 6B). However, dync1h1^+/− retinas exhibited numerous swollen ER structures within the IS (see Fig. 6B). Overall, our results suggest that partial dync1h1-deficiency may lead to mild, progressive retinal degeneration primarily associated with ER dysfunction.

Figure 4.

Mild but progressive retinal degeneration in adult dync1h^+/− zebrafish. (A) Immunofluorescence-stained cryosections at indicated ages showing the PRC IS/OS stained with anti-Nok and PNA; nuclei stained by DAPI. (B) Quantitative analysis of ONL thickness in WT and dync1h1^+/− retinas over time. +/−, dync1h1 heterozygote. Scale bar = 20 µm. Data are shown as mean ± SD. Not significant (ns), P > 0.05, *, P < 0.05; **, P < 0.01, n = 3 biologically independent samples per group. DAPI, 4′,6-diamidino-2-phenylin-dole; IS/OS, inner segment/outer segment; Nok, nagie oko; ONL, outer nuclear layer; PNA, peanut agglutinin; PRC, photoreceptor cell; SD, standard deviation; WT, wild type.

Figure 5.

Dync1h1-deficiency maintains OLM integrity and PRC survival in adult dync1h^+/− retina. (A) Immunofluorescence-stained cryosections at indicated ages showing OLM stained with anti-ZO1; nuclei stained by DAPI. (B) TUNEL assay images at indicated ages with DAPI-stained nuclei. +/−, dync1h1 heterozygote. Scale bar = 50 µm. DAPI, 4′,6-diamidino-2-phenylin-dole; OLM, outer limiting membrane; PRC, photoreceptor cell.

Figure 6.

Prioritization of ER impairment in dync1h1^+/− retinas at 20 mpf. (A) TEM images showing well-developed mitochondria in both WT and dync1h1^+/− retinas. (B) TEM image highlighting well-developed Golgi (white arrow) in both WT and dync1h1^+/− retinas. The black asterisks point to abnormally swollen ER scattering in IS. +/−, dync1h1 heterozygote. Scale bar = 1 µm (A), 500 nm (B). ER, endoplasmic reticulum; mpf, months post fertilization; WT, wild type.

Dync1h1-Deficiency Impairs Cilium Biogenesis

Severe retinal degeneration in dync1h1^−/− retinas, particularly with OS ablation, suggests that dync1h1-deficiency has adverse effects on cilium biogenesis. Immunofluorescent staining with anti-α-acetylated tubulin, a marker of axonemal microtubules, revealed a complete loss of ciliary structures in dync1h1^−/− PRCs (Fig. 7A), highlighting dync1h1’s role in the early cilium formation in PRCs. Additionally, mean microtubule lengths were significantly shorter in 5, 7, and 10 mpf dync1h1^+/− PRCs (Figs. 7B, 7C), with mutant and control lengths measured as 22.31 ± 1.69 µm vs. 28.87 ± 0.54 µm at 5 mpf, 24.43 ± 1.31 µm vs. 28.51 ± 1.52 µm at 7 mpf, and 26.35 ± 0.07 µm vs. 32.09 ± 1.63 µm at 10 mpf, respectively. These results indicate that partial dync1h1-deficiency progressively hinder microtubule development and maintenance, leading to ciliary hypoplasia in developed retinas.

Figure 7.

Dync1h1-deficiency impairs cilium biogenesis. (A) PRC cilium biogenesis in WT and dync1h1^−/− retinas at 5 dpf, detected by immunofluorescence for axonemal microtubules using anti-α-acetylated tubulin antibody; nuclei stained by DAPI. (B) PRC development in WT and dync1h1^+/− retinas at indicated ages, with axonemal microtubules by anti-α-acetylated tubulin; nuclei stained by DAPI. The lengths of microtube were labeled with white vertical lines. (C) Quantitative analysis of microtubule length in WT and dync1h1^+/− retinas across ages. +/−, dync1h1 heterozygote; −/−, dync1h1 homozygote. Scale bar = 50 µm. Data are shown as mean ± SD. *, P < 0.05; **, P < 0.01, n = 3 biologically independent samples per group. DAPI, 4′,6-diamidino-2-phenylin-dole; dpf, days post fertilization; PRC, photoreceptor cell; SD, standard deviation; WT, wild type.

Dync1h1-Deficiency Leads to the Transport Disorder of Various Phototransduction Proteins in PRCs

In healthy retinas, membranous OS disks are densely packed with phototransduction proteins and transported from the IS to optimize photon capture and signal transduction.^23,24 To assess trafficking disruptions from ciliary hypoplasia, we labeled visual pigments, including cone opsins and rhodopsin, with specific antibodies. In dync1h1^−/− PRCs, immunofluorescence showed complete loss of blue cone opsin (anti-Opn1sw2), whereas red/green-cone opsin (anti-RG-opsin) and rhodopsin (anti-Rhodopsin) were partially preserved but mislocalized near the basal body in a disordered arrangement (Fig. 8A). Based on these results, we speculated that dync1h1 was essential for phototransduction protein distribution and OS development in PRCs. In dync1h1^−/− retina, B-cone PRCs were most severely affected, followed by rods, with RG-cones being the least affected. Furthermore, we evaluated the distribution of these proteins by examining the length of the OS in the dync1h1^+/− retinas at 5, 7, and 10 mpf. Although OS morphology in mutants was visually similar to controls, quantitative measurements showed shorter OS lengths of all ages (Figs. 8B, 8C). The Table shows the OS lengths at indicated ages. These findings of reduced OS suggest that partially preserved dync1h1 in dync1h1^+/− is insufficient to fully maintain the transport function of OS proteins, which impairs trafficking system and limits the protein supply from the IS to the OS.

Figure 8.

Transport disorder of phototransduction proteins in PRCs due to dync1h1-deficiency. (A, B) Immunofluorescence images of WT and dync1h1^−/− retinas at 5 dpf, and WT and dync1h1^+/− retinas at indicated ages, with anti-RG-opsin antibody, anti-Opn1sw2, and anti-Rhodopsin antibodies marking blue-cone opsin, red/green-cone opsin, and rhodopsin, respectively. Nuclei are stained by DAPI. Zoom-in pictures were shown in the red rectangle. The lengths of OS were labeled with white vertical lines. (C) Quantitative analysis of the lengths of RG-cone OS, B-cone OS, and rod OS in WT and dync1h1^+/− retinas at indicated ages. +/−, dync1h1 heterozygote; −/−, dync1h1 homozygote. Scale bar = 50 µm (A), 20 µm (B). Data are shown as mean ± SD. Not significant (ns), P > 0.05; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001, n = 3 biologically independent samples per group. DAPI, 4′,6-diamidino-2-phenylin-dole; dpf, days post fertilization; H&E, hematoxylin and eosin; OS, outer segment; PRC, photoreceptor cell; WT, wild type; SD, standard deviation.

Table.

The OS Lengths of PRCs in WT and dync1h1^+/− Retinas Across Ages

PRC	Age, mpf	+/−, µm	WT, µm
RG-cone	5	15.12 ± 0.94	18.64 ± 0.38
	7	16.37 ± 0.43	18.98 ± 0.93
	10	17.03 ± 0.72	20.28 ± 1.53
B-cone	5	8.25 ± 0.95	11.42 ± 0.81
	7	8.39 ± 0.12	11.32 ± 0.09
	10	9.63 ± 0.48	9.63 ± 0.48
Rod	5	7.82 ± 0.24	11.61 ± 0.76
	7	8.43 ± 0.57	11.95 ± 0.71
	10	8.98 ± 0.16	13.12 ± 0.48

OS, outer segment; PRC, photoreceptor cells; RG, red-green; WT, wild type.

Bioinformatics Analysis Based on RNA Sequencing Reveals Phototransduction Disorder and Apoptosis in dync1h1^−/− Zebrafish

To explore the molecular mechanisms underlying developmental impairment in dync1h1^−/− zebrafish, we conducted RNA sequencing at 5 dpf. Pearson correlation analysis confirmed high genotype consistency among samples (Fig. 9A). The volcano plot (Fig. 9B) identified 1652 upregulated and 1856 downregulated genes. Gene ontology analysis revealed significant downregulation of genes associated with visual perception, phototransduction, non-motile cilia, and transmembrane transporter functions (Fig. 9C, Supplementary Fig. S1A). Additionally, genes related to the ER lumen were markedly upregulated (see Fig. 9C, Supplementary Fig. S1B). These findings align with observed retinal degeneration in dync1h1 mutants. KEGG pathway analysis showed that numerous differentially expressed genes were linked to phototransduction and apoptosis pathways (Fig. 9D).

Figure 9.

Bioinformatic analysis reveals phototransduction disruption and apoptosis in dync1h1^−/− zebrafish at 5 dpf. (A) Pearson correlation analysis comparing WT and dync1h1^−/− samples. (B) Volcano plot showing differential gene expression, with 1652 upregulated and 1856 downregulated genes (|log2 (FoldChange)|≥1 and adjusted P value ≤ 0.05). (C) Top Gene Ontology (GO) terms enriched among all differentially expressed genes. (D) Top KEGG pathways enriched among all differentially expressed genes. W1 to W3, three independent samples from the WT group; D1 to D3, three independent samples from dync1h1^−/− group. WT, wild type.

Mislocalization of Phototransduction Proteins Triggers ER Stress-induced Apoptosis in PRCs of dync1h1^−/− Zebrafish

TEM revealed that dync1h1-deficiency contributed to primary ER impairment in PRCs (see Fig. 6B). Bioinformatic analyses suggest that mislocalization of phototransduction proteins in IS creates excessive metabolic load, triggering ER stress-induced apoptosis, and leading to retinal degeneration.

To confirm ER stress as a pathogenic factor in dync1h1^−/− retinas, we analyzed ER stress markers at 5 dpf. Western blot showed significant upregulation of binding immunoglobulin protein (BiP, also known as GRP78) and activating transcription factor 4 (ATF4) in dync1h1^−/− (Figs. 10A–D). With limited zebrafish-specific antibodies, qRT-PCR was used to assess C/EBP homologous protein (CHOP; also known as GADD153) and apoptosis-related genes (bax, caspase3, and caspase8), which were also significantly upregulated in dync1h1^−/− (Figs. 10E–H). These results suggest that dync1h1-deficiency induces ER stress-triggered apoptosis via BiP-ATF4-CHOP pathway in dync1h1^−/−.

Figure 10.

Validation of ER stress-induced apoptosis pathway in vivo. (A) Western blot analysis of BiP expression in WT and dync1h1^−/− zebrafish at 5 dpf, with β-actin as the endogenous control. (B) Quantitative analysis showing elevated BiP levels in dync1h1^−/− compared to WT. (C) Western blot analysis of ATF4 protein expression in WT and dync1h1^−/− zebrafish at 5 dpf, with Gapdh as endogenous control. (D) Quantitative analysis showing elevated ATF4 protein levels in dync1h1^−/− compared to WT. (E–H) The qRT-PCR analysis showing increased mRNA levels of chop, bax, caspase3, and caspase8 in dync1h1^−/− relative to WT. β-actin was used as the endogenous control. −/−, dync1h1 homozygote. Data are shown as mean ± SD. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001, n = 3 biologically independent samples per group. ATF 4, activating transcription factor 4; WT, wild type; BiP, binding immunoglobulin protein; ER, endoplasmic reticulum; qRT-PCR, quantitative real-time polymerase chain reaction; SD, standard deviation.

Discussion

Photoreceptor OS lacks biosynthetic machinery. All OS proteins are synthesized by ribosomes attached to the ER near the nucleus and transported through Golgi stacks to OS via dynein-dependent mechanisms.^23,25–27 Approximately 10% of the OS is renewed daily through disk shedding, which is vital for maintaining visual perception. The OS length remains stable as new disks form at the base, pushing older disks apically, where they are phagocytosed by RPE cells.²⁸ This rapid turnover necessitates an organized system for transporting a large volume of proteins from their synthesis site in the IS to their functional site in the OS.

Dync1h1 is vital for neuronal function and survival across various organisms.^6,29 Early lethality associated with homozygous deletion of Dync1h1 has limited investigation into molecular mechanisms underlying protein transport disorders and retinal degeneration.^12–14,17 Studies on other subunits focused on their roles in ciliary development and OS protein trafficking.^30,31 Deletion of dynein light intermediate chain (Dlic1) in the mouse leads to ectopic accumulation of OS proteins, including rhodopsin and arrestin, and impaired cilium biogenesis and OS growth caused by abnormal Rab11-vesicle transport.³⁰ Knockout of dlic1 in zebrafish results in abnormal Rab8 aggregation in the outer plexiform layer (OPL), which exhibits progressive abnormalities in cone opsin expression, and PRC apoptosis.³¹ These findings prompt further exploration of Dync1h1’s influence on the axoneme backbone and cargo trafficking.

Currently, comprehensive reports on the pathological mechanisms of retinal degeneration involving dync1h1-deficient zebrafish are lacking. Whereas mice are commonly used as models for retinal degeneration, they do not possess a cone-rich macula lutea. Therefore, studies of an axonal transport system in PRCs have largely centered on rod-specific visual pigments like rhodopsin. However, lower vertebrates, particularly zebrafish, offer distinct advantages, including practicality and reduced costs. The cone-to-rod ratio in zebrafish closely resembles that of the human retina, with zebrafish retinas comprising 60% cones and 40% rods, whereas rodent retinas consist of only 3% cones and 97% rods.³² This makes zebrafish retinas a more suitable model for studying human retinal degenerative diseases.

This study investigated the effects of dync1h1-deficiency in zebrafish. Our dync1h1^−/− zebrafish exhibited widespread developmental defects and died within 6 to 8 dpf. At 5 dpf, dync1h1^−/− retinas displayed microphthalmia and severely disorganized lamination with no polarized IS/OS structures. The absence of ciliary structures led to the accumulation of phototransduction proteins around the basal body. Overall, the disrupted polarized trafficking of phototransduction proteins to their specific OS domains contribute to PRC apoptosis, a common feature of retinal degenerative diseases. However, dync1h1^+/− exhibited mild and progressive PRC defects without dramatic global consequences. In adult dync1h1^+/−, impaired retrograde axonal transport in PRCs was primarily attributed to the ER dysfunction, characterized by ciliary hypoplasia and insufficient supply of phototransduction proteins from the IS to the OS. RNA sequencing revealed significant upregulation of ER stress-associated genes in dync1h1^−/−. Western blot and quantitative RT-PCR confirmed that dync1h1-deficiency activates BiP-ATF4-CHOP signaling pathway, leading to PRC apoptosis. We hypothesize that the molecular mechanism of PRC degeneration involves ectopic accumulation of phototransduction proteins, possibly accompanied by toxic byproducts that cannot be cleared from the cell body and IS over time. Excessive accumulation of misfolded proteins may increase ER stress in PRCs, leading to pathological apoptosis.

Unresolved ER stress is crucial in the pathology of various ophthalmic diseases, including glaucoma, cataract, achromatopsia, RP, diabetic retinopathy, optic nerve degeneration, and age-related macular degeneration.^33,34 During normal conditions, BiP binds to ER stress sensors, including ATF6, inositol-requiring enzyme 1α, and protein kinase R-like ER kinase (PERK), which detect misfolded proteins at critically high concentrations. During ER stress, BiP dissociates from these sensors, activating them and triggering the unfolded protein response (UPR), which accelerates protein degradation through ER-associated protein degradation (ERAD). The activation of PERK enhances ATF4 and CHOP expression, promoting transcription of apoptosis-related genes that ultimately lead to cell death.³⁵ The CHOP pathway plays a critical role in retinal degeneration induced by ER stress. For instance, T17M mice show elevated ATF4 and CHOP levels, contributing to PRC degeneration.³⁶ Similarly, the P23H rat exhibited UPR activation, as shown by an elevation in BiP and Chop mRNA expression.³⁷ Moreover, rd16 mouse experienced persistent upregulation of ATF4 and CHOP.³⁸ All of them shared one similarity that rhodopsin abnormally accumulated in the cell body of PRCs,³⁹ providing evidence for the pathogenic roles of ER stress in ophthalmic diseases. Targeting ER stress pathways might provide valuable insights for exploiting new therapeutic interventions. However, because dync1h1 is ubiquitously expressed in various tissues, we cannot exclude systemic lethal factors associated with loss of dync1h1, such as the dysfunction of motor system and blood supply. Our results suggest that apoptosis of PRCs is highly correlated with ER stress. However, it is still necessary to develop conditional knockout models in the future to further comprehend with retina-specific phenotype.

In conclusion, our findings point out that dync1h1 may be an important candidate for microphthalmia and retinal degeneration, which plays critical roles in the polarized trafficking of phototransduction proteins. Dync1h1-deficiency disrupts the retrograde axonal transport system, leading to ectopic accumulation of phototransduction proteins in the cell body or IS. This accumulation increases ER stress and upregulates BiP-ATF4-CHOP signaling, thereby disrupting PRC morphogenesis and maintenance, ultimately leading to apoptosis. Our study offers valuable insights into the pathogenesis of retinal degenerative diseases and opens new clues for precise diagnosis and therapeutic targeting.