The Role of TRiC-enhanced Actin Folding in Leber Congenital Amaurosis

Abstract Purpose Mutations in TCP-1 ring complex (TRiC) have been associated with Leber Congenital Amaurosis (LCA). TRiC is involved in protein folding and has 8 essential subunits including CCT5. Herein, we studied the retina of TRiC mutant zebrafish to evaluate the possible role of impaired actin and tubulin folding in LCA. Methods The cct5 tf212b retina was histologically studied using Toluidine Blue staining as well as TUNEL, BrdU-labeling, and Phalloidin assays. Retinal organisation was assessed by quantification of the cellularity utilising DAPI. Results Laminar organization of cct5 tf212b retinas was intact. Enhanced apoptosis throughout the cct5 tf212b retina was not compensated by higher proliferation rates, leaving the cct5 tf212b retina smaller in size. Quantification of retinal layer cellularity demonstrated that specifically the numbers of the amacrine and the retinal ganglion cells were depleted, suggesting that the cct5 tf212b retina was not uniformly affected by the reduced actin folding. Conclusion Whereas the current literature suggests that LCA is predominantly affecting retinal photoreceptor cells and the retinal pigment epithelium, cct5 tf212b analyses demonstrated the important role of folding of actin by TRiC, suggesting that cct5 tf212b is a useful tool to specifically analyze the role of F-actin filaments in the context of LCA.


INTRODUCTION
Leber Congenital Amaurosis (LCA) is an early-onset congenital retinal dystrophy causing severe visual impairment. Over 38 disease-causing variants involving various pathological mechanisms have been associated with LCA, which consequently presents with a broad clinical spectrum, typically including poor to near-absent pupillary responses, nystagmus, photophobia, and severe visual impairment. [1] Diagnosis is made within the first year of life usually based on extinguished electroretinograms, detecting little if any retinal activity.
Compound heterozygous missense mutations in CCT2 have been identified in individuals suffering from LCA. [2] Computational modeling combined with biochemical studies suggested that these mutations combined induce partial, and not complete, loss of TRiC functionality. [2] CCT2 is a subunit of the eukaryotic TCP-1 ring complex (TRiC, also called chaperonin containing TCP-1 [CCT]); an ATP-driven chaperonin that aids mis-or unfolded proteins in their folding. TRiC has a barrel-shaped structure with two back-toback rings, each comprising of eight paralogous subunits (CCT1 -8). [3] Whereas several substrates have been described for TRiC, [4] actin as well as all -and -tubulin are the main folding substrates. [5,6] Accordingly, genetic analysis of TRiC in the zebrafish vertebrate model system in vivo revealed that loss of TRiC function leads to severe muscle and neuronal defects provoked by deficits in folding of actin and tubulin. [7,8] The notion from structural data that loss of a single subunit leads to full loss of TRiC function was confirmed by genetic analysis of single and compound mutants. [3,7] In a forward genetic screen, the zebrafish mutant tf212b was isolated based on its abnormal locomotion behavior and later reported to harbor a missense mutation leading to G422V replacement within the Cct5 subunit of TRiC. [7,9] In contrast to mutants with a full loss of TRiC function, the missense mutation in cct5 212 leads to actin misfolding by TRiC, whereas tubulin folding is not affected. [7] In order to assess the role of impaired actin folding in retinal development, the retina of cct5 212 was studied. As tubulin folding is intact in cct5 212 homozygotes, defects of the retina within cct5 212 homozygotes can be attributed to impaired actin folding. According to the impaired folding of actin within cct5 212 , retina analyses revealed that the formation of F-actin filaments is reduced in the cct5 212 retina. Although proliferation was enhanced within the ciliary marginal zone, the retina in cct5 212 mutants was smaller in size, likely due to enhanced apoptosis. Further cellularity quantification of the retinal layers revealed a reduction number of retinal ganglion cells and amacrine cells, indicating that these cell types are specifically affected by reduced actin folding.

Immunohistochemistry and Histological Analysis
Immunohistochemistry as well as the stains with Toluidine Blue, DAPI, and H&E were performed according to standard methods. Only retinal cross sections with optic nerve were analyzed. Retinal areas were measured on H&E or Toluidine Bluestained cross sections using the software ImageJ. BrdU labeling was performed by soaking larvae in 1% bromodeoxyuridine (BrdU) dissolved in fish water followed by 14 µm cryostat cross sectioning and treatment with fluorescein conjugated BrdU monoclonal antibodies (BMC9318, Merck) as described. [10] Apoptosis was detected using the in situ Cell Death Detection Kit as recommended by the manufacturer (Roche). All fluorescence images were recorded on a Zeiss ImagerZ1 fluorescence microscope (Zeiss, Germany). Images within figures show representatives of a minimum of six analyzed larvae per genotype.

Comparison of F-actin Levels
F-actin was detected on 14 µm cryostat cross sections with phalloidin conjugated with AlexaFluor-568 (A12380, Life Technologies). To compare F-actin levels, phalloidin-stained sections were imaged on a Zeiss ImagerZ1 fluorescence microscope (Zeiss, Germany) under constant conditions and subsequently analyzed for their brightness values. Utilizing the software Fiji, the area of the inner plexiform layer (IPL) was selected and the mean of all grey values of the pixels within this area was measured, resulting in a single grey value per quantified IPL. To enable comparison of the brightness of the IPL from different genotypes, obtained grey values were rescaled to siblings set to 100%. To rescale grey values of siblings, measured grey values (A 1 to A ) of individual IPL were multiplied by 100 and divided by the average of all measured grey values of siblings using ×100 ∑ =1 / . To normalize values of mutants, measured grey values (B 1 to B ) of individual IPL were multiplied by 100 and divided by the average of the measured grey values of the siblings using ×100 ∑ =1 / . Only IPL from retinal cross sections were analyzed that showed the optic nerve. Ten IPL were analyzed per genotype (n = 10).

Quantification of Cellularity of Retinal Layers
At six days post fertilization (dpf), transverse sections were stained with DAPI to label cell nuclei. Subsequent to imaging using a Zeiss ImagerZ1 fluorescence microscope (Zeiss, Germany), images were converted to 8-bit format and adjusted for contrast, brightness, and threshold. The outer nuclear layer, the inner nuclear layer, and the ganglion cell layer were distinguished based on their separation by the plexiform layers. The inner nuclear layer was further subdivided into a basal and an apical layer based on the difference in the fluorescence intensity of their nuclei. All DAPImarked cell nuclei in individual layers were counted as described earlier. [11] Five sections per genotype on the level of the optic nerve were included in the quantification (n = 5).

Statistical Analysis
Significance between two groups was determined by Student's t-test. Statistical significance was calculated using the software Prism (GraphPad Software). Presented data are mean ± standard error of the mean (SEM).

RESULTS
To document the retinal degeneration at larval stages within zebrafish, cct4 −14 mutants were utilized that feature loss of TRiC function due to their lack of the Cct4 subunit. [7] Semi-thin cross sections of retinas from cct4 −14 homozygotes were generated and stained with Toluidine Blue at six dpf [ Figure 1A]. In contrast to the highly organized layers of the sibling retina, within cct4 −14 homozygous only a rudimentary lens and dispersed pigments mainly accumulated in a band shape were present. Defined cell structures were absent between a rudimentary lens and clustered pigment, indicating a full degeneration of the cct4 −14 retina. In contrast, the retina of cct5 212 homozygotes, in which folding of actin but not tubulin is impaired, showed defined nuclei and largely intact layering comparable to their siblings [ Figure 1B]. However, whereas signs of retinal degeneration were not obvious on Toluidine Bluestained sections, the cct5 212 homozygous retina seemed reduced in size compared to their sibling retina, indicating that minor defects could be apparent within the cct5 212 retina.
To study the retina of cct5 212 homozygotes in more detail, the retinal area was quantified on H&E-stained cross sections. Quantification of the retinal area revealed that the retina of cct5 212 homozygotes was significantly reduced in size compared to their siblings at three dpf as well as six dpf [ Figure 2A]. To further study the reduced retinal size of cct5 212 homozygotes, cells in the S phase were labeled using a single 1-hr pulse of the thymidine analogue bromodeoxyuridine (BrdU) prior to fixation at three dpf. Subsequent analysis of BrdU-labeled cells on cross sections revealed that although the location of the proliferating cells within the retinal ciliary marginal zone was not altered in cct5 212 homozygotes [ Figure 2B], significantly more cells were proliferating during the 1-hr BrdU pulse within cct5 212 homozygotes compared to siblings [ Figure 2C]. To assess if the significantly smaller cct5 212 retinal area despite their higher proliferation rate could be attributed to increased cell death, apoptotic cells were labeled on sections using the TUNEL assay (TdTmediated dUTP-biotin nick end labeling). With an  sizes revealed that the retinal area of cct5 121 homozygotes was significantly smaller compared to siblings. The mean retinal area of cct5 121 homozygotes was 17.8 ± 0.8 (×10 3 µm 2 ) at three dpf and 23.8 ± 0.2 (×10 3 µm 2 ) at six dpf compared to siblings that featured mean retinal areas of 24.7 ± 0.5 (×10 3 µm 2 ) at three dpf and 28.4 ± 0.5 (×10 3 µm 2 ) at six dpf; n = 10 per genotype and stage. (B) In contrast to three-dpf-old siblings, more BrdU-positive cells (green) were detected at the periphery of the retina within the ciliary marginal zone of cct5 121 homozygotes. (C) Quantification of the number of BrdU-positive cells within individual retinas revealed that significantly more cells were proliferating within the ciliary marginal zone of cct5 121 homozygotes at three dpf. Whereas cct5 121 had 48 ± 3 BrdU-positive cells per retinal cross section, 18 ± 1 were proliferating in siblings (n = 5). (D) In contrast to the retina of three-dpf-old siblings, in which apoptotic cells were rarely labelled by the TUNEL assay (green), apoptosis was frequently observed throughout the entire retina of cct5 121 homozygotes. Data are mean ± SEM; ***P < 0.001 by Student's t-test. Arrowheads indicate optic nerves (ON). average of 1.2 ± 0.3 TUNEL-positive cells per retinal cross section, apoptotic cells were rarely found on retinal sections of three-dpf-old siblings [ Figure 2D]. In contrast, sections from cct5 212 homozygotes harbored significantly more TUNELpositive cells with an average of 60 ± 6 per section, indicating that apoptosis was greatly enhanced in cct5 212 mutants (n = 10, P < 0.0001). In summary, the significantly reduced size of the retinal area of cct5 212 homozygotes could be attributed to the greatly enhanced apoptosis that could not be compensated by the significantly higher proliferation within the ciliary marginal zone.
Residual actin-based thin filaments are formed within the sarcomeres of cct5 212 skeletal muscle cells, as TRiC only enhances actin folding but is not absolutely required. [7] In order to unveil F-actin filament assembly within the cct5 212 retina, cross sections were stained with the Factin marker phalloidin and counterstained with DAPI at three dpf. As F-actin mainly locates on the inner surface of cell membranes, the inner and the outer plexiform layers of the sibling retina are strongly marked by phalloidin due to the high neurite densities within these regions [ Figure 3A]. In comparison, the phalloidin signal of both plexiform layers appeared weaker in Phalloidin also marked the optic nerve (ON). (B) Quantification of the brightness of the phalloidin signal within the IPL revealed that cct5 121 homozygotes showed a significant reduction in signal intensity to 54 ± 4% in relation to their siblings, which were rescaled to 100 ± 3%. Data are mean ± SEM; ***P < 0.001 by Student's t-test; n = 10. Per retinal cross section, siblings had 240 ± 10 RGC and 199 ± 9 AC in contrast to 180 ± 9 RGC and 142 ± 6 AC in cct5 121 homozygotes. The number of HC and BC as well as PR remained unchanged. Siblings had 273 ± 20 HC and BC as well as 104 ± 24 PR and cct5 121 homozygotes had 276 ± 6 HC and BC as well as 97 ± 3 PR. Data are mean ± SEM; **P < 0.01 by Student's t-test; n.s., not significant; n, 5. cct5 212 homozygotes [ Figure 3A]. To quantify the signal intensity of phalloidin from both genotypes, staining of retinal sections and imaging were performed under constant conditions for both cct5 212 homozygotes and siblings. Subsequent quantification of the brightness of the IPL revealed that the mean signal intensity from cct5 212 homozygotes was significantly reduced compared to their siblings, indicating a reduced level of Factin filaments within cct5 212 mutants. Taken together, in accordance with the reduced actin folding detected in skeletal muscle of cct5 212 homozygotes, [7] biogenesis of F-actin filaments was also impaired within the retina, although F-actin filaments were still assembled within cct5 212 mutants.
Using the nuclear stain DAPI, the laminar organization of the retina can be visualized that includes the outer nuclear layer, the inner nuclear layer, and the ganglion cell layer [ Figure 4A]. The inner nuclear layer can further be subdivided into the basal positioned amacrine cells and the apical compartment with the bipolar and horizontal cells. [12] To further characterize the retinal defects, the cell number within individual retinal layers were counted at six dpf. Quantification of the cellularity of the outer nuclear layer with photoreceptor cells as well as the apical inner nuclear layer with the horizontal cells and the bipolar cells revealed that the number of these cell types remained unchanged between siblings and cct5 212 mutants [ Figure 4B]. However, quantification of the cellularity of the basal inner nuclear layer and the ganglion cell layer revealed that both layers of cct5 212 homozygotes contained significantly less nuclei compared to their siblings, suggesting that amacrine cells as well as retinal ganglion cells were diminished within cct5 212 homozygotes [ Figure 4B]. Taken together, these results indicate that specifically the amacrine and the retinal ganglion cells of cct5 212 mutants are affected by the reduced biogenesis of F-actin filaments.

DISCUSSION
Analysis of zebrafish cct5 212 mutants revealed that their retinas were reduced in size and their level of F-actin filaments. Enhanced proliferation was not able to compensate for the enhanced apoptosis and numbers of the amacrine and the retinal ganglion cells were depleted.
LCA is a severe congenital retinal dystrophy with over 38 disease-causing variants, including CCT2. [1,2] Mutations in the zebrafish ortholog Cct2 as well as other TRiC subunits have been reported to result in retinal degeneration. [7,8,13] Retinal ganglion and amacrine cells are reduced in number in cct5 121 mutants. (A) At six dpf, the retinal nuclei of siblings and cct5 121 mutants were marked at the level of the optic nerve using the nuclear stain DAPI. Indicated by the dotted lines, the outer nuclear layer contains the photoreceptor cells (PR). The inner nuclear layer harbours the apical horizontal cells (HC) and the bipolar cells (BC) as well as the basally located amacrine cells (AC). The most basal layer contains the retinal ganglion cells (RGC). Arrowheads indicate optic nerves (ON). (B) Quantification of the cellularity of the different retinal layers within siblings and cct5 121 homozygotes at six dpf. The cell number of the RGC and the AC were significantly reduced in cct5 121 homozygotes. Per retinal cross section, siblings had 240 ± 10 RGC and 199 ± 9 AC in contrast to 180 ± 9 RGC and 142 ± 6 AC in cct5 121 homozygotes. The number of HC and BC as well as PR remained unchanged. Siblings had 273 ± 20 HC and BC as well as 104 ± 24 PR and cct5 121 homozygotes had 276 ± 6 HC and BC as well as 97 ± 3 PR. Data are mean ± SEM; **P < 0.01 by Student's t-test; n.s., not significant; n, 5. However, these mutants were full or partial loss-of-function mutants of individual TRiC subunits resulting in full or partial loss of TRiC function. Consequently, these studied mutants featured deficiencies in folding of both actin and tubulin, the main substrates of TRiC. Accordingly, retinal degeneration within cct4-deficient mutants, with only a rudimentary lens and clustered pigment remaining, was confirmed at six dpf in this study.
In order to specifically distinguish the effects of impaired actin folding from the other functions of TRiC in the context of LCA, the retina of the mutant cct5 212 was assessed. Although the size of the retina was reduced at three and six dpf, the six-dpfold cct5 212 retina was organized in defined layers and appeared largely intact. Although proliferation of the cct5 212 retina was enhanced, abundant apoptosis was detected as well, suggesting that the smaller retinal area might be attributed to the greatly increased apoptosis. These findings are in accordance with zebrafish cct2 394 −7 mutants that showed a higher number of retinal S phase cells. [13]