Molecular and Clinical Characterization of CNGA3 and CNGB3 Genes in Brazilian Patients Affected with Achromatopsia

Achromatopsia (ACHM) is a congenital cone photoreceptor disorder characterized by reduced visual acuity, nystagmus, photophobia, and very poor or absent color vision. Pathogenic variants in six genes encoding proteins composing the cone phototransduction cascade (CNGA3, CNGB3, PDE6C, PDE6H, GNAT2) and of the unfolded protein response (ATF6) have been related to ACHM cases, while CNGA3 and CNGB3 alone are responsible for most cases. Herein, we provide a clinical and molecular overview of 42 Brazilian patients from 38 families affected with ACHM related to biallelic pathogenic variants in the CNGA3 and CNGB3 genes. Patients’ genotype and phenotype were retrospectively evaluated. The majority of CNGA3 variants were missense, and the most prevalent CNGB3 variant was c.1148delC (p.Thr383Ilefs*13), resulting in a frameshift and premature stop codon, which is compatible with previous publications in the literature. A novel variant c.1893T>A (p.Tyr631*) in the CNGB3 gene is reported for the first time in this study. A great variability in morphologic findings was observed in our patients, although no consistent correlation with age and disease stage in OCT foveal morphology was found. The better understanding of the genetic variants landscape in the Brazilian population will help in the diagnosis of this disease.


Introduction
Achromatopsia (ACHM) is a rare genetic retinal disease that is inherited in an autosomal recessive; it is estimated to affect one in 30,000 people [1,2]. Clinical symptoms usually present after birth or early infancy, and are typically characterized by reduced visual acuity, nystagmus, photophobia, and very poor or absent color vision. These symptoms are due to a primary functional defect of the cone photoreceptors that is reflected in a severely reduced or absent light-adapted electroretinogram (ERG) and a preserved scotopic ERG signal. While the fundus appearance is often normal, abnormal foveal reflex, pigmentary mottling, and atrophic changes may be found in the macula area, especially in advanced cases [3,4]. Patients do not report progression of symptoms, and the disease was initially thought to be nonprogressive. However, previous studies established structural alterations and foveal findings that can emerge and are compatible with a slow progressive degeneration and loss of cone photoreceptor cells [5,6].
ACHM can be defined as complete or incomplete depending on the extent of cone photoreceptor dysfunction and resulting severity of symptoms [7]. Patients with incomplete ACHM present with a milder phenotype, residual color discrimination, better visual acuity, and/or residual photopic ERG responses [2,3]. In these cases, the diagnosis is occasionally made even later in childhood.
Pathogenic variants in six genes encoding components of the cone phototransduction cascade (CNGA3, CNGB3, PDE6C, PDE6H, GNAT2) and of the unfolded protein response (ATF6) account over 90% of ACHM cases, while CNGA3 and CNGB3 alone are responsible for most cases [1,2]. CNGB3 pathogenic variants constitute approximately 40-50% of cases and are more common in the Caucasian population (Europe and the United States) [8][9][10]. CNGA3 pathogenic variants underlie approximately 30-40% of cases and are more common in the Asian population (Middle East and China), accounting for 80% of all cases in this region [11].
The majority of CNGB3 pathogenic variants are nonsense, frameshift, or splicing mutations that result in truncated or loss of function channel proteins [8][9][10][11]. In contrast, most CNGA3 pathogenic variants are missense mutations that affect only single amino acid residues of the protein [6,11,12].
In 2018, our group published the frequency of inherited retinal dystrophies in Brazil [13]. At that time, we reported on six patients with ACHM-associated genes. Now, we are able to expand our study on CNGA3 and CNGB3 pathogenic variants related to ACHM in Brazilian patients.

Patient Selection
A retrospective study of medical records from two centers specializing in inherited retinal dystrophies was performed, one located in São Paulo (Federal University of São Paulo and Instituto de Genética Ocular) and one in Rio de Janeiro (Instituto Brasileiro de Oftalmologia). Forty-six Brazilian patients with clinical diagnosis of ACHM were identified. Among them, 42 patients from 38 families had conclusive genetic testing with pathogenic variants in CNGA3 and CNGB3 genes. Medical and family histories were collected, as were genetic data. Only patients in whom the diagnosis could be genetically confirmed were included.

Ophthalmic Examination
The clinical diagnosis of ACHM was based on detailed clinical examination, visual function, signs/symptoms, ophthalmologic features, and age of onset. Patients underwent detailed ophthalmic exams, including best-corrected visual acuity (BCVA), contrast sensitivity (CS), slit-lamp exams, and multimodal retinal imaging using color fundus images, fundus autofluorescence (FAF), and optical coherence tomography (OCT). BCVA was assessed with the Early Treatment Diabetic Retinopathy Study (ETDRS) chart, while CS was measured with the Pelli Robson chart at 1 m.
Fundus photographs were reviewed to confirm the findings reported in the medical record.
The severity of degeneration shown on the OCT was graded in different stages: preserved inner segment ellipsoid, disrupted inner segment ellipsoid, inner segment ellipsoid loss, presence of a hyporeflective zone, inner segment, and retinal pigment epithelium (RPE) loss.

Genotyping
The records of patients who underwent genetic testing for ACHM causative variants in CNG genes were reviewed. Genes associated with ACHM were included in a larger panel of genes associated with inherited retinal disease. ACHM was considered genetically confirmed if two pathogenic or likely pathogenic variants in one of the six known genes were identified in the patient. Segregation was performed when possible.
Genetic analysis was performed using commercial next-generation sequence (NGS) panels for inherited retinal disorders with either 224 or 330 genes (see Supplemental Table S1 for the list of genes analyzed). Genomic DNA obtained from the submitted sample was enriched for targeted regions using a hybridization-based protocol and sequenced using Illumina technology. Confirmation of the presence and location of reportable variants was performed based on stringent criteria established by each accredited diagnostic laboratory.
The standards and guidelines provided by the American College of Medical Genetics and Genomics (ACMG) and the Association for Molecular Pathology (AMP) [14] were applied in order to classify the identified variants. Novel variants were classified as pathogenic or likely pathogenic when representing a loss of function variant (frameshift or nonsense or copy number variation or affecting a canonical splice site). In addition, pathogenic score was evaluated when allele frequency in the gnomAD population databases was extremely low. Two platforms were assessed that combine computational predictions with clinical support, segregation, or functional studies in order to assist with variant calling; both of which use sets of rules that follow ACMG criteria: Franklin (https://franklin. genoox.com) and Varsome (https://varsome.com), both last accessed on 28 April 2023. Variants found were compared with variations listed in ClinVar (https://www.ncbi.nlm. nih.gov/clinvar/ accessed on 28 April 2023).
This study was performed in accordance with the Declaration of Helsinki and the protection of patient identity and was approved by the Research Ethics Committee of the Federal University of São Paulo (protocol number 5.113.810). Written informed consent was obtained whenever it was necessary to perform molecular tests. When DNA samples were collected for molecular tests, all patients and/or their legal guardians provided written informed consent for the use of the personal medical data for scientific purposes and publication.

Results and Discussion
We identified 46 patients with clinical diagnosis of ACHM. Forty-two patients presented biallelic pathogenic variants in the CNGA3 and CNGB3 genes, while four patients presented biallelic pathogenic variants in the PDE6C gene. In our cohort, we did not find any ACHM patients related to the other described genes (PDE6H, GNAT2, or ATF6). Table 1 summarizes the genotype identified in the affected individuals with CNGA3 variants, while Table 2 shows the same results for CNGB3 variants. Both tables show the classification according to the ACMG guidelines.  In this Brazilian sample of 42 patients from 38 families with ACHM related to biallelic variants in CNG genes (19 patients in CNGA3 and 23 patients in CNGB3), 20 pathogenic variants in CNGA3 gene and 11 pathogenic variants in CNGB3 gene were identified, including one novel variant in CNGB3 gene. The pathogenicity of this novel variant was "likely pathogenic" according to the ACMG classification.
In this cohort, among CNGA3 variants, sixteen were missense variants (80%), three were nonsense (15%), and one was an initiation codon variant (5%). CNGA3 pathogenic variants were found in 46% of patients. These findings are in line with previously conducted studies in the literature [6,12], where the most prevalent variants were found to be missense followed by nonsense.
Considering CNGB3 variants, three were frameshift (27%), four were nonsense (36%), three were splice-site variants (27%), and one was an initiation codon variant (9%). The most common CNGB3 variant found in this cohort was the deletion c.1148delC (p.Thr383Ilefs*13), resulting in a frameshift and premature stop codon most prevalent in homozygosity; this accounted for 45% (11 patients) of all CNGB3-linked genotypes, while six patients (25%) presented this variant in at least one allele in heterozygosis. One nonsense novel variant (c.1893T > A; p.Tyr631*) is described here for the first time. This variant meets a very strong (PVS1) and a supporting (PM2) ACMG criteria, with extremely low frequency in gnomAD databases. This sequence change creates a premature stop signal in the CNGB3 gene. It is expected to result in an absent or disrupted protein product. Loss of function variants in CNGB3 are known to be a mechanism of disease; 185 pathogenic null variants were reported in ClinVar for this gene across 18 different exons. This variant was classified as deleterious by three pathogenicity predictors (MutationTaster, DANN and BayesDel), was classified as "likely pathogenic" by the reporting lab and was absent in ClinVar. Table 3 resumes variant data, showing the allele count in this cohort and the total allele frequency from all populations in gnomAD. The prevalence of CNGA3 or CNGB3 variants varies globally [9][10][11]. Mayer et al. evaluated CNGB3 pathogenic variants in 485 independent families with ACHM, mainly of Western Europe and North America descent [10]. The c.1148delC variant was by far the most common variant in that cohort, accounting for 66% of all CNGB3-linked ACHM alleles. It was demonstrated that the high prevalence of this variant was due to a founder effect, and the presence of this variant in European population is most likely due to a single mutation event [10].
The cone CNG channel is composed of three CNGA3 and one CNGB3 subunits located exclusively in the plasma membrane of the outer segment of cone photoreceptors [15]. To date, more than 230 pathogenic variants in CNGA3 [12] and around 200 pathogenic variants in CNGB3 [10] have been found to cause inherited ACHM in humans. All known diseasecausing variants are inherited in an autosomal recessive manner, and only homozygous or compound heterozygous patients show the typical symptoms of ACHM.
Tables 1 and 2 summarize BCVA and CS exams. BCVA ranged from 20/100 to 5/400. The only patient presenting CF in one eye (P31.1) had been submitted to vitrectomy surgery in the right eye due to retinal detachment after contuse trauma. When available, CS ranged from 0.55 to 1.65 logCS. These findings did not show any correlation with age. Another patient presented reduced visual acuity and CS (P15.2) with a severe phenotype that differs from other patients, including his affected sister (P15.1).
One clinical trial with ACHM associated with the CNGA3 gene has published 1-year [16] and 3-year follow-ups [17], showing improvements in secondary endpoints in visual acuity and contrast sensitivity compared to baseline data.
Fundoscopy findings varied from normal fundus appearance to atrophy in the foveal area. Correlating to color fundus photos, FAF presented different findings, varying from a normal fundus autofluorescence to a reduced autofluorescence with subtle hyperautofluorescence ring around the fovea. Figure 1 exemplifies the retinal findings. On morphologic exams, OCT shows varying degrees of foveal abnormalities in the inner segment ellipsoid zone. Representative OCT images are shown in Figure 2. Because certain patient had poor fixation, it was not possible to obtain good horizontal scans for all patients. The severity of degeneration shown on the OCT was graded in different On morphologic exams, OCT shows varying degrees of foveal abnormalities in the inner segment ellipsoid zone. Representative OCT images are shown in Figure 2. Because certain patient had poor fixation, it was not possible to obtain good horizontal scans for all patients. The severity of degeneration shown on the OCT was graded in different stages.  Several clinical studies have investigated outer retinal and foveal morphology in detail by using high resolution OCT in ACHM [18]. The macular appearance in OCT can either show normal anatomy architecture or variable degrees of disruption of the photoreceptor layers and loss of RPE. Previously published cross-sectional studies have described conflicting findings with respect to the age dependency of progression in OCT [18]. Aboshiba et al. suggested that retinal structure alterations in ACHM may be slowly progressive and subtle in most patients and may not be correlated with age or genotype [19]. Triantafylla et al. showed longitudinal changes in foveal structure, mainly in children, though in adults with ACHM as well, over a long follow-up period [20]. Four stages of morphological degenerative changes have been described in ACHM: preserved inner segment ellipsoid, disrupted inner segment ellipsoid, inner segment ellipsoid loss, and inner segment and RPE loss [21]. However, whether morphological changes over time follow the proposed four-stage linear pattern needs to be confirmed through long-term studies.
Great variability in OCT findings was observed in our patients ( Figure 2). Considering disease onset at birth or early childhood, we did not find a consistent correlation with age or disease stage in OCT foveal morphology. There were varying degrees of abnormalities in the inner segment ellipsoid observed in both young and elderly patients. This suggests that progression may not be age dependent.
Limitations of our study include its retrospective nature and the consequent fact that not all data were available for every patient. In addition, data were acquired using various methods and protocols. Finally, segregation data and detailed clinical information were limited.

Conclusions
This paper has presented a considerable cohort of patients with ACHM related to the CNGA3 and CNGB3 genes. The current development of gene therapy for ACHM requires characterization of these patients in detail in order to better understand disease evolution.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/genes14061296/s1, Table S1: Genes included in the NGS panels used.  Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. Written informed consent has been obtained from patients and their parents to publish this paper. Data Availability Statement: Not applicable.

Conflicts of Interest:
The authors declare no conflict of interest.