Mutation Spectrum and De Novo Mutation Analysis in Stickler Syndrome Patients with High Myopia or Retinal Detachment

Stickler syndrome is a connective tissue disorder that affects multiple systems, including the visual system. Seven genes were reported to cause Stickler syndrome in patients with different phenotypes. In this study, we aimed to evaluate the mutation features of the phenotypes of high myopia and retinal detachment. Forty-two probands diagnosed with Stickler syndrome were included. Comprehensive ocular examinations were performed. A targeted gene panel test or whole exome sequencing was used to detect the mutations, and Sanger sequencing was conducted for verification and segregation analysis. Among the 42 probands, 32 (76%) presented with high myopia and 29 (69%), with retinal detachment. Pathogenic mutations were detected in 35 (83%) probands: 27 (64%) probands had COL2A1 mutations, and eight (19%) probands had COL11A1 mutations. Truncational mutations in COL2A1 were present in 21 (78%) probands. Missense mutations in COL2A1 were present in six probands, five of which presented with retinal detachment. De novo COL2A1 mutations were detected in 10 (37%) probands, with a mean paternal childbearing age of 29.64 ± 4.97 years old. The mutation features of probands with high myopia or retinal detachment showed that the probands had a high prevalence of COL2A1 mutations, truncational mutations, and de novo mutations.


Introduction
Stickler syndrome is a connective tissue disorder that was first reported by Stickler et al. in 1965 [1]. The characteristics of Stickler syndrome include ocular findings of myopia, cataracts, and retinal detachment; hearing loss that is both conductive and sensorineural; midfacial underdevelopment and cleft palate (either alone or as part of the Robin sequence); and mild spondyloepiphyseal dysplasia and/or precocious arthritis [2]. The estimated prevalence of Stickler syndrome among neonates is approximately 1:7500 to 1:9000 [3].
Ocular involvement is the main indication for the diagnosis of Stickler syndrome [2,11]. In patients with COL2A1 mutations, myopia is the most common sign, which is present in 90% of patients, followed by vitreous changes, detected in 40% of patients, and cataracts, in 30% of patients. Retinal abnormalities, such as lattice degeneration, retinal holes, retinal detachment, or retinal tears, are present in nearly 20% of patients [12].
In this study, we recruited Stickler syndrome patients with initial symptoms of high myopia or retinal detachment to evaluate the molecular genetic spectrum and the relationship between the genetic findings and the phenotypes of high myopia and retinal detachment.

Materials and Methods
A total of 84 eyes of 42 probands from 42 unrelated families with Stickler syndrome were enrolled in our study. This study was performed in accordance with the guidelines of the Declaration of Helsinki and was approved by the Institutional Review Board of the Zhongshan Ophthalmic Center (approval code 2014MEKY048). Informed consent was obtained from all participants or their guardians prior to the collection of the clinical data and genomic samples.
Genomic DNA was extracted from the leukocytes of venous blood using previously reported methods [13]. Previously published diagnostic criteria for Stickler syndrome [11]-based on assigning points for clinical features, family history data, and molecular data [2]-were used as follows: orofacial abnormalities, including cleft palate (2 points) and characteristic facial features (1 point); characteristic vitreous changes or retinal abnormalities (2 points); auditory abnormalities, including high-frequency sensorineural hearing loss (2 points) and hypermobile tympanic membranes (1 point); skeletal abnormality (1 point); and family history or molecular data (1 point). If the score was ≥5 points, a diagnosis of Stickler syndrome was established. For the patients with a score <5 but vitreous abnormalities and characteristic facial features making them highly suspect for Stickler syndrome, we will consider the following criteria presented by Snead et al. [11]: (1) a congenital vitreous anomaly and, in addition, any three of the following: (2) myopia with onset before 6 years of age, usually stable; (3) retinal detachment or para-vascular pigmented lattice degeneration; (4) joint hypermobility with an abnormal Beighton score, with or without radiological evidence of joint degeneration; (5) audiometric confirmation of a sensorineural hearing defect; and (6) midline clefting.
All carriers underwent a complete ophthalmic investigation, including best corrected visual acuity (BCVA) tests, intraocular pressure measurements, slit lamp examinations of the anterior segment (cornea, iris, and lens), and funduscopic examination. The retinas of the affected eyes were assessed under mydriatic conditions with binocular indirect ophthalmoscopy. Ophthalmic findings at presentation included the lens and vitreous status, type and location of retinal holes, status of the macula, and extent of retinal detachment, which were recorded. Refractive errors were measured by subjective refraction combined with objective refraction. The axial length was measured with an IOLMaster (Zeiss, Berlin, Germany). An ultrasound A system (Compact Ttouch 4.0, Quantel Medical, Cournon-d'Auvergne, France) was used for young children who could not cooperate with the IOLMaster examination and patients with retinal detachment. For fundus examinations, a fundus camera (Zeiss, Oberkochen, Germany), a scanning laser ophthalmoscope (SLO, Opto Inc., Marlborough, MA, USA), or a RetCam (Natus Medical, Inc., Pleasanton, CA, USA) was used, depending on the patient's age and cooperation ability. High myopia was defined as an axial length greater than 26 mm or a spherical equivalent less than −6.00 D in at least one eye.

Demographic and Clinical Characteristics
A total of 84 eyes of 42 probands, 23 (55%) male and 19 (45%) female, from 42 unrelated families were included in this study. The Stickler score for each proband is listed in Supplementary Table S1. The diagnosis of Stickler syndrome was based on the sores and especially based on the typical vitreous changes and retinal abnormalities (Supplementary Figure S1). The median age at the first-time examination was eight years old. The P25 and P75 were 5 and 10.25 years, respectively. The median axial length was 27.72 mm, the P25 being 24.15 mm and the P75 being 29.04 mm. The mean spherical equivalent was −11.91 ± 5.37 D.

De Novo Mutations in COL2A1 and COL11A1
Segregation analysis in the available family members revealed de novo mutations in 37% (10/27) of the families with COL2A1 mutations ( Figure 3D) and 12% (one in eight) of the families with COL11A1 mutations ( Figure 3E). The family histories and family trees of these families were double-checked to verify the biological parenthood. The reproductive ages of the families with de novo mutations are displayed in Table 3. The mean paternal childbearing age was 29.64 ± 4.97 years, and the mean maternal childbearing age was 27.36 ± 4.97 years (Table 3). Except for two parents with mild myopia, all the other parents of the probands with de novo mutations had normal fundi and a BCVA better than 1.0.

De Novo Mutations in COL2A1 and COL11A1
Segregation analysis in the available family members revealed de novo mutations in 37% (10/27) of the families with COL2A1 mutations ( Figure 3D) and 12% (one in eight) of the families with COL11A1 mutations ( Figure 3E). The family histories and family trees of these families were doublechecked to verify the biological parenthood. The reproductive ages of the families with de novo mutations are displayed in Table 3. The mean paternal childbearing age was 29.64 ± 4.97 years, and the mean maternal childbearing age was 27.36 ± 4.97 years (Table 3). Except for two parents with mild myopia, all the other parents of the probands with de novo mutations had normal fundi and a BCVA better than 1.0.

Clinical Manifestation in Probands with COL2A1 Missense Mutations
Four different missense mutations (two known and two novel) were detected in six probands in COL2A1. Proband 20, with c.4385C > T (p.Arg1459Cys), developed retinal detachment in the right eye at the age of 19, following which they underwent pars plana vitrectomy (PPV), and glaucoma in their left eye at the age of 30. Proband 18, with c.3674G > C (p.Gly1225Ala), presented with bilateral leukocoria at two months of age and was subsequently diagnosed with retinal detachment bilaterally. Proband 7, with c.1648C > T (p.Arg550Cys), had high myopia and vitreous degeneration with no detected complications at the age of six. Three probands with c.1693C > T (p.Arg565Cys) mutations developed retinal detachment during early childhood bilaterally at 6, 8, and 10 years of age. One of them underwent timely PPV operations, and the other two developed ophthalmatrophia in one eye. In total, there were five out of six probands with missense COL2A1 mutations that had retinal detachment, whereas 15 out of 21 probands with truncated COL2A1 mutations had retinal detachment. There was no significant difference (Fisher's exact test, p = 0.498) in the numbers with retinal detachment between the probands with missense and truncated COL2A1 mutations. Table 3. Childbearing ages of the parents whose children had de novo mutations.

Clinical Manifestation in Probands with COL2A1 Missense Mutations
Four different missense mutations (two known and two novel) were detected in six probands in COL2A1. Proband 20, with c.4385C > T (p.Arg1459Cys), developed retinal detachment in the right eye at the age of 19, following which they underwent pars plana vitrectomy (PPV), and glaucoma in their left eye at the age of 30. Proband 18, with c.3674G > C (p.Gly1225Ala), presented with bilateral leukocoria at two months of age and was subsequently diagnosed with retinal detachment bilaterally. Proband 7, with c.1648C > T (p.Arg550Cys), had high myopia and vitreous degeneration with no detected complications at the age of six. Three probands with c.1693C > T (p.Arg565Cys) mutations developed retinal detachment during early childhood bilaterally at 6, 8, and 10 years of age. One of them underwent timely PPV operations, and the other two developed ophthalmatrophia in one eye. In total, there were five out of six probands with missense COL2A1 mutations that had retinal detachment, whereas 15 out of 21 probands with truncated COL2A1 mutations had retinal detachment. There was no significant difference (Fisher's exact test, p = 0.498) in the numbers with retinal detachment between the probands with missense and truncated COL2A1 mutations.

Discussion
In this study, we evaluated the mutation spectrum of patients with Stickler syndrome whose initial symptom was high myopia or retinal detachment.
To date, mutations in seven genes (COL2A1 [4], COL11A1 [5], COL11A2 [6], COL9A1 [7], COL9A2 [8], COL9A3 [9], and LOXL3 [10]) have been reported to cause Stickler syndrome. COL2A1 mutations explained the majority (80-90%) of cases [18,23], followed by COL11A1 mutations (10-20%) [23]. COL11A2, COL9A1, COL9A2, COL9A3, and LOXL3 explained only a few cases, mostly documented in case reports or case series [8,10,[24][25][26]. These observations are consistent with this study's findings. We detected mutations in only COL2A1 (64%) and COL11A2 (19%) and no mutations in the other genes. There may be several reasons for this. First, this study included patients with initial signs of high myopia or retinal detachment; however, COL11A2 mutations have been reported in non-ocular Stickler syndrome [25,27]. Second, patients with COL9A1 and COL9A3 mutations mainly present with moderate to severe myopia and rarely with retinal detachment [24,28]. Finally, the rest of the genes are inherited as autosomal recessive traits, whereas the families recruited in this study were autosomal dominant or sporadic cases. In this study, the first examination age was statistically significantly different among the three groups (p = 0.048). However, no statistically significant difference was found in the patient numbers with retinal detachment among the three groups. Perhaps longer periods of follow up and larger numbers of patients would provide additional insights into the different clinical trajectories and outcomes of COL2A1 and COL11A1 mutant patients.
Various types of mutations in COL2A1 have been reported, and truncation mutations (nonsense, frameshift, or splicing) that result in a premature stop codon are present in most patients with Stickler syndrome [1,16]. This is in line with the findings in our study, where truncation mutations were detected in 78% of the probands with COL2A1 mutations. We also detected four missense mutations in six probands, five of whom presented with retinal detachment, which is a fairly severe ocular complication.
De novo mutations in Stickler syndrome have only been reported in a few cases [29]. The proportion of cases caused by de novo pathogenic variants is unknown. In this study, we observed a large proportion of probands with de novo mutations (37% COL2A1 and 12% COL11A1) because these mutations could not be detected in the leukocyte DNA of either parent. The actual proportion of probands with de novo mutations may have been higher as 10 probands were of unknown inheritance. De novo mutations have been established as the cause of early-onset disorders ranging from rare congenital diseases to common neurodevelopment disorders [30][31][32]. Studies have demonstrated that the majority of germline de novo mutations have a paternal origin and that a higher paternal age and the number of de novo mutation increases with paternal childbearing age [33,34]. In this study, we evaluated both the paternal and maternal ages. The paternal childbearing age (29.64 ± 4.97 years old) might be one of the reasons for the high proportion of de novo mutations in this study. Further studies are needed to ascertain the origins of the mutations.
This study was a real-world single-center study. Its sample size was limited due to the rarity of Stickler syndrome and all patients being recruited in a single department (the Pediatric Ophthalmology Department of Zhongshan Ophthalmic Center). The participants presented with an initial sign of high myopia or retinal detachment, which means that the identified mutation spectrum can only be considered representative of the specific phenotypes of Stickler syndrome patients. In this study, we did not use the genomic markers to verify parentage, and the parentage of families with de novo mutations will be further confirmed by genomic markers.

Conclusions
We investigated the mutation features in Stickler syndrome patients with high myopia or retinal detachment. Mutations were detected only in COL2A1 (64%) and COL11A1 (19%). No mutations were detected in other genes known to be associated with Stickler syndrome. De novo mutations accounted for 37% of all COL2A1 mutations.