Review of Genotype-Phenotype Correlations in Usher Syndrome

Abstract

Usher Syndrome (USH) encompasses a group of clinically and genetically heterogenous disorders defined by the triad of sensorineural hearing loss (SNHL), vestibular dysfunction, and vision loss. USH is the most common cause of deaf blindness. USH is divided clinically into three subtypes - USH1, USH2, and USH3 – based on symptom severity, progression and age of onset. The underlying genetics of these USH forms are however significantly more complex, with over a dozen genes linked to the 3 primary clinical subtypes and other atypical USH phenotypes. Several of these genes are associated with other deaf-blindness syndromes that share significant clinical overlap with USH, pointing to the limits of a clinically based classification system. The genotype-phenotype relationships among USH forms also may vary significantly based on the location and type of mutation in the gene of interest. Understanding these genotype-phenotype relationships and associated natural disease histories is necessary for the successful development and application of gene-based therapies and precision medicine approaches to USH. Currently, the state of knowledge varies widely depending on the gene of interest. Recent studies utilizing Next Generation Sequencing technology have expanded the list of known pathogenic mutations in USH genes, identified new genes associated with USH-like phenotypes, and proposed algorithms to predict the phenotypic effects of specific categories of allelic variants. Further work is required to validate USH gene causality, and better define USH genotype-phenotype relationships and disease natural histories – particularly for rare mutations – in order to lay the groundwork for the future of USH treatment.

Introduction

The term Usher Syndrome (USH) encompasses a group of clinically and genetically heterogenous disorders defined by the triad of sensorineural hearing loss (SNHL), vision loss, and vestibular dysfunction. With a combined prevalence of over 400,000 cases worldwide, USH represents the most common cause of deaf-blindness. USH is associated with significant morbidity, impeding the ability of affected individuals to communicate with and navigate the world around them (Ehn et al., 2019). USH is divided clinically into three subtypes - USH1, USH2, and USH3 – based on symptom severity, progression and age of onset. However, this classification system only partially captures the complexity of USH. Twelve genes have been identified as corresponding to the three clinical subtypes, although three of these genes (CIB2, PDZD7, and HARS) have been called into question (http://hereditaryhearingloss.org). A growing list of additional genes have been associated with atypical USH phenotypes (Fuster-García et al., 2019; Mathur & Yang, 2015). For each associated gene, the clinical phenotype can also vary based on the type and location of the causative mutation, further complicating the clinical picture (Table 1) (Ahmed et al., 2008; Lin et al., 2020; Perez-Carro et al., 2018; Ramzan et al., 2018).

Table 1:

Genes Associated with Usher Syndrome

Gene	Protein	Usher Phenotypes
MYO7A	Myosin IIVA	USH1, Atypical USH
CDH23	Cadherin 23	USH1, USH2
PCDH15	Protocadherin 15	USH1
USH1C	Harmonin	USH1
USH1G	Scaffold protein containing ankyrin repeats and SAM domain	USH1, Atypical USH
CIB2	Calcium- and integrin-binding protein 2	USH1
USH2A	Usherin	USH 2, Atypical USH
GPR98	Very large G protein-coupled receptor 1	USH 2, Atypical USH
WHRN	Whirlin	USH 2
CLRN1	Clarin-1	USH 3
HARS	Histidyl-tRNA synthetase	USH 3, Atypical USH
ABHD12	alpha/beta-Hydrolase domain containing 12	USH 2, USH 3, Atypical USH
CEP250	C-Nap1	Atypical USH
CEP78	Centrosomal protein 78	Atypical USH
ARSG	Arylsulfatase g	Atypical USH
PDZD7	PDZ domain-containing protein 7	USH Disease Modifier

Current treatment of USH is entirely symptom directed. Depending on severity, hearing loss can be partially mitigated with hearing aids or cochlear implants, which have been shown to have significant quality of life benefits (Hartel et al., 2017). Management for vision loss – which is secondary to retinitis pigmentosa (RP) – includes visual aids and retinal prostheses, though effectiveness is limited (Fahim, 2018). The primary treatment option for vestibular dysfunction is vestibular therapy, though development of vestibular protheses is ongoing (Guyot & Perez Fornos, 2019). However, none of these treatment modalities address the pathophysiologic mechanisms underlying USH. A variety of gene therapy approaches to the inner ear and retina are being investigated, and research is ongoing to develop mutation-specific treatment modalities (Geleoc & El-Amraoui, 2020). A currently recruiting trial for RP is investigating an intravitreal injection of an antisense oligonucleotide for mutation-specific treatment of the most common variants of exon 13 in the USH2A gene (ProQR, Stellar; Trial #NCT03780257). Another ongoing European clinical trial for vision loss in USH1B patients is employing an adeno-associated virus to deliver vectors to the retina to potentially increase the activity of the affected gene, MYO7A (UshTher, Horizon, 2020; Trial #NCT02065011). The outcomes of these studies have not yet been reported. As with any precision medicine approach, the successful development and application of gene therapies for USH requires a move beyond clinical classifications to a comprehensive understanding of the involved genes, specific pathogenic variants, and genotype-phenotype relationships. Furthermore, a better understanding of the natural histories of specific USH genotypes will be necessary for conducting gene therapy clinical trials. As such, we reviewed the English language literature pertaining to the genes and causative mutations associated with USH from the years 2000-2020, with a focus on the current state of knowledge regarding genotype-phenotype relationships and natural disease histories of the known USH-associated genes.

Usher Syndrome Type I

Usher syndrome type I (USH1), comprising about one-third of USH patients, is the most severe of the three clinical subtypes. It is characterized by severe to profound congenital deafness, vestibular dysfunction, and childhood onset of RP. USH1 patients are born with vestibular areflexia, which may present with gross motor delays; they are also either born with complete SNHL or develop severe deafness in the first twelve months of life (Millan et al., 2011). RP progresses with increased night blindness and decreased field of vision, usually resulting in legal blindness by age 30 (Testa et al., 2017). Compared to the general population, patients with USH1 have decreased health-related quality of life, including poorer physical and psychological health, as well as higher rates of unemployment (Dean et al., 2017; Ehn et al., 2018).

Six genes have been associated with the USH1 clinical phenotype. These genes and their corresponding subtypes are as follows: MYO7A (USH1B), USH1C (USH1C), CDH23 (USH1D), PCDH15 (USH1F), USH1G (USH1G), and CIB2 (USH1J). These genes code for a variety of interrelated proteins found within the retina and the inner and outer hair cells of the inner ear (Mathur & Yang, 2015). Many of these genes are also associated with other genetic deafness disorders such as the nonsyndromic recessive deafness (DFNB) group (Ahmed et al., 2002; Schultz et al., 2011). The USH1 phenotype typically results from either frameshift, nonsense, or splice-site severe pathogenic variants of the aforementioned genes (Toms et al., 2015). In comparison, DFNB is usually a result of either mild missense mutations or leaky splice sites (Riazuddin et al., 2008).

Mutations of MYO7A – which codes for myosin VIIA - are the most common cause of USH1 (USH1B), accounting for 29-50% of USH1 patients (Millan et al., 2011). Myosin VIIA is an actin-based motor protein found in the cytoplasm and stereocilia of inner ear hair cells and within the connecting cilium and periciliary membranes of retinal photoreceptors (Hasson et al., 1995). Mutations in MYO7A are also associated with nonsyndromic recessive deafness 2 (DFNB2), nonsyndromic dominant deafness 11 (DFNA11), and atypical USH (Mathur & Yang, 2015). Several mutations of MYO7A have been associated with USH1B. The nonsense mutation c.6070C>T (p.R2024X) is thought to lead to reduced levels of myosin VIIA or to a dysfunctional resultant protein. A splice-site mutation of MYO7A, c.5168+1G>A, has been reported to contribute to USH1B by removal of the splice donor site and subsequent insertion of a premature stop codon, resulting in protein truncation. Both of these mutations introduce premature stop codons and result in dysfunctional myosin VIIA, and thus both mutations lead to the same observed phenotype in USH1 patients (Cheng et al., 2018). A number of recent studies utilizing next generation sequencing (NGS) have also identified compound heterozygous variants in MYO7A that result in an USH1 phenotype (Lin et al., 2020; Ramzan et al., 2018). In a study of the clinical phenotype of USH1B, patients developed severe bilateral hearing loss at an average age of 5 months. On average, patients were walking consistently by 18 months of age (95% CI ±1 month), with all patients reporting vestibular symptoms (Lenassi et al., 2014; Testa et al., 2017). They began to experience night blindness at an average age of 13 years (95% CI ±1 year) with decreases in visual acuity reported at an average age of 16 (95% CI ±3 years). Fifty percent of patients reached legal blindness by age 40 (Lenassi et al., 2014; Testa et al., 2017).

Mutations of CDH23 – which codes for cadherin 23 – are the second most common cause of USH1 (USH1D), accounting for 19-35% of USH1 patients (Liu et al., 2001; Millan et al., 2011). Cadherin 23 is a cell adhesion molecule found in the hair bundle tip-links and transient interstereocilia and kinociliary links, which connect the kinocilium to nearby stereocilia (Siemens et al., 2004). USH1D patients have been reported to have nonsense, frameshift, splice site, missense, insertions, deletions, small indels, and compound heterozygous mutations of CDH23 (Menghini et al., 2019; Neuhaus et al., 2017; Okano et al., 2019; Oshima et al., 2008). Genotype-phenotype relationships in USH1D vary with mutation type, with some homozygous moderate missense and splice site mutations resulting in hearing impairment or atypical USH features (i.e. variable vestibular dysfunction and progressive SNHL), while homozygous nonsense or frameshift mutations result in a classical USH1 phenotype (Liu et al., 2001; Pennings et al., 2004; Valero et al., 2019). In one recent study, a patient was initially diagnosed with USH2 due to a lack of vestibular symptoms, before genetic testing revealed a causative noncanonical splice site mutation in CDH23 (Valero et al., 2019). Mutations of CDH23 are also associated with nonsyndromic recessive deafness 12 (DFNB12) (Pennings et al., 2004; Schultz et al., 2011).

Mutations in PCDH15 – which codes for protocadherin 15 – account for 11-19% of USH1 patients (USH1F) (Millan et al., 2011). Protocadherin is a cell adhesion molecule similar to cadherin and is a component of hair cell stereocilia tip links and kinociliary links (Ahmed et al., 2006; Ahmed et al., 2001; Kazmierczak et al., 2007). In retinal photoreceptor cells, protocadherin 15 is colocalized to the synaptic terminal with cadherin 23, although the exact function of these two cadherins is still being investigated (Reiners et al., 2005; Zhang et al., 2020). Several mutations have been associated with USH1F, including missense, splice-site, nonsense, and deletion mutations. Some missense pathogenic variants (p.R134G, p.D178G, p.G262D) of PCDH15 are also associated with nonsyndromic recessive deafness 23 (DFNB23) (Ahmed et al., 2008; Jaijo et al., 2012). The clinical phenotype of USH1F patients does not differ significantly from the classical USH1 phenotype, with the majority of USH1F patients diagnosed with profound SNHL in childhood and with RP by age 10 (Jaijo et al., 2012).

Mutations in USH1C – which codes for harmonin – are responsible for 6-7% of USH1 cases (USH1C) (Millan et al., 2011). Harmonin is a PDZ scaffold protein found within upper tip link densities (UTLD) of stereocilia in hair cells; the UTLD, which contains the mechanoelectrical transduction machinery of the stereocilia, is the insertion point for cadherin 23 at the stereociliary membrane of hair cells (Grillet et al., 2009). USH1C is associated with several mutations, including nonsense (p.C224X) and frameshift (p.D124TfsX7) pathogenic variants; a founder effect for the c. 216G>A mutation in this gene is seen in the Acadian population. USH1C mutations are also associated with nonsyndromic recessive deafness 18 (DFNB18) (Johnson et al., 2003).

Mutations in USH1G - which codes for scaffold protein containing ankyrin repeats and SAM domain (SANS) - are responsible for an additional 7% of cases, and are especially prevalent in patients from the United States and United Kingdom (Aller et al., 2007). Like harmonin, SANS is part of the UTLD. A wide variety of mutation types have been described in USH1G patients including frameshift, missense, and nonsense mutations, as well as small deletions, insertions, and gross deletions (Abdi et al., 2016; Imtiaz et al., 2012; Ouyang et al., 2005; Riahi et al., 2015; Weil et al., 2003). While the clinical phenotype of USH1G is typically that of classical USH1, there is some variability. Both a frameshift mutation (c.163 164+13del15) and a homozygous missense mutation (c.1187T>A, p.Leu396Gln) have been associated with an atypical Usher syndrome involving late-onset RP, with the former mutation also being associated with normal vestibular function (Bashir et al., 2010; D’Esposito et al., 2019).

Mutations in CIB2—which codes for calcium- and integrin-binding protein 2 in stereocilia and is thought to maintain calcium concentrations in cochlear hair cells— had previously been associated with an USH1 clinical phenotype which was assigned the designation USH1J (Toms et al., 2015). However, recent genetic studies of families carrying mutations in CIB2 associated this gene only with nonsyndromic deafness 48 (DFNB48), and the authors concluded that variants of CIB2 may not be associated with an USH1 phenotype (Booth et al., 2018). Of note, the evidence from these studies does not definitively preclude the possibility that some variants of CIB2 may be a rare cause of USH.

Usher Syndrome Type II

Usher syndrome type II (USH2) is the most prevalent clinical form of USH, accounting for approximately two-thirds of USH patients. USH2 patients display less severe congenital hearing loss than USH1 patients and have normal vestibular function. Hearing loss in USH2 varies across frequencies, with most patients showing moderate loss at low frequencies and severe loss at high frequencies. Additionally, USH2 patients develop RP later than in USH1, with symptoms typically beginning in the second decade of life or later (Millan et al., 2011).

Three genes have been associated with the USH2 clinical phenotype. The three genes and their corresponding subtypes are: USH2A (USH2A), ADGRV1 (USH2C), and WHRN (USH2D) (Mathur & Yang, 2015). The three proteins produced by the known USH2 genes all localize to the ankle links and synapses of cochlear hair cells and the periciliary membrane complexes of retinal photoreceptors (Michalski et al., 2007; Sahly et al., 2012).

Mutations of USH2A—which codes for the transmembrane cell adhesion protein usherin - are the most common cause of USH2 (USH2A) - accounting for 55-90% of USH2 patients - and the most common cause of USH overall (Millan et al., 2011). Over 130 variants in USH2A have been associated with this subtype, with the most common mutation c.2299delG accounting for up to 45% of cases (Aller et al., 2006; Dreyer et al., 2008; Kuang et al., 2020). Other USH2A pathogenic variants include nonsense, frameshift, missense, deletions, duplications, splice-site mutations, and compound heterozygotes; some of these mutations have been associated with non-syndromic RP (Austin-Tse et al., 2018; Dai et al., 2008; He et al., 2020; Kuang et al., 2020; Nakanishi et al., 2009; Qu et al., 2020; Rivolta et al., 2000; Zhu et al., 2020). In French Canadian populations within Quebec, the deletion c.4338_4339delCT (p.C1447QfsX29) is responsible for over 50% of USH2 cases (Ebermann et al., 2009). Studies of clinical phenotype for USH2A patients report onset of bilateral hearing loss at an average age of 14 and night blindness at an average age of 21; this is consistent with the expected phenotype of USH2 overall. By age 37, mean visual acuity was reported to be 20/40 (Testa et al., 2017). USH2A patients may also demonstrate dysfunctional tactile and vibratory sensation, a phenomenon that has not been established in other USH2 subtypes (Frenzel et al., 2012). However, there is a wide range of phenotypes seen in USH2A patients depending on the specific underlying mutation. The p.(Cys759Phe) mutation - pathologic variant in the Spanish population - displays consistent genotype-phenotype correlations in for both non-syndromic RP and USH2A. These patients display different age at diagnosis of RP and of hypoacusis depending on whether they are homozygous for this variant, compound heterozygous for this variant and a missense variant, or compound heterozygous for this variant with an additional truncating variant; legal blindness and hearing loss occurs earlier in heterozygous patients than in homozygous patients (49 vs. 62 years for blindness and 53 vs. 70 years for hearing loss, respectively) (Perez-Carro et al., 2018). Studies of two newly identified USH2A variants in two Korean families have demonstrated more profound and progressive hearing loss than expected for USH2, requiring cochlear implantation (S. Y. Lee et al., 2019). Furthermore, some USH2A associated genotypes show variable expressivity, with phenotypic variance seen both inter- and intra-familial (Austin-Tse et al., 2018; Bernal et al., 2005; Liu et al., 1999; Qu et al., 2020; Zhu et al., 2020).

Mutations of ADGRV1 (previously known as GPR98) - which codes for VLGR1 (Very large G protein-coupled receptor 1) - account for 3-6% of USH2 cases (USH2C). VLGR1 is one of the largest identified human proteins, consisting of 6306 amino acids and seven transmembrane segments (Mathur & Yang, 2015; Weston et al., 2004; Zhang et al., 2018). Over 51 variants of ADGRV1 have been reported in association with USH2C, including frameshift and missense mutations (Kahrizi et al., 2014; Wei et al., 2018; Zhang et al., 2018). USH2C patients demonstrate moderate to severe hearing loss with little progression before age 40 and RP symptoms beginning in the second decade of life. Genotype-phenotype correlations have been studied for several specific mutations, revealing similar clinical phenotype across all USH2C patients (Hilgert et al., 2009). Unlike USH2A patients who display a wide variation in clinical phenotype, USH2C patients show comparatively little variation (Schwartz et al., 2005). One exception is reported in a recently studied Chinese family, with three affected siblings with novel mutations in ADGRV1 (p.Gly5003fs and p.His6130fs) displaying post-lingual deafness and absence of night blindness, straying from the expected USH2 phenotype (Zhang et al., 2018).

Mutations of WHRN—which codes for the cytoskeletal PDZ scaffolding protein whirlin—are rare and were originally identified in a single German family with USH2 (USH2D) (Ebermann et al., 2007). Most WHRN mutations have been associated with DFNB31.(Yang et al., 2010) There appears to be a correlation between whether the implicated nonsense pathogenic variant occurs at the N- (specific to long isoforms) or C- (common region) terminal and whether it results in USH2D or DFNB1, respectively (Mathur et al., 2015).

Usher Syndrome Type III

Usher syndrome type III (USH3), the least common form of USH, is found mainly in Ashkenazi Jewish and Finnish populations. Patients display progressive, post-lingual SNHL which begins in the high frequency range; most cases of hearing loss are recognized by age 10, but this varies from childhood to 35 years (Geng et al., 2017). The age of onset of RP varies widely but typically begins by the second decade of life. Vestibular dysfunction is present in half of USH3 patients (Millan et al., 2011).

At least one gene, CLRN1, has been associated with the USH3 clinical phenotype (USH3A) (Mathur & Yang, 2015). Mutations in CLRN1—which codes for clarin-1, a protein with four transmembrane domains found in hair bundles and connecting cilium—account for the majority of USH3 cases (Toms et al., 2015). Mutations in CLRN1 are also associated with non-syndromic RP (Khan et al., 2011). Notably, in Ashkenazi Jewish populations, the mutation c.143T>C (p.N48K) is responsible for over 40% of USH cases overall (Ness et al., 2003). Similarly, the mutation c.300T>C (p.Y176X) accounts over 40% of USH cases in the Finnish population (Joensuu et al., 2001). Mutations in an additional gene HARS—which codes for histidyl-tRNA synthetase—have also been linked to USH3, though the association is questionable, with only two patients documented in the literature (Mathur & Yang, 2015). This subtype has been associated with episodes of psychosis in addition to the typical USH3 clinical phenotype, and thus it has been suggested that these patients may actually have other rare genetic syndromes (Mathur & Yang, 2015; Puffenberger et al., 2012).

Other Usher Clinical Phenotypes

A number of atypical USH genes which result in phenotypes that do not fit into the other USH clinical classifications have been identified in the literature (Nolen et al., 2020; Toms et al., 2015). Many of these additional “atypical” gene variants (including variants of PDZD7, HARS, ABHD12, CIB2, CEP250, CEP78, ARSG, and ESPN) and phenotypic presentations have been described as “ultra-rare USH” (Nolen et al., 2020).

In several studies of patients initially given a clinical diagnosis of USH2 or USH3, comprehensive molecular screening subsequently revealed pathogenic mutations in ABHD12, a gene which codes for the enzyme ABHD12 (Eisenberger et al., 2012; T. Li et al., 2019; Sun et al., 2018). Mutations in ABHD12 are primarily associated with PHARC (polyneuropathy, hearing loss, ataxia, RP, and cataracts) syndrome; however none of the patients clinically diagnosed with USH demonstrated polyneuropathy at the time of the studies (Eisenberger et al., 2012; Lerat et al., 2017; T. Li et al., 2019; Nishiguchi et al., 2014; Sun et al., 2018; Yoshimura et al., 2015). While a number of pathogenic missense, nonsense, and splice site mutations have been identified in ABHD12, no consistent genotype-phenotype relationships have been identified, further blurring the distinction between atypical USH and PHARC (Lerat et al., 2017; T. Li et al., 2019; Nishiguchi et al., 2014; Yoshimura et al., 2015).

Patients who are homozygous for a nonsense mutation in CEP250, which codes for ciliary protein C-Nap1, display an atypical USH phenotype with mild RP symptoms and childhood-onset SNHL. These atypical USH patients are also heterozygous for a mutation in C2ORF71, which codes for a ciliary protein and is associated with autosomal recessive RP (Khateb et al., 2014). Two novel nonsense mutations in CEP250 have been identified in a Japanese patient with features similar to Usher syndrome, with bilateral, progressive SNHL beginning at age 13 and late-onset cone-rod dystrophy - rather than RP - causing vision loss at age 44 (Fuster-Garcia et al., 2018).

Several different missense and truncating mutations in the related gene CEP78 have also been associated with an atypical USH phenotype consisting of adolescent or adult onset of progressive cone-rod dystrophy and SNHL without vestibular dysfunction (Ascari et al., 2020; Fu et al., 2017; Namburi et al., 2016; Nikopoulos et al., 2016). CEP78 codes for ciliary protein CEP78, the function of which is not well delineated but which is known to interact with C-Nap1 (Ascari et al., 2020; Fu et al., 2017; Namburi et al., 2016; Nikopoulos et al., 2016). There is disagreement over whether this clinical phenotype meets criteria for atypical USH or is rather a separate entity given the presence of cone-rod dystrophy rather than RP, with one group coining the term “cone-rod dystrophy with hearing loss” to distinguish the phenotype from USH (Ascari et al., 2020; Fu et al., 2017; Namburi et al., 2016; Nikopoulos et al., 2016).

A missense mutation in ARSG (c.133G>T), which codes for the lysosomal sulfatase arylsulfatase g, has been associated with an atypical USH phenotype in three families of Yemenite Jewish origin (Khateb et al., 2018). Patients presented with SNHL and rod-cone dystrophy with a distinctive ring shaped distribution, both of which appeared around age 40 (Khateb et al., 2018).

PDZD7 (PDZ domain-containing 7), which codes for the PDZ scaffold protein found in hair cell ankle link complexes, has been identified as a possible USH disease modifier and contributor to digenic USH in select patients (Ebermann et al., 2010). This protein is a known homolog of the proteins involved in USH1C and USH2D - harmonin and whirlin – and has been shown experimentally to interact with the protein products of USH2A and ADGRV1. PDZD7 mutations have been found in combination with USH2A and ADGRV1 mutations in some families with USH, suggesting a form of digenic inheritance or non-Mendelian genetics that may account for observed intrafamilial differences in phenotype. (Ebermann et al., 2010). However, this association is disputed, as some investigators have proposed that PDZD7 is only associated with NSHL and not USH (Booth et al., 2015). Biallelic PDZD7 pathogenic variants have been associated with moderate to severe autosomal recessive non-syndromic SNHL (S.-Y. Lee et al., 2019).

New Directions in Usher Genotype-Phenotype Classifications

The traditional, clinical phenotype-based classification system for USH – by grouping patients by disease severity – provides a useful framework for patient counseling. However, as our understanding of the genetic basis of USH continues to expand, this system has become increasingly insufficient to encompass the current state of knowledge. With the advent of NGS based screening strategies, the number of causative mutations identified for each USH associated gene has grown rapidly (Fuster-Garcia et al., 2018; Jouret et al., 2019; Y. Li et al., 2019; Okano et al., 2019; Pater et al., 2019; Qu et al., 2020; Santana et al., 2019; Wei et al., 2018; Zhang et al., 2018). This includes variants in common USH genes which result in atypical USH phenotypes, complicating the paradigm of assigning certain USH genes to either USH1, USH2, or USH3 (Bashir et al., 2010; D’Esposito et al., 2019; S. Y. Lee et al., 2019; Pennings et al., 2004; Perez-Carro et al., 2018; Valero et al., 2019; Zhang et al., 2018). Another study which used whole exome sequencing to screen clinically diagnosed USH patients who lacked pathogenic mutations in known USH genes found potentially causative mutations in a variety of hearing and retina related non-USH genes, again highlighting the disconnection between clinical and molecular classifications (Fuster-García et al., 2019). This is in line with data from the University of Iowa’s Genetic Eye-Ear Clinic, which found that the combination of expert clinical evaluation and molecular testing led to a change in diagnosis in 52% of deaf-blind families who presented to the clinic, 29% of who received molecularly confirmed final diagnoses of USH (Stiff et al., 2020). To help clarify the genetic classification of USH, Jouret et al. conducted a meta-analysis of 11 NGS studies of USH patients, and classified the genes involved with USH based on the frequency in which they occurred among the 684 USH patients included, ranging from 50% for USH2A to 0.1% for PDZD7 (Jouret et al., 2019). Likewise, to better clarify genotype-phenotype relationships in USH2A, Molina-Ramirez et al created an allelic hierarchy model of USH2A based on a cohort of isolated RP and USH2A patients with known genotypes for the purpose of predicting the phenotypic effects of specific categories of allelic variants (Molina-Ramírez et al., 2020). When applied to an external dataset of USH2A patients, the model was able to correctly predict clinical phenotype based on genotype in 95% of cases; this represents a novel approach to exploring genotype-phenotype relationships that could potentially be replicated for other USH genes.

In addition to allowing for better patient screening, counseling, and monitoring, a better understanding of genotype-phenotype relationships in USH is critical for better predictions of USH gene mutation outcomes, and for moving forward with the development of the precision medicine-based therapies that represent the future of USH treatment. To successfully design and implement these therapies, a detailed understanding of the involved genes and the pathophysiologic effects of specific causative variants is necessary. Understanding the disease natural history associated with each USH variant is critical both to determining how and when to best intervene in a patient, and to evaluating efficacy of therapeutic interventions. To successfully treat USH, new advances in therapeutic strategies and tools must be accompanied by advances in our understanding of the ever-expanding set of genotype-phenotype relationships defining the disease.