De Novo ACTG1 Variant Expands the Phenotype and Genotype of Partial Deafness and Baraitser–Winter Syndrome

Actin molecules are fundamental for embryonic structural and functional differentiation; γ-actin is specifically required for the maintenance and function of cytoskeletal structures in the ear, resulting in hearing. Baraitser–Winter Syndrome (B-WS, OMIM #243310, #614583) is a rare, multiple-anomaly genetic disorder caused by mutations in either cytoplasmically expressed actin gene, ACTB (β-actin) or ACTG1 (γ-actin). The resulting actinopathies cause characteristic cerebrofrontofacial and developmental traits, including progressive sensorineural deafness. Both ACTG1-related non-syndromic A20/A26 deafness and B-WS diagnoses are characterized by hypervariable penetrance in phenotype. Here, we identify a 28th patient worldwide carrying a mutated γ-actin ACTG1 allele, with mildly manifested cerebrofrontofacial B-WS traits, hypervariable penetrance of developmental traits and sensorineural hearing loss. This patient also displays brachycephaly and a complete absence of speech faculty, previously unreported for ACTG1-related B-WS or DFNA20/26 deafness, representing phenotypic expansion. The patient’s exome sequence analyses (ES) confirms a de novo ACTG1 variant previously unlinked to the pathology. Additional microarray analysis uncover no further mutational basis for dual molecular diagnosis in our patient. We conclude that γ-actin c.542C > T, p.Ala181Val is a dominant pathogenic variant, associated with mildly manifested facial and cerebral traits typical of B-WS, hypervariable penetrance of developmental traits and sensorineural deafness. We further posit and present argument and evidence suggesting ACTG1-related non-syndromic DFNA20/A26 deafness is a manifestation of undiagnosed ACTG1-related B-WS.


Introduction
Actin isoforms β and γ are identical except for four amino acid substitutions within the first 10 N-terminal residues. Both isoforms are vital to embryonic differentiation and interact in some tissues, with each having hundreds of first-level interactions with other cellular molecules [1]. Nonetheless, these forms have different specific expression patterns and are involved in multiple, largely separate processes, with little functional redundancy [1,2]. γ-Actin expression predominates in the auditory hair cells [3] and is required for the formation and maintenance of the structures involved in the mechanotransduction of sound waves to electrical signals which are relayed to the brain as hearing [4][5][6][7]. Consistent  [52] In the literature to date, these variants do not overlap with the 17/51 confirmed B-WSrelated ACTG1 variants [53]. The two conditions are therefore considered to be separate molecular and clinical diagnoses. Pathology that is neither B-WS nor deafness has been reported for 13 other listed ACTG1 variants (2,12,13,14,23,27,31,37,40,46,47,49 and 69; Table 1). Pathogenic or possibly pathogenic variants listed for ACTG1 therefore currently fall into three non-overlapping subsets according to the reported phenotype, with causal implications. However, the accumulated literature [10,24,50] reveals a snowballing degree of wobble in the molecular classification of ACTG1 variants into subsets causing different pathologies. As individual ACTG1-related B-WS cases and familial DFNA20/26 studies accumulate, as well as ACTG1-linked pathologies reported as unassociated with either, the overlap of the symptoms and molecular etiology has become increasingly evident. A compilation of the data to date in Table 1 presents this information, which is taken up in our discussion. This has considerable diagnostic, clinical and genetic counseling impact, as well as biomedical interest. Here we present a 28th B-WS patient carrying a previously unassociated mutation in ACTG1 that encapsulates and informs this issue.

Results
Our patient is a boy born at term in the 39th week of the mother's third pregnancy by natural delivery. Parents (father: 40 years, healthy; mother: 39 years at date of birth, treated with Plaquenil for lupus erythematosus diagnosed 3 years earlier) are non-consanguineous Caucasians of Polish descent. During pregnancy, nuchal translucency at the 95th percentile and a higher risk of trisomy 21 was determined by a non-invasive test (PAPP-A), and a cell-free fetal DNA (cffDNA) test was performed. This showed a low risk of Down syndrome.

Patient Parameters at Birth
Patient parameters at birth were as follows: weight, 3350 g (50 c); length, 56 cm (>95 c); occipital frontal circumference (OFC), 34 cm (25 c); Apgar score, 10; head circumference, 34 cm (−2 SD of the normal mean). Distinctive dysmorphic features and cryptorchidism were noted; an ultrasound examination revealed unilateral duplication of the pelvicalyceal system and echocardiography showed a membranous ventricular septal defect (VSD). Ultrasound images of all other internal organs were normal. The patient tested negative at birth for toxoplasmosis and cytomegaly.
Facial features noted at birth and as the child grew ( Figure 1) show wide-set eyes (hypertelorism), microcephaly and a wide face. ence, 34 cm (−2 SD of the normal mean). Distinctive dysmorphic features and cryptorchidism were noted; an ultrasound examination revealed unilateral duplication of the pelvicalyceal system and echocardiography showed a membranous ventricular septal defect (VSD). Ultrasound images of all other internal organs were normal. The patient tested negative at birth for toxoplasmosis and cytomegaly.
Facial features noted at birth and as the child grew ( Figure 1) show wide-set eyes (hypertelorism), microcephaly and a wide face. Clinically, these specific facial features presenting together are strongly indicative of B-WS. Brachycephaly, a short nose and a long philtrum were also noted at birth and were present at all later examination age points. Progressively, we saw the following: at age 3 months: hypertelorism, short nose and round face; at age 8 months: small chin, retrognathia and posteriorly rotated ears; at age 12 months: progressive and more pronounced arching of the eyebrows, elongated palpebral fissures, mouth corners directed down, face becoming less round and more elongated, and unilateral ptosis; at age 2 years 2 months: high forehead with frontal bossing evident; at age 4years: strong unilateral ptosis, more pronounced arching of the eyebrows and face shape changing with age, becoming increasingly less round.

Sensorineural Deafness
Screening tests after birth suggested deafness. Hearing tests between the ages of 2 to 4 months were inconclusive. At 4 months, severe deafness was diagnosed using the Clinically, these specific facial features presenting together are strongly indicative of B-WS. Brachycephaly, a short nose and a long philtrum were also noted at birth and were present at all later examination age points. Progressively, we saw the following: at age 3 months: hypertelorism, short nose and round face; at age 8 months: small chin, retrognathia and posteriorly rotated ears; at age 12 months: progressive and more pronounced arching of the eyebrows, elongated palpebral fissures, mouth corners directed down, face becoming less round and more elongated, and unilateral ptosis; at age 2 years 2 months: high forehead with frontal bossing evident; at age 4years: strong unilateral ptosis, more pronounced arching of the eyebrows and face shape changing with age, becoming increasingly less round.

Sensorineural Deafness
Screening tests after birth suggested deafness. Hearing tests between the ages of 2 to 4 months were inconclusive. At 4 months, severe deafness was diagnosed using the Auditory Brainstem Response (ABR) test, with 70 dB (left) and 80-90 dB (right) auditory deficits. At age 3 and a half months, CT imaging revealed shadowing of the entire pneumatic structures of both pyramids, suggesting inner ear inflammation. His middle ear canals were drained, anti-inflammatory treatment was applied and bilateral hearing aids were used from 6 months. Brainstem auditory evoked potentials (BAEP) at 5 months and at 16 months were positive. Hearing tests at 3 years 4 months showed sensorineural hearing loss of medium degree with a reaction to sound at 35 dB and 40-45 dB with and without hearing aids, respectively. This reflects an appreciable regression of deafness, out of character for B-WS or isolated ACTG1-related deafness, which are both normally progressive. We conclude that early infection of the ear canals and subsequent healing were responsible for the regressive nature of deafness in our patient, since hearing improved after drainage and anti-inflammatory treatment. We note that congenital malformations of the skull (microcephaly, brachycephaly) may contribute to an anomalous inner ear structure causing secretion and fluid accumulation, and a predisposition to ear canal infections exacerbating hearing loss.
Magnetic resonance imaging (MRI) of the brain at 3 months (Figure 2a-h) revealed hypoplasia/fenestration of the anterior falx cerebri with the gyri of the right cerebral hemisphere crossing the brain midline (Figure 2a (Figure 2f-h). The anatomical line normally separating the caudate and lentiform nuclei from the anterior limbs of the internal capsules (ALICs) remains undetectable at the structural and myelination levels (Figure 2b-h, white arrows), indicating severely disrupted/delayed myelination. This is currently considered rare in B-WS patients [8].
A brain MRI at 18 months (Figure 2i-n) showed that opercularization remained incomplete (Figure 2i, red arrows). The anterior limbs of the internal capsules (ALICs) that are normally appreciable at 10-11 months were not detected. The basal ganglia appear to be fused bilaterally (Figure 2j-n, arrows showing fusion lines).
A CT of the skull at age 4 months (Figure 2o-q) showed a brachycephalic skull with a cephalic index of 91.9, indicating pathologically disturbed proportions of the skull. This exceeded the range of 76 to 81 compatible with a mesaticephalic skull in normal males. A SD greater than 3.5 between the length and width of the skull was noted at 33 months (−2.29 SD vs. 1.32 SD). We note that the initial physical examination suggested positional brachycephaly with a discrete asymmetry of the occiput (plagiocephaly), but a later CT examination ruled out positional skull deformation because there were clear differences in the width of the lambdoid suture: it was narrow on the left side and wide on the right (Figure 2r,s).
A chest X-ray at 2 months showed an abnormal shape of the ribs (third to eighth).

Molecular Analyses
We used array comparative genomic hybridization (aCGH) to compare the patient's and reference human genomes, and assess the genomic content variation and gene copy number aberrations. Our patient shows a normal genomic copy number and genomic content.
Patient exome screening for possible pathological mutations using exome sequencing (ES) identified a missense variant c.542C > T (NM_001614.5), p.Ala181Val (NP_001605.1) in the patient's ACTG1 gene (Figure 3c  Family segregation analyses (Figure 3a,b) showed that the variant c.542C > T, p.Ala181Val arose de novo in one allele in our patient. In silico pathogenic prediction analysis resulted in 23 out of 25 prediction algorithms classifying this variant as deleterious, with high scores produced by the CADD and DANN algorithms ( Table 2). The patient's ES data revealed no additional pathogenic or likely pathogenic mutations to support dual molecular diagnoses for the patient's phenotype.  Family segregation analyses (Figure 3a,b) showed that the variant c.542C > T, p.Ala181Val arose de novo in one allele in our patient. In silico pathogenic prediction analysis resulted in 23 out of 25 prediction algorithms classifying this variant as deleterious, with high scores produced by the CADD and DANN algorithms ( Table 2). The patient's ES data revealed no additional pathogenic or likely pathogenic mutations to support dual molecular diagnoses for the patient's phenotype.

Discussion
Our patient displays cranial, facial and developmental traits which together are clinically typical for B-WS [8]. These include microcephaly, specific dysmorphic facial features, intellectual disability, developmental defects of speech and brain microstructure, myelination defects and sensorineural hearing loss (Table 3, #14).
He also carries a de novo point mutation in the γ-actin gene ATCG1 but no other mutation that could support a dual molecular diagnosis [14]. The specific point mutation identified here, c.542C > T, p.Ala181Val, was previously recorded in the dbSNP database (rs797044730) and the ClinVar clinical database (VCV000197198.4) as a variant with conflicting interpretations of pathogenicity, with no specific phenotype described; it is currently unlinked to any pathology. There is no further mention in the general literature, and it is not present in population genomic databases gnomAD v2 and v3. A pathogenic likelihood score of "potentially deleterious" was obtained for this variant when assessed by 23/25 predictive algorithms (Table 2). According to the American College of Medical Genetics and Genomics (ACMG) classification, this variant is classified as pathogenic (PS2, PM2, PM5, PP2, PP3). Furthermore, basal rate germline nucleotide point mutations per generation result in only one or two coding sequence mutations in a given exome [14], and our exome analyses detected only c.542C > T, p.Ala181Val, a coding missense point mutation in ACTG1. All B-WS-associated alleles for ACTG1 reported to date are missense point mutations (Table 1), and are usually de novo, although an inherited ACTG1 mutation linked to non-syndromic DFNA20/A26 deafness and B-WS in the same family has been reported [10]. Taken together, these data indicate that the variant identified here is responsible for the characteristic B-WS traits present, including deafness, and contributes to the phenotype expansion observed. Neither brachycephaly nor a complete absence of speech faculty have been previously described in ACTG1-related B-WS patients, nor for ACTG1-related isolated sensorineural deafness cases.
We therefore identify and reclassify the variant c.542C > T, p.Ala181Val as a dominant pathogenic allele of the ACTG1 gene that is causative for BW-S and also sensorineural deafness. Table 3. List of all 28 published patients with deleterious mutations in the ACTG1 gene and B-WS syndrome. Nd-no data; na-not applicable; R-regressive; CC-corpus callosum; F-female; M-male; F*-female proband in the study; F∆-mother of F*; M -father of F*; M1 and M2-first and second patients from one study, red and bold-patient first described in this study. Our patient is the third person documented as carrying this particular variant of ACTG1 and the 28th patient worldwide documented with an array of symptoms resulting from the expression of defective γ-actin molecules providing a clear diagnosis of B-WS.
Currently, the ACTG1 variants listed in HGMD (69 including the variant presented here) fall into three non-overlapping groups associated with different symptomology: sensorineural deafness DFNA20/26, B-WS and symptomology apparently unrelated to either deafness or B-WS (Table 1, white background, grey background and green background, respectively). This leads to distinct molecular and clinical diagnoses, prognoses, therapies and genetic counseling for each group.
There are also nine positions in the amino acid sequence of ACTG1 ( This leads to two different variants with an associated pathology listed in the HGMD. In most cases (7/9), the associated pathology is the same for both variants (deafness in 5/9 positions; B-WS in 2/9 positions); in two further cases (Table 1, positions p.181 and p.256), we see B-WS for one variant (#34, p.Ala181Val; #50 p.Arg256Trp), with either deafness (#33) or pachygyria (#49), respectively, for the alternative variant. Variant #34, (p.Ala181Val) represents our patient; different substitutions at the same position in ACTG1 result in B-WS in our patient rather than isolated deafness, an apparently different phenotype. Deafness and pachygyria, however, are present in 76% and 89% respectively, of formally diagnosed ACTG1-related B-WS cases where an appropriate examination was performed (Table 1). We note that while a highly specific intragenic correlation of ACTG1 mutation with a particular trait cannot at present be formally excluded without rigorous molecular mapping, the phenotype is identical for both variants in 7/9 of positions dispersed between positions p.82 and p.332 (77%) of the ACTG1 protein sequence (375 residues). Further, in 9/9 positions, the phenotype is entirely consistent with hypervariable penetrance of traits within the spectrum for B-WS.
In short, all mutations in ACTG1 listed in the HGMD are associated with phenotypes consistent with hypervariable expression of B-WS traits. An extensive cellular network of first-level interaction partners exists for γ-actin molecules [1]. Together with normal variation in the patient's genetic landscape, this implies considerable variability inherent in the mutation load impacting actin physiology and, therefore, actinopathies arising from a single pathogenic variant in different patient backgrounds. A simple interpretation is, therefore, that all these differing phenotypes, including deafness, reflect variable trait manifestation of undiagnosed B-WS. We therefore contend the phenotypes reported in 37/69 deafness-associated ACTG1 variants, as well as the 18/69 confirmed variants with a pathology consistent with isolated B-WS traits (Table 1), though currently associated with neither B-WS nor deafness, result from variable penetrance of molecular variants underlying actinopathies capable of causing B-WS.
Given the progressive nature and hypervariable trait penetrance of ACTG1-related B-WS and DFNA20/26 deafness, which are particularly well showcased in the familial context of segregating ACTG1 pathogenic alleles, we hope that this perspective of the overlapping etiology may be clinically useful and, crucially, inform specific molecular diagnoses and genetic counseling, in addition to therapeutic handling.

Array Comparative Genomic Hybridization (aCGH)
We used the CGX 3x720K Whole-Genome array (Roche NimbleGen, Madison, WI, USA) as per the manufacturer's instructions and scanned slides into image files using an MS 200 Microarray Scanner (Roche NimbleGen, Madison, WI, USA). Data were analyzed using DEVA (Roche NimbleGen, Madison, WI, USA) and Genoglyphix Analysis Software (Perkin Elmer, Waltham, MA, USA).

Exome sequencing (ES)
Genomic DNA was extracted from peripheral blood leukocytes of the patient and his parents using an automatic magnetic bead-based method using MagNA Pure 96 system, Roche, Germany. Whole-exome sequencing libraries were prepared using Agilent SureSelect XT Human All Exon V6 sample preparation kits and the Illumina NovaSeq 6000 sequencer (Illumina, CA, USA), via 2 × 100 bp reads. Genomic data processing was based on an in-house pipeline, and the reads were aligned with Burrows-Wheeler Aligner (BWA) v0.7.16 software to the GRCh38 reference genome. The Genome Analysis Toolkit (GATK) HaplotypeCaller v4.0b4 was used for variant calling and the Ensembl Variant Effect Predictor (VEP) v96 was used to annotate the variants.
For variant prioritization, the ES data were first analyzed for the known pathogenic or likely pathogenic variants reported in the ClinVar database. The in silico gene panel was used for rare variants (under 0.01% in the gnomAD v2 database) in genes associated with microcephaly consisting of over 800 genes. Those genes were manually selected from several gene panels customized for patients with microcephaly, and additional genes were selected from various databases such as OMIM or DECIPHER. The detailed gene list is included in Supplementary Table S1. Variant pathogenicity prediction and ACMG classification were carried out with the use of the VarSome website [55] and the dnNSFP v4.2 database [56].
Polymerase chain reaction was performed by use of the FastStart Taq DNA Polymerase, dNTPack kit (Roche) as per the manufacturer's instructions and the following primers: forward: TCCAGGTTTCTCATTTGGTTTCT; reverse: CCCGACAGCACCGTGTT (100 ng template DNA; annealing temperature, 58 • C; polymerase activity time, 1 min; 35 cycles; product length, 759 nt).