Expanding The Clinical Phenotype and Genetic Spectrum of PURA-Related Neurodevelopmental Disorders

Background: PURA-related neurodevelopmental disorders (PURA-NDDs) include 5q31.3 deletion syndrome and PURA syndrome. PURA-NDDs are characterized by neonatal hypotonia, moderate to severe global developmental delay/intellectual disability (GDD/ID), facial dysmorphism, epileptic seizures, nonepileptic movement disorders, and ophthalmological problems. PURA-NDDs have recently been identied and underestimated in neurodevelopmental cohorts, but their diagnosis is still challenging. We retrospectively reviewed the clinical characteristics, genetic spectrum, and diagnostic journey of patients with PURA-NDDs. Results: We report 2 patients with 5q31.3 microdeletion and 5 with PURA pathogenic variants. They demonstrated hypotonia (7/7, 100%), feeding diculties (4/5, 80%), and respiratory problems (4/7, 57%) in the neonatal period. All of them had severe GDD/ID and could not achieve independent waking and verbal responses. Distinctive facial features of open-tented upper vermilion, long philtrum, and anteverted nares and poor visual xation and tracking with or without nystagmus were most commonly found (5/7, 71.4%). There were no signicant differences in clinical phenotypes between 5q31.3 microdeletion syndrome and PURA syndrome. PURA-NDDs need to be considered as a differential diagnosis in individuals who show severe hypotonia, including feeding diculties since birth and severe developmental retardation with distinctive facial and ophthalmological features. Conclusions: Our data expands the phenotypic and genetic spectrum of PURA-NDD. Next-generation sequencing methods based on the detailed phenotypic evaluation would shorten the diagnostic delay and would help this rare disorder become a recognizable cause of neurodevelopmental delay.

There is no doubt that chromosomal microarray (CMA) as a rst-tier diagnostic test enhances the diagnostic yield for patients with GDD/ID [9,10]. Wholeexome sequencing (WES) also provided new opportunities to diagnose sporadic de novo variants in neurodevelopmental disorders [11,12]. PURA-NDDs have recently been identi ed and may be underestimated, accounting for less than 1% of the cause of neurodevelopmental delay [13][14][15]. There are universal clinical features of PURA-NDD; however, associated features expanding the phenotypes of PURA-NDDs are increasingly reported [16,17]. Furthermore, most reported PURA-NDD patients are in their infancy or childhood, and the natural clinical course of PURA-NDDs is rarely reported.
We report 2 individuals with the 5q31.3 deletion and 5 individuals with de novo pathogenic variants in PURA. We described their clinical features, including the detailed neurodevelopmental history and genetic spectrum. We retrospectively reviewed the diagnostic journey that individuals went through for etiologic and differential diagnoses.

Patients and study approval
We identi ed 7 patients with PURA-NDD from our cohort who were referred to the pediatric neurology clinic for unexplained GDD/ID. Six of them participated in the Korean Undiagnosed Disease Program and were diagnosed with PURA-NDDs [18]. All medical records, including birth history, detailed developmental history, and brain magnetic resonance imaging (MRI), were reviewed retrospectively. This study followed the Declaration of Helsinki's principles and was approved by the institutional review board of Seoul National University Hospital (IRB No. 1912-136-1091.

Molecular analysis
Blood samples were obtained from enrolled patients and their parents who provided informed consent. Genomic DNA was extracted from peripheral blood samples using a QIAamp DNA Blood Midi Kit (Qiagen, Valencia, CA, USA). CMA testing was performed with either the Agilent Human Genome oligonucleotide comparative genomic hybridization (CGH) microarray 8x60K (Agilent Technologies, Santa Clara, CA, USA) or the Cytoscan Dx Assay (Affymetrix, Thermo Fisher Scienti c) according to the manufacturer's instructions. Array data were analyzed using Genomic Workbench software (version 7.0.4.0, Agilent Technologies). Whole-exome sequencing (WES), including exome capture and sequencing, was performed at the Theragen Etex Bio Institute (Suwon, Korea) following the standard protocol. Sequenced reads were aligned to the hg19 reference genome with Burrows-Wheeler Aligner (v.0.7.15) and were processed with Picard software (v2.8.0), SAMtools (v.1.8) and Genome Analysis Toolkit (GATK, v.4.1.4). Variants were annotated using ANNOVAR. The pathogenicity of variants was evaluated according to the American College of Medical Genetics (ACMG) standard guidelines and annotated by InterVar [19,20]. All of the PURA variants identi ed by WES were con rmed by Sanger sequencing. Segregation tests were performed to con rm de novo occurrence of the pathogenic PURA variants.

Results
Clinical features with developmental outcome There were 2 patients with 5q31.3 deletion syndrome and 5 patients with PURA variants. Detailed clinical features and genetic information are summarized in Table 1. Neonatal problems were evident in all individuals immediately after birth. All of them had profound neonatal hypotonia (7/7, 100%). Feeding di culties requiring tube feeding (4/5, 80%) and apnea and/or respiratory distress (3/7, 43%) were associated in the neonatal period. All patients had severe gross motor delay, and they never attained independent walking (7/7, 100%) at the last follow-up (age range 3-18 years). The maximal motor function in our patients was standing with support at age 3 years or older. All individuals had severe language impairment in that they could not speak any single words (7/7, 100%). Non-epileptic movements, including exaggerated startling responses and jerky movements, were associated in 43% (3/7) of individuals. Two patients (2/7, 28.5%) had epileptic seizures that were well controlled with conventional anti-seizure medications. On reviewing brain MRIs, the most frequently observed feature was widening of the subarachnoid spaces, suggesting decreased cerebral volume (6/7, 86%). Delayed myelination was also observed on MRI in 29% (3/7) of patients. Five of 7 patients (71.4%) presented distinctive facial features including open-tented upper vermilion, long philtrum, and depressed nasal bridge with anteverted nares. They also had ophthalmological problems such as poor visual xation and tracking with or without nystagmus. On polysomnography, patient No.1 had central apnea that required bilevel positive airway pressure. Patient No.2 was the oldest patient in our cohort and was found to have primary amenorrhea. Her pubertal development at 17 years of age was Stage 4 (breasts and pubic hair) in Tanner staging; however, she did not have menarche.

Diagnostic journey
We reviewed all patients comprehensively and shared their diagnostic journey and their developmental outcome (Table 1). Prior to the identi cation of PURA-NDD, patients underwent evaluation related to Prader-Willi syndrome (4/7, 57%), metabolic screening (4/7, 57%), or spinal muscular atrophy (2/7, 29%) in their neonatal or infancy periods. Five of seven patients (71.4%) underwent muscle biopsies, electromyography, or nerve conduction studies due to severe hypotonia with feeding di culty and facial weakness. All patients underwent initial diagnostic evaluation in their neonatal period and were referred to our center before age of 2 years. However, it took an average of 5 years (range 1-16 years) to obtain the genetic diagnosis of PURA-NDDs from the time they were referred to our center. For example, we could not diagnose patient No. 2 in the neonatal period, which eventually caused the patient to be lost to follow-up for 15 years. Prior to the identi cation of PURA-NDD, she was considered to have cerebral palsy and continued physical therapy. New-onset epileptic seizures at age 17 brought her to our clinic. Finally, we diagnosed her with 5q31.3 microdeletion syndrome based on a chromosomal microarray.  Figure 1A) We found four missense and two nonsense variants from ve patients, which were all de novo ( Figure 1B). Patient No. 5 interestingly had two pathogenic PURA variants (c.228C>G, c.331C>T) that were located in cis.

Discussion
PURA-related NDD has been recently recognized as a cause of ID with advances in genetic technologies. To date, 11 patients with 5q31.3 deletion syndrome and 85 patients with PURA syndrome, including our patients, have been reported. The universal clinical features of PURA-NDDs are moderate to severe ID with verbal language delay and profound neonatal hypotonia [15]. We con rmed that all of our patients with PURA-NDDs displayed hypotonia in the neonatal period. Their developmental milestones regarding gross motor and language were severely retarded beyond the neonatal period. All patients in our cohort were not able to attain independent gait even with appropriate physical therapy, similar to most previously reported patients (1). Some earlier studies described patients with PURA variants who were able to achieve their rst steps from 28 months to 7 years and even maintained independent gait at age 20 [13,16]. Language delays were evident in all patients, as they were nonverbal. Cinquina et al. summarized other associated clinical features, including skeletal, cardiac, urogenital, and dermatological abnormalities, in PURA syndrome [15]. They mentioned that distinctive facial traits were reported in 84.5% of patients with PURA syndrome. As demonstrated by computational analysis, dysmorphic facial features include myopathic faces, full cheeks, high anterior hairlines, shorter palpebral ssures, and prominent philtrums [13]. Concordant with earlier studies, 71.4% of our patients had distinctive facial features with typically opentented upper vermillion, long philtrum, and anteverted nares. Ophthalmological problems such as nystagmus, poor visual xation and tracking were also commonly associated in our patients. The prevalence was higher than that in the pre-existing literature, in which strabismus was an accompanying symptom in approximately 30-40% of individuals with PURA-NDDs [6, 13,14]. Exaggerated startling responses were associated in 43% of patients, which was similar to previous reports [15]. Endocrine abnormalities such as aberrant sex and thyroid hormone levels have also been reported in PURA-NDD [13,21]. Bonaglia et al. PURA-NDD was rst described and identi ed by microarray-based CGH [2]. As WES has been gradually expanded in the clinical setting, a number of PURA syndromes caused by de novo pathogenic variants have been increasingly reported [5,6,14]. PURA-NDDs do have core clinical phenotypes. Reijnders et al. depicted a striking example of diagnosing PURA syndrome in a neonate through targeted Sanger sequencing based on clinical phenotype [13]. However, there is still a broad spectrum of clinical features and variability in clinical severity within PURA-NDD. These variabilities challenge the diagnostic process even for specialized clinicians. When we retrospectively reviewed the diagnostic process, more than half of patients underwent evaluations under the impression of central or neuromuscular hypotonia such as Prader-Willi syndrome, inherited myopathies and metabolic disorders in their neonatal period due to hypotonia, feeding di culty, and respiratory problems. In their infancy, more than 70% of patients had muscle biopsies and electrophysiological studies to rule out neuromuscular disorders. Reviewing other works, clinicians frequently performed single-gene testing of Prader-Willi syndrome, Fragile X syndrome (FMR1), metabolic testing and occasional mitochondrial studies [1,16]. Similar to our study, muscle biopsies and myotonic dystrophy testing were also frequently performed. Due to respiratory problems or irregular breathing patterns, a central hypoventilation panel or PHOX2 sequencing was applied in a few patients. We suggest that PURA-NDDs need to be considered as a differential diagnosis in patients who have severe neonatal hypotonia with feeding and respiratory di culty, followed by profound global developmental delay, particularly in verbal expression.
We did not observe differences in the severity of clinical features between individuals with 5q31.3 microdeletions and single-nucleotide pathogenic variants in PURA. Our study supports that PURA is a primary causative gene for the core neurodevelopmental features of 5q31.3 microdeletion syndrome [5]. Similar to previous studies, we could not nd correlations between the types or locations of variants and clinical severity and variability [13,23]. Patient No.5 had two pathogenic variants, and the phenotypic severity was not different from the others. This mechanism is suggested by a dominant-negative effect for structural variants and functional haploinsu ciency for truncating mutations. This might be explained by the involvement of other genetic or biological mechanisms.

Conclusions
PURA-NDDs have universal clinical features of neonatal hypotonia and severe GDD/ID with severely impaired verbal expression. Based on the detailed clinical phenotypes, PURA-NDD is gradually being illustrated in neurodevelopmental disorders. Thus, we should be able to obtain increasing amounts of information about the natural course of disease, which would lead to earlier diagnosis and better patient care.

Declarations
Ethics approval and consent to participate This study protocol was in accordance with the tenets of Declaration of Helsinki and was approved by the Institutional Review Board of Seoul National University.

Consent for publication
Not applicable Availability of data and materials All data generated or analyzed for the study are available from the corresponding author upon reasonable request.

Competing interests
The authors have no con ict of interest of nancial relationship to disclose. Author's contributions CSA and HSL prepared the rst draft of the manuscript and edited the manuscript drafts. TJP and YJK contributed to the data acquisition. SP and SYK managed all the process of CMA, WES and Sanger sequencing. SP and HSL analyzed the WES data. BCL and KJK supervised the study and critically reviewed the manuscript. JHC conceived of the study and reviewed edited the manuscript drafts until the nal draft was produced. All authors read and approved the nal manuscript.