Global developmental delay and a de novo deletion of the 16p13.13 region

Abstract

Many rare genetic variants are associated with the risk of atypical neurodevelopmental trajectories. In this study, we report a patient with developmental delay, autistic traits and multiple congenital anomalies, including congenital heart anomalies and orofacial cleft, with a 0.832 Mb de novo deletion of the 16p13.13 region classified as a variant of uncertain significance. Comparison of similar sized deletions and duplications overlapping the same genes in the DECIPHER database, revealed seven reports of copy number variants (CNVs), four duplications and three deletions. A neurodevelopmental phenotype including learning disability and intellectual disability was noted in some of the DECIPHER entries where phenotype was provided. Although the association between a deletion in this region and an atypical neurodevelopmental trajectory remains to be elucidated, the overlapping CNVs with neurodevelopmental phenotypes suggests possible candidate genes within the 16p13.13 region.

Background

For an increasing number of patients with neurodevelopmental disorders (NDDs), including autism and intellectual disability, a genetic aetiology can be identified.^{1 2} In many countries, microarray analysis is a recommended standard genetic test for patients with unexplained developmental delay and/or autism spectrum disorder.³ Microarray testing detects copy number variants (CNVs), which consist of deletions and duplications of genetic content. Some CNVs have been implicated in NDDs (ie, pathogenic variant), while others are common and have no impact on health or developmental trajectories (ie, benign variant). For many genetic variants, as a result of their rarity, there is insufficient evidence to either rule out or support their contribution to the risk of an NDD. As a result, clinicians are increasingly confronted with patients in whom an NDD prompts genetic testing, which reveals a result classified as a variant of uncertain significance (VUS). The annotation VUS ‘represents a broad category and may include findings that are later demonstrated with additional evidence to be either pathogenic or benign’.⁴ It is therefore important for clinicians to report on patients who present with rare and unique genetic variants and NDDs, as this contributes to evolving evidence towards helping elucidate the possible role of these variants in the aetiology of NDD. In particular, researching the clinical relevance of genes included within the CNV region and assessing the complex overlap in NDD presentations may provide some insight into links between phenotype and the rare genetic variant. In this study, we report a patient with developmental delay and a de novo deletion of the 16p13.13 region overlapping 10 Refseq genes, previously unreported in the literature. This novel deletion is proximal to the well characterised 16p13.3 deletion syndrome, which has been associated with mild to moderate intellectual disability.⁵

Case presentation

The patient was seen in early childhood and was born to healthy, non-consanguineous parents who have two older sons. It is unknown whether there were any maternal infections, or use of medications, alcohol, drugs or smoking during pregnancy. One brother had been diagnosed with a mild learning disability that affects reading, writing and math and the other brother had been diagnosed with a moderate-severe learning disability that affects reading and writing.

The medical record of the patient was reviewed for phenotypic assessments and genetic testing results. Genetic data was extracted from the patient’s clinical records, primarily from reports and consultation notes from the genetics clinic.

An anatomical ultrasound of the patient at 19.5 weeks showed a heart defect and cleft lip and palate. The patient was born prematurely, at 35 weeks with a birth weight of 1.8 kg. His APGAR scores at 1 and 5 min were 8 and 9, respectively. Congenital anomalies consisted of a hypoplastic right ventricle, tricuspid atresia, a cleft lip and palate, and bilateral inguinal hernias. The patient underwent three cardiac surgeries shortly after birth (pulmonary artery banding, pulmonary artery tightening and bidirectional cavopulmonary connection) along with surgery to repair his cleft lip and palate.

When the patient presented at a genetic clinic at 17 months of age, physical examination revealed a height in the 3rd–10th percentile, weight in the 50th percentile and head circumference below the 3rd percentile. Facial characteristics included bilateral epicanthal folds, hypertelorism, and mildly low-set, squarish-shaped pinna. He had normal digits, hands and feet. The remainder of his physical examination, including spine and genitalia, was unremarkable. An MRI of his brain revealed no structural abnormalities. His vision was determined to be in the normal range. He had mild hearing loss in his left ear but normal hearing in his right ear. Developmental assessment suggested global delay, including motor and speech development, the extent of which was not further specified.

Given the patient’s multiple congenital anomalies and developmental delay, a microarray analysis was ordered, which detected a 0.832 Mb deletion in chromosome region 16p13.13, at position g.11,591,626–12,423,578 (hg38). The deleted region includes ten RefSeq genes: LITAF, SNN, TXNDC11, ZC3H7A, BCAR4, RSL1D1, GSPT1, NPIPB2, TNFRSF17 and SNX29. The genes in the region are not OMIM Morbid Map genes, and have not been associated with any medical conditions. The 16p13.3 region was examined by fluorescence in situ hybridisation in the patient and parental samples, which confirmed the deletion in the patient and indicated de novo status. Based on this evidence, the deletion was classified by the clinical lab as a VUS. Given the patient’s medical presentation being highly suggestive of a genetic syndrome, clinical whole exome sequencing (WES) was approved to look for smaller variants beyond deletions and duplications, including singe nucleotide variants. The WES analysis of the trio (child and parents) confirmed the deletion at 16p13.13 (the analysis did not provide refined breakpoints) and evaluated variants that are de novo, compound heterozygous, homozygous, heterozygous and X-linked. No causative variants in known disease genes associated with the reported phenotypes (ie, microcephaly, bilateral cleft lip and palate, inguinal hernia, complex congenital heart defect and developmental delay) were identified. Variant calling in WES data was performed with 103× mean depth coverage and quality threshold of 98.7% (percentage of exome covered by at least 10 sequence reads).

Phenotypic data was primarily informed by a recent assessment by our psychiatry clinic, which provides a comprehensive, interdisciplinary, developmental and cognitive assessment to children with genetic variants known or suspected to be associated with neurodevelopmental outcomes. The assessment included a semi-structured psychiatric interview with the patient and his parent(s) by a senior child and adolescent psychiatry resident, a staff child and adolescent psychiatrist and a psychologist. Psychoeducational testing was subsequently performed by a psychometrist using standardised assessments, and the results were interpreted by a psychologist.

At the time of presentation in our psychiatry clinic, the boy was preschool aged. He had evidence of both expressive and receptive speech delay. Although he had not learnt to walk independently until 2 years of age, his motor skills were appropriate for his age at time of the appointment. He also had difficulties with daily living skills for his age. Psychiatric assessment revealed some inattention and hyperactive/impulsive behaviours, as well as mildly atypical communication (eg, using another’s body to communicate), circumscribed interests, sensory sensitivities and unusual sensory interests, suggesting the possibility of autism spectrum disorder. Psychological test results are summarised in table 1.

Table 1.

Summary of clinical features and psychological assessment data for the proband (16p13.13 deletion)

Genetic result	16p13.13 de novo deletion
Age range	Preschooler (3–5 years of age)
Sex	Male
Physical/medical characteristics
	Microcephaly
	Bilateral cleft lip and palate
	Hypoplastic right heart
	Tricuspid atresia
	Bilateral inguinal hernia
	Mild hearing loss in left ear
Psychological assessments
Intelligence test: WPPSI-IV	Full Scale IQ: 54 (<1st)* General Ability Index: 57 (<1st) Verbal Comprehension Index: 58 (<1st)
Language test: OWLS-II	Listening comprehension: 69 (2nd) Oral expression: 63 (1st) Oral language composite: 64 (1st)
Adaptive functioning: VABS-3	Communication: 70 (2nd) Daily living: 80 (9th) Socialisation: 83 (13th) Adaptive Behaviour Composite: 76 (5th)
Autism spectrum disorder symptoms
ADOS-2 Module 1 (total scores >8 indicate possible ASD diagnosis)	Social affect: 2 Restricted and repetitive behaviour: 3 Total score: 5
ADI-R	Total score below algorithm cut-offs for all three domains Social interaction (cut-off=10): 7 Communication and language (non-verbal cut-off=7): 1 Restricted and repetitive behaviour (cut-off=3): 1
SRS-2 (parent report)	No deficits in reciprocal social behaviour
CBCL (parent report)	Slightly elevated score for attention problems

*Norm referenced standard scores and percentiles in parentheses. Scores have a mean of 100 and SD of 15.

ADI-R, Autism Diagnostic Interview-Revised; ADOS-2, Autism Diagnostic Observation Schedule, Second Edition; ASD, autism spectrum disorder; CBCL, Child Behaviour Checklist; OWLS-II, Oral and Written Language Scales, Second Edition; SRS-2, Social Responsiveness Scale, Second Edition; VABS-3, Vineland Adaptive Behaviour Scales, Third Edition; WPPSI-IV, Wechsler Preschool & Primary Scale of Intelligence, Fourth Edition.

During standardised testing, the patient had extreme difficulty focusing on tasks and understanding directions and required constant redirection. The patient scored below cut-off on both the Autism Diagnostic Observation Schedule-Second Edition Module 1, and the Autism Diagnostic Interview-Revised. Endorsed items on the Social Responsiveness Scale-Second Edition did not indicate clinically significant deficiencies in reciprocal social behaviour. Endorsed items on the ASEBA Child Behaviour Checklist resulted in slightly elevated scores in attention problems. No other areas of concern were identified.

CNV comparison

To compare the patient’s genetic result with known cases, we searched for similar-sized deletions and duplications at 16p13.13, overlapping the same genes in several databases: Database of Genomic Variants (DGV),⁶ gnomAD structural variant (SV),⁷ ClinVar⁸ and DECIPHER.⁹ These databases catalogue CNVs identified through clinical and research testing, as well as in general population control data. As our patient presented with some autistic traits, we also searched for overlapping CNVs in whole genome sequencing (WGS) data from the Autism Speaks MSSNG project^{10 11} and the Simons Simplex Collection (SSC).¹² Specifically, the hg38 coordinates were checked for overlap with rare CNVs (<1% frequency in MSSNG parents and in 1000 Genomes Project population controls) in MSSNG and SSC. Furthermore, for each gene included in the 16p13.13 deleted region, we determined the number of rare loss of function (LoF) (stop gain, frameshift or splice site-disrupting) single nucleotide variants and insertions/deletions (indels) in individuals with autism from the MSSNG (n=5123), SSC (n=2419) and SPARK (n=21 900)¹³ cohorts. We also determined the number of rare deletions and duplications overlapping coding exons of each gene (or any exon for the ncRNA BCAR4) in individuals with autism from MSSNG and SSC.

There are no similar-sized deletions or duplications spanning the 16p13.13 region reported in our patient in DGV or ClinVar or gnomAD (structural variants callset). In DECIPHER there are three deletion cases and four duplication cases overlapping some of the same genes. Deletions and duplications that overlapped with our patient, but included additional genes were not counted. Among the seven CNVs, consent was obtained for four (figure 1). Case ID#249 701 is a deletion smaller (674.15 kb) than our patient’s deletion, which overlaps the GSPT1, NPIPB2, TNFRSF17 and SNX29 genes. The phenotype for this case included intellectual disability and facial dysmorphisms. Of note, the patient had a diagnosis of macrocephaly-cutis marmorata, which is a congenital condition of unknown aetiology associated with connective tissue abnormalities and in some patients, intellectual disability.¹⁴ Unlike our patient, this case had macrocephaly. Case ID#252656 is a 656.4 kb duplication with breakpoints within the TXNDC11 and SNX29 genes with a history of general psychomotor delays and behavioural problems. Unlike our patient, this case did not have any dysmorphisms. The last two cases are similar duplications of 206.97 kb (ID#331295) and 217.37 kb (ID#277824) with breakpoints within the GSPT1 and SNX29 genes. Case ID#331 295 is noted to have a specific learning disability and case ID#277 824 has microcephaly.

Figure 1.

Similar-sized deletions or duplications spanning the 16p13.13 region reported in our patient in DECIPHER.

Our analysis of WGS data from MSSNG and SSC found no evidence of similar-sized deletions or duplications in this region. For each gene included in the patient’s deleted region, the number of rare loss-of-function variants, deletions and duplications in individuals with autism spectrum disorder (ASD) from the MSSNG, SSC and SPARK cohorts are listed in table 2. However, none of these genes were enriched in autism cases compared with controls or identified as autism related candidate genes based on the transmission and de novo association analysis.¹⁵ Additionally, the 16p13.13 deletion does not overlap any of the pathogenic structural variants reported in the recent MSSNG analysis,¹⁵ and none of the 10 genes have been implicated in autism in previous large-scale gene discovery studies.^{10 16}

Table 2.

Counts of rare (<1% population frequency) loss-of-function (LoF) SNVs/indels and copy number variants (deletions or duplications overlapping coding exons, or any exons for the ncRNA BCAR4) in individuals with autism spectrum disorder (cases) from the MSSNG, Simons Simplex Collection and SPARK cohorts

Gene	pLI	/e	LoF SNVs/indels	Deletions	Duplications
LITAF	0.03	0.5	1	0	1
SNN	0.14	0.51	1	0	1
TXNDC11	0	0.57	42	0	1
ZC3H7A	1	0.04	7	0	0
BCAR4	N/A*	N/A*	N/A	12	2
RSL1D1	0	0.63	5	0	0
GSPT1	1	0.03	6	0	1
NPIPB2	0	1.1	7	0	0
TNFRSF17	0	1.27	6	0	1
SNX29	0	0.52	11	4	2

For variants in cases for which parental sequencing data were available, none of the variants were found to be de novo.

*Probability of being loss-of-function intolerant (pLI), observed/expected number of loss-of-function variants (o/e) and LoF counts are not applicable for this gene because it is a non-coding RNA.

Differential diagnosis

Neurodevelopmental delay, microcephaly, congenital heart abnormalities and orofacial cleft are each frequent findings in suspected inherited disease cases referred for arrays and/or diagnostic sequencing. However, the co-occurrence of these congenital anomalies is less frequent. The differential diagnoses for the co-occurrence of these findings include prenatal infection (eg, cytomegalovirus, toxoplasmosis), exposure to drugs of abuse (eg, alcohol) or medications (eg, anticonvulsant drugs).¹⁷

Outcome and follow-up

Overall, the patient met diagnostic criteria for global developmental delay. This is supported by his language delay and earlier motor delays. Formal cognitive testing suggests intellectual disability but was hampered by his limited attention span; as such, we were hesitant to give this diagnosis. Although the patient had a short attention span, his level of attention was consistent with his developmental level, indicating a mental age equivalent to approximately 2.5 years; therefore, he did not meet criteria for attention deficit hyperactivity disorder. Despite the mild autistic behaviours, the patient also presented with many social strengths. Based on the current assessment, autism spectrum disorder was thought to be unlikely but could not be definitively ruled out. There were no symptoms indicating a disorder in the domain of mood, anxiety or psychosis. A repeated reassessment in one or 2 years was recommended to re-evaluate these clinical impressions.

Discussion

We report on a patient with developmental delay and multiple congenital anomalies with a de novo deletion in 16p13.13, and no other pathogenic variants on genomic analysis. The ten Refseq genes in the deletion region have not been associated with any neurodevelopmental or medical conditions to date. To assess whether any of the genes are likely candidates, we determined whether the gene has a high probability of being loss-of-function intolerant (pLI score close to 1) and whether it has a low observed/expected number of loss-of-function variants (o/e<0.35) (table 2). Two out of the ten genes had a pLI score of 1 and an o/e ratio of less than 0.35: ZC3H7A and GSPT1. These constraint parameters suggest that disruptions to these genes could impact the expression or function of the resulting proteins. The ZC3H7A gene encodes a protein that enables miRNA binding activity (OMIM* 619819), while the GSPT1 gene encodes a protein that is involved in regulation of translational activity (OMIM* 619819). Comparing our patient’s 16p13.13 deletion to CNVs in the DECIPHER database, we see that two of the duplications’ breakpoints disrupt the GSPT1 gene (ID#331 295 and ID#277824), possibly resulting in a loss-of-function effect. These two cases have neurodevelopmental phenotypes including learning disability and microcephaly, respectively.

Although the constraint scores for SNX29 do not support it being haploinsufficient (pLI=0; o/e=0.52), we took a closer look at this gene given that our patient and all four CNVs in DECIPHER intercepted this gene. SNX29 encodes sorting nexin 29, a member of the sorting nexin family, which is thought to play a role in membrane trafficking and protein sorting.¹⁸ Interestingly, dysfunction of sorting nexins has been implicated in neurodegenerative diseases such as Alzheimer’s disease and Down’s syndrome.¹⁹ Recent studies have implicated the role of common variants in the SNX29 gene with cognitive functioning and psychiatric disorders. A small-scale genome-wide association study found that a single nucleotide polymorphism (SNP) at SNX29 was associated with autism.²⁰ Additional genome-wide association studies found that SNPs in SNX29 was associated with intelligence.^21–23 Chen et al found that certain SNPs in the SNX29 gene were associated with depression and bipolar disorder.²⁴ Of note, our patient has a rare deletion of most of the SNX29 gene rather than a SNP; therefore, should this gene play a role in cognitive, behavioural and/or emotional functioning, a deletion of the entire gene may be significant.

To date, two studies have reported on clinical phenotype in carriers of CNVs in 16p13.13.^{25 26} Carlo and colleagues²⁵ presented a case of a young boy with a duplication in chromosome 16p13 and chromosome 9q34 with autism spectrum disorder. As this young boy carried two different duplications, it is unclear which duplication, if any, was associated with autism. Interestingly, the duplication at 16p13 for this patient is larger and fully encompasses the deleted region in our patient and likely all ten Refseq genes in hg38. The duplication is noted to partially disrupt the CLEC16A gene upstream of the genomic region in our deletion case, but shares a breakpoint within the SNX29 gene, which may correlate with LoF effects, similar to those predicted from deletion (like in our patient). Digilio et al ²⁶ presented a case of a child with multiple congenital anomalies and a duplication on 16p13.3p13.13; however, the exact boundaries of the duplication are not specified. In this case the duplication includes a region at 16p13.3, which is considered part of the critical region for 16p13.3 deletion syndrome.²⁶ In both published cases, the involvement of other CNVs is more likely contributing to the phenotype than the shared genomic content with our patient.

The additional seven reports of 16p13.13 CNVs in DECIPHER highlight the uniqueness and rarity of most CNVs. Along with absent or limited phenotypic data, which is particularly true for NDDs, attributing clinical significance is especially challenging. Detailed genomic and phenotypic reports of patients is required to amount enough evidence to reclassify VUSs. The evaluation of CNVs by clinical labs is guided by The American College of Medical Genetics and Genomics, and Clinical Genomic Resource groups’ quantitative, evidence-based scoring system. Individual case evidence can contribute to the score (section 4 of scoring standard), and a higher score reflects higher likelihood of pathogenicity.³ Cases with phenotypic consistency and specificity score more points. Patients with NDDs, which can present quite broadly, score lower and require additional evidence to reach the pathogenicity threshold.

It is worth nothing that in our study both chromosomal microarray analysis and WES were performed to identify variants potentially involved in our patient’s phenotype. It is possible that additional candidate variants could have been identified had WGS been performed. However, a recent comparison among different genetic tests in ASD and fetal structural anomalies suggests that WGS currently provides only a modest gain in diagnostic yield when compared with the combination of both chromosomal microarray analysis and WES, although the size of this difference may be phenotype-dependent.²⁷

In conclusion, we report for the first time a de novo deletion in the 16p13.13 region in a child with developmental delay. The link between a deletion in this region and developmental delay remains to be elucidated. Based on gene constraint scores and overlapping CNVs in available databases, the ZC3H7A, GSPT1 and SNX29 genes may be possible candidate genes for neurodevelopment.

Learning points.

• It is important for clinicians to report on patients who present with rare genetic variants and neurodevelopmental disorders (NDDs), as this contributes to evolving evidence towards helping elucidate the possible role of these variants in the aetiology of NDD.

• Researching the clinical relevance of genes included within the copy number variant region may provide some insight into links between phenotype and the rare genetic variant.

• ZC3H7A, GSPT1 and SNX29 may be candidate genes for neurodevelopment and warrant further exploration.

Ethics statements

Patient consent for publication

Consent obtained from parent(s)/guardian(s).