Molecular Case Study Research Report

Title: KBG syndrome involving a single base insertion in ANKRD11

Authors: Robert Kleyner1*, Janet Malcolmson1,2*, David Tegay1, Kenneth Ward3, Justine

Coppinger3, Annette Maughan4, Glenn Maughan5, Lesa Nelson3, Kai Wang6,7,8, Reid

Robison8, Gholson J. Lyon1,8,#

1Stanley Institute for Cognitive Genomics, One Bungtown Road, Cold Spring Harbor

Laboratory, Cold Spring Harbor, NY, USA; 2Genetic Counseling Graduate Program,

Long Island University (LIU), 720 Northern Boulevard, Brookville, NY 11548; 3Affiliated

Genetics, Inc. 2749 East Parleys Way Suite 100 Salt Lake City, UT 84109; 4Epilepsy

Association of Utah, 8539 So. Redwood Rd., West Jordan, UT. 84088; 5KBG Syndrome

Foundation, 8539 So. Redwood Rd., West Jordan, UT. 84088; 6Zilkha Neurogenetic

Institute, University of Southern California, Los Angeles, CA 90089, US; 7Department of

Psychiatry & Behavioral Sciences, Keck School of Medicine, University of Southern

California, Los Angeles, CA 90033, USA; 8Utah Foundation for Biomedical Research,

Salt Lake City, UT

*These authors contributed equally to this work.

#To whom correspondence should be addressed:

Email: gholsonjlyon@gmail.com

ABSTRACT KBG syndrome is a rare autosomal dominant genetic condition characterized by

neurological involvement, macrodontia and distinct facial, hand and skeletal features.

Over 70 cases have been reported; however it is likely that KBG syndrome is

underdiagnosed due to lack of comprehensive characterization of the heterogeneous

phenotypic features. We describe the clinical manifestations in a male currently at 13

years of age, who exhibited symptoms including epilepsy, severe developmental delay,

distinct facial features and hand anomalies, without positive genetic diagnosis.

Subsequent exome sequencing identified a novel de novo heterozygous single base pair

insertion (c.6015dupA) in ANKRD11, which was validated by Sanger sequencing. This

insertion is predicted to lead to a premature stop codon and loss of function in

ANKRD11, thereby implicating it as contributing to the proband’s symptoms and yielding

a molecular diagnosis of KBG syndrome for the case.

INTRODUCTION Whole exome sequencing (WES) is a method that sequences only regions of the

genome that code for proteins and is more comprehensive than other testing methods

such as microarray and CNV analyses. WES is useful for detecting disease-contributing

variants in genes associated with rare genetic syndromes. Here we report our efforts in

phenotypic characterization and molecular diagnosis of a previously undiagnosed

pediatric patient. This case demonstrates the utility of whole exome sequencing for

finding rare disease-contributing mutations that can then lead to the diagnosis of rare,

previously unrecognized syndromes. In this case, we report in a single family the

identification of a de novo mutation in ANKRD11, which led to the recognition of KBG

syndrome in the sequenced proband.

RESULTS Clinical presentation and family history

The proband was born to a non-consanguineous couple, who had an

unremarkable pregnancy history; however, at birth a large fontanel was reported.

Parents and siblings were healthy and no significant family history was reported (Figure 1). The proband had his first epileptic episode at three years of age. After this episode,

he lost all speech, began exhibiting autistic behavior, and also started to have frequent

generalized tonic-clonic seizures. Over time, tonic, atonic, mild clonic, complex partial,

myoclonic and gelastic seizures were reported in the proband. Other developmental

skills, including throwing a ball, responding to his name, feeding himself with utensils

and self-care skills were lost by 4-years of age. No significant conductive hearing loss,

heart abnormalities or delayed bone age were found in the proband at that age.

The proband was evaluated (by G.J.L.) at eleven years of age. He presented

with several neurological and craniofacial abnormalities including epilepsy,

ventriculomegaly, relative macrocephaly, prominent forehead, low hairline, thick

eyebrows, wide-set eyes (Figure 2), macrodontia of upper central incisors, and full lips

(Figure 3). Hand and foot abnormalities included clinodactyly of the fifth digit, bilateral

single transverse palmar creases, brachydactyly (Figure 4) and flat feet. He also had a

diagnosis of cerebral folate deficiency due to the presence of folate receptor

autoantibodies.

Genomic Analyses Blood and saliva samples from the proband as well as his parents and siblings

were used as samples to be sequenced. These samples were sent to Affiliated Genetics

in Salt Lake City, Utah, where genomic DNA was extracted and exons sequenced using

the Life Technologies Ampliseq Exome RDY kit and the Life Technologies Proton

sequencing system (see Methods). These targeted regions were sequenced using the

Ion Proton sequencing system using Ion Hi-Q Chemistry with 200 bp reads. The DNA

sequencing data was compared to the UCSC hg19 reference sequence using several

methods of analysis (see Methods). These analyses included in-house protocols and

several commercial software packages including Tute Genomics, Omicia Opal, and

Cartagenia v4.1, along with the use of an OTG-SNP Caller pipeline (see Methods). The

various analyses helped to provide a more comprehensive and in-depth approach to the

data1; 2.

As one example, for the OTG-SNP Caller pipeline, for each individual, the final

VCF file contained 20,000 to 25,000 variants, of which around 300-400 variants were

found to be autosomal recessive, i.e. heterozygous in both parents, and homozygous

only in the proband. However, over a thousand variants were recognized as de novo,

which is notably above the expected number of de novo mutations found in WES3; 4.

Therefore, even with an optimized variant calling pipeline, there were still a significant

number of false positives called.

Autosomal variants were examined, and there was no evidence found to support

any of them as possible contributing mutations. These variants are provided in

supplementary files (as described in Methods). For the de novo mutations, a single base

insertion of adenine (A) at position 6015 in exon 10 of ANKRD11 (c.6015dupA ,

p.Gly2006Argfs*26) (Figure 5) was identified as the most relevant mutation. All

phenotypic analysis software, including Phenolyzer5, wAnnovar6, and PhenIX7 indicated

that a heterozygous frame-shift mutation in ANKRD11, or the Ankyrin Repeat Domain 11

gene, might be a contributing factor in this individual’s disease The presence of the

mutation was confirmed using Sanger sequencing (Figure 5).This mutation has a CADD

score of 32, and is considered to be ‘Deleterious’ by SIFT with a confidence score of

0.858, and is therefore predicted to have a severe effect on protein structure. Although

insertions have not yet been reported, other mutations in this gene have been previously

identified as contributing to KBG syndrome, a rare disease that affects around 60-70

people worldwide8-11.

DISCUSSION Many syndromes affecting neurological development present with heterogeneous

and non-distinct phenotypes12 and therefore remain undiagnosed or are misdiagnosed.

The combination of whole exome sequencing combined with detailed and standardized

phenotypic documentation is a powerful method to achieve accurate diagnosis.

KBG syndrome (OMIM #148050) is a rare, but increasingly recognized,

autosomal dominant genetic condition. It was first described in 1975, and is

characterized by craniofacial features, hand abnormalities, macrodontia, and

neurological involvement including developmental delay and epilepsy13. The syndrome’s

name was derived from the last names of the first three families found to have this

syndrome8. Over 70 cases have been reported, however it is likely that KBG syndrome

is underdiagnosed due to the fact that dysmorphic features may be subtle and cognitive

delay can vary from mild to moderate9. Skjei et al. suggested that a clinical diagnosis of

KBG syndrome can be made if the individual meets four out of the following eight major

criteria: characteristic facial features, macrodontia of upper central incisors, short

stature, delayed bone age, neurological involvement, hand abnormalities, costovertebral

anomalies and the presence of a family member affected with the syndrome14. Facial

features include hypertelorism, short nose with broad base and bulbous nasal tip, and

broad bushy eyebrows15. Although the shape of the face is often described as being

round, it has been noted that the shape evolves as affected children develop14. Hand

abnormalities typically include brachydactyly, clinodactyly of the fifth digit, small hands

and nail anomalies14. Skeletal anomalies frequently involve the pelvis, thorax, limbs and

skull with abnormal curvature of the spine, including kyphosis and scoliosis, being

reported in some cases14. Minor features of KBG syndrome include cutaneous

syndactyly, conductive hearing loss, palatal abnormalities, cryptorchidism, webbed/short

neck, strabismus and congenital heart defects8. There is some phenotypic overlap with

Cornelia de Lange syndrome (CdLS)16; 17.

Individuals with KBG syndrome have been found to have heterozygous

mutations leading to haploinsufficiency of the ankyrin repeat domain 11 (ANKRD11)

gene or a 16q24 microdeletion that encompasses ANKRD1115. When mutations in the

gene affect the highly conserved region of the domain for transcriptional regression, they

are predicted to lead to premature stop codons which could result in haploinsufficiency

and when a 16q24 microdeletion is present the haploinsufficency of ANKRD11 is

confirmed to be the pathogenic mechanism of KBG syndrome9. Sporadic and familial

cases of KBG syndrome have been reported, with familial cases following an autosomal

dominant inheritance pattern14; 15; 18-25. ANKRD11 (ankyrin repeat domain containing

protein 11) is a chromatin regulator that controls histone acetylation and gene

expression during neural development26. There are two functional domains that act as

transcriptional repressors and one domain that functions as a transcriptional promoter27.

The majority of reported mutations in KBG syndrome result in a truncated protein that

affects a domain for transcriptional repression9. ANKRD11 interacts with the p160

coactivator and the nuclear receptor complex and it functions to inhibit ligand-dependent

transcriptional activation by recruiting histone deacytelases (HDACs)11. Additionally,

ANKRD11 was also found to play a role in enhancing the transcriptional activity of p5328.

Homozygosity for a missense mutation in ANKRD11 is embryonic lethal in mice,

whereas the heterozygous mice have an osteopenia-like phenotype and craniofacial

abnormalities29.

This sporadic case of KBG syndrome demonstrates the importance of ongoing

investigations of rare conditions. Each case reported in the literature will help to

delineate the phenotype so that we may better identify cases in the future and determine

appropriate recommendations for clinical management. Current recommendations for

management of KBG syndrome include hearing tests, ophthalmologic assessments,

echocardiography, an EEG, orthodontic evaluation and skeletal investigation with special

attention to spine curvatures and limb asymmetry 8; 15.

Seizures are an often reported feature of KBG syndrome, found in up to 28% of

patients14. Typically, seizures with KBG syndrome are characterized as tonic-clonic,

responsive to treatment, transient and benign8; 15. However, in this proband, seizures

have been persistent, mixed generalized and partial, treatment resistant and temporally

associated with developmental regression at seizure onset. This may be due to the

combined effects of Cerebral Folate Deficiency (CFD) and KBG syndrome, both present

in this proband. Cerebral folate deficiency (CFD) is caused by the interruption of folate

transport across the blood brain barrier due to folate receptor autoantibodies (FRAs) 30.

CFD often presents with neurological findings including seizures, spastic paraplegia,

cerebellar ataxia, dyskinesia, and developmental regression, and recent cases have

described autism spectrum disorder30. Patients with CFD have low 5-

methyltetrahydrofolate (5MTHF) levels, but when treated with oral folinic acid these

levels can normalize. The proband receives folinic acid (Leucovorin – 50 mg/day), which

has helped. He has also been treated with various anti-eplileptic drugs (AEDs) including

Lamotrigine (Lamicatal - 400mg/day), which has been the most effective anti-epileptic

drug to control his seizures, and recently 1.5 mL twice daily of 100 mg/mL Epidiolex

(cannabidiol) has also been reported by the family to reduce his frequency of seizures31;

32. There are no other cases reported of combined CFD and KBG syndrome but we

hypothesize that this has contributed to his atypical history of developmental regression

with persistent and resistant epilepsy in addition to his non-verbal intellectual disability

and autistic behaviors.

Although there is currently no cure for this disease, identifying individuals with

this syndrome will not only provide a method to track the outcome of these individuals,

but also helps provide support. For instance, the family of the affected individual created

a social media group, and a KBG nonprofit foundation was specifically developed to

connect families with children affected with KBG syndrome. Tracking these individuals

might also identify information regarding the progression of the disease, and any shared,

or individual, phenotypes that might be relevant to study in the future.

METHODS

DNA ISOLATION AND SEQUENCING: Genomic DNA was extracted using standard methods (Pure Gene, Qiagen,

Valencia, CA). The Life Technologies Ampliseq Exome RDY kit (Thermo Fisher,

Carlsbad, CA) was used to target the exon regions. 97% of CDCs with 5bp exon

padding were amplified using 294,000 primer pairs. These products were sequenced

using the Life Technologies Proton sequencing system with 200 bp reads using a P1V3

chip.

VARIANT CALLING: The DNA sequence was aligned to the UCSC hg19 reference sequence and

variants were called using the Torrent Suite software and the Torrent Variant

caller. Only exonic variants and variants at the intron-exon boundary (1 or 2 nucleotides

into the intron and 1 nucleotide into the exon) were reviewed. For each variant

considered, depth of coverage was >10X and the quality score was >30. Ethnicity and

variant frequency were considered during analysis. Analysis of the variants was

conducted by two independent reviews using in-house protocols and two commercial

software packages, Tute Genomics and Cartagenia v4.1. Pathogenic variants were

confirmed by Sanger sequencing. ACMG reporting criteria were used to evaluate

variants33.

In additional analyses, binary alignment (BAM) files from the Ion

Torrent Personal Genome Machine (PGM) platform were converted to FASTQ

files. Variants were called using the OTG-SNP Caller pipeline, which has been reported

to map a higher proportion of sequencing reads to the reference genome in comparison

to other methods, and result in lower error rates when analyzing sequences coming from

the PGM platform 34. Unlike other sequencing software and pipelines such as the

Genome-Analysis-Toolkit (GATK) and Freebayes 35; 36, OTG-SNP Caller is specifically

designed to take into account errors associated with PGM data, such as errors

around homopolymers, thus increasing overall accuracy. Variants were aligned to the

GRCh37 assembly, as several downstream analysis tools do not yet support the new

GRCh38 assembly. A variant call format (VCF) file containing information about each

mutation was then output 37.

VARIANT SELECTION AND PRIORITIZATION The resulting VCF file for each individual in the family (see Supplementary File 2)

was converted into ANNOVAR files (avinput) using ANNOVAR. Avinput provides

information regarding chromosome number, start position, end position, reference

nucleotide, alternate nucleotide, and quality scores for each variant 38. All avinput files

for a particular family were then loaded into a Python program (see Supplementary File

3), which performs set intersections using DataFrame functions from the Pandas library,

and set functions using the Numpy library to identify de novo and autosomal recessive

variants 39; 40. Autosomal recessive variants were identified by isolating homozygous

variants in the affected child, intersecting these variants with variants that were

heterozygous in both parents, and subtracting variants that were homozygous in the

siblings. De novo variants were identified by subtracting variants found in the parents

and siblings from variants found in the proband.

The columns examined included the chromosome number, start point, end point,

and zygosity of each called variant. The resulting avinput files were then output as BED

files, which contain columns providing chromosome number, start point, and end point of

the mutation. This process ensured that the resulting BED files contained all autosomal

recessive and de novo variants that could be determined from the VCF.

Using the GATK SelectVariants tool, these two BED files were intersected with

the original VCF file two separate times, creating two VCF files, one containing only

autosomal recessive variants, and one containing only de novo variants.

Both these VCF files were then annotated with the Variant Effect Predictor (VEP)

software 41, which provided additional information about the variants. This annotated

VCF file was then used with 42, which is a powerful, yet flexible network that allows for

organization, sorting, and filtering of variants based on VEP and additional annotations.

Variants were then filtered using rarity, deleteriousness, and read quality as filter criteria.

Rarity was determined using the Exome Aggregation Consortum (ExAC)

database, which contains population allele frequencies for exonic variants gathered from

60706 unrelated individuals with no history of severe pediatric disease (EXAC

2015). Rare variants were considered to be variants not found in EXAC.

Deleteriousness was determined by Combined Annotation Dependent Depletion (CADD)

scores, which encompasses 63 annotations to determine a variant’s deleteriousness.

Autosomal Recessive=[(Mhet ∩Fhet) ∩ Phom] - SIBhom]

de novo = Pall – Mall – Fall - SIVall

Where M refers to the mother’s variants, F refers to the father’s variants, P refers to

the proband’s variants, and SIB refers to the sibling’s variants. The subscript het refers to heterozygous variants, hom refers to homozygous variants, and all refers to

all variants.

CADD scores are based off PHRED quality scores; therefore a minimum CADD score of

>= 20 or zero (as CADD was not calculated for indels), corresponding to the top 1% most

deleterious variants, was selected as a cutoff 43. While the resulting quality scores from

the OTG-SNP Caller pipeline did not correspond to the standard PHRED quality score, a

minimum cutoff score of >=120 was decided after comparing variant calls with their

corresponding binary sequence alignment (BAM) files. Chromosome number, start point,

and end point columns of variants that met these three requirements were

obtained using GEMINI, and the output was saved as a BED file (see Supplementary

File 5). The GEMINI query used was: gemini query -q "select chrom, start, end from

variants where qual>=120 AND (cadd_scaled>20 OR cadd_scaled is NULL) AND

in_exac=0 order by chrom, start" denovo.db This BED file, along with human phenotype

ontology (HPO) numbers corresponding to the proband’s phenotype were used in

conjunction with Phenolyzer 5, which is desiged to determine and prioritize which

mutations contribute most to the phenotype by comparing the provided HPO numbers to

the phenotypes attributed to the gene in which the proband’s mutation is located. A VCF

file containing the same variants as the BED file used with Phenolyzer was then input

into similar programs such as wANNOVAR and PhenIX in order to utilize several

sources of analysis 6; 7. These same VCF files were input into the Omicia Opal system,

with similar results (see Supplementary File 4)44-46.

CONFIRMATION OF VARIANTS Once a possible disease-contributory mutation was identified, its location was

then input into GoldenHelix GenomeBrowser, which displayed read information from the

BAM files corresponding to each family. All variants of interest were also researched,

and ruled out as major contributing mutations due to no association with a relevant

phenotype. The genic locations of each variant were identified using GEMINI42. Initially,

the known functions, phenotypes, and diseases associated with each gene, would be

researched using the GeneCards online database47-49, which contains information

compiled from over 100 sources. These results were then confirmed by researching the

gene in other databases, such as NCBI, PubMed, and OMIM 50; 51. These findings were

also compared to the output of the phenotype analysis software. No additional

contributing mutations were identified in this individual. The GEMINI query selected no

autosomal recessive mutations of interest, and sixteen rare de novo mutations as

mutations of interest. The Phenolyzer, wAnnovar, and PhenIX outputs all identified a

heterozygous missense mutation in ANKRD11 as the most likely contributing

mutation. Sanger sequencing confirmed this variant.

Additional Information Ethics Statement

Research was carried out in compliance with the Helsinki Declaration. The family

was recruited to this study at the Utah Foundation for Biomedical Research (UFBR)

where extensive clinical evaluation was performed. Written consent was obtained for

phenotyping, use of facial photography, and whole-exome sequencing through Protocol

#100 at the Utah Foundation for Biomedical Research, approved by the Independent

Investigational Review Board, Inc.

Acknowledgments

G.J.L. is supported by funds from the Stanley Institute for Cognitive Genomics at

Cold Spring Harbor Laboratory. K.W. is supported by NIH grant HG006465. The authors

would like to thank the ExomeAggregation Consortium and the groups that provided

exome variant data for comparison. A full list of contributing groups can be found at

http://exac.broadinstitute.org/about. G.J.L. would like to thank Omicia for access to the

Omicia Opal system. The authors acknowledge Jason O’Rawe for bioinformatics support

and comments on the manuscript.

Conflict of interest G.J.L serves on advisory boards for GenePeeks, Inc. and Omicia, Inc., and is a

consultant to Good Start Genetics. K.W. is board member and shareholder of Tute

Genomics, Inc. R.R. is employee, chief executive officer and shareholder of Tute

Genomics, Inc.

SUPPLEMENTARY FILES

File 1: Videos in a zip file

VIDEO DESCRIPTIONS

KBG_HDV_0105.mp4 - Video of parents discussing the proband’s abnormalities with

Dr. Gholson Lyon.

0:00-0:13 – The proband had an abnormally large fontanelle, which resolved without

treatment.

0:13-0:27 – The proband does not appear to have a sacral dimple.

0:27-0:35 – Other than the presence of flat arches, there are no obvious signs of foot

abnormalities.

0:35-1:05 – The proband does not look like his other siblings, although there was a

resemblance between him and his sister when they were the same age.

1:05-1:12 - Features of proband’s face can be seen, including bushy eyebrows, broad

nasal tip, short philtrum, full lips and cupid bow of upper lip.

KBG_HDV_0107.mp4:

Video of proband’s behavior. Proband is non-verbal, and hyperactive. He repetitively

spins his toy. While playing, he gets up from his chair, walks a few steps, stomps his

feet, and sits back down.

KBG_HDV_0147.mp4 - Additional interview in August 2015, after the parents received

the proband’s diagnosis of KBG syndrome.

00:00 –00:25 The proband develops new mannerisms every four to six months, the most

recent being short, hard breaths through the nose and head turning.

00:25 – 1:14 The proband has had a substantial decrease in the number of seizures

after starting an Epidiolex (cannabidiol) treatment (70-80% decrease as described by the

parents). The frequency of seizures increased after the proband fell and fractured his

jaw.

KBG_HDV_0148.mp4:

The mother describes the proband’s macrodontia. Although the mother and several

siblings have large teeth, the severe macrodentia in the proband does not appear in any

other member of the family.

KBG_HDV_0149.mp4 - The proband’s features are compared to other characteristics

usually found in other KBG patients. Variation might be a result of an insertion, and not a

deletion in ANKRD11.

0:00-0:12 Unlike most KBG patients, the proband has full lips.

0:12–0:32 Like most KBG patients, the proband has curved pinkies (diagnosed as

clinodactyly), which are often found in KBG patients.

KBG_HDV_0150.mp4:

Although the proband has relatively short toes, this trait may have been inherited from

the father. The proband also has curved toenails, which commonly appear in autistic

children.

File 2. VCF FILES in a zip folder

proband.vcf

Unaffected_brother.vcf

Unaffected_father.vcf

Unaffected_mother.vcf

Unaffected_sister1.vcf

Unaffected_sister2.vcf

File 3. Program Scripts and Outputs used in the analysis, uploaded as a zip file,

containing the below information.

Python Folder:

iPython_Notebook.html – HTML version of the annotated iPython Notebook used to create the program, which can be viewed by opening the file in a web browser such as Safari or Firefox.

iPython_Notebook.pdf – PDF version of the annotated iPython Notebook.

Python Program To Find De Novo and AutRec.py – Program used to identify de novo and autosomal recessive variants.

Files in ‘Inputs’ Folder – Files input into the Python program.

Files in ‘Outputs’ Folder – Files output from the Python program.

Variant Calling Folder:

Sample_Pipeline.sh – Annotated script used to analyze raw read (fastq) files to

VCF files.

Final_VCF Prep.sh – Script used to create a single VCF file from the analysis

pipeline’s outputs.

File 4. Excel SPREADSHEETS ILLUSTRATING VAAST RESULTS AMONG THE

VARIOUS QUADS in a zip folder, run using the Omicia Opal software.

File 5. GEMINI_Analysis.zip

de_novo_interest.bed – BED file including variants which were selected using the

GEMINI query discussed in the paper (gemini query -q "select chrom, start, end

from variants where qual>=120 AND (cadd_scaled>20 OR cadd_scaled is NULL)

AND in_exac=0 order by chrom, start" denovo.db).

de_novo_interest.vcf – BED file including variants which were selected using the

GEMINI query discussed in the paper (gemini query -q "select chrom, start, end

from variants where qual>=120 AND (cadd_scaled>20 OR cadd_scaled is NULL)

AND in_exac=0 order by chrom, start" denovo.db).

Denovo_Autrec_VCF

autrec.vcf – VCF including all autosomal recessive variants.

Denovo.vcf – VCF including all de novo variants.

Gemini_Upload_Files - Includes a script used to ‘upload’ files to GEMINI,

through which a database of variants. Also includes the files created when

variant databases were created using GEMINI.

Variant Database:

autrec.db – GEMINI database of autosomal recessive variants.

denovo.db – GEMINI database of de novo variants.

File 6. ANKRD11_Mutations_Web_Viewable.svg

A more detailed version of the mutaions can be viewed by opening this file in a web

browser, and holding the cursor over the domain and the mutations.

FIGURES

Figure 1: Pedigree: II-1, the affected proband (age 13), is the son of an unaffected, non-consanguineous couple. The proband has two younger unaffected sisters (10-years-old and 4-years-old) and one younger unaffected brother (2-years-old).

Figure 2: Pictures of phenotype of proband throughout childhood (a) 6-months-old, (b) 4-years-old, (c) 9-years-old and (d) 13-years old. Facial characteristics include rounded face, bushy eyebrows, broad nasal tip, short philtrum, thick lips and prognathism.

Figure 3: Macrodontia present in proband at age 12. (a) Right upper central incisor is large and left upper central incisor has been chipped. (b) Lower teeth seen.

Figure 4: Hand anomalies include bilateral clinodactyly of the 5th finger, brachydactyly and bilateral single transverse palmar creases.

Figure 5: GenomeBrowse output for the proband. The orange ‘T’ in the proband’s

nucleotide indicates a heterozygous thymine insertion in chromosome 16, position

89,346,935. This insertion appears to be supported by >20 reads, and is likely a true-

positive mutation. None of the other family members appear to have this mutation,

indicating that it is likely de novo. The small orange bar on the father and sister indicates

one read supporting an adenine insertion, and a thiamine insertion respectively, and are

considered to be sequencing errors. The red box on the cytoband shows the mutation’s

location on chromosome 16 (a). Sanger sequencing confirms the thymine insertion (b).

a

b

Figure 6: Venn diagram showing the eight major criteria suggested for KBG syndrome

and the five characteristics present in the proband. For a clinical diagnosis of KBG

syndrome it has been suggested that four of the eight criteria be present.

!! !!!  Large!Fontanel!At!

Birth!!  Ventriculomegaly!!  Cerebral!Folate!

Deficiency!!

Proband( Both( Major(Criteria!

!  Delayed!Bone!Age!

!  Costovertebral!Anomalies!

!  First<Degree!Rela>ve!With!Syndrome!

!

!  Neurological!Involvement!

!  Macrodon>a!!  Characteris>c!

Facial!Appearance!

!  Short!Stature!!  Hand!Findings!

Figure 7: Some of the published loss of function (LOF) mutations in ANKRD11. Mutations represented by the black circles were those documented in individuals with KBG syndrome. The mutations represented by the gray circles are those reported in ExAC. The mutation represented by the large black circle was the one found in the proband. Only the LOF mutations in ExAC were plotted. The height of each mutation varies only for the ease of showing the mutations in the figure. The image was created using the ‘lollipops’ tool (https://github.com/pbnjay/lollipops), which retrieved domains from Pfam. Unless otherwise noted in parentheses, all mutations were found in only one individual. None of the exact same mutations have been found in more than one family. A more detailed version of this figure can be viewed by opening ANKRD11_Mutations_Web_Viewable.svg in a web browser, and holding the cursor over the domain and the mutations.

Table 1. Summary of the Clinical Features in this proband

Features (Human Phenotype Ontology Nos.) Proband

FACIAL DYSMORPHISM

Large fontanelle (HP:0000239) +

Rounded Face (HP:0000311) +

Bushy Eyebrows (HP:0000574) +

Broad Nasal Tip (HP:0000455) +

Short Philtrum (HP:0000322) +

Full/Thick Lips (HP:0012471) +

Cupid Bow Upper Lip (HP:0002263) +

Macrodontia of Upper Central Incisors (HP:0000675) +

Prognathism (HP:0000303) +

DEVELOPMENTAL/INTELLECTUAL DISABILITY

Intellectual Disability (HP:0001249) +

Developmental Regression +

Developmental Delay Prior to Regression +

Absent Speech (HP:0001344) +

SKELETAL

Clinodactyly of the 5th finger (HP:0004209) +

Brachydactyly (HP:0009803) +

Bilateral single transverse palmar creases (HP:0007598) +

Short toes (HP:0001831) +

Pes planus (HP:0001763) +

NEUROLOGICAL

Epilepsy Mixed (T/C, Atonic, Complex, Partial, Tonic, Gelastic)

(HP:0001250)

+

GROWTH

Currently short stature (10th percentile) (HP:0004322) +

BEHAVIORAL

Autistic behavior (HP:0000729) +

CONGENITAL BIRTH DEFECTS

Congenital Heart Defect -

Surgeries: ear tubes, broken jaw +

Cryptorchidism -

Palatal Defects -

MISCELLANEOUS

Low CSF 5-methyltetrahydrofolate (HP:0012446) +

Hearing Loss -

REFERENCES

1.O'Rawe,J.,Jiang,T.,Sun,G.,Wu,Y.,Wang,W.,Hu,J.,Bodily,P.,Tian,L.,Hakonarson,H.,Johnson,W.E.,etal.(2013).Lowconcordanceofmultiplevariant-callingpipelines:practicalimplicationsforexomeandgenomesequencing.GenomeMedicine5,28.

2.O'Rawe,J.A.,Ferson,S.,andLyon,G.J.(2015).AccountingforuncertaintyinDNAsequencingdata.Trendsingenetics:TIG31,61-66.

3.Bamshad,M.J.,Ng,S.B.,Bigham,A.W.,Tabor,H.K.,Emond,M.J.,Nickerson,D.a.,andShendure,J.(2011).ExomesequencingasatoolforMendeliandiseasegenediscovery.NatureReviewsGenetics12,745-755.

4.Kong,A.,Frigge,M.L.,Masson,G.,Besenbacher,S.,Sulem,P.,Magnusson,G.,Gudjonsson,S.a.,Sigurdsson,A.,Jonasdottir,A.,Jonasdottir,A.,etal.(2012).Rateofdenovomutationsandtheimportanceoffather’sagetodiseaserisk.Nature488,471-475.

5.Yang,H.,Robinson,P.N.,andWang,K.(2015).Phenolyzer:phenotype-basedprioritizationofcandidategenesforhumandiseases.NatMethods12,841-843.

6.Chang,X.,andWang,K.(2012).wANNOVAR:annotatinggeneticvariantsforpersonalgenomesviatheweb.JMedGenet49,433-436.

7.Zemojtel,T.,Köhler,S.,Mackenroth,L.,Jäger,M.,Hecht,J.,Krawitz,P.,Graul-Neumann,L.,Doelken,S.,Ehmke,N.,andSpielmann,M.(2014).Effectivediagnosisofgeneticdiseasebycomputationalphenotypeanalysisofthedisease-associatedgenome.Sciencetranslationalmedicine6,252ra123-252ra123.

8.Brancati,F.,Sarkozy,A.,andDallapiccola,B.(2006).KBGsyndrome.OrphanetJRareDis1,50.

9.Crippa,M.,Rusconi,D.,Castronovo,C.,Bestetti,I.,Russo,S.,Cereda,A.,Selicorni,A.,Larizza,L.,andFinelli,P.(2015).FamilialintragenicduplicationofANKRD11underlyingthreepatientsofKBGsyndrome.MolCytogenet8,20.

10.Walz,K.,Cohen,D.,Neilsen,P.M.,Foster,J.,2nd,Brancati,F.,Demir,K.,Fisher,R.,Moffat,M.,Verbeek,N.E.,Bjorgo,K.,etal.(2015).CharacterizationofANKRD11mutationsinhumansandmicerelatedtoKBGsyndrome.HumGenet134,181-190.

11.Sirmaci,A.,Spiliopoulos,M.,Brancati,F.,Powell,E.,Duman,D.,Abrams,A.,Bademci,G.,Agolini,E.,Guo,S.,Konuk,B.,etal.(2011).MutationsinANKRD11causeKBGsyndrome,characterizedbyintellectualdisability,skeletalmalformations,andmacrodontia.AmJHumGenet89,289-294.

12.Lyon,G.J.,andO'Rawe,J.(2015).Humangeneticsandclinicalaspectsofneurodevelopmentaldisorders.InTheGeneticsofNeurodevelopmentalDisorders,K.Mitchell,ed.(Wiley).

13.Herrmann,J.,Pallister,P.D.,Tiddy,W.,andOpitz,J.M.(1975).TheKBGsyndrome-asyndromeofshortstature,characteristicfacies,mentalretardation,macrodontiaandskeletalanomalies.BirthDefectsOrigArticSer11,7-18.

14.Skjei,K.L.,Martin,M.M.,andSlavotinek,A.M.(2007).KBGsyndrome:reportoftwins,neurologicalcharacteristics,anddelineationofdiagnosticcriteria.AmJMedGenetA143A,292-300.

15.Ockeloen,C.W.,Willemsen,M.H.,deMunnik,S.,vanBon,B.W.,deLeeuw,N.,Verrips,A.,Kant,S.G.,Jones,E.A.,Brunner,H.G.,vanLoon,R.L.,etal.(2015).FurtherdelineationoftheKBGsyndromecausedbyANKRD11aberrations.EurJHumGenet23,1270.

16.Ansari,M.,Poke,G.,Ferry,Q.,Williamson,K.,Aldridge,R.,Meynert,A.M.,Bengani,H.,Chan,C.Y.,Kayserili,H.,Avci,S.,etal.(2014).GeneticheterogeneityinCorneliadeLangesyndrome(CdLS)andCdLS-likephenotypeswithobservedandpredictedlevelsofmosaicism.JMedGenet51,659-668.

17.Parenti,I.,Gervasini,C.,Pozojevic,J.,Graul-Neumann,L.,Azzollini,J.,Braunholz,D.,Watrin,E.,Wendt,K.S.,Cereda,A.,Cittaro,D.,etal.(2015).Broadeningofcohesinopathies:exomesequencingidentifiesmutationsinANKRD11intwopatientswithCorneliadeLange-overlappingphenotype.ClinGenet.

18.Zollino,M.,Battaglia,A.,D'Avanzo,M.G.,DellaBruna,M.M.,Marini,R.,Scarano,G.,Cappa,M.,andNeri,G.(1994).SixadditionalcasesoftheKBGsyndrome:clinicalreportsandoutlineofthediagnosticcriteria.AmJMedGenet52,302-307.

19.Brancati,F.,D'Avanzo,M.G.,Digilio,M.C.,Sarkozy,A.,Biondi,M.,DeBrasi,D.,Mingarelli,R.,andDallapiccola,B.(2004).KBGsyndromeinacohortofItalianpatients.AmJMedGenetA131,144-149.

20.Maegawa,G.H.,Leite,J.C.,Felix,T.M.,daSilveira,H.L.,anddaSilveira,H.E.(2004).ClinicalvariabilityinKBGsyndrome:reportofthreeunrelatedfamilies.AmJMedGenetA131,150-154.

21.Davanzo,A.M.,Rosalia,G.,Biondi,M.,DeBrasi,D.,Colucci,A.R.,Panetta,A.,Zaccagnino,P.,Andreoli,G.,andRoggini,M.(2005).EightisolatedcasesofKBGsyndrome:anewhypothesisofstudy.EurRevMedPharmacolSci9,49-52.

22.Lim,J.H.,Seo,E.J.,Kim,Y.M.,Cho,H.J.,Lee,J.O.,Cheon,C.K.,andYoo,H.W.(2014).AdenovomicrodeletionofANKRD11geneinaKoreanpatientwithKBGsyndrome.AnnLabMed34,390-394.

23.Hafiz,A.,Mufeed,A.,Ismael,M.,andAlam,M.(2015).AnunusualcaseofKBGsyndromewithuniqueoralfindings.BMJCaseRep2015.

24.Kim,H.J.,Cho,E.,Park,J.B.,Im,W.Y.,andKim,H.J.(2015).AKoreanfamilywithKBGsyndromeidentifiedbyANKRD11mutation,andphenotypiccomparisonofANKRD11mutationand16q24.3microdeletion.EurJMedGenet58,86-94.

25.Tunovic,S.,Barkovich,J.,Sherr,E.H.,andSlavotinek,A.M.(2014).DenovoANKRD11andKDM1AgenemutationsinamalewithfeaturesofKBGsyndromeandKabukisyndrome.AmJMedGenetA164A,1744-1749.

26.Gallagher,D.,Voronova,A.,Zander,M.A.,Cancino,G.I.,Bramall,A.,Krause,M.P.,Abad,C.,Tekin,M.,Neilsen,P.M.,Callen,D.F.,etal.(2015).Ankrd11isachromatinregulatorinvolvedinautismthatisessentialforneuraldevelopment.DevCell32,31-42.

27.Zhang,A.,Li,C.W.,andChen,J.D.(2007).Characterizationoftranscriptionalregulatorydomainsofankyrinrepeatcofactor-1.BiochemBiophysResCommun358,1034-1040.

28.Neilsen,P.M.,Cheney,K.M.,Li,C.W.,Chen,J.D.,Cawrse,J.E.,Schulz,R.B.,Powell,J.A.,Kumar,R.,andCallen,D.F.(2008).IdentificationofANKRD11asap53coactivator.JCellSci121,3541-3552.

29.Barbaric,I.,Perry,M.J.,Dear,T.N.,RodriguesDaCosta,A.,Salopek,D.,Marusic,A.,Hough,T.,Wells,S.,Hunter,A.J.,Cheeseman,M.,etal.(2008).AnENU-inducedmutationintheAnkrd11generesultsinanosteopenia-likephenotypeinthemousemutantYoda.PhysiolGenomics32,311-321.

30.Frye,R.E.,Sequeira,J.M.,Quadros,E.V.,James,S.J.,andRossignol,D.A.(2013).Cerebralfolatereceptorautoantibodiesinautismspectrumdisorder.MolPsychiatry18,369-381.

31.Filloux,F.M.(2015).Cannabinoidsforpediatricepilepsy?Upinsmokeorrealscience?TranslPediatr4,271-282.

32.Devinsky,O.,Thiele,E.,Laux,L.,Friedman,D.,Patel,A.,Bluvstein,J.,Chez,M.,Joshi,C.,Cilio,R.,Filloux,F.,etal.(2015).EfficacyandSafetyofEpidiolex(Cannabidiol)inChildrenandYoungAdultswithTreatment-ResistantEpilepsy:UpdatefromtheExpandedAccessProgram.InAmericanEpilepsySocietyAnnualMeeting.(

33.Richards,S.,Aziz,N.,Bale,S.,Bick,D.,Das,S.,Gastier-Foster,J.,Grody,W.W.,Hegde,M.,Lyon,E.,Spector,E.,etal.(2015).Standardsandguidelinesfortheinterpretationofsequencevariants:ajointconsensusrecommendationoftheAmericanCollegeofMedicalGeneticsandGenomicsandtheAssociationforMolecularPathology.GenetMed17,405-424.

34.Zhu,P.,He,L.,Li,Y.,Huang,W.,Xi,F.,Lin,L.,Zhi,Q.,Zhang,W.,Tang,Y.T.,Geng,C.,etal.(2014).OTG-snpcaller:anoptimizedpipelinebasedonTMAPandGATKforSNPcallingfromiontorrentdata.PLoSOne9,e97507.

35.Garrison,E.,andMarth,G.(2012).Haplotype-basedvariantdetectionfromshort-readsequencing.arXivpreprintarXiv:12073907.

36.McKenna,A.,Hanna,M.,Banks,E.,Sivachenko,A.,Cibulskis,K.,Kernytsky,A.,Garimella,K.,Altshuler,D.,Gabriel,S.,andDaly,M.(2010).TheGenomeAnalysisToolkit:aMapReduceframeworkforanalyzingnext-generationDNAsequencingdata.Genomeresearch20,1297-1303.

37.Danecek,P.,Auton,A.,Abecasis,G.,Albers,C.A.,Banks,E.,DePristo,M.A.,Handsaker,R.E.,Lunter,G.,Marth,G.T.,Sherry,S.T.,etal.(2011).ThevariantcallformatandVCFtools.Bioinformatics27,2156-2158.

38.Wang,K.,Li,M.,andHakonarson,H.(2010).ANNOVAR:functionalannotationofgeneticvariantsfromhigh-throughputsequencingdata.NucleicAcidsRes38,e164.

39.McKinney,W.(2010).DatastructuresforstatisticalcomputinginPython.InProceedingsofthe9th.pp51-56.

40.VanDerWalt,S.,Colbert,S.C.,andVaroquaux,G.(2011).TheNumPyarray:astructureforefficientnumericalcomputation.ComputinginScience&Engineering13,22-30.

41.McLaren,W.,Pritchard,B.,Rios,D.,Chen,Y.,Flicek,P.,andCunningham,F.(2010).DerivingtheconsequencesofgenomicvariantswiththeEnsemblAPIandSNPEffectPredictor.Bioinformatics26,2069-2070.

42.Paila,U.,Chapman,B.a.,Kirchner,R.,andQuinlan,A.R.(2013).GEMINI:IntegrativeExplorationofGeneticVariationandGenomeAnnotations.PLoSComputationalBiology9.

43.Kircher,M.,Witten,D.M.,Jain,P.,O'Roak,B.J.,Cooper,G.M.,andShendure,J.(2014).Ageneralframeworkforestimatingtherelativepathogenicityofhumangeneticvariants.NatGenet46,310-315.

44.Hu,H.,Huff,C.D.,Moore,B.,Flygare,S.,Reese,M.G.,andYandell,M.(2013).VAAST2.0:improvedvariantclassificationanddisease-geneidentificationusingaconservation-controlledaminoacidsubstitutionmatrix.GenetEpidemiol37,622-634.

45.Kennedy,B.,Kronenberg,Z.,Hu,H.,Moore,B.,Flygare,S.,Reese,M.G.,Jorde,L.B.,Yandell,M.,andHuff,C.(2014).UsingVAASTtoIdentifyDisease-AssociatedVariantsinNext-GenerationSequencingData.CurrProtocHumGenet81,61411-161425.

46.Rope,A.F.,Wang,K.,Evjenth,R.,Xing,J.,Johnston,J.J.,Swensen,J.J.,Johnson,W.E.,Moore,B.,Huff,C.D.,Bird,L.M.,etal.(2011).UsingVAASTtoidentifyanX-linkeddisorderresultinginlethalityinmaleinfantsduetoN-terminalacetyltransferasedeficiency.AmJHumGenet89,28-43.

47.Dierking,A.,andSchmidtke,J.(2014).ThefutureofClinicalUtilityGeneCardsinthecontextofnext-generationsequencingdiagnosticpanels.EurJHumGenet22,1247.

48.Rebhan,M.,Chalifa-Caspi,V.,Prilusky,J.,andLancet,D.(1998).GeneCards:anovelfunctionalgenomicscompendiumwithautomateddataminingandqueryreformulationsupport.Bioinformatics14,656-664.

49.Stelzer,G.,Dalah,I.,Stein,T.I.,Satanower,Y.,Rosen,N.,Nativ,N.,Oz-Levi,D.,Olender,T.,Belinky,F.,andBahir,I.(2011).In-silicohumangenomicswithGeneCards.HumGenomics5,709-717.

50.Brown,G.R.,Hem,V.,Katz,K.S.,Ovetsky,M.,Wallin,C.,Ermolaeva,O.,Tolstoy,I.,Tatusova,T.,Pruitt,K.D.,andMaglott,D.R.(2015).Gene:agene-centeredinformationresourceatNCBI.Nucleicacidsresearch43,D36-D42.

51.Hamosh,A.,Scott,A.F.,Amberger,J.S.,Bocchini,C.A.,andMcKusick,V.A.(2005).OnlineMendelianInheritanceinMan(OMIM),aknowledgebaseofhumangenesandgeneticdisorders.Nucleicacidsresearch33,D514-D517.