Phenotypic expansion of the BPTF‐related neurodevelopmental disorder with dysmorphic facies and distal limb anomalies

Abstract Neurodevelopmental disorder with dysmorphic facies and distal limb anomalies (NEDDFL), defined primarily by developmental delay/intellectual disability, speech delay, postnatal microcephaly, and dysmorphic features, is a syndrome resulting from heterozygous variants in the dosage‐sensitive bromodomain PHD finger chromatin remodeler transcription factor BPTF gene. To date, only 11 individuals with NEDDFL due to de novo BPTF variants have been described. To expand the NEDDFL phenotypic spectrum, we describe the clinical features in 25 novel individuals with 20 distinct, clinically relevant variants in BPTF, including four individuals with inherited changes in BPTF. In addition to the previously described features, individuals in this cohort exhibited mild brain abnormalities, seizures, scoliosis, and a variety of ophthalmologic complications. These results further support the broad and multi‐faceted complications due to haploinsufficiency of BPTF.


| INTRODUCTION
The bromodomain PHD finger transcription factor (BPTF) is the largest subunit of the nucleosome remodeling factor (NURF), a member of the imitation switch (ISWI) chromatin remodeling complex family (Bartholomew, 2014). NURF complexes have been shown to catalyze ATP-dependent nucleosome sliding and facilitate transcription of chromatin (Hamiche et al., 1999), thereby regulating various genes and helping govern higher-order chromatin structures. BPTF in particular facilitates this interaction by the preferential binding of its plant homeodomain (PHD) finger to the tails of the trimethylated lysine 4 of histone H3 (H3K4me3) and binding of its bromodomain to the acetylated lysine 16 of histone H4 (H4K16ac) (Filippakopoulos et al., 2012;Wysocka et al., 2006). Animal studies have demonstrated the essential role of Bptf in the differentiation of the primary germ layers and establishment of the embryonal anterior-posterior axis (Landry et al., 2008). Recent unpublished studies showed that Emx1-Cre driven inactivation of Bptf in the developing mouse forebrain results in animals with a severe reduction in cortical mass with little to no effect in heterozygous animals (David Picketts, personal communication). Human studies have detailed extensively the role of BPTF in the development of several malignancies (Dai et al., 2015;Green et al., 2020;Lee et al., 2016;Richart et al., 2016;Zhao et al., 2019) and in T-cell homeostasis and function (Mayes et al., 2016;Wu et al., 2016).
The significance of constitutional, pathogenic variants in BPTF was reported in 2017, when the first 10 individuals with heterozygous, pathogenic single nucleotide or copy-number deletion variants were described (Stankiewicz et al., 2017). Haploinsufficiency of BPTF in humans leads to the clinical entity known as neurodevelopmental disorder with dysmorphic facies and distal limb anomalies (NEDDFL, MIM#617755). The syndrome consists primarily of developmental delay (DD)/intellectual disability (ID), speech delay, postnatal microcephaly, and dysmorphic features. The disorder remains, however, incompletely differentiated with few further individuals described in the medical literature (Deciphering Developmental Disorders, 2015Midro et al., 2019;Popp et al., 2017). All individuals identified thus far, have carried heterozygous de novo changes with mostly copy-number variant (CNV) deletions and frameshift, or nonsense single-nucleotide variants (SNVs). Based on these publications, it has remained unclear whether pathogenic variants may be inherited and whether deleterious variants are fully or incompletely penetrant.
Here, we describe 25 novel patients with NEDDFL due to 20 distinct variants in BPTF, including, for the first time, four patients found to have inherited a causative variant from their apparently nonmosaic, affected parents. The study provides additional insight into the phenotypic features of this disorder and expands on our knowledge of the inheritance and penetrance of this still new disorder.

| Editorial policies and ethical considerations
This study was conducted in accordance with the ethical standards of the Baylor College of Medicine Committee on Human Research.

| Patients and recruitment
Subjects were identified and recruited either through their treating clinicians, self-referral, or GeneMatcher (Sobreira et al., 2015). Molecular testing results from exome sequencing, chromosomal microarray, or next-generation sequencing panel were submitted by patients' healthcare providers (See "Supplementary Materials" for details of each test's composition). Basic medical information including birth parameters, developmental histories, and physical examinations were collected from healthcare providers and/or patient families. Percentiles and z-scores for height, weight and head circumference were calculated based on the Centers for Disease Control and Prevention (CDC) growth charts using the PediTools (https://peditools.org) (Chou et al., 2020) or SimulConsult (https://simulconsult.com/resources/ measurement.html) online calculators. All photographs submitted are used with the written consent of patients or guardians as appropriate.
In the case of patients lost to follow-up or for whom no clinical data were available, only basic demographic and molecular findings are reported.

| Molecular findings
Our analysis identified 20 novel, distinct variants in BPTF (Table 1) distributed throughout the gene (Figure 1), including 14 de novo and four inherited. Unique exonic variants include nine frameshift, four nonsense, three splicing, two in-frame deletions, one missense, and one single exon truncating deletion (Tables 1 and S1). We also describe one previously published chromosomal translocation and CNV deletion disrupting BPTF (Patient 22) (Midro et al., 1993;Midro et al., 2019). Variants were interpreted as pathogenic (11), likely pathogenic (7), or VUS (2) based on the current ACMG criteria. There does not appear to be a predominant genotype-phenotype correlation within our cohort though it is worth noting that manifestations were milder in the lone individual with a missense variant (Patient 23) when compared to the rest of the cohort.

| Birth history and growth parameters
Most individuals in our cohort were born at term (median 38, range 36-42 weeks) following reportedly uncomplicated pregnancies (Table 1). When available, APGAR scores were within normal limits (median at 1 min = 9 (range 2-10); median at 5 min = 10 (range 8-10)). Available birth parameters showed that 13 individuals were Previously reported (Midro et al., 1993(Midro et al., , 2019 Parental samples were negative by Sanger sequencing. Paternity and maternity were not confirmed. small for gestational age (i.e., birth weight less than the 10th percentile) though only one patient had a length below the third percentile.
Of the individuals with measurements available, 4/10 (40%) had a head circumference less than the third percentile for gestational age at birth.
At the time of their most recent clinical assessment, 5/20 (25%) individuals exhibited short stature for age and 10/19 (53%) individuals had decreased weight for age (based on measurements less than the third percentiles for height and weight respectively). Though head circumferences were not provided for every individual, consistent with previous observations (Stankiewicz et al., 2017), 12/20 (60%) individuals demonstrated a head circumference less than the third percentile for age.

| Development and neurologic findings
In our cohort, 23/26 (88%) patients demonstrated some form of DD, ranging from severe to only mild delays/deficits (

| Dysmorphic features and other findings
As noted previously (Stankiewicz et al., 2017), mild dysmorphic fea- We identified skeletal abnormalities in a total of 13 individuals.

| DISCUSSION
The rapid advancements of genomic technologies have led to an almost continuous stream of new diseases and syndromes based on a common defect gene. This "genotype-first" approach has gained prominence, particularly in complex diseases like autism spectrum disorders or DD where affected individuals often display a range of nonspecific symptoms and manifestations (Stessman et al., 2014 (Midro et al., 1993), (l) Patient 22 at age 35 years (Midro et al., 2019), (m) Patient 23, (n) Subject 5 from Stankiewicz et al., 2017(Stankiewicz et al., 2017. Note the presence of prominent nasal ridge (a, b, c, e, g, h, i, k, l, n)  mother of Patient 15) as having microcephaly; however, it is unclear whether this was ever clinically ascertained. Patient 10, the mother of Patient 9, was found to have microcephaly though she does not have a history of reported delays or ID. Interestingly, Patient 2 (the mother of Patient 1) was described by her own mother as being small for gestational age, having a "small head" as a child and exhibiting significant feeding difficulties. She had substantial learning disabilities in school and though she is able to function reasonably independently as an adult, she continues to live with her own mother who assists her with childcare and some day-to-day activities. The histories reported in these patients and the presence of inherited pathogenic variants in BPTF do suggest a pattern of variable expressivity in NEDDFL. It is possible that this is related to patients' genotypes; however, additional clinical information is needed to definitively characterize these relationships. Given its role in chromatin remodeling, it is likely that the severity of NEDDFL is heavily influenced by additional epigenetic factors.
The present study has some inherent limitations. Subjects in our study have been recruited from different clinical settings and in several different countries. It is thus unsurprising that the availability of records and willingness of families to participate varied significantly.
These factors may, therefore, result in an underestimation of the frequency of some of the clinical features, particularly as some patients have been lost to follow-up since initial enrollment. Additionally, because of the different molecular diagnostic techniques used (i.e., exome sequencing, panel testing, etc.) and differences in coverage, it is possible that some patients may harbor variants in other neurodevelopmental genes, which may be modifying their presentations.
Finally, as the recognition of this disorder increases, it is anticipated that the finding of missense variants in BPTF may present a diagnostic challenge for practitioners, particularly in individuals with mild phenotypes. Though there is not a well-validated functional assay for the disorder at this time, the use of RNA-sequencing or proteomics may have some utility in such cases.
The present study does, however, add significantly to the clinical description of NEDDFL and more than doubles the number of affected individuals in the literature. The clinical and molecular information gathered may allow for increased recognition of this disorder on the part of practitioners and shorten the diagnostic odyssey experienced by many patients. Also, the descriptions here may provide additional diagnostic, prognostic, and management information for both providers and caregivers.