Recurrent NFIA K125E substitution represents a loss‐of‐function allele: Sensitive in vitro and in vivo assays for nontruncating alleles

Abstract Nuclear factor I A (NFIA) is a transcription factor that belongs to the NFI family. Truncating variants or intragenic deletion of the NFIA gene are known to cause the human neurodevelopmental disorder known as NFIA‐related disorder, but no patient heterozygous for a missense mutation has been reported. Here, we document two unrelated patients with typical phenotypic features of the NFIA‐related disorder who shared a missense variant p.Lys125Glu (K125E) in the NFIA gene. Patient 1 was a 6‐year‐old female with global developmental delay, corpus callosum anomaly, macrocephaly, and dysmorphic facial features. Patient 2 was a 14‐month‐old male with corpus callosum anomaly and macrocephaly. By using Drosophila and zebrafish models, we functionally evaluated the effect of the K125E substitution. Ectopic expression of wild‐type human NFIA in Drosophila caused developmental defects such as eye malformation and premature death, while that of human NFIA K125E variant allele did not. nfia‐deficient zebrafish embryos showed defects of midline‐crossing axons in the midbrain/hindbrain boundary. This impairment of commissural neurons was rescued by expression of wild‐type human NFIA, but not by that of mutant variant harboring K125E substitution. In accordance with these in vivo functional analyses, we showed that the K125E mutation impaired the transcriptional regulation of HES1 promoter in cultured cells. Taken together, we concluded that the K125E variant in the NFIA gene is a loss‐of‐function mutation.


| INTRODUCTION
The prevalence of intragenic deletions in patients with aplasia or hypoplasia of the corpus callosum and developmental delay suggests that haploinsufficiency of the Nuclear factor I A (NFIA) gene is a primary cause of chromosome 1p32-p31 deletion syndrome or brain malformations with or without urinary tract defects (MIM 613735) (Bayat et al., 2017;Hollenbeck et al., 2017;Mikhail et al., 2011;Nyboe et al., 2015;Rao et al., 2014). Identification of frameshift and nonsense mutations in NFIA further supports this notion (Negishi et al., 2015;Revah-Politi et al., 2017;Zhang et al., 2020). Indeed, the mouse ortholog of this conserved transcription factor, Nfia, is required for differentiation and maturation of astrocyte and oligodendrocyte and its loss results in the aplasia/hypoplasia of corpus callosum and urinary tract defects (das Neves et al., 1999;Lu et al., 2007).
Despite growing recognition of the impact of NFIA haploinsufficiency on the neurodevelopmental disorder, there is no patient heterozygous for a pathogenic missense variant to date (but see also Zenker et al., 2019 for three candidate pathogenic variants).
Here, we report the same de novo missense mutation K125E in the NFIA gene in two unrelated patients. By using Drosophila and zebrafish models (Suzuki et al., 2019;Uehara et al., 2020) as well as cell culture system, we unambiguously demonstrated that this K125E missense variant represents a loss-of-function pathogenic allele.
Presently reported in vivo assays will be useful for functional evaluation of other missense variants of NFIA.

| CLINICAL REPORT
Patient 1 was a 6-year-old female who was the first child of healthy and nonconsanguineous Japanese parents. She was born after an uncomplicated pregnancy at 35 weeks and 4 days of gestation. Her weight at birth was 2142 g (À0.5 SD), length was 43 cm (À1.2 SD), and head circumstance was 33 cm (+0.8 SD). After birth, she showed tachycardia. Head ultrasound showed ventricular enlargement and intraventricular hemorrhage. She had been in neonatal intensive care unit since 34-day-old. After discharge, she attended her hospital regularly due to cerebral palsy and global developmental delay. At 2 years of age, she underwent an operation for exotropia. She had congenital hearing loss and wore hearing aids at 5 years of age. Head magnetic resonance imaging (MRI) at 5 years of age showed thin corpus callosum, cyst of septi pellucidi, ventricular wall irregularity, and periventricular leukomalacia (Figure 1a). She showed distinctive dysmorphic features with high hairline, small eyes, anteverted nares, a depressed nasal bridge, a broad columella, a thin upper-lip, and higharched palate (Figure 1b). Her developmental milestones were delayed. She started to walk at the age of 3 years. She also stated to F I G U R E 1 Clinical characteristics of two patients with the same NFIA variant. (a) Results of head MRI of Patient 1 at 5 years of age. The picture above shows sagittal T1-weighted image. The picture below shows axial T2-weighted fluid-attenuated inversion recovery image. Note thin corpus callosum, cyst of septi pellucidi, ventricular wall irregularity, and decreased white matter volume. (b) Pictures of Patient 1 at 6 years of age. Note the high hairline, small eyes, anteverted nares, a depressed nasal bridge, a broad columella, and a thin upper-lip. (c) Results of head MRI of Patient 2 at 1 month of age. The picture above shows sagittal-T2-weighted image. The picture below shows axial diffusion-weighted image. Note polycerebral gyrus at parasylvius fissures, cortical dysplasia of bilateral cerebral hemisphere, partial myelination delay, and hypoplasia of corpus callosum [Color figure can be viewed at wileyonlinelibrary.com] speak her words at the age of 3 years and spoke only few words at the age of 6 years. Her developmental quotient as assessed using the WISC-IV test was 23. Her physical growth was also delayed. At 6 years of age, her weight was 15.9 kg (À1.3 SD), height was 102.4 cm (À2.4 SD), and head circumference was 54.5 cm (+2.4 SD).
She had no urogenital anomalies.
Patient 2 was a 14-month-old boy who was born at 34 weeks gestation. He had been diagnosed at 34 weeks gestation with a head enlargement. His weight at birth was 2635 g (+2.1 SD), length was 47.5 cm (+1.6 SD), and head circumference was 35.3 cm (+3.0 SD). After birth, a head MRI showed polycerebral gyrus at parasylvius fissures, cortical dysplasia of bilateral cerebral hemisphere, partial myelination delay, and hypoplasia of corpus callosum ( Figure 1c). He had mild congenital hearing impairment. An electrocardiogram and a renal echogram showed no anomalies. He showed distinctive dysmorphic features with high hairline, thick eyebrow, short nose, anteverted nares, long philtrum, thin upper-lip vermilion, and a retrognathia. His developmental milestones were delayed; he gained head control and rolling over at 8 months of age, sat without support at 11 months of age, and slithering at 1 year of age. He was able to stand with support at 1 year of age. At 1 year and 1 month of age, his weight was 10.42 kg (+0.3 SD), height was 81.0 cm (+1.9 SD), and head circumference was 52.0 cm (+3.9 SD).

| Functional assays in Drosophila
We PCR-amplified the NFIA open reading frame sequence from a human cDNA clone (Kazusa DNA Research Institute, Chiba, Japan, ORK00836) and subcloned it into a modified pENTR221 vector. The K125E mutation was introduced into the subclone by site-directed mutagenesis. Both subcloned fragments were verified by sequencing, and transferred into a destination vector pUASg-attB (Bischof et al., 2013)

| Functional assays in zebrafish
Zebrafish (Danio rerio) were reared and maintained under a 14 h light and 10 h dark photoperiod according to the standard protocol.
Zebrafish carrying the nfia Q232X nonsense mutant allele (nfia sa16768 ) was obtained from Zebrafish International Resource Center and was used for rescue experiments with wild-type and K125E mRNA.
nfia sa16768 was previously generated by a targeting induced local lesions in genomes project (Kettleborough et al., 2013). For rescue experiments, wild-type human NFIA coding sequence was generated by a DNA synthesis service (Fasmac, Japan) and subcloned into an expression vector pCS2+. The K125E point mutation was introduced into human NFIA by the QuickChange method (Agilent Technologies) using following two primers; 5 0 -GCTGCACAAACTCTTTAAGCATTTC TTGGGGATGGTATCTAATG-3 0 and 5 0 -CATTAGATACCATCCCCAAG AAATGCTTAAAGAGTTTGTGCAGC-3 0 . These constructs were used to generate wild-type and mutant NFIA mRNA using the mMESSAGE mMACHINE SP6 Kit (Thermo Fisher Scientific) according to the manufacture's protocol. The human NFIA mRNAs (100 pg) were injected into 1-2-cell stage zebrafish embryos produced by crossing nfia heterozygous mutant carrier fish. Embryos were fixed at 72 h postfertilization (hpf) and subjected to immunolabeling using antiacetylated α-tubulin (clone 6-11B-1; Sigma), HRP-conjugated anti-mouse IgG (Invitrogen), and ImmPACT DAB Substrate (Vector Laboratories). For genotyping of immunolabeled embryos, the region surrounding the Q232X mutation site was amplified by genomic PCR using following two primers; 5 0 -CTGTATTCTGTCATGTTCATTCAGATAACAGTC-3 0 and 5 0 -GCTCAATGATGTCCCAAAAGGAAG-3 0 . PCR products were digested with Fai I restriction enzyme (SibEnzyme, Russia) and separated by 15% polyacrylamide gel electrophoresis. This zebrafish study was approved by Animal Care and Use Committee of Aoyama Gakuin University (A9/2020) and carried out according to the Aoyama Gakuin University Animal Care and Use Guideline.

| NanoLuc reporter assay
The human 0.6-kb (À602 to +44) GFAP and 1.2-kb (À1039 to +135) HES1 promoter regions were amplified by PCR from human genome and were subcloned into the reporter vector pNL2.2 (Promega Corp., Madison, Wisconsin). The pNL2.2 reporter constructs were cotransfected with pCAGGS-Luc2, and pCAGGS-NFIA-3xFlag plasmids into 293T cells using the calcium phosphate-mediated method. Two days later, the cells were lysed in passive lysis buffer (Promega Corp.), and the luciferase and NanoLuc activities were measured using the Nano-Glo Dual-Luciferase reporter assay system (Promega Corp.) according to the manufacturer's instructions.

| Variants identified in two patients
Trio exome analysis showed that both patients had the same de novo nonsynonymous mutation in the NFIA gene (NM_001134673.4), c.373A>G (p.Lys125Glu). The DNA-binding domains of NFIA and its paralogs are highly conserved and the K125 residue is identical across vertebrates and invertebrates studied (Figure 2). Indeed, the combined annotation-dependent depletion scores (Kircher et al., 2014) (Tadaka et al., 2019) and also absent in the Genome Aggregation Database (gnomAD) (http://gnomad.broadinstitute.org/).

| Functional assays of K125E variant in Drosophila
To assess the functional significance of the K125E mutation in vivo, we introduced the wild-type and mutant human NFIA transgenes into Drosophila and expressed them in the nervous system and imaginal discs using the GAL4/UAS system. We found that ectopic expression

| Functional assays of K125E variant in zebrafish
To further address the functional consequences of the K125E mutation in a vertebrate model, zebrafish, we employed the nfia nonsense mutant (nfia sa16768 ) as an NFIA-deficient animal model and performed an mRNA rescue assay. As described previously (Chitnis & Kuwada, 1990;Wilson et al., 1990), commissural axons crossed the midline in the midbrain/hindbrain boundary region in wild-type and nfia heterozygous mutant embryos at 3 dpf (Figure 4a

| K125E mutation impaired the transcriptional regulation ability of NFIA
Because NFIA is known to repress transcription of Hes1 and to activate that of Gfap by direct binding to the promoter regions in mouse (Miura et al., 1990;Piper et al., 2010), we performed luciferase reporter assay using the human GFAP and the HES1 promoters to determine whether NFIA K125E has the ability to regulate transcription

| DISCUSSION
In this article, we report the recurrent heterozygous missense mutation K125E in the NFIA gene in two unrelated patients with an intellectual disability, corpus callosum anomaly, and macrocephaly. Our F I G U R E 3 Phenotypes of ectopic expression of NFIA gene in Drosophila. Ectopic expression of NFIA WT (a) but not NFIA K125E (b) during retinal development (GMR>NFIA WT ) caused severe neurodegeneration (rough eye phenotype). Ectopic expression of NFIA WT (c and d) but not NFIA K125E (e and f) under the control of the ey-GAL4 driver caused antenna-to-leg transformation. Note that the ey gene is expressed in the eye-antennal disc primordia and ubiquitously in the first instar larval disc (Quiring et al., 1994;Urbach & Technau, 2003) [Color figure can be viewed at wileyonlinelibrary.com] Chen et al., 2011;Hollenbeck et al., 2017;Koehler et al., 2010;Lu et al., 2007;Mikhail et al., 2011;Negishi et al., 2015;Nyboe et al., 2015;Rao et al., 2014;Revah-Politi et al., 2017;Zhang et al., 2020). A few missense mutations have also been reported to be pathogenic or likely pathogenic (Zenker et al., 2019), but their pathogenicity has not been experimentally verified yet. Therefore, this is F I G U R E 4 Commissural defects in nfia mutant zebrafish. Commissural axons (arrows) cross the midline in the midbrain/hindbrain boundary in wild-type (a) (n = 4/4) and nfia heterozygous mutant (b) (n = 6/6) embryos at 3 dpf, but many of them failed to do so in nfia homozygous mutants (c) (n = 5/5). The midline crossing defects in nfia homozygous mutants were rescued by injection of NFIA WT mRNA (d) (n = 5/5), but not by that of NFIA K125E one (e) (n = 5/5). Axons were labeled with anti-acetylated α-tubulin antibody [Color figure can be viewed at wileyonlinelibrary.com] F I G U R E 5 HES1 and GFAP promoter assay in HEK293T cells. (a) Schematic representation of the human HES1 promoter-NlucP and the human GFAP promoter-NlucP constructs. Potential NFI-binding sites (NNTTGGCNNNNNNCCNNN) predicted by TFBIND (http://tfbind.hgc.jp/; Tsunoda & Takagi, 1999) are shown by triangles. (b) HES1 promoter repression by NFIA. Dose-dependent repression of human HES1 promoter by NFIA WT was not observed in NFIA K125E . Data are mean ± SD from triplicate experiments. **p < 0.01 by lower-tailed Dunnett multiple comparisons test (α = 0.05). ns, not significant. (c) GFAP promoter activation by NFIA. Dose-dependent activation of human GFAP promoter by NFIA WT was weakened in NFIA K125E . Data are mean ± SD from triplicate experiments. **p < 0.01 by upper-tailed Dunnett multiple comparisons test (α = 0.05). ##p < 0.01 by unpaired t-test. ns, not significant the first case report of NFIA missense variant associated with the neurodevelopmental disorder.
Nuclear factor I family proteins are found to bind the palindromic consensus sequence as homo-or heterodimers (Gronostajski, 2000). Although dimerization is essential for DNA binding, these two activities can be separated by mutations (Armentero et al., 1994). K125 residue is sandwiched between two mutants, 6th and 7th mutations in Armentero et al. (1994); Figure 2), that disrupt the DNA-binding activity, but the former does not impair the dimerization activity. If the K125E mutant protein can still dimerizes, but cannot bind the target sequences, its detrimental effect may be even stronger than that of truncating variants.
As in Figure 2, the K125 residue is conserved in all four human NFI family genes and the missense mutations in NFIB and NFIX are also classified as pathogenic or probably pathogenic in the ClinVar database (Landrum et al., 2018). In particular, NFIB K126E mutation (at the site corresponding to K125 residue in NFIA) causes a severe loss of transcriptional activity and is one of the variants associated with intellectual disability and macrocephaly (Schanze et al., 2018). In addition, two mutations at the same K125 residue have been reported in NFIX. One is K125E in a patient with Malan syndrome (Gurrieri et al., 2015) and another is K125N in a patient with developmental disabilities (Lu et al., 2017); both patients had macrocephaly. Taken together, all these findings underscore the importance of the K125 residue for NFI function.
The hypoplasia of corpus callosum and macrocephaly may represent diagnostic clues to the NFIA-related disorder. Indeed, macrocephaly was shown in 14 of 14 reported patients with truncating variants or intragenic deletions in the NFIA gene (Bayat et al., 2017;Mikhail et al., 2011;Negishi et al., 2015;Nyboe et al., 2015;Rao et al., 2014;Revah-Politi et al., 2017;Zhang et al., 2020) (Table 1). Consistent with this shared feature, knockout mice for Nfia, Nfib, and Nfix all exhibit severe brain malformations including megalencephaly (Campbell et al., 2008;Chang et al., 2013;das Neves et al., 1999). This megalencephaly is hypothesized to be due to delayed radial glia differentiation, which promotes extended self-renewal and results in an increased number of T A B L E 1 Summary of the patients with neurodevelopmental disorder and heterozygous variants in NFIA neural progenitors (Zenker et al., 2019). Corpus callosum hypoplasia was also shown in 13 of these 14 patients (Table 1) and is indeed an important feature of the NFIA-related disorder. In mouse, formation of the corpus callosum requires astroglial-mediated remodeling of the interhemispheric midline (das Neves et al., 1999;Gobius et al., 2016). Knockout of Nfia and Nfib delays differentiation of midline zipper glia cells from radial glia, which prevents normal interhemispheric remodeling and affects subsequent callosal tract formation (Gobius et al., 2016). Our in vitro experiments clearly show that the repressive effect of NFIA on the HES1 promoter was severely impaired by the K125E mutation ( Figure 5b).
Therefore, it is possible that cellular differentiation from radial glia was delayed and overgrowth of progenitor cells caused the subsequent macrocephaly in the present patients (Piper et al., 2010). The observation of commissural defects in our zebrafish model deficient for nfia also substantiates the association of NFIA disruption with hypoplasia of corpus callosum.
Finally, we anticipate that other NFIA missense variants may also be associated with the neurodevelopmental disease. If so, our in vitro and in vivo assays would be a valuable tool for diagnosis, especially for evaluating whether a missense mutation is a loss-offunction.

ACKNOWLEDGMENTS
We thank Mrs. Chika Kanoe, Mrs. Keiko Tsukue, and Mrs. Yumi Obayashi for their technical assistance in the preparation of this article. This work was supported by the Japan Agency for Medical Research and Development under Grant Numbers 18ek0109301, 18ek0109288h0002, 19gk0110038h, 19ek0109288h0003, and 20ek0109484h0001.

CONFLICT OF INTEREST
The authors declared no potential conflicts of interest. All authors agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are openly available in