A Mouse Model with a Frameshift Mutation in the Nuclear Factor I/X ( NFIX ) Gene Has Phenotypic Features of Marshall‐Smith Syndrome

The nuclear factor I/X (NFIX) gene encodes a ubiquitously expressed transcription factor whose mutations lead to two allelic disorders characterized by developmental, skeletal, and neural abnormalities, namely, Malan syndrome (MAL) and Marshall–Smith syndrome (MSS). NFIX mutations associated with MAL mainly cluster in exon 2 and are cleared by nonsense‐mediated decay (NMD) leading to NFIX haploinsufficiency, whereas NFIX mutations associated with MSS are clustered in exons 6–10 and escape NMD and result in the production of dominant‐negative mutant NFIX proteins. Thus, different NFIX mutations have distinct consequences on NFIX expression. To elucidate the in vivo effects of MSS‐associated NFIX exon 7 mutations, we used CRISPR‐Cas9 to generate mouse models with exon 7 deletions that comprised: a frameshift deletion of two nucleotides (Nfix Del2); in‐frame deletion of 24 nucleotides (Nfix Del24); and deletion of 140 nucleotides (Nfix Del140). Nfix +/Del2 , Nfix +/Del24 , Nfix +/Del140 , Nfix Del24/Del24 , and Nfix Del140/Del140 mice were viable, normal, and fertile, with no skeletal abnormalities, but Nfix Del2/Del2 mice had significantly reduced viability (p < 0.002) and died at 2–3 weeks of age. Nfix Del2 was not cleared by NMD, and NfixDel2/Del2 mice, when compared to Nfix +/+ and Nfix +/Del2 mice, had: growth retardation; short stature with kyphosis; reduced skull length; marked porosity of the vertebrae with decreased vertebral and femoral bone mineral content; and reduced caudal vertebrae height and femur length. Plasma biochemistry analysis revealed Nfix Del2/Del2 mice to have increased total alkaline phosphatase activity but decreased C‐terminal telopeptide and procollagen‐type‐1‐N‐terminal propeptide concentrations compared to Nfix +/+ and Nfix +/Del2 mice. Nfix Del2/Del2 mice were also found to have enlarged cerebral cortices and ventricular areas but smaller dentate gyrus compared to Nfix +/+ mice. Thus, Nfix Del2/Del2 mice provide a model for studying the in vivo effects of NFIX mutants that escape NMD and result in developmental abnormalities of the skeletal and neural tissues that are associated with MSS. © 2023 The Authors. JBMR Plus published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research.


Introduction
T he nuclear factor I/X (NFIX) gene (MIM #164005), (1)(2)(3) located on chromosome 19p13.2, (4) consists of 11 exons (Fig. S1) that encode 14 transcripts, of which 11 are protein coding. NFIX encodes a ubiquitously expressed transcription factor that forms part of the NFI gene family, which in mammals consists of NFIA, NFIB, NFIC, and NFIX. These transcription factors share a highly conserved N-terminal DNA binding and dimerization domain, which bind as homo-or heterodimers to the consensus palindromic sequence 5 0 -TTGGC(N5)GCCAA-3 0 present in the promoter regions of viral and cellular genes, (5) and a variable C-terminal transactivation/repression domain, which can potentially provide a range of preferential interactions with other proteins to either activate or suppress transcription. (6)(7)(8)(9) NFI transcription factors play important roles in the regulation of stem cell differentiation, quiescence, and differentiation during the development of organs that include lung, kidney, liver, blood, heart, skeleton, and the nervous system. (6,10) Heterozygous mutations in the NFIX gene can lead to two rare allelic disorders, Malan syndrome (MAL; MIM #614753) and Marshall-Smith syndrome (MSS; MIM #602535). (1,9) MAL is an overgrowth disorder, characterized by a slender habitus, long hands and advanced bone age, moderate to severe intellectual disability, unusual facial phenotype consisting of a long, triangular face with a prominent forehead, everted lower lip and prominent chin, and behavioral problems, which are usually dominated by anxieties and, less frequently, by aggression (9)(10)(11)(12) ( Table S1). The missense, nonsense, and frameshift NFIX variants reported in MAL patients predominantly affect exon 2 (Fig. S1), which encodes the highly conserved N-terminal DNA binding and dimerization domain of the NFIX protein. (1,3,13,14) Entire gene deletions and NFIX mutations observed in MAL patients are predicted to be cleared by nonsense-mediated mRNA decay (NMD) and lead to NFIX haploinsufficiency. (1-3, 11, 13, 14) MSS is characterized by short stature with skeletal abnormalities that may include kyphoscoliosis, abnormal bone maturation, craniofacial defects, and osteopenia and be associated with delays in motor and neural development that lead to moderate to severe mental retardation, limited or absent speech, and postnatal failure to thrive. (15,16) In addition, MSS patients may have distinctive facial features that include a high forehead, proptosis, blue sclera, anteverted nares, small and retracted mandible, gingival hypertrophy, and hypertrichosis (Table S1). MSS patients may also suffer from respiratory difficulties with upper-airway obstruction and apneas. The de novo frameshift NFIX mutations reported in MSS patients are all clustered in exons 6-10 of the NFIX gene, which encode the variable C-terminal transactivation/repression domain (Fig. S1). The different mutations result in the production of aberrant transcripts that escape NMD and lead to the production of dysfunctional truncated NFIX proteins, which are predicted to behave in a dominant-negative manner. (1)(2)(3) Thus, mutations that affect different regions of the NFIX gene have distinct consequences on the resulting transcripts and encoded proteins.
To date, only the in vivo consequences of Nfix exon 2 deletion, which encodes the conserved N-terminal DNA binding and dimerization domain, have been studied in mouse models. (17,18) ) In one study wherein Nfix exon 2 was replaced with an in-frame lacZ reporter gene, Nfix +/lacZ mice were reported to have normal survival, but reduced body weight, while Nfix lacZ/lacZ mice developed skeletal abnormalities due to defects in ossification that resulted in kyphosis and neurological abnormalities such as partial agenesis of the corpus callosum that was associated with hydrocephalus. (17) In other studies wherein an Nfix null allele was initially generated via Cre-recombinase-mediated excision of Nfix exon 2, the heterozygous Nfix +/À mice also had normal survival but with neurological abnormalities, (19) and the homozygous Nfix À/À mice had neurological defects that included dysgenesis of the corpus callosum but did not have skeletal abnormalities. (18) Moreover, Nfix À/À mice are reported to have severe delay in intermediate progenitor cells during forebrain development (20) and smaller muscle fibers with impairment of muscle regeneration despite the lack of skeletal defects (21) (Table S1). These Nfix-deficient mice with targeted deletions of exon 2 are reported to be representative of MAL. Therefore, to establish potential representative models for MSS, we generated Nfix mouse models with frameshift mutations in exon 7, which is the most commonly mutated exon in MSS patients. (1)(2)(3)

Study approval
All animal studies were approved by the Medical Research Council Harwell Institute Ethical Review Committee and were licensed under the Animal (Scientific Procedures) Act 1986, issued by the UK Government Home Office Department (PPL30/2433 and PPL30/3271).

Generation of mutant mice and genotyping analysis
Mice were generated using the CRISPR/Cas9 system, (22) and genotyping was performed by PCR amplification using genomic DNA and confirmed by RT-PCR using total RNA extracted, as described in Data S2 Materials and Methods.

Cell lines and in vitro expression assays
Murine embryonic fibroblast (MEF) cells and monkey kidney fibroblast (COS-7) cells that were used for RNA sequencing analysis or transiently transfected with wild-type (WT) or mutant murine Nfix cDNA expression constructs and luciferase reporter constructs were utilized for qRT-PCR, Western blot, and immunofluorescence analyses, as detailed in Data S2 Materials and Methods.

Phenotype analysis
Blood samples were collected and used for plasma biochemical analysis, (22) and skeletons and tissues of WT and mutant mice were prepared and used for imaging and histological analyses, as detailed in Data S2 Materials and Methods. (23) Results Establishment of mutant Nfix mouse models with targeted mutations of exon 7 To derive mouse models with frameshift mutations that affect the variable C-terminal transactivation or repression domain of the NFIX gene, the CRISPR-Cas9 system was used to target exon 7 of the murine Nfix gene. Following injection of Cas9 mRNA and Nfix guide RNA into C57BL/6J embryos, founder mice were generated from which three mutant Nfix lines comprising deletions of two nucleotides (Del2), 24 nucleotides (Del24), and 140 nucleotides (Del140) were established. More specifically, Nfix Del2 consists of a frameshift two-nucleotide deletion from position +49,580 to +49,581 relative to the translation start site (TSS), Nfix Del24 contains an in-frame 24-nucleotide deletion (from position +49,561 to +49,584 relative to the TSS), and Nfix Del140 contains a 140-nucleotide deletion (from position +49,577 to +49,716 relative to the TSS) and comprised 53 nucleotides from exon 7 and 87 nucleotides of intron 7 (Fig. S2A).
Effects of three exon 7 mutations (Nfix Del2, Nfix Del24, and Nfix Del140) on Nfix transcription and translation The differences in viability between the homozygous Nfix Del2/Del2 mutant mice and the homozygous Nfix Del24/Del24 and Nfix Del140/Del140 mutant mice suggested that the Nfix allelic variants may have different effects on the expression of this transcription factor. We therefore investigated the effects of these Nfix allele variants on the transcription and translation of Nfix. Murine Nfix contains 11 exons that encode eight transcripts (five of which are protein coding), due to alternative splicing of exons 7 and 9 and the use of different transcription initiation sites (ENSMUSG00000001911.16). Thus, alternative splicing may produce WT Nfix transcripts that retain exon 7 (Nfix long isoform) or shorter conserved isoforms that lack exon 7 (Nfix ΔEx7) (Fig. S3A). To study the effects of Nfix Del2, Nfix Del24, and Nfix Del140 deletions on splicing of exon 7, RT-PCR using primers located in exons 6 and 8 and Sanger sequencing were performed on total RNA obtained from MEFs derived from Nfix +/+ , Nfix +/Del2 , Nfix Del2/Del2 , Nfix +/Del24 , Nfix Del24/Del24 , Nfix +/Del140 , and Nfix Del140/Del140 mice. This revealed that Nfix +/+ MEFs had the Nfix WT long (317 bp) and short WT ΔEx7 (194 bp) isoforms ( Fig. S3B-D), but the Nfix Del2/Del2 and Nfix Del24/Del24 MEFs had mutant Nfix long isoforms of 315 and 293 bp, respectively, and the Nfix WT short isoform (ΔEx7 of 194 bp) (Fig. S3B,C). Nfix +/Del2 and Nfix +/Del24 MEFs were confirmed to express a WT Nfix long isoform, a mutant Nfix long isoform, and the Nfix WT ΔEx7 short isoform (Fig. S3B,C). In contrast, Nfix Del140/ Del140 MEFs had only the Nfix WT ΔEx7 short isoform, thereby suggesting that the 140-nucleotide deletion, which comprised 53 nucleotides of the 3 0 end of exon 7 along with 87 nucleotides from intron 7 that included the donor splice site, led to exon 7 skipping. Sanger DNA sequence analysis confirmed that the sequence of this Nfix short isoform from the Nfix Del140/Del140 MEFs matched the consensus murine sequence of the Nfix WT ΔEx7 short isoforms (Fig. S3D). Therefore, the 140-nucleotide deletion in the Nfix Del140 MEFs led to skipping of exon 7 and alternative splicing of exon 6 to exon 8 due to loss of a donor splice site, resulting in a  frameshift and the introduction of a stop codon after 81 amino acids, which corresponded to the WT short NFIX isoforms. However, the two-nucleotide deletion in Nfix Del2 MEFs resulted in a frameshift and the introduction of a premature stop codon after 65 amino acids, and the 24-nucleotide in-frame deletion in Nfix Del24 MEFs predicted the loss of eight amino acids (QGSSPRMA).
To further investigate the effects of the Nfix Del2 and Nfix Del24 mutations on Nfix transcription, translation, and cellular localization, in vitro expression assays in COS-7 cells transiently transfected with N-terminal-FLAG-tagged WT or mutant (Del2 or Del24) Nfix cDNA constructs that retain exon 7 were undertaken. Analysis by qRT-PCR showed that there was no significant difference in the amount of Nfix Del2 or Nfix Del24 expression compared to Nfix WT, suggesting that these mutations affecting the C-terminal part of the Nfix transcripts were not cleared by NMD mechanisms (Fig. 1A). Furthermore, Western blot analysis demonstrated the production of Nfix Del2 and Nfix Del24 smaller NFIX mutant proteins (<55 kDa), as expected, compared to WT (55 kDa) (Fig. 1B), thereby confirming that mutations in the C-terminal part of the Nfix gene produced truncated NFIX proteins. In addition, the expression of the NFIX Del2 protein was significantly decreased (p < 0.05), whereas that of the NFIX Del24 protein was significantly increased (p < 0.01) compared to NFIX WT, thereby revealing differences in the stabilities and likely degradations of the mutant proteins ( Fig. 1C). Immunofluorescence analysis showed that the cellular localization of NFIX Del2 and NFIX Del24 proteins was similar to the predominantly nuclear localization of NFIX WT (Fig. 1D). The Nfix Del140 mutation, which causes skipping of exon 7 to produce the Nfix WT ΔEx7 isoform, was not investigated in vitro.
To further assess the effects of the Nfix Del2 and Nfix Del24 mutations on NFIX transcription factor function, given that NFIX is reported to activate GFAP expression, (24) reporter constructs comprising the luciferase reporter gene downstream of the GFAP promoter were transiently cotransfected with WT or mutant Nfix cDNA constructs into COS-7 cells. WT NFIX activated the GFAP promoter and caused an approximately ninefold increase (n = 4, p < 0.0001, Fig. 1E) in luciferase activity in cells with the GFAP promoter cloned in the forward orientation compared to cells with the GFAP promoter cloned in the reverse orientation. Luciferase reporter activity was unaffected by the Nfix Del24 mutation (Fig. 1E,G) compared to WT NFIX (Fig. 1E,F). In contrast, the Nfix Del2 mutation caused a significant increase (n = 6, p < 0.05, Fig. 1E) in luciferase activity, in a threshold-dependent manner ( Fig. 1H) compared to WT NFIX, suggesting that the Nfix Del2 mutation had aberrant NFIX transactivation activity at the GFAP locus. Overall, these findings suggest that different frameshift mutations or in-frame deletions affecting the C-terminal part of the Nfix gene have distinct consequences on the activity of the resulting mutant NFIX proteins. Therefore, the phenotypes of the three mouse models harboring the Nfix allelic variants-Del2, Del24, and Del140-were further characterized for features of MSS.
Phenotypic characterization of Nfix Del2, Del24, and Del140 mice Heterozygous and homozygous Nfix mutant and WT littermates were characterized for features of MSS that included abnormalities of growth, skeleton, central nervous system (CNS), viscera, and plasma biochemistry.

Discussion
Our study reports the phenotypic characterization of three CRISPR-Cas9 generated mouse models with three allelic variants in Nfix exon 7, which is the most commonly mutated exon in MSS patients. These three allelic mutations, all of which affect the Cterminal regions of NFIX, have different effects on the phenotypes and on the expression of Nfix transcripts and proteins. Thus, of the Nfix Del2, Nfix Del24, and Nfix Del140 mouse models, only the Nfix Del2/Del2 mice developed postnatal skeletal and cranial defects, brain abnormalities, and likely dysfunction of the kidney and liver, which might potentially account for their premature deaths by 2-3 weeks of age, whereas the Nfix +/Del2 , Nfix +/Del24 , Nfix Del24/Del24 , Nfix +/Del140 , and Nfix Del140/Del140 mice were viable, normal, and fertile and survived to adulthood. These observations indicate that allelic variation, rather than potential off-target effects of the CRISPR-Cas9 system, could be responsible for the differences in phenotypes in the three Nfix mouse models, despite similar allelic mutations, the same environmental conditions, and identical genetic background, which is a common occurrence even on inbred backgrounds. (27,28) Only the two-nucleotide deletion in Nfix Del2 mice caused a frameshift and the introduction of a premature stop codon, which led to the production of intermediate levels of mutant NFIX Del2 protein with aberrant NFIX protein function that might potentially account for the more severe phenotype observed in the Nfix Del2/Del2 mice, while the 24-nucleotide in-frame deletion in Nfix Del24 mice caused the loss of eight amino acids, which could potentially be tolerated and is probably not as damaging as a frameshift mutation, whereas the 140-nucleotide deletion in the Nfix Del140 mice, comprising 53 nucleotides from exon 7 and 87 nucleotides of intron 7 including the splice donor site, caused skipping of exon 7 and alternative splicing of exon 6 to exon 8 to produce WT (normal) Nfix isoforms. This suggests that different frameshift mutations or in-frame deletions affecting the C-terminal part of the Nfix gene have different consequences on the transcripts and activity of the resulting proteins, thereby accounting for the different phenotypes in the three mouse models. Cell autonomous, monoallelic, and stochastic variation in gene expression, as well as functionally redundant paralogs, could also account for phenotypic variability. (25,26,(29)(30)(31) Redundant paralogs that are ubiquitously expressed in a partially overlapping manner and that recognize similar motifs may provide backup for one another in case of mutation by changing their expression pattern and acquiring new regulatory capabilities in order to compensate for the mutation. For example, NFIA, NFIB, NFIC, and NFIX have the same conserved N-terminal DNA binding and dimerization domain that enables all four related genes to recognize the same consensus sequence present in the promoter region of genes expressed in almost every organ, including the brain, lung, liver, intestine, and skeleton. Nfia À/À mice have CNS and kidney abnormalities (32) and die perinatally, (33) Nfib À/À mice have CNS and lung anomalies and die at birth, (34) while Nfic À/À mice have only a mild phenotype involving abnormal tooth development of the incisors and molars. (35) More recently, overlapping patterns of NFIA, NFIB, and NFIX expression have been reported in the brain. (36) NFIA, NFIB, NFIC, and NFIX, which were previously shown to interact with each other as well as other cofactors, bind the same regulatory motif of promoters of genes, such as brain fatty acid binding protein (B-FABP), GFAP, and inscuteable (INSC), and the NFIs or the ratio of the four NFIs have been shown to act either antagonistically or synergistically to regulate transcription in a promoter and context dependent manner. (37)(38)(39) Moreover, knockdown of one NFI member can affect the expression levels of other NFI members, suggesting cross-talks and possible compensation within the NFI family. (37) NFIX was also recently shown to act sequentially after NFIA and NFIB during gliogenesis within the spinal cord, and NFIB was reported to be able to activate Nfix expression in vitro, thereby suggesting autoregulatory mechanisms within the NFI gene family. (38) In this study, we have shown that the combination of NFI family expression might potentially influence the phenotypes of the Nfix mouse models. Nfia and Nfib, but not Nfic, change their expression pattern in order to possibly compensate for their respective Nfix Del2 and Del140 frameshift mutations in the unaffected Nfix +/Del2 , Nfix +/Del140 , and Nfix Del140/NfixDel140 mice, while Nfia, Nfib, and Nfic expression was unaltered in the unaffected Nfix +/Del24 and Nfix Del24/Del24 mice, thus suggesting that the in-frame Nfix Del24 mutations might potentially be tolerated and are probably not as damaging as a frameshift mutation. Moreover, the lack of functional redundancy from the Nfix paralogs in the Nfix Del2/Del2 mice as well as the presence of intermediate levels of aberrant mutant NFIX Del2 protein might possibly account for the more severe phenotype observed in the Nfix Del2/Del2 mice. Nfix Del2/Del2 mice represent a mouse model for MSS in which patients commonly have: reduced growth rate; short stature; craniofacial defects; osteopenia with increased fracture rate and kyphosis that normally worsens in puberty and adolescence and that is possibly aggravated by decreased bone density; (15) and anxiety and intellectual disability due to nonspecific rain abnormalities. (1,2) Thus, the Nfix Del2/Del2 mice had; short stature; reduced growth and TTM; kyphosis; shortened skull; marked porosity of the vertebrae; reduced BMC; shorter vertebrae height and femur length; reduced plasma CTX and P1NP concentrations but increased total ALP activity, indicative of abnormal bone function; and raised plasma urea and total bilirubin levels, suggestive of renal and hepatic dysfunction, which merits further investigation. Furthermore, Nfix Del2/Del2 mice had enlarged anterior cingulate, somatosensory and retrosplenial cortices, and ventricles but reduced dentate gyrus (Table S1). However, other features present in MSS patients, which include intellectual disability, airway obstruction leading to respiratory problem, umbilical hernia, cardiac anomalies, and abnormal bone maturation, (1-3, 15, 16, 40) were not assessed in the Nfix Del2/Del2 mice in this study. Plasma biochemistry in MSS patients is reported to be usually normal, and our findings of elevated urea and bilirubin concentrations and ALP activity in association with reductions in plasma CTX and P1NP concentrations in the Nfix Del2/Del2 mice may represent important differences to MSS patients, or it may be that such abnormalities do occur in MSS patients but have hitherto not been found. This latter notion is a possibility as exemplified by our experience. Thus, following our identification of likely renal dysfunction in the Nfix Del2/Del2 mice, ultrasound scan investigations were undertaken in two MSS patients and revealed the occurrence of renal cysts in both patients and nephrocalcinosis in one (Hennekam-personal communication). Moreover, the reduction in plasma CTX concentrations in the Nfix Del2/Del2 mice may suggest abnormal osteoclast activity and function, which merits further investigation. Moreover, the paradoxical increased plasma ALP activity, which is a marker of bone turnover, in association with reduced plasma concentrations of CTX and P1NP, which are markers of bone resorption and bone formation respectively, in the Nfix Del2/Del2 mice suggests additional extraskeletal origin for the raised ALP activity such as the kidneys or intestine, but not liver as mice, in contrast to humans, express little or no ALP in the liver, (41) and a search for additional renal or intestinal abnormalities in MSS may be warranted. Thus, it seems possible that MSS patients may have renal, intestinal, and hepatic dysfunction, and that there may be more similarities with the Nfix Del2/Del2 mice.
Our Nfix Del2/Del2 mice have similarities and differences when compared to two previous homozygous Nfix-deficient mouse models that had targeted deletions of exon 2 (17,18) (Table S1). Thus, homozygous Nfix-deficient mice (Nfix lacZ/lacZ ) were viable and had: growth retardation; an inability to fully open eyes; ataxic gait; feet-clasping posture when lifted by their tail indicating neurological abnormalities; gastrointestinal defects; brain malformations consisting of hydrocephalus and partial agenesis of the corpus callosum; defects in endochondral ossification, reduction in trabecular bone formation and calcification; thinning of cranial bones; kyphotic deformation of the spine; and early postnatal death between 3 and 4 weeks of age. (17) The other homozygous Nfix À/À mice showed; failure to thrive and grow when on a standard lab chow diet; delayed eye and ear opening; leg-clasping phenotypes indicating neuroanatomical defects; increased brain weight due to expansion of the cortex and entire brain along the dorsal ventral axis; aberrant neocortex, cerebellum, hippocampus, and spinal cord formation; and an abnormal ventricular cell population due to excessive generation of Pax6-expressing ventricular cells with hydrocephalus. (18,18,20,38,39,(42)(43)(44)(45)(46) Liver and kidney phenotypes were not assessed in these two previously reported Nfix-deficient mouse models, although it is important to note that Nfix lacZ/lacZ mice had gastrointestinal defects. (17) Importantly, the Nfix Del2/Del2 mice are not Nfixdeficient but instead have aberrant Nfix transcripts that escape NMD and lead to the production of mutant truncated NFIX protein, which is representative of MSS. Interestingly, MSS patients are heterozygous for NFIX mutations, and this contrasts with Nfix +/Del2 mice, which are normal, while developmental, skeletal, cranial, neural, hepatic, and renal abnormalities are observed in Nfix Del2/Del2 mice, which could account for their reduced viability. However, phenotypic differences between organisms are not uncommon and can be attributed to allelic variation, modifier genes, genetic variations, genetic background, environmental conditions, and reduced sensitivity of assays, such as behavioral assays, in animal models versus in patients. (29)(30)(31) For example, the autosomal dominant disorder spondyloepimetaphyseal dysplasia, Missouri type (SEMD MO ) in humans, is due to a heterozygous matrix metalloproteinase 13 (MMP13) missense F56S mutation, whereas heterozygous Mmp13 +/À mice deleted for exons 3, 4, and 5 have normal growth plates, but the homozygous Mmp13 À/À mice have defects in growth plate cartilage and delayed endochondral ossification. (47) In summary, in this study we report three Nfix mouse models with three different targeted mutations in exon 7 of the Nfix gene, which are representative of the most frequent NFIX mutations observed in MSS patients. The three mouse models, although being on the same genetic background, have differing phenotypes and viability. While the Nfix Del2/Del2 mice have some similarities to previously reported Nfix deficient mouse models, they also have a number of other phenotypes that are consistent with MSS. Further studies of the Nfix Del2/Del2 mice will help better understand the role of NFIX mutations that result in dominantnegative NFIX proteins and give rise to MSS, as well as provide useful resources for testing potential future treatments.