CRISPR Activation Reveals the Spliceogenicity of an Intronic NEB Variant in Fetuses With Arthrogryposis Multiplex Congenita 6

ABSTRACT

Whole‐genome sequencing identifies intronic variants whose pathogenicity can be predicted with tools like SpliceAI. However, an actionable classification of such variants may require RNA‐based validation, which can be limited by low expression in clinically accessible tissues. We report two fetuses from one family with Arthrogryposis multiplex congenita 6 (AMC6 [OMIM # 619334]) and biallelic NEB variants: a paternally inherited likely pathogenic frameshift variant, Chr2(GRCh38):g.151579391del; NM_001164508.2:c.16653del; NP_001157980.2:p.(Asp5552ThrfsTer5), and a maternally inherited intronic variant of uncertain clinical significance, Chr2(GRCh38):g.151496267G>A; NM_001164508.2:c.24486+9C>T; NP_001157980.2:p.(?). Because NEB is poorly expressed in fibroblasts, we used CRISPR activation to induce its expression in fibroblasts from the heterozygous mother. RNA‐sequencing subsequently confirmed that the intronic variant generated a novel splice donor site associated with inferred loss of splicing at the canonical donor site. After NMD‐inhibition, we could thus identify 45.5% of NEB transcripts with a 7 bp exon extension, predicted to result in a protein‐coding frameshift. The intronic variant was classified as likely pathogenic, allowing a genetic diagnosis.

CRISPR activation of NEB in fibroblasts, followed by RNA‐sequencing, documents spliceogenic effects of a NEB intronic variant. The assay enabled variant reclassification as likely pathogenic, providing molecular diagnosis in fetuses with Arthrogryposis multiplex congenita 6.

1. Introduction

Intronic variants, increasingly recognized as contributors to Mendelian diseases, are well identified by Whole‐genome sequencing (WGS), but without functional validation for example using RNA‐sequencing (RNA‐seq), these often remain variants of uncertain significance (VUS).

The Genotype‐Tissue Expression project indicates that nearly 20% of disease‐associated genes exhibit very low expression (transcript per million [TPM] lower than 0.5) across easily accessible cell types [1], which hinders reliable RNA‐seq analysis. CRISPR activation (CRISPRa) uses a catalytically inactive Cas9 (dCas9) fused to transcriptional activators to upregulate target genes, thus overcoming low endogenous expression [2]. This approach followed by RNA‐seq allows studies of the effect of potential splice‐altering variants [1, 3].

Pathogenic NEB variants cause Nemaline myopathy 2, autosomal recessive (NEM2 [OMIM # 256030]) and Arthrogryposis multiplex congenita 6 (AMC6 [OMIM #619334]), which differ in severity and timing of onset. NEB, encoding Nebulin, is selectively expressed in skeletal muscle.

We identified two fetuses with arthrogryposis from one family with NEB biallelic variants: a paternally inherited likely pathogenic variant Chr2(GRCh38):g.151579391del; NM_001164508.2:c.16653del; NP_001157980.2:p.(Asp5552ThrfsTer5) and a maternally inherited VUS Chr2(GRCh38):g.151496267G>A; NM_001164508.2:c.24486+9C>T; NP_001157980.2:p.(?). We used CRISPRa to induce NEB expression in fibroblasts of the mother, followed by RNA‐seq to resolve the pathogenicity of the VUS.

2. Case History

Subject 1 was a 27‐year‐old primigravida, who conceived via in vitro fertilization (IVF). Ultrasound examination at gestational week 18 + 3 revealed a female fetus with arthrogryposis; growth was age‐appropriate. There was generalized hydrops fetalis with marked nuchal edema (thickness 30 mm, abnormal > 6.0 mm), septations, ascites, and diffuse subcutaneous edema. During the scan fetal bradycardia was observed. The pregnancy was terminated at week 19. Autopsy confirmed AMC affecting all limbs (Figure 1A,B). Findings included fixed flexion contractures of the elbows with pterygium, flexion deformities of the wrists, hips, and knees, bilateral pes cavus, underdeveloped calf muscles.

FIGURE 1.

Pathological findings in Fetus 1. (A) AMC with facial dysmorphisms. (B) Nuchal edema and hyperflexion of large joints with webbing of skin behind the elbows and to some extent behind the knees. The fetus had pes cavus and slim calf muscles.

Hydrops fetalis was present, with fluid in the thoracic cavity and a prominent subcutaneous edema. The nuchal region showed marked thickening and edema. The fetus displayed anteverted nares, a long philtrum, retrognathia, hypertelorism, low‐set ears, and a short neck.

Subject 1 (27 years) subsequently had an IVF twin pregnancy; one male fetus showed increased nuchal translucency (4.0 mm, > 95th centile), multiple contractures, pes equinovarus, and hypokinesia, leading to selective reduction at 14 weeks without autopsy.

3. Materials and Methods

3.1. Genomics Analysis

Fetal DNA was extracted from umbilical cord tissue and parental DNA was extracted from peripheral blood. WGS libraries were prepared using the Illumina TruSeq DNA PCR‐Free kit and sequenced on an Illumina platform. Fastq files were analyzed as described in the Supporting Information.

3.2. RNA Analysis

Skin‐derived fibroblasts of Subject 1 and from an anonymous normal male neonate donor (ATCC CRL‐2097) were cultured in RPMI‐1640 medium supplemented with fetal bovine serum (10%), l‐glutamine (2 mM), penicillin (50 U/mL) and streptomycin (50 mg/mL). CRISPRa was performed as previously described [1] with minor modifications. Briefly, the fibroblasts were electroporated using program EO‐114 on the Amaxa 4D‐Nucleofector (Lonza) with in vitro‐transcribed Sp‐dCas9‐VPR mRNA (95 μg/mL) and sgRNA (50 μg/mL, Synthego) in P3 Primary Cell Nucleofector Solution (Lonza). The sgRNAs were designed using CRISPick [4] (https://portals.broadinstitute.org/gppx/crispick/public) (Table S1). The fibroblasts were harvested for RNA extraction 24 h after the electroporation. To inhibit nonsense‐mediated decay (NMD), the fibroblasts were treated with 100 μg/mL cycloheximide (CHX) for 6 h before RNA extraction [5]. Total RNA was extracted using the TRIzol Plus RNA Purification Kit with on‐column DNase treatment (Invitrogen).

RT‐qPCR was performed on a Lightcycler 480 using the iScript cDNA synthesis kit (BioRad) and RealQ Plus 2x Master Mix Green kit (Ampliqon) with intron‐spanning primer sets for NEB and GAPDH (Table S2). Relative RNA expression was estimated using the 2^−∆∆C _t method [6].

RNA‐seq libraries were prepared from ribosomal RNA‐depleted total RNA and sequenced on an Illumina platform. The RNA‐seq data were analyzed as described in the Supporting Information. Splice junctions were visualized with the Integrative Genomics Viewer version (IGV) 2.19.1.

4. Results

WGS data analysis revealed that the fetuses were compound heterozygous for two variants in NEB : (1) a paternally inherited frameshift variant Chr2(GRCh38):g.151579391del; NM_001164508.2:c.16653del; NP_001157980.2:p.(Asp5552ThrfsTer5) in exon 105, CADD PHRED 26.2, allele frequency 1.31 × 10⁻⁶, only in heterozygosity in gnomAD v4.1.0; (2) a maternally inherited intronic variant Chr2(GRCh38):g.151496267G>A; NM_001164508.2:c.24486+9C>T; NP_001157980.2:p.(?) in intron 173, CADD PHRED 18.2, allele frequency of 5.65 × 10⁻⁶, only in heterozygosity (Figure 2 A–C). The variants were submitted to ClinVar: ID SCV006099533 and SCV006099534. The list of variants left after filtering is in Table S3. The SpliceAI algorithm predicted the intronic variant to create a novel splice donor site, ∆‐score 0.86 (range 0–1, 0 low and 1 high), at chr2:151496269, 7 bp downstream of the canonical donor site, and to strengthen the splice acceptor site of exon 173, ∆‐score 0.35. The paternal frameshift variant was a Class 4 likely pathogenic variant (PVS1 and PM2), while the maternal intronic variant was a VUS (criteria PP3, PM2, PM3), according to current guidelines [7, 8]. To assess the pathogenicity of this VUS, we performed CRISPRa to induce NEB expression in fibroblasts (Figure 3 A). In Subject 1, heterozygous for the VUS NM_001164508.2(NEB):c.24486+9C>T, RNA‐seq showed a 7 bp insertion (NM_001164508.2(NEB):r.24486_24487ins24486+1_24 486+7), corresponding to the gain of a novel splice donor site 7 bp downstream of the canonical donor site in exon 173 (Figure 3 B). Use of this novel donor site was not observed in wild‐type control fibroblasts. NEB includes 183 exons with four regions of alternative splicing, and alone 20 different NEB transcripts have been identified from alternative splicing of exons 166–177 [9]. Thus, we assessed the use of the novel splice donor site in comparison with both canonical splicing and with alternative splicing. Here, use of the novel donor was supported by 45.5% of junction‐spanning reads in the heterozygous Subject 1 after CHX treatment, while both canonical and alternative splicing were reduced compared to the wild‐type control (Figure 3 C). With the caveat of low read counts and the possibility of incomplete‐NMD inhibition, this indicated that the variant may potentially have resulted in a gain of splicing replacing both canonical and alternative splicing. The effect of the resulting exon extension was a shift in the reading frame predicted to trigger NMD, due to the introduction of a premature termination codon in exon 174 (NP_001157980.2(NEB):p.Tyr8165GlyfsTer18). Because no SNV was located within the range of paired‐end short‐read sequencing, we were unable to quantify the allele‐specific splicing pattern and therefore could not rule out residual canonical splicing associated with the variant. However, we inferred from our observations that the variant most likely resulted in a near‐complete or complete loss of canonical splicing. Thus, the PVS1 (RNA) criterion was applied with reduced weight to strong, and the variant was re‐classified as likely pathogenic (criteria PVS1_S (RNA), PM2, PM3).

FIGURE 2.

Family and genetic findings. (A) Pedigree. (B and C) Bam files of the WGS data in IGV showing the paternally inherited frameshift and the maternally inherited intronic variant.

FIGURE 3.

RNA‐seq studies. (A) RT‐qPCR analysis of NEB expression after CRISPRa showing strong upregulation of gene expression in control fibroblasts including after CHX treatment. (B) Sashimi plots showing RNA‐seq of fibroblasts from control and Subject 1. Upper plots: Results without CHX treatment; lower plots: Results after CHX treatment. The number of junction‐spanning reads are shown for the canonical (black) and the novel splice donor site (red) 7 bp after exon 173. Junction‐spanning reads corresponding to alternative splicing of NEB are also visible, but their counts are not shown. (C) Illustration of NEB alternative splicing in Subject 1 and the control (top) and the relative frequencies of canonical splicing, normal alternative splicing and novel donor site usage quantified from the junction‐spanning reads (bottom).

5. Discussion

The clinical significance of intronic variants is challenging to establish in the absence of functional evidence. The ACMG/AMP and the ClinGen Sequence Variant Interpretation Splicing Subgroup made detailed recommendations on the application of ACMG/AMP codes to splicing‐related variants, while underscoring the importance of experimental RNA‐based assays, preferably performed on cells from affected individuals or carriers [7, 8]. However, tissue‐specific gene expression presents a barrier to such RNA‐based assays. To address this, strategies such as CRISPRa, transdifferentiation or the differentiation of patient‐derived induced pluripotent stem cells (iPSCs) may be employed. Clinical examples of the application of CRISPRa to assess the pathogenicity of genetic variants are so far few; however, CRISPRa seemed to work for most disease genes using one of the available transactivation platforms [3]. This makes CRISPRa a promising strategy to assess putative spliceogenic variants, while an advantage of applying mRNA‐based delivery of the transcriptional activator is that the same reagents can be applied to any gene by simply replacing the readily available target guide RNAs [1].

We acknowledge certain limitations of the study. (1) Although we achieved more than a 100‐fold upregulation of NEB using the CRISPRa system based on mRNA delivery of the transcriptional activator, the RNA‐seq read depth of NEB was quite limited, reducing the quantitative accuracy of the analysis. (2) Ex vivo RNA analysis in fibroblasts may not always reflect in vivo splicing patterns in skeletal muscle. (3) We were unable to perform allele‐specific quantification to discriminate between the variant and wild‐type alleles, which weakens the functional evidence. Thus, we could not rule out any residual canonical splicing associated with the variant allele and downgraded the strength of the ACMG/AMP loss‐of‐function criterium accordingly. (4) Ideally, multiple sex‐ and age‐matched wild‐type controls should have been included [8]. However, observing usage of this novel splice donor site would be highly unlikely in wild‐type controls, because the intronic variant generated a novel canonical splice donor motif absent from the human reference genome. With CRISPRa, wild‐type controls are a particular challenge since RNA‐seq data cannot be reused for new target genes. (5) The assessment of variant impact was complicated by alternative splicing of NEB. However, disease‐associated frameshift variants are described in the alternatively spliced regions of NEB, including several variants in exons 105 and 173 registered in ClinVar. These variants are expected to cause disease via a reduction in nebulin isoform diversity. In fact, homozygous frameshift variants in ubiquitously expressed exons of NEB are considered not tolerated unless at the very end of the gene [10].

In conclusion, by utilizing the CRISPRa system, we demonstrate the spliceogenicity of a novel intronic NEB variant, establishing a molecular diagnosis of AMC6 in two fetuses. Given the profound impact of AMC on lifespan and quality of life, early detection through prenatal ultrasound combined with genetic diagnosis provides essential information to support genetic counselling and informed parental decision‐making.

Author Contributions

Concept and design: D.M., T.T., U.B.J., and E.F. Drafting the manuscript: D.M. and T.T. Clinical examinations: S.M.B. and V.C.O. WGS study: D.M., P.M.B., and M.E.S. RNA‐sequencing: T.T. and UBJ.

Ethics Statement

The study was approved by the Norwegian Regional Ethical Committee (REK 2020/73834).

Consent

The parents gave written informed consent to perform genetic testing, publish clinical information and photos.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: Guide RNAs for CRISPRa.

Table S2: Primers for RT‐qPCR.

Table S3: cge70097‐sup‐0002‐TableS3.xlsx.

Misceo D., Terkelsen T., Bøen Keim S. M., et al., “ CRISPR Activation Reveals the Spliceogenicity of an Intronic NEB Variant in Fetuses With Arthrogryposis Multiplex Congenita 6,” Clinical Genetics 109, no. 4 (2026): 772–777, 10.1111/cge.70097.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.