Childhood-Onset Refractory Hypertension Results from Neurofibromatosis Type 1 Caused by a Splicing NF1 Mutation

Neurofibromatosis type 1 (NF-1; OMIM #162200) is one of the most common autosomal dominant heritable cancer predisposition syndromes, affecting about 1 in 2,000–3,000 individuals worldwide [1‒3]. NF-1 results from inactivating mutations in the NF1 gene that encodes neurofibromin, a large protein (approximately 250 kDa) with 2,818 amino acids that plays a role in upstream negative regulation of RAS-guanosine triphosphatases. The pathophysiology of NF1 is attributed to the loss of the negative RAS regulator neurofibromin, activating downstream cascade-signaling pathways including the MAPK/extracellular-signal-regulated kinase/mitogenic extracellular signal-regulated kinase (MEK/ERK) pathway and the phosphoinositide 3-kinase/protein kinase B/mammalian target of rapamycin (PI3K/Akt/mTOR) pathway [4].

NF1, located at 17q11.2, is a large gene that spans approximately 350 kb of genomic DNA and contains 60 exons. Complex and alternative splicing patterns in NF1 have resulted in many isoforms of neurofibromin [5]. To date, nearly 3,890 DNA variants of NF1 are listed in the Human Gene Mutation Database (HGMD, Institute of Medical Genetics, Cardiff, UK; http://www.hgmd.org/). Approximately 50% of the reported mutations are de novo mutations, and postzygotic NF1 mutations that can lead to a mosaic phenotype have also been reported [6]. Most of the pathogenic mutations have produced truncated neurofibromin proteins, and 30% of them are splicing mutations that affect mRNA processing [7, 8]. Destruction of the 3′ or 5′ splice site caused by splicing variants results in abnormal neurofibromins including exon skipping and intron retention [9]. Although in silico tools have been used to predict the pathogenicity of splicing mutations, the results need to be confirmed by, for example, in vivo splicing analysis or in vitro minigene assay experiments.

The clinical phenotype of NF1 is progressive with complete penetrance. Café-au-lait macules (CALMs), axillary and inguinal freckling, and subcutaneous neurofibromas are common features of NF-1, and brain tumors (optic pathway gliomas and glioblastoma), skeletal dysplasia, and malignant peripheral nerve sheath tumors are less common but more serious complications [10]. Cardiovascular abnormalities, including arterial stenosis and aneurysms, are under-recognized complications that are associated with increased mortality and deserve close attention in patients with NF-1 [11]. Early diagnosis of NF-1 and regular screening of high-risk complications can help improve the prognosis and decrease the heavy disease burden. Conventionally, the clinical diagnosis of NF-1 has been based on diagnostic criteria outlined in a National Institutes of Health (NIH) Consensus Development Conference [12]. However, the variable and complex phenotypes of NF-1 make the clinical diagnosis of this disease challenging. More recently, next-generation sequencing methods have facilitated the development of molecular diagnosis approaches that have been applied to the diagnosis and management of NF-1.

Here, we identified a splicing mutation c.479+1G>C in NF1 in a Chinese pedigree that we detected using next-generation sequencing combined with Sanger sequencing. We used a minigene splicing assay to verify the consequence of the mutation, further contributing to the understanding of the pathogenesis of NF-1.

The proband, an 18-year-old male, was referred to Fuwai Hospital (Beijing, China) for refractory hypertension. Previously, he was diagnosed with hypertension, left renal artery, and superior mesenteric artery stenosis and underwent a percutaneous transluminal renal angioplasty (PTRA) procedure. However, his blood pressure (BP) still fluctuated in the range 150–180/100–110 mm Hg. During hospitalization, a detailed physical examination, imaging, and laboratory tests were performed. Six of the proband’s family members were also recruited for genetic and clinical characteristics analyses (Fig. 1a).

This study was approved by the Ethics Committee of Fuwai Hospital and carried out in accordance with the Declaration of Helsinki. Each participant or authorized representative signed written informed consent forms in advance.

Peripheral blood samples were retrieved from the enrolled subjects and genomic DNA was extracted using a QIAamp DNA Mini Kit (Qiagen, Valencia, CA, USA). Next-generation sequencing of the proband’s sample was performed using a monogenic hypertension-related-gene panel (online suppl. Table 1; for all online suppl. material, see https://doi.org/10.1159/000533144). The sequenced data were aligned to the reference human genome (GRCh37/hg19, https://www.ncbi.nlm.nih.gov/assembly/GCF_000001405.13/), then single-nucleotide variants, insertion/deletion variants, and copy number variants were identified and analyzed as described previously [13, 14]. SpliceSiteFinder-like (v3.1; http://umd.be/Redirect.html), NNSPLICE (v0.9; https://www.fruitfly.org/seq_tools/splice.html), MaxEntScan [15], Human Splicing Finder [16], and NetGene2 (https://services.healthtech.dtu.dk/service.php?NetGene2-2.42) were used to predict the effect of the genetic variant on splicing. Sanger sequencing was used to verify the possible pathogenic mutations in the family members’ samples.

Briefly, the amplified target sequences contained exons 3–6 of human NF1 (exon 3-intron 3-exon 4-intron 4-exon 5-intron 5-exon 6) and the mutant fragment of NF1 was induced by site-directed mutagenesis (JiKai Gene, Shanghai, China). These fragments were inserted into GV219 reporter vectors with restriction endonucleases Xhol/Kpnl. Sanger sequencing was performed to verify the successful construction of the recombinant wild-type (WT) and mutant-type (MT) expression vectors (Fig. 2).

Human HEK293T cells (Cell Resource from National Center for Cardiovascular Diseases, Beijing, China) were routinely cultured in Dulbecco’s Modified Eagle’s Medium (Invitrogen, Waltham, MA, USA) containing 10% fetal bovine serum (Gibco, Carlsbad, CA, USA). Lipofectamine 3000 reagent (Invitrogen, Waltham, MA, USA) was used for transient plasmid transfection in accordance with the manufacturer’s protocol. Cells were incubated for 24 h after transfection, then harvested. Total RNA was extracted from the cells using TRIzol reagent (Invitrogen) and reverse-transcribed using PrimeScriptRT Master Mix (Takara, RR036A). Complementary DNA (cDNA) sequences were amplified using 2× PCR Master Mix (Beyotime, Shanghai, China). The products were purified by 2.4% agarose DNA gel electrophoresis and sequenced. The primers for the PCR amplification were as follows: forward, 5′-AGA​ATA​TTT​GGA​GAA​GCT​GCT​G-3′; reverse, 5′-ATA​ACT​GCT​AAC​TGC​GCA​ACC-3′.

We conducted a homology search in NCBI database using the WT sequence of NF1 as the query sequence and found a 100% homologous template of NF1 (Protein Data Bank ID 7R03) [17]. A three-dimensional structure model of WT neurofibromin was generated using SWISS-MODEL [18], and RoseTTAFold was used to predict the three-dimensional structure of MT neurofibromin [19]. We used PROCHECK to evaluate the optimized WT and MT neurofibromin models [20].

The proband was born with ten CALMs that were distributed in the armpit, groin, neck; the largest spot was 4 mm in diameter after puberty. A few freckles were scattered around the armpits, groin, and neck area. At 6 years of age, the proband complained of intermittent headache, had nausea and vomiting symptoms, and was diagnosed with hypertension (BP 200/160 mm Hg). After screening the etiology of early onset hypertension, he was found to have renal artery and superior mesenteric artery stenosis. At 12 years of age, the proband underwent a PTRA procedure to reduce his BP. More recently, despite of taking metoprolol and amlodipine, his BP fluctuated in the range 150–180/100–110 mm Hg. During hospitalization, an aortic computerized tomography scan showed moderate stenosis in the proximal segment of the left renal artery and severe stenosis in the initial segment of the superior mesenteric artery (Fig. 3). The echocardiographic findings suggested left ventricular wall hypertrophy, positive motor stimulation test, left ventricular outflow tract, and left ventricular midsection obstruction. The ophthalmologic examination revealed bilateral Lisch nodules and hypertensive retinal artery stenosis. BP monitoring of extremities found BP readings of 178/97 mm Hg (right upper limb), 170/85 mm Hg (left upper limb), 181/87 mm Hg (right ankle), and 186/81 mm Hg (left ankle). The results of the biochemical examinations of the proband’s sample are list in Table 1. Mental retardation, hearing impairment, skeletal abnormalities, or subcutaneous neurofibromas were not observed in the proband. Essentially, the patient met the NIH clinical diagnostic criteria for NF-1 [12].

No abnormal clinical features were detected among the proband’s family members, except for his mother who had freckles on both cheeks and had developed glioma at 22 years of age. She died of glioma when she was 25 years old.

A heterozygous splice site mutation c.479+1G>C in NF1 was found in the proband but not in any of the other family members. The Sanger sequencing results are shown in Figure 1b. This mutation resulted in a nucleotide G>C transversion in the donor splice site of intron 4. The SpliceSiteFinder-like, NNSPLICE, MaxEntScan, Human Splicing Finder, and NetGene2 in silico analysis predicted that c.479+1G>C may affect splicing products of pre-mRNA.

We performed a minigene splicing assay after the splice site mutation was confirmed by Sanger sequencing. After 2.4% gel extraction, we sequenced various PCR products of WT and MT minigenes, and the results showed that exon 4 was missing in the MT minigene product (Fig. 4a, b). The mutation disrupted the normal splice donor site, resulting in the skipping of exon 4 (c.289_479del), a shift in the open reading frame, a premature termination code, and finally a truncated protein (p.Q97Vfs13).

The predicted three-dimensional models of WT and MT neurofibromin are shown in Figure 4c. The mutation (p.Q97Vfs13) resulted in the absence of most of the domains that were present in WT neurofibromin, namely the cysteine/serine-rich, GAP-related, Sec14-homologous, and pleckstrin homology domains. This finding provides new evidence for the loss-of-function variant of neurofibromin.

We identified a splicing mutation c.479+1G>C in the NF1 gene of a Chinese male patient with NF-1 who presented with multiple CALMs, scattered freckles, and refractory renovascular hypertension by next-generation sequencing. We performed in silico analysis to predict the potential harm of the mutation, then conducted a minigene splicing assay together with three-dimensional structural modeling to elucidate the pathogenesis of NF-1. We found that the c.479+1G>C mutation disrupted a normal splice donor site and led to the skipping of exon 4, which resulted in a truncated protein product (p.Q97Vfs13) in which most neurofibromin domains were missing.

Cutaneous presentations such as dermal neurofibromas and pigmentation are the most common features of NF-1 with early penetrance, whereas cardiovascular complications are non-neoplastic and easily overlooked but have significant manifestations that can potentially lead to catastrophic outcomes and reduced average life expectancy [21‒23]. Hypertension is frequently detected in patients with NF-1 and is associated with essential hypertension, renal or aortic vasculopathy, and pheochromocytoma/paraganglioma [24]. Renal vasculopathy is the primary cause of refractory renovascular hypertension in patients with NF-1, accounting for approximately 1% of patients [25, 26]. Oderich et al. [27] studied vascular abnormalities in 31 patients with clinical NF-1 and found that aneurysms or stenoses lesions in the aortic, renal, and mesenteric arteries were the most common of these abnormalities. Although the relative success rate of endovascular intervention was lower in patients with NF-1 than in those without NF-1, PTRA is still the first-line therapy for patients with NF-1 and refractory hypertension caused by renal artery vasculopathy whose BP was not controlled by oral antihypertensives, and repeated procedure is advised if initially unsuccessful [22, 28]. In this study, the proband initially presented with CALMs and refractory hypertension secondary from right renal artery stenosis. Superior mesenteric artery and hypertrophic cardiomyopathy were also detected; the latter may have resulted from long-term uncontrolled hypertension. However, despite the restenosis of renal artery appeared after PTRA procedure, the proband refused to accept a repeat PTRA because of the high cost. For NF-1 children or adolescents presenting with refractory hypertension, we recommend considering renal artery Doppler ultrasound or computed tomography angiography to detect renal vasculopathy. Timely medication or endovascular intervention can alleviate the disease, and regular follow-ups are necessary to monitor progression.

Neurofibromin is a cytoplasmic guanosine triphosphatase-activating protein that acts as a negative regulator of the RAS proto-oncogene. The correlation between vessel wall homeostasis and downstream signaling pathway activity can partially explain the pathogenesis of vasculopathy induced by NF-1. Neurofibromin-deficient vascular smooth muscle cells (VSMCs) or monocytes/macrophages were shown to increase cell proliferation and migration in response to growth factor or inflammatory states, and this response was regulated by over-activation of the RAS-MEK/ERK signaling axis, consistent with the histology of vascular disorders in NF-1 [29, 30]. Xu et al. [31] generated mice with homozygous inactivation of NF1 in VSMCs and identified increased activity of mitogen-activated protein kinase after vascular injury followed by enhanced VSMC proliferation and marked intimal hyperproliferation. In 2010, Lasater et al. [32] demonstrated that specific NF1 heterozygous bone marrow-derived hematopoietic cells were major drivers of the amplification of neointima formation by accelerating the accumulation of macrophages in NF1^+/− mice. NF1^+/− monocytes/macrophages produced increased numbers of reactive oxygen species, which induced the proliferation of VSMCs [33]. Moreover, NF1^+/− monocytes/macrophages were found to mobilize the enhanced release of the chemokine receptor CCR2 into peripheral blood [34]. Chemokine ligand (CCL2; also known as monocyte chemotactic protein-1) is the ligand of CCR2, and elevated CCL2 was shown to promote the proliferation and migration of macrophages and activate RAS kinases [34]. By targeting these mechanisms using MEK inhibition [35], NADPH oxidase-2 inhibitor [33], or CCR2 antagonist [34], the neointima was alleviated with satisfactory effects in animal models. Understanding the exact mechanisms of vasculopathy in NF1 will contribute to the development of precision prevention and therapy of this disease; however, such an understanding requires further research.

In this study, we identified a splicing mutation c.479+1G>C in NF1 in the proband. Although his mother died of glioma without genetic testing, her freckle feature and malignant tumor history make it reasonable to speculate that the pathogenic mutation was inherited from his mother. Although this splicing mutation has been reported in a French cohort with NF-1 [36], we carried out an in vitro functional analysis for the first time and found that the mutation resulted in the skipping of exon 4. This frame shift mutation led to the production of a truncated neurofibromin protein that was 108 amino acids long and was absent of major domains, including the cysteine/serine-rich and GAP-related domains. Splicing mutations in NF1 are common, accounting for approximately 30% of the pathogenic variants identified in patients with NF-1 [7, 37, 38]. All splicing mutations that led to exon 4 skipping are summarized in Table 2. Alkindy et al. [38] summarized genotype-phenotype associations in 149 patients with NF1 and suggested there was an increased risk of brain gliomas and malignant peripheral nerve sheath tumors in the patients who carried splice site mutations, which confirms our recommendation that the progression of tumorigenesis should be closely monitored in the proband.

Molecular diagnosis has important implications in the diagnosis of NF-1, especially in patients who are paucisymptomatic or do not satisfy the clinical diagnostic criteria [41, 46]. Precise diagnosis is needed to improve the prognosis and for genetic counseling. If either parent carries a pathogenic NF1 mutation, prenatal diagnosis is recommended to prevent the birth of infants with the same mutation [10]. Moreover, identification of germline NF1 mutations could be valuable for genotype-phenotype correlation analysis to better guide the management of NF-1. Cost-effective methods and protocols for genetic diagnosis of NF1 are now available, including whole-genome sequencing [47], NF1-targeted panel sequencing [48], cDNA analysis combined with multiplex ligation-dependent probe amplification of genomic DNA [49], and NF1 transcriptome analysis [50]. However, because of the complexity and large size of the NF1 gene, sensitive localization of pathogenic variants is challenging and will probably require the combination of multiple detection technologies [47].

In conclusion, we identified a splicing mutation c.479+1G>C in a Chinese patient with NF-1 by next-generation sequencing. A minigene splicing assay showed that this mutation led to the skipping of exon 4, an open reading frame shift, and termination codon at 108, resulting in a truncated neurofibromin protein (p.Q97Vfs13). Taken together, our results highlight the importance of genetic testing in the diagnosis of NF-1 and confirm that the minigene splicing assay is a useful strategy for analyzing mRNA expression outcomes caused by donor splicing site variation.

We thank all participants for taking part in this study.

This study was approved by the Ethics Committee of Fuwai Hospital, approval number 2020-1321. All of the participants provided written informed consent.

The authors have no conflicts of interest to declare.

This work was supported by CAMS Innovation Fund for Medical Sciences (CIFMS, 2022-I2M-C&T-A-010 and 2022-I2M-C&T-B-041).

Conception and design: Yi-Ting Lu, Xian-Liang Zhou, and Peng Fan; administrative support: Xian-Liang Zhou and Peng Fan; provision of study materials or patients: Di Zhang, Kun-Qi Yang, and Tao Tian; collection and assembly of data: Yi-Ting Lu, Di Zhang, Buweimairemu·Rejiepu, and Dong-Cheng Cai; data analysis and interpretation: Yi-Ting Lu, Buweimairemu·Rejiepu, Dong-Cheng Cai, and Kun-Qi Yang; manuscript writing and final approval of the manuscript: all authors.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.