Clinical characteristics and spectrum of NF1 mutations in 12 unrelated Chinese families with neurofibromatosis type 1

Neurofibromatosis type 1 (NF1; MIM: 162200) is one of the most common autosomal dominant inherited diseases with an incidence of 1 in 2500–3000 individuals [1]. Caused by a germline heterozygous mutation in the tumor suppressor gene neurofibromin 1 (NF1; MIM: 613113) located on chromosome 17q11.2, NF1 is characterized by typical café-au-lait spots and cutaneous neurofibromas [2]. Individuals with NF1 are predisposed to plexiform neurofibromas, axillary and inguinal freckling, Lisch nodules of the iris, benign and malignant tumors, and renal artery stenosis, among a list of other abnormalities [3]. Although NF1 is a classical monogenic disease with complete penetrance by adulthood, clinical symptoms may vary in patients who come from the same family, or even for the same patient at different life stages. Complex though the clinical manifestations of patients may be, individuals in this study were diagnosed as NF1 only when they met two or more of the National Institutes of Health Diagnostic Criteria for NF1 [4].

NF1 is one of the largest known genes with a genomic size of 282 kb, consisting of 57 constitutive exons and three alternatively spliced exons [5]. Owing to its extremely frequent incidence of mutation (circa 1 in 10,000 gametes per generation) without obvious mutational hot spots, over 2600 NF1 mutations have hitherto been reported in the Human Gene Mutation Database (HGMD). Single nucleotide substitutions and small deletions (20 bp or less) account for 71% of currently known mutations. Moreover, approximately half of all NF1 cases are de novo mutations [6]. In addition, the large size of the NF1 gene, the existence of homologous pseudogenes dispersed on other chromosomes [7], the diversity of mutation types and positions, and the great variety of lesions make traditional mutation detection in patients with NF1 a complicated, time-consuming and laborious process [8]. With the superiority of being high throughput and its rapidity, the next-generation sequencing can make up for any deficiency in the single Sanger sequencing method to some extent. In addition, multiplex ligation-dependent probe amplification (MLPA) for the detection of copy number was incorporated in our methods. Hence, we adopted various approaches such as Sanger sequencing, targeted next-generation sequencing, and MLPA so as to overcome challenges in the detection of NF1 mutations in patients.

Despite several reports with regard to genotype–phenotype correlations in patients [9‐11], the underlying causes of sophisticated clinical manifestations among patients have not yet been elucidated [12]. Nevertheless, the causes of polymorphisms in genotype–phenotype correlations may be assigned to modifier genes, gender, loss of heterozygosity (based on the two-hit hypothesis) [13], tumor microenvironment, and heterogeneity in the regulation of signaling pathways [3]. Consequently, it is of great significance to identify the causative mutation and assess the prognosis of NF1 patients, if genotype–phenotype correlations can be clarified, before the onset of symptoms.

In brief, we examined 12 non-consanguineous Chinese families from which patients were diagnosed with NF1. A molecular diagnosis and clinical characterization of NF1 patients were undertaken to identify causative mutations and evaluate any correlations between genotype and phenotype.

In this study, we utilized a synthetic method of Sanger sequencing, targeted next-generation sequencing, and MLPA to detect potential mutations in patients. We also investigated the clinical presentations of each patient with NF1 to elucidate the factors associated with the severity of the disease phenotype.

As a consequence, 11 different mutations scattered in different exons of NF1 were identified in 12 unrelated Chinese families with NF1, suggesting a positive detection rate of 92% (11/12). Though no mutational hot spot was discovered, our attention was drawn to the observation that about three quarters of the mutations identified were truncating mutations. Consequently, a truncated neurofibromin with a partial or absolute functional loss may be produced, or the protein may become degraded as a result of its abnormal termination, resulting in the inactivation of the negative regulatory protein. Therefore, the pathogenicity of the four novel truncating mutations that we identified is proverbially acknowledged because of a prematurely disrupted reading frame. Furthermore, we detected three different missense mutations that had been previously reported [21‐26]. Nonetheless, multiple sequence alignment and in silico analysis were still performed to authenticate their pathogenicity (Additional file 1: Figure S3 and Table S2). Moreover, for Family 12 in whom a causative mutation was not found by either targeted next-generation sequencing or MLPA, we conjectured the existence of a deep intronic mutation.

Maruoka et al. [27] described a c.6853_6854insA (hg19; NM_001042492.2) NF1 mutation that was identical to the mutation we found in Family 11 (Table 2). However, we thought a description of c.6791dupA (hg19; NM_000267.3) rather than c.6790_6791insA may be more appropriate in keeping with the ​standard human sequence variant nomenclature of the ​Human Genome Sequence Variation Society (HGVS) [28] and Mutalyzer website (version 2.0.23; Leiden University Medical Center, Leiden, the Netherlands) [29]. Likewise, the amino acid change in the c.1754_1757delTAAC mutation in Family 9 (Table 2) should be depicted as p.T586Vfs*18 [30], instead of p.L585fs*18 [31].

According to our observations, clinical manifestations in patients who had the same mutation within a family, and the same patient at different stages of their life, may be highly discrepant. Additionally, with regard to the polymorphisms in genotype–phenotype correlations, possible causes may be modifier genes, gender, or heterogeneity in the regulation of signaling pathways, to name a few.

Alterable though the phenotype of NF1 may be, the progression of NF1 was ascribed to age and pregnancy according to a long-term follow-up of several contactable patients. In general, it was older patients who usually manifested severer symptoms, while prepubertal patients displayed a comparatively mild phenotype with only café-au-lait spots. For example, Patient 7’s symptoms gradually worsened in the form of an increase in the number and size of cutaneous neurofibromas during a 10-year observation period (from 2007 to 2017; Fig. 2). Patients 1, 7, 16, 20 and 22 recalled that cutaneous neurofibromas appeared and symptoms became worse in their adolescence. Furthermore, some females with NF1 (Patients 1, 8, 10 and 18) complained of a marked exacerbation of disease after pregnancy (Fig. 2), as also described by Griffiths et al. [32]. These observations indicated that a patient’s physical condition, particularly their hormone levels, played a vital role in the development of NF1.