Background

Noonan syndrome (NS) is a genetic multisystem disorder with a prevalence of 1 in 1000–2500 live births (1). This condition is characterised by varying developmental delays, distinctive facial features, congenital heart defects and short stature; clinical diagnosis is often based on these features (2). NS is part of a group of phenotypically similar developmental disorders (RASopathies), caused by germline variants in the genes within the RAS/MAPK signalling pathway (3).

There are multiple genes that cause NS, all linked to the RAS/MAPK signalling pathway (4, 5). Fifty percent of individuals with NS have a germline pathogenic variant (PV) in PTPN11 (2). Other reported genes include SOS1, RAF1, ROT1 and KRAS, occurring in 13%, 5%, 5% and < 5% of cases, respectively (2). PVs in other genes, including BRAF, MAP2K1 and NRAS, have been identified in less than 1% of affected individuals (2). Another gene, LZTR1, has been recently identified as a causative gene in RASopathies (6); however, its precise role in the RAS/MAPK signalling pathway is less defined. Additionally, germline pathogenic variants in LZTR1 have been identified in patients with schwannomatosis and is thought to be a distinct entity as schwannomas are not frequently seen in NS (7, 8). There is a paucity of cases to address this genetic and phenotypic heterogeneity in LZTR1 carriers.

LZTR1 (OMIM 600574), encoding leucine zipper-like transcription regulator 1 (LZTR1), was proposed as a tumor suppressor gene belonging to the BTB-Kelch superfamily (9). LZTR1 is a Golgi protein and belongs to the BTB-Kelch superfamily (7, 9), where it is reported to be involved in apoptosis (9) and acts as a substrate-specific adaptor for the Cullin-3 (Cul3) ubiquitin ligase (10, 11). Using a computational platform, LZTR1 was identified as a tumor suppressor gene, with somatic mutations in this gene driving glioblastoma (10). Somatic mutations in LZTR1 have also been associated with liver cancer (12).

Recent investigations have provided more insight into its potential role in the RAS/MAPK signalling pathway. LZTR1 binds to the RAF1/SHOC2/PP1CB complex and promotes RAF1 Ser259 phosphorylation, leading to MAPK signalling pathway inactivation (13). Additional data reports that LZTR1 enables the polyubiquitination and degradation of endogenous RAS, which ultimately inhibits RAS/MAPK signalling (14). Two other studies suggest that LZTR1 mediates RAS ubiquitination and MAPK pathway activation, contributing to the development of human disease (15, 16). Examination of LZTR1 variants associated with NS suggest this gene is functionally-linked to the RAS/MAPK pathway by negatively controlling RAS protein levels and MAPK signalling (17). In addition, a biological relationship has been proposed between LZTR1 and RIT1, whereby pathogenic mutations affecting RIT1 or LZTR1 leads to RIT1 accumulation and contributes to hyperactivation of MAPK signalling (18).

While the association of germline LZTR1 variants with human disease is still being elucidated, germline loss-of-function mutations in LZTR1 predispose to schwannomatosis (7, 19, 20) and NS (2). NS has wide genetic heterogeneity and clinical variability (21). Inheritance most frequently occurs in an autosomal dominant (AD) manner, however, PVs in LZTR1 leading to NS can also be inherited in an autosomal recessive (AR) manner (2). Expression experiments suggest that LZTR1 variants that cause AD NS may not be gain-of-function, whereas variants in patients with AR NS may have a loss-of-function effect (13). Loss-of-function mutations in biallelic LZTR1 variants include splice site, frameshift, nonsense and missense modifications (17). Experiments assessing the functionality of missense LZTR1 mutations suggests that schwannomatosis-associated LZTR1 mutations act heterogeneously in modifying RAS-MAPK signalling, similar to variants causing dominant NS (17). The pleotropic function of LZTR1 may explain the multiple inheritance patterns (AR, AD) of LZTR1-associated NS. NS has been associated with Neurofibromatosis type 1 (NF1) [Neurofibromatosis-Noonan syndrome (NFNS)], a rare disorder where individuals present with phenotypic characteristics of these two autosomal dominant conditions (22). While there was an earlier debate about whether NFNS is a separate genetic condition, more recent reports illustrated clinical findings of both disorders, providing support that this is a new syndrome (23). NFNS is mainly caused due to mutations in NF1 (24–28), however, there has occasionally been a PTPN11 mutation reported in addition to the NF1 gene mutation (29, 30). The involvement of other genes within the RAS-MAPK signalling pathway that cause NS have not yet been explored in NFNS (31). As such, NS should be considered part of the differential diagnosis in NF1 patients.

In this report, we describe a patient with clinically diagnosed NS who presented with sacral plexiform nerve sheath tumors, suggestive of neurofibromas or schwannomas. Whole exome sequencing (WES) revealed a de novo likely pathogenic heterozygous variant in the LZTR1 gene. Here we demonstrate the clinical overlap of NS, schwannomatosis and NF1 and support the inclusion of LZTR1 in gene panels for NS. In addition, we have reviewed the published literature on NS patients and variants in the LZTR1 gene, noting zygosity, inheritance pattern and clinical characteristics, to inform the management of individuals with a heterozygous LZTR1 mutation.

Case presentation

Case description

A Caucasian female received a clinical diagnosis of NS at 8 months of age, prompted by an extra nuchal skin fold (Fig. 1). Facial features were consistent with NS (large eyes, low-set cupped ears and prominent lips) and she was also found to have congenital heart defects (subaortic ridge with mild left ventricular outflow tract obstruction, mitral valve prolapse with mild mitral regurgitation and tricuspid aortic valve with mild aortic insufficiency), recurrent migraines, pectus excavatum and a history of learning disability.

Fig. 1
figure 1

Timeline of the patient’s diagnosis, symptoms and treatment

Full size image

The patient received daily growth hormone (GH) injections as a child between ages 9–16 to address short stature. Her pre-growth hormone height was 118.6 cm (< 5%ile according to CDC growth charts and 50%ile in Noonan Syndrome growth charts) (32). No significant adverse effects were reported. Her height at adulthood (~ 17 years of age) was 158.4 cm, placing her in the 25%ile on the CDC Growth Charts and 90%ile on Noonan Syndrome growth charts.

Due to delayed puberty, the patient had a trial of transdermal estradiol at the age of 14. This was transitioned to Premarin with good gonadotropic effect, and she has been on oral contraceptives since 16 years of age.

At 20 years of age, a magnetic resonance imaging (MRI) brain and intracranial magnetic resonance angiography (MRA) was performed for chronic frontal headaches. MRA was normal. MRI of the brain and subsequent MRIs of the whole spine revealed a cerebellar ectopia related to a Chiari type 1 malformation and septated syringomelia extending from C1-T6 with no evidence of an underlying mass. There were multiple plexiform neurofibromas in the sacral spine, prompting a referral to neurosurgery for suspicion of NF1. At 23 years of age, a discontinuous expansile syrinx present from C6-T6 was noted, with enlargement of the dorsal root ganglia in the cervical spine.

The plexiform neurofibromas involved all nerve roots and caused marked expansion of the sacral foramina. As a biopsy was unlikely to change management, these lesions were not biopsied and thus imaging findings were not pathologically confirmed (Fig. 2A–B). Serial MRIs have shown stable imaging findings. Upon clinical examination, the patient did not present with café-au-lait macules (CALMs), skinfold freckling, or cutaneous or subcutaneous neurofibromas. Due to the initial question of NF1, genetic testing was conducted in a step wise fashion to establish a molecular genetic diagnosis.

Fig. 2
figure 2

De-identified images of the patient’s plexiform neurofibromas. Panel A shows cornal T1 sequence, with multiple hypodense lesions arising from the sacral nerve roota, in keeping with neurofibromas (white arrows). Panel B shows sagittal T2 sequence, with neurofibromas seen as hyperintense lesions (white arrows)

Full size image

Next generation sequencing

Exon-level germline analysis of NF1 [NCBI RefSeq NM_000267.3], SPRED1 [NM_152594.2] and a multigene NS panel (BRAF [NM_004333.4], CBL [NM_005188.3], HRAS [NM_005343.2], KRAS [NM_004985.3], MAP2K1 [NM_002755.3], MAP2K2 [NM_030662.3], NRAS [NM_002524.4], PTPN11 [NM_002834.3], RAF1 [NM_002880.3], RIT1 [NM_006912.5], SHOC2 [NM_007373.3], SOS1 [NM_005633.3]) was completed on DNA extracted from blood leukocytes at The Hospital for Sick Children Genome Diagnostics Laboratory (Toronto, ON). Next Generation Sequencing (NGS) was performed using a targeted Agilent SureSelect custom capture followed by paired-end sequencing using the Illumina sequencing platform. Variant calls were generated using Genomic Analysis Tool Kit (GATK) after read alignment with the Burrows-Wheeler Aligner (BWA). Genome build NCBI37/hg19 with decoy and data analysis software SK High Coverage Clinical Pipeline was used. Multiplex ligation-dependent probe amplification (MLPA) was used to test gene dosage.

Clinical exome sequencing analysis (NGS with copy number variant [CNV]) was completed by the GeneDx Molecular Laboratory (Gaithersburg, USA) on DNA obtained via a buccal swab. The enriched targets were simultaneously sequenced with paired-end reads on an Illumina platform and genome build GRCh37/UCSC hg19. Using a custom-developed analysis tool (XomeAnalyzer), data were filtered and analysed to identify sequence variants and most deletions and duplications involving three or more coding exons (33). Sequence and CNVs are reported according to the Human Genome Variation Society (HGVS) or International System for Human Cytogenetic Nomenclature (ISCN) guidelines, respectively. Reportable PVs, likely PVs and variants of uncertain significance were reported as per AMP/ACMG guidelines (34). Secondary findings were examined in the coding regions of the gene list provided by ACMG SF v2.0 (September 2016) (35).

Molecular genetic approach

Exon-level germline genetic testing of the NF1, SPRED1, and a multigene NS panel that did not include the LZTR1 gene did not reveal any PVs in the genes tested. A single nucleotide polymorphism (SNP) microarray at the same laboratory was in-keeping with a normal female, arr(1-22,X)×2. As schwannomas and neurofibromas are not typically reported in patients with NS, the molecular genetic approach in patients with nerve sheath tumors often begins with NF1. For this patient, multi-gene panel testing for NS, NF1 and Legius syndrome were negative and a negative SNP chromosomal microarray. Of note, due to the emerging literature for LZTR1 in NS, LZTR1 was not yet included in the laboratory’s NS gene panel.

Clinical exome sequencing analysis identified a likely pathogenic heterozygous c.743G>A (p.Gly248Glu) in exon 8 of the LZTR1 gene [NM_006767.3]. Variants at this residue are normally associated with an autosomal dominant disorder. These findings are consistent with the patient’s reported clinical features. No reportable secondary findings were identified and was in accordance to the reporting structure recommended by the American College of Medical Genetics (ACMG) (35, 36).

The p.Gly248Glu variant identified in this study is located in the Kelch domain of LZTR1; heterozygous missense mutations in this region have been seen in patients with a clinical diagnosis of NS (6), NS patients with heterozygous PVs in LZTR1 (37), as well as in NS patients with a bleeding phenotype (8, 38). This variant has previously been identified in a patient with fetal pleural effusion and NS (39) and has not been observed in larger population cohorts (40).

Parental targeted variant testing confirmed this was a de novo variant in the patient. Her parents have no features of NS, supporting the causality of this variant. The patient and her family have undergone genetic counselling regarding implications of the results for her, her family and any future children. She is being followed clinically by cardiology, neurology and medical genetics.

Discussion and conclusions

Monoallelic and biallelic LZTR1 variants in NS are not well described, however, previous studies have reported germline variants causative for NS (Table 1) in either an autosomal recessive or autosomal dominant manner (6, 8, 13, 37, 38, 41–47). Of note, a different missense mutation at the same protein residue as our patient (p.Gly248Arg) was reported as pathogenic or likely pathogenic in other individuals with NS (6, 13, 38, 47), with the same designations on ClinVar. In silico analysis of variants (p.Ala116Val, p.Arg284Cys, p.Arg688Cys, p.Gly248Arg, p.His287Tyr, p.Pro520Leu, p.Ser247Asn, p.Ser122Leu, p.Tyr119Cys and p.Val456Gly) found in the LZTR1 gene support a deleterious effect (6, 7, 42).

Table 1 Previous published reports of Noonan syndrome and LZTR1 variants
Full size table

Our patient’s phenotype is unique, as not many patients with NS present with plexiform nerve sheath tumors. These are typically benign tumors, with neurofibromas characteristic of NF1 (48) and schwannomas often occurring in patients with NF2 or schwannomatosis (49). This case highlights one of the main issues in identifying NS patients—there is a degree of phenotypic overlap and uncertainty between other similar conditions; therefore, the focus of genetic testing may be on this rare manifestation with a misdiagnosis of NF1, NF2 or schwannomatosis, rather than NS. A recent study performed an extensive literature review to estimate the number of conditions that may mimic NF1, as well as examining data from 40 pediatric patients with NF1-like syndromes (50). Phenotypic overlap, particularly with skin manifestations or tissue overgrowth, was observed between NF1, NF1-like syndromes, the RASopathies and other disorders associated with higher tumor development, including phosphatase and tensin homolog (PTEN) hamartoma tumor syndromes, constitutional mismatch repair deficiency (CMMRD) syndromes, chromosomal abnormalities and multiple endocrine neoplasia (MEN) syndromes (50). Furthermore, genotypic overlap can also be seen between NF1-like syndrome and other disorders, including NS (50).

Besides our patient, there have been a few reports of neurofibromas/schwannomas in patients with a RASopathy phenotype. One patient presented with an overgrowth of peripheral nerve sheaths, suggestive of a schwannoma or neurofibroma (51). The final diagnosis was multiple diffuse schwannomas; although this is suggestive of NF2 or schwannomatosis, a germline KRAS variant (p.Lys5Glu) was identified (51). In one family with unaffected parents and four affected children with NS, a c.2220-17C>A (p.Tyr741Hisfs*89) maternally-inherited splice variant was found; several members of this family had suggestive signs of schwannomas on MRI analysis (37). Another individual with a c.740C>A (p.Ser247Asn) variant in LZTR1 developed multiple schwannomas in the right arm, however, no material from the schwannomas was available for molecular testing (6). This individual was the mother of a patient with NS (6).

Carriers of LZTR1 have phenotypic heterogeneity and management of LZTR1-NS patients has not been well described. Guidelines regarding the management of individuals with NS varies and is based on the clinical manifestations. Some common treatments include: treatment for cardiovascular anomalies, early intervention programs for developmental disabilities, treatment for serious bleeding conditions, GH treatment for short stature and further monitoring for abnormalities (2). In comparison, germline or mosaic mutations in the LZTR1 (7, 19, 20) and the SMARCB1 (52) genes have been associated with schwannomatosis, although the link between LZTR1 schwannomatosis and the development of other tumors has not been clearly defined. Current clinical management for schwannomatosis recommends that individuals have a baseline brain and spine MRI in late childhood/early adulthood to monitor the disease and management in adulthood should be performed by a neurologist or neurofibromatosis specialist to manage pain (53). Whole body MRI and increased surveillance can be considered if the patient is symptomatic (53). The guidelines on how to manage LZTR1-NS and LZTR1-schwannamatosis patients are not clear; therefore, a reasonable approach would be management of patients under both NS and schwannomatosis guidelines. Clarification of the genetic etiology of NS should be pursued early in the course of NS management as this may have implications on the use of GH therapy in a patient with LZTR1-NS. This includes consideration of tumor development and cardiac abnormalities, as individuals with NS are at increased risk, and there have been reports of tumors (54–57) and adverse cardiac reactions (58) following GH therapy.

Our proband’s likely PV in the LZTR1 gene was not detected on the original NF1, SPRED1 and multigene NS panel. Eventually, clinical exome sequencing analysis was able to identify the causative gene, which was later confirmed to be absent in her clinically unaffected parents. In 2018, the Clinical Genome Resource (ClinGen) RASopathy expert panel assessed 19 genes, including LZTR1, and found a strong association between LZTR1 and AD NS (59). As of 2020, the RASopathy expert panel definitively associates LZTR1 with AD NS and has stated there is strong evidence between LZTR1 and AR NS. Since then, LZTR1 has been a common new addition to many commercial and academic and laboratory NS panels. Based on our findings, inclusion of additional NS genes to schwannomatosis panels could also be considered. Due to the high de novo rate of variants causing NF1, NS and schwannomatosis, an alterative approach may be trio whole exome sequencing to aid in the interpretation of variants.

Based on previous reports and our findings, individuals with NS and a monoallelic or biallelic germline LZTR1 mutation may be at an increased risk of developing schwannomas and may meet the diagnostic criteria for schwannomatosis. Additional case reports of NS and variants in LZTR1, as well as functional studies showing the involvement of LZTR1 in the RAS/MAPK pathway would provide support for the creation of management guidelines for LZTR1-NS individuals. Due to the difficulties with overlapping phenotypes and genotypes, we believe that all NF1-like syndromes should be assessed in young individuals when identifying a causative gene and disorder. As clinical guidelines and gene panels change, clinicians should perform annual assessments of these patients to re-evaluate their original diagnosis and management. Due to the clinical implications for this patient presenting with schwannomas and NS, our patient will be followed as per NS and schwannomatosis surveillance guidelines to screen for nervous system tumors and monitor for tumor development.