Silver–Russell syndrome (SRS, MIM #180860) is a rare imprinting disorder (1:30.000–1:100.000), characterized by severe prenatal and postnatal growth retardation, relative macrocephaly at birth, body asymmetry, prominent forehead, and feeding difficulties. Clinical diagnosis is based on the occurrence of at least four out of six clinical signs, in accordance with the Netchine–Harbison clinical scoring system (NH-CSS), but molecular testing is recommended in patients with ≥3/6 criteria. The most common underlying mechanisms are loss of methylation of the paternal allele at H19/IGF2:IG-DMR in the 11p15.5 region (IC1_LoM and 30%–60% of cases) and maternal uniparental disomy of chromosome 7 (UPD(7)mat and 5%–10% of cases) involving GRB10:alt-TSS-DMR, PEG10:TSS-DMR, and MEST:alt-TSS-DMR. Other rare abnormalities, such as UPD(14)mat, UPD(16)mat, UPD(20)mat, and pathogenic variants within the CDKN1C, IGF2, PLAG1, and HMGA2 genes have been described in a minor fraction of SRS patients. To date, the diagnosis remains clinically based in about 40% of SRS patients (Wakeling et al., 2017; Abi Habib et al., 2018).

Among the “non-cardinal” clinical features of SRS, Isolated LO Myoclonus-Dystonia (M-D) has been reported in six SRS patients, all carrying maternal uniparental disomy of chromosome 7 (UPD(7)mat) (Guettard et al., 2008; El-Shewy and Luttrell, 2009; Stark et al., 2010; Sheridan et al., 2013; Martins et al., 2020). M-D is a rare movement disorder with onset in the first 2 decades of life, characterized by truncal and upper-limb myoclonus associated with cervical dystonia and/or writer’s cramp. Depression, panic attacks, and obsessive–compulsive disorders are also included in the non-motor symptoms (Peall et al., 2011). Loss-of-function mutations in SGCE (7q21.3) have been identified in 20% of M-D patients (Peall et al., 2014), where the disorder shows autosomal dominant inheritance with reduced penetrance when transmitted by the mother, with the gene being maternally imprinted. This finding led to the hypothesis that the lack of gene expression in UPD(7)mat SRS patients exhibiting M-D might be the driver mechanism of the neurological manifestation. However, the prevalence of M-D in SRS cohorts with UPD(7)mat is low (Guettard et al., 2008; Sheridan et al., 2013; Lokulo-Sodipe et al., 2020; Martins et al., 2020). To our knowledge, unlike UPD(7)mat SRS patients, SRS patients exhibiting dystonia due to hypomethylation of the paternal allele at H19/IGF2:IG-DMR have not yet been described.

Here, we report the clinical history of a patient referred for apparently Isolated Lateralized Overgrowth (ILO) and molecularly diagnosed with SRS. Interestingly, since her 16th birthday, our SRS girl has displayed symptoms suggestive of hand dystonia.

In accordance with the manufacturer’s protocols, DNA was extracted from peripheral blood lymphocytes (automated extractor, Tecan Group Ltd.) of sibs and parents, from the proband’s epithelial buccal cells (Oragene tubes OG-575 and DNA kit, DNA Genotek Inc.), and from proband’s urine samples (Wizard Genomic DNA Purification Kit, Promega).

The MRC-Holland kit (MRC Holland, Amsterdam, Netherlands) ME-030 BWS/RSS C3-1,017 was used for the molecular diagnosis, in accordance with the kit instructions. Raw data were analyzed by using Coffalyser.Net software (version 140,701- MRC Holland).

Whole-exome sequencing (WES) was performed on genomic DNA from proband and parents (trio) at BIODIVERSA srl service (Milan, Italy) using the SureSelect Human All Exon V7 library and the Illumina HiSeq X platform. Bioinformatic analyses were carried out at the Bioinformatics Unit of the IRCCS Istituto Auxologico Italiano (Milano, Italy). FASTQ data were aligned to the reference genome assembly GRCh37/hg19 using the maximal exact matches algorithm in the Burrows–Wheeler Aligner (BWA) (v.0.7.10). PCR duplicate removal was performed by using Picard (v1.119) (picard.sourceforge.net/). UnifiedGenotyper (GATK v3.7) was used to locally realign insertion/deletions (InDels) and recalibrate base quality scores. Variants were visually inspected using the Integrative Genomics Viewer (IGV, Broad Institute) and further annotated using wANNOVAR (http://wannovar.wglab.org/). Variants with a coverage of >20x and a variant allele frequency of more than or equal to 0.35 were selected for quality filtering. Next, we designed the following two virtual panels of genes to scrutinize putative variants: one related to overgrowth syndromes and the other to dystonia (Supplementary Table S1). Variants were filtered by minor allele frequency (MAF) < 1% in the 1,000 Genomes, Genome Aggregation Databases (gnomAD), and Exome Aggregation Consortium (ExAC) databases. In silico prediction of missense variants’ pathogenicity was performed by combining the PolyPhen-2, SIFT, and CADD algorithms. The interpretation of the variants was based on the classification by the VarSome and InterVar databases (Li and Wang, 2017; Kopanos et al., 2019) in accordance with the ACMG–AMP (American College of Medical Genetics and Genomics/Association for Molecular Pathology) guidelines (Richards et al., 2015; Rehder et al., 2021). Selected variants were validated by Sanger sequencing.

Array CGH analysis was performed using the high-resolution Human Genome CGH Microarray Kits 2 × 400K (median resolution of 16 kb) (Agilent Technologies) in accordance with the manufacturer’s instructions. Data were analyzed using Agilent CytoGenomics 3.0. Detected copy number variations (CNVs) were interpreted in accordance with the work of Crippa et al. (2019).

Silver–Russell syndrome is a heterogeneous clinical and genetic condition, mainly characterized by lower growth parameters than the population. The growth rate and the final height in adulthood are highly heritable continuous quantitative traits. A number of genetic, hormonal, environmental, and nutritional factors can concur to influence the final phenotype. Given the complex etiology of this trait, suspicion of growth retardation is particularly difficult (Gudbjartsson et al., 2008). Here, we describe a patient initially diagnosed as BWSp by her ILO, in accordance with the diagnostic consensus of the syndrome (Brioude et al., 2018). Despite asymmetric body growth being one of the six specific clinical criteria listed in the NH-CSS for SRS (Wakeling et al., 2017), our patient’s growth parameters at birth and postnatally did not raise the suspicion of SRS (Table 1). However, her growth chart was below the familial target, since her mid-parental target height is higher than the reference one in the population. To our knowledge, few SRS patients have been reported with normal growth parameters (height > −2 SDS) (Bliek et al., 2006; Zeschnigk et al., 2008; Russo et al., 2016; Patti et al., 2018; Mackay et al., 2019; Lokulo-Sodipe et al., 2020), and only two patients have been described as having a final height of about 170 cm without receiving growth hormone treatment (Bliek et al., 2006; Lokulo-Sodipe et al., 2020). However, in both studies, parental height was not indicated. The patient reported by Bliek and others differed from our case for SGA (the birth length was unknown) and feeding difficulties, while that reported by Lokulo-Sodipe et al. (2020) shares two NH-CSS diagnostic criteria with our case. This suggests that, in evaluating patients with an ILO phenotype without marked growth delay, the presence of NH-CSS criteria (such as relative macrocephaly) and the mid-parental height should be considered, as the genetic background can influence the syndrome’s phenotypic expression.

Moreover, in the distinctive growth chart, our patient shows a neurological disorder, manifesting hand-writer’s cramp, and psychiatric disorders such as bipolar disorder and depression, which may be part of an M-D’s phenotype (Peall et al., 2011). To our knowledge, this patient is the first SRS exhibiting this disorder, associated with hypomethylation of the paternal allele at H19/IGF2:IG-DMR, and we hypothesized that an additional hit might be involved. WES analysis did not reveal pathogenic variants in candidate genes previously associated with M-D; out of the few identified VUSs (Supplementary Table S2), we focused on those with a putative interaction with the IGF2 pathway. Specifically, we focused on the VPS16 variant (p.Leu729Val). The VPS16 gene encodes a protein belonging to the HOPS (homotypic fusion and protein sorting) complex, a structural bridge composed of four Vps-C core proteins (VPS11, VPS16, VPS18, and VPS33A), and two additional subunits (VPS39 and VPS41), which are necessary for the fusion of late endosomes and autophagosomes with the lysosomes in the cytoplasm (Luzio et al., 2007; Ostrowicz et al., 2010; Monfrini et al., 2021). Figure 2A summarizes the connection between the degradation of IGF2 and the recovery of extracellular M6P-lysosomal enzymes through the plasma membrane cation-independent mannose-6-phosphate receptor (CI-M6PR/IGF2R) (Oka et al., 1985; Braulke et al., 1990; El-Shewy and Luttrell, 2009; Zavorka et al., 2016; Wang et al., 2017; 2020; Bajaj et al., 2019). Dysregulation of the endolysosomal and autophagic systems is linked to the pathogenesis of dystonia, which is a feature of the clinical presentation of many lysosomal storage disorders (Monfrini et al., 2021).

Several heterozygous VPS16 variants (17 loss-of-function and three missense variants, Supplementary Table S3) were found to co-segregate with dystonia in multigenerational families displaying a dominant pattern of inheritance with incomplete penetrance and wide clinical expressivity of the age of onset and dystonia severity (Steel et al., 2020; Ostrozovicova et al., 2021; Park et al., 2022. Patients with heterozygous VPS16 variants often display psychiatric features such as depression or bipolar disorder (Monfrini et al., 2021). Our patient fits the clinical phenotype of the reported VPS16-mutated patients and the segregation of their variants (overviewed in Supplementary Table S4). Clinical data were collected from all 32 heterozygous VPS16 patients described in the literature. Many patients displayed early-onset dystonia (0–20 years, 25/32), and the most frequent first symptom was isolated writer’s cramp (10/32) (Cai et al., 2016; Steel et al., 2020; Li L.-X. et al., 2021; Li X. et al., 2021; Ostrozovicova et al., 2021; Pott et al., 2021; Gu et al., 2022; Park et al., 2022; Petry-Schmelzer et al., 2022).

Despite few missense variants in the C-terminal domain of VPS16 being reported, in vitro mutagenesis experiments (Ala669Asp and Leu725Glu) support the relevant role of the HOPS domain by showing that variants in this domain may severely affect the VPS16-VPS33A binding and mutations proximal to the interaction surface could interfere in the binding (Graham et al., 2013). Furthermore, Park et al. identified 11 missense variants in a wide dystonic cohort. Although missense variants were not enriched in this cohort compared to non-dystonic patients, the authors did not exclude their roles in the disease, or other genetic, epigenetic, and environmental factors that may contribute to disease manifestation and severity (Park et al., 2022). Otherwise, since no other patients affected by dystonia have so far been reported with the VPS16 (NM_022575.4):c.2185C>G:p.Leu729Val, the proband’s hand dystonia might be one of the rare clinical phenotypes of the SRS spectrum, whose pathomechanism still needs to be elucidated. Based on this hypothesis, the patient’s psychiatric manifestations might be interpreted as a more severe manifestation of familial depressive disorder.

To sum up, we hypothesize that in our proband, the low extracellular levels of IGF2, due to the IC1_LoM, may worsen the effect of the co-present VPS16 variant and contribute to the expression of the neurological phenotype (Figure 2B).

Discrepant phenotypes have been reported for SRS and BWSp (Mackay et al., 2019), but confounding factors such as mid-parental height should be considered, as well as the presence of the diagnostic criteria of either syndrome. The path to the molecular diagnosis in our patient has been tortuous, and recognition of SRS came late. It is true to say that the result of the genetic test in childhood could have led to clinical diagnosis and the proper follow-up. Indeed, as an adult, the patient has expressed disappointment about the long-lasting and ineffective follow-up for the timely diagnosis of embryonic tumors. On the other hand, SRS patients are susceptible to developing metabolic syndrome in adulthood (Patti et al., 2018), whereas BWS/ILO patients can interrupt the follow-up. Interestingly, the deficit of IGF2 may correspond to the expression of a VPS16-related dystonia phenotype.

The description and discussion of this unique SRS patient, in addition to being useful in the clinical setting to reach the proper diagnosis and management, may enhance the understanding of the molecular basis of the broad SRS spectrum.