Pelizaeus-Merzbacher disease (PMD, OMIM:312080) is a severe X-linked recessive hypomyelinating leukodystrophy, with initial clinical presentation of nystagmus and developmental delay, progressing to spastic quadriplegia and ataxia (Sistermans et al., 1998). PMD is caused by disease-causing variants in the proteolipid protein gene (PLP1, OMIM:300401), resulting in the improper formation of myelin in the central nervous system (Wolf et al., 1993).

Gene duplications of PLP1 represent a common cause of PMD and account for 50%–75% of all clinically manifest disease-causing variants and yield the classic PMD phenotype (Garbern et al., 1999; Osório and Goldman, 2018). Both triplosensitivity and haploinsufficiency of the PLP1 gene have been well-established (Raskind et al., 1991; Woodward et al., 1998; Mimault et al., 1999; Cailloux et al., 2000; Hobson et al., 2000; Inoue et al., 2002; Wolf et al., 2005; Combes et al., 2006; Regis et al., 2009; Grossi et al., 2011). Molecular genetic testing of PLP1 disorders is usually performed by gene-targeted deletion/duplication analysis using multiplex ligation-dependent probe amplification (MLPA), microarray, quantitative PCR, long-range PCR, fluorescent in situ hybridization, and combinations thereof (Wolf et al., 1993), however these methods are not designed to detect both the orientation and additional genomic context of the affected region. While male PMD patients typically do not reproduce, female carriers of pathogenic variants carry a 50% chance of transmitting the variant in each pregnancy, and prenatal testing for a pregnancy at increased risk is available if the pathogenic variant is known (Wolf et al., 1993).

We now present our prenatal PLP1 duplication case, where the older male child of the carrier mother had an atypical neuro-developmental clinical presentation inconsistent with the microarray-detected PMD pathogenic duplication, which prompted the analyses of this structural variant using a novel laboratory technology, optical genome mapping (OGM) (Dremsek et al., 2021; Mantere et al., 2021). Based on the findings of OGM which showed a complex variant involving an inverted duplication of PLP1, and extended segregation analysis, this novel structural variant was reclassified, and a healthy child was delivered.

A woman planning a pregnancy was referred for antenatal genetic consultation to the Clinical Institute of Genomic Medicine (CIGM), University Medical Centre Ljubljana (UMCL), Slovenia, because of clinical problems in her older child. Because of the pandemic, she received a teleconsultation. Her first child, a 7-year-old boy reportedly had an unspecific short attention span, mild central, and axial hypotonia, elements of autistic features, and joint hypermobility. The perinatal history of the boy was unremarkable as reported by the parents via teleconsultation and the boy was referred for genetic testing using microarray.

The mother was already in the first trimester of pregnancy by the time microarray was performed on her son. The microarray analysis revealed a duplication at arr[GRCh37] Xq22.2(103008605_103172424)×2, approximately 163.8 kb in size (Figure 1). The duplication included the disease-associated PLP1 gene (OMIM:300401), the duplication of which was previously reported as pathogenic and associated with PMD, and the RAB9 (OMIM:300285) gene, that has not yet been associated with any disease. The reported clinical presentation in the proband was atypical of PMD, therefore, we performed segregation by microarray on the parents, which revealed the mother to be a carrier of the PLP1 duplication (164.7 kb(arr[GRCh37] Xq22.2(103003034_103167770)×3]. Subsequently, amniocentesis was performed, and the presence of the duplication was also confirmed in a male fetus (Figure 1). Due to the difference between postnatal and prenatal array design, the reported size of the duplication in the fetus was approx. 285.0 kb (arr[GRCh37] Xq22.2(103003034_103288063)×2 or arr[GRCh38] Xq22.2(103748106_104033500)×2], and additionally includes gene TMSB15B (OMIM:301011), that has not yet been associated with any disease. The equal size of the variant in the proband was confirmed by a control array of the proband on the prenatal chip design (Figure 1).

However, the atypical clinical presentation of the proband made us further question the validity of the interpretation of this familial variant as pathogenic.

This presented a unique prenatal counseling challenge, as the non-specific and non-severe clinical manifestation in the older child did not indicate the need for a pregnancy intervention despite the detection of a pathogenic variant for a severe, progressive, early-onset disorder.

Indeed, the review of the proband’s medical history, the clinical re-examination, and the repeated MRI scan confirmed an atypical clinical presentation, inconsistent with PMD. Although behavioral issues were reported in kindergarten, and initial evaluation for autism spectrum disorder confirmed elements of autistic behavior, at the time of examination the proband displayed a short attention span and joint hypermobility, but apart from this, successfully attended regular primary school. Furthermore, the MRI displayed sequelae of T2 and FLAIR hyperintensities in the right hemisphere, resulting from a head injury at age 8 months that required surgical intervention, but showed no signs of hypomyelination that would be consistent with PMD.

To investigate the nature of the identified PLP1 duplication further, we tested the proband by using a novel laboratory technology, optical genome mapping (OGM). OGM has been available for routine genetic testing in Slovenia since 2021 and enables the detection of structural genomic variants, such as inversions, duplications, insertions, deletions, translocations, and complex rearrangements in DNA size ranges previously undetectable by other technologies.

Since the carrier mother was from a large family with six neurologically normal brothers, we were able, in parallel, to perform segregation analysis to find possible additional unaffected male carriers of the familial variant.

Optical genome mapping analysis of the proband revealed the genomic gain in the Xq22.2, to be approximately 432.7 kb in size. The variant is a previously undescribed complex structural variant of PLP1—an inverted duplication of region [GRCh38] Xq22.2(103734669_103958435)×2 approx. 223.8 kb in size, followed by tandem duplication of region [GRCh38] Xq22.2(103871512_104080423)×2 approx. 208.9 kb in size. This inverted duplication, followed by the tandem duplication, is inserted at approximate position [GRCh38] Xq22.2(104080423_104085533).

The inverted duplication [GRCh38] Xq22.2(103734669_103958435)×2 contains the genes PLP1 (OMIM:300401), RAB9 (OMIM:300285), TMSB15B (OMIM:301011) and a lamina-associated domain (LAD), while the tandem duplication of the region [GRCh38] Xq22.2(103871512_104080423)×2, contains the gene TMSB15B (OMIM:301011) and same LAD domain, thus the inverted additional copy of the PLP1 gene contains LAD domains at both ends (Figure 2, Supplementary Figure S2). The copy number gain of the TMSB15B was not detected by initial microarray of the proband due to lack of probes in the region.

The extended segregation analysis revealed one of the mother’s brothers, and the grandmother to be additional carriers of the inverted duplication (Figure 2, Supplementary Figure S1). None of the three adult carriers (mother, uncle, grandmother) had any signs or symptoms of neurological disease.

Based on this additional information, we estimated this PLP1 inverted duplication variant is not associated with clinical observations in the proband and does not represent an increased risk for classic PMD. Accordingly, we have classified it as a variant of unknown significance (ACMG Criteria 1A, 2A, 4J, total points: 0.70) and deposited it in the ClinVar database (https://www.ncbi.nlm.nih.gov/clinvar/variation/1810757/). The mother continued with the pregnancy and gave birth at full term to a healthy boy. At 6 months old, her second son had a normal neurological examination, while the neurological status of the proband at his last examination at 9 years of age remains within normal, except for joint hypermobility, and he successfully attends regular school.

The pathogenicity of PLP1 duplication has been well established in humans (Woodward et al., 1998; Mimault et al., 1999; Wolf et al., 2005; Regis et al., 2009) and PLP1 overexpression as the underlying mechanism has been proposed based on animal models (Readhead et al., 1994; Peyron et al., 1997). All previously described PLP1 duplications in males were reported to lead to central hypomyelination and progressive neurological deterioration with cognitive decline and spasticity, which are the hallmarks of PMD.

The discovery, by microarray, of a previously reported pathogenic duplication variant for a severe, progressive, early-onset disorder but an atypical clinical presentation in the proband, presented a unique prenatal counseling challenge.

Based on the atypical clinical presentation in the proband, we hypothesized that the microarray-identified duplication variant itself may not be typical. PLP1-related disorders of central nervous system myelin formation include a range of phenotypes from Pelizaeus-Merzbacher disease to spastic paraplegia 2 (Wolf et al., 1993). While all previous PLP1 duplications in males were considered pathogenic, PLP1 deletions in the context of various genomic rearrangements have already been reported to contribute to different dysmyelinating phenotypes (Inoue et al., 2002). So far, only one smaller, partially-overlapping inversion without an associated copy-number change, containing only the PLP1 gene, has been identified in a presumably normal individual (Database of genomic variants, esv7229) (Ahn et al., 2009).

Because microarray analysis does not enable us to detect structural changes other than copy number variants, we examined the DNA of the patient using a novel technology, optical genome mapping. Optical genome mapping is based on the imaging of labeled ultra-high molecular weight DNA molecules and provides a high-resolution detection of structural variants such as inversions, insertions, and even balanced translocations. Indeed, optical genome mapping showed the PLP1 variant in this family to be a previously undescribed complex structural variant—an inverted duplication, followed by a tandem duplication, located adjacent to the normal copy of the gene, and possessing putative lamina-associated domains at both ends (Figure 2, Supplementary Figure S2).

Although it remains unknown why this rare variant does not result in PMD, several genetic mechanisms are known to influence the expression of genes through topological means or rearrangement of regulatory regions, among which lamina-associated domains have been reported to actively inhibit transcription (Phillips-Cremins and Corces, 2013; van Steensel and Belmont, 2017; Huang et al., 2021; Hong and Cohen, 2022). While the mechanism by which the detected inverted duplication apparently avoids leading to the development of PMD currently remains unknown, the segregation analysis (Figure 2), showing three male carriers of this structural variant lacking clinical features associated with PMD, allows us to reclassify this variant from pathogenic to a variant of unknown significance on clinical criteria alone, with important implications in prenatal counseling and care. As the current molecular genetic testing of PLP1 disorders consists of techniques not designed to detect orientation and not able to detect structural variants other than copy number variants, it is possible that similar rearrangements will be detected with increased use of novel technologies, such as optical genome mapping.

To conclude, we suggest that optical genome mapping may help both reclassify and reinterpret structural variants presenting as typical duplications on microarray, as it provides additional information regarding the orientation and genomic context of the copy-number change. Examining structural variants with such novel methods is warranted especially in cases with atypical clinical presentation and may in these cases lead to improved prenatal and postnatal genetic counseling.